User login
‘Impressive’ SOLO3 results should influence practice
In this edition of “Applying research to practice,” I highlight a study suggesting olaparib is helpful in patients BRCA mutations experiencing multiple relapses of ovarian cancer.
SOLO3 was the first phase 3 trial comparing the oral PARP inhibitor olaparib (OLA; 300 mg twice daily) with physician’s choice of intravenous single-agent chemotherapy (TPC) in relapsed high-grade serous or endometroid ovarian, fallopian tube, or primary peritoneal cancer (J Clin Oncol. 2020 Feb 19. doi: 10.1200/JCO.19.02745).
The trial involved 266 BRCA-mutated patients who had received two (approximately 50%) or more lines of platinum-based TPC. All patients were required to be completely platinum sensitive (progression beyond 12 months from last platinum exposure) or partially platinum sensitive (progression within 6-12 months).
Women were randomized to receive either OLA or nonplatinum TPC (pegylated liposomal doxorubicin, paclitaxel, gemcitabine, or topotecan). After an amendment to the study in 2017, the primary endpoint was objective response rate, determined by blinded independent central review, with a variety of secondary endpoints.
Among 223 patients with measurable disease, the objective response rate was 72.2% with OLA and 51.4% with TPC (odds ratio, 2.53; P = .002). Across all patients, the median progression-free survival was significantly better with OLA (13.4 months) than with TPC (9.2 months; P = .013). Overall survival data were immature.
The superiority of OLA for the primary endpoint was maintained in multiple subgroups of patients, including those who had received only two prior lines of therapy (OR, 3.44) and those who had three or more prior lines (OR, 2.21). Time to first subsequent therapy (HR, 0.48) and time to treatment discontinuation or death (HR, 0.17) were significantly longer for OLA than for TPC.
Adverse events were consistent with the established safety profiles of OLA and chemotherapy. The most common grade 3 or higher adverse events were anemia (21.3%) with OLA and neutropenia (15.8%) and hand-foot syndrome (11.8%) with TPC.
However, median treatment durations were substantially and consistently longer for OLA than for TPC, and there were fewer treatment discontinuations because of toxicity for OLA than for TPC. At the time of data cutoff, 43 patients in the OLA group and 1 patient in the TPC cohort remained on treatment.
How these results influence practice
The results of the SOLO3 trial are clear: Treatment with OLA is a reasonable alternative to nonplatinum-containing chemotherapy for women with BRCA mutations and platinum-sensitive ovarian cancer. OLA is a “chemotherapy-free” option for these patients in the second- and later-line settings.
Less clear are the following:
- How many patients with BRCA mutations will not have already received a PARP inhibitor in the frontline maintenance setting in the future? SOLO3 required modification in the accrual target and endpoint because of challenges in patient recruitment from the entry of PARP inhibitors into routine clinical practice.
- Would OLA be superior to a carboplatin doublet rather than a nonplatinum single agent in patients with two prior relapses of platinum-sensitive ovarian cancer? Standard practice would be for patients in the second-line setting to receive a platinum doublet.
- Is extending the platinum-free interval a worthwhile objective, or would some patients prefer a finite interval of a platinum doublet over an indefinite period of treatment with OLA?
All phase 3 clinical trials have limitations since they require years to complete and the applicability of the results are challenged by intercurrent advances in treatment options and diagnostic tests.
However, overall, the results of SOLO3 are impressive and should influence clinical practice for the subset of relapsed ovarian cancer patients who would have qualified to participate in it. OLA represents an important treatment advance for a group of patients who are trying to string together remission after remission, with limited negative impact on quality of life.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
In this edition of “Applying research to practice,” I highlight a study suggesting olaparib is helpful in patients BRCA mutations experiencing multiple relapses of ovarian cancer.
SOLO3 was the first phase 3 trial comparing the oral PARP inhibitor olaparib (OLA; 300 mg twice daily) with physician’s choice of intravenous single-agent chemotherapy (TPC) in relapsed high-grade serous or endometroid ovarian, fallopian tube, or primary peritoneal cancer (J Clin Oncol. 2020 Feb 19. doi: 10.1200/JCO.19.02745).
The trial involved 266 BRCA-mutated patients who had received two (approximately 50%) or more lines of platinum-based TPC. All patients were required to be completely platinum sensitive (progression beyond 12 months from last platinum exposure) or partially platinum sensitive (progression within 6-12 months).
Women were randomized to receive either OLA or nonplatinum TPC (pegylated liposomal doxorubicin, paclitaxel, gemcitabine, or topotecan). After an amendment to the study in 2017, the primary endpoint was objective response rate, determined by blinded independent central review, with a variety of secondary endpoints.
Among 223 patients with measurable disease, the objective response rate was 72.2% with OLA and 51.4% with TPC (odds ratio, 2.53; P = .002). Across all patients, the median progression-free survival was significantly better with OLA (13.4 months) than with TPC (9.2 months; P = .013). Overall survival data were immature.
The superiority of OLA for the primary endpoint was maintained in multiple subgroups of patients, including those who had received only two prior lines of therapy (OR, 3.44) and those who had three or more prior lines (OR, 2.21). Time to first subsequent therapy (HR, 0.48) and time to treatment discontinuation or death (HR, 0.17) were significantly longer for OLA than for TPC.
Adverse events were consistent with the established safety profiles of OLA and chemotherapy. The most common grade 3 or higher adverse events were anemia (21.3%) with OLA and neutropenia (15.8%) and hand-foot syndrome (11.8%) with TPC.
However, median treatment durations were substantially and consistently longer for OLA than for TPC, and there were fewer treatment discontinuations because of toxicity for OLA than for TPC. At the time of data cutoff, 43 patients in the OLA group and 1 patient in the TPC cohort remained on treatment.
How these results influence practice
The results of the SOLO3 trial are clear: Treatment with OLA is a reasonable alternative to nonplatinum-containing chemotherapy for women with BRCA mutations and platinum-sensitive ovarian cancer. OLA is a “chemotherapy-free” option for these patients in the second- and later-line settings.
Less clear are the following:
- How many patients with BRCA mutations will not have already received a PARP inhibitor in the frontline maintenance setting in the future? SOLO3 required modification in the accrual target and endpoint because of challenges in patient recruitment from the entry of PARP inhibitors into routine clinical practice.
- Would OLA be superior to a carboplatin doublet rather than a nonplatinum single agent in patients with two prior relapses of platinum-sensitive ovarian cancer? Standard practice would be for patients in the second-line setting to receive a platinum doublet.
- Is extending the platinum-free interval a worthwhile objective, or would some patients prefer a finite interval of a platinum doublet over an indefinite period of treatment with OLA?
All phase 3 clinical trials have limitations since they require years to complete and the applicability of the results are challenged by intercurrent advances in treatment options and diagnostic tests.
However, overall, the results of SOLO3 are impressive and should influence clinical practice for the subset of relapsed ovarian cancer patients who would have qualified to participate in it. OLA represents an important treatment advance for a group of patients who are trying to string together remission after remission, with limited negative impact on quality of life.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
In this edition of “Applying research to practice,” I highlight a study suggesting olaparib is helpful in patients BRCA mutations experiencing multiple relapses of ovarian cancer.
SOLO3 was the first phase 3 trial comparing the oral PARP inhibitor olaparib (OLA; 300 mg twice daily) with physician’s choice of intravenous single-agent chemotherapy (TPC) in relapsed high-grade serous or endometroid ovarian, fallopian tube, or primary peritoneal cancer (J Clin Oncol. 2020 Feb 19. doi: 10.1200/JCO.19.02745).
The trial involved 266 BRCA-mutated patients who had received two (approximately 50%) or more lines of platinum-based TPC. All patients were required to be completely platinum sensitive (progression beyond 12 months from last platinum exposure) or partially platinum sensitive (progression within 6-12 months).
Women were randomized to receive either OLA or nonplatinum TPC (pegylated liposomal doxorubicin, paclitaxel, gemcitabine, or topotecan). After an amendment to the study in 2017, the primary endpoint was objective response rate, determined by blinded independent central review, with a variety of secondary endpoints.
Among 223 patients with measurable disease, the objective response rate was 72.2% with OLA and 51.4% with TPC (odds ratio, 2.53; P = .002). Across all patients, the median progression-free survival was significantly better with OLA (13.4 months) than with TPC (9.2 months; P = .013). Overall survival data were immature.
The superiority of OLA for the primary endpoint was maintained in multiple subgroups of patients, including those who had received only two prior lines of therapy (OR, 3.44) and those who had three or more prior lines (OR, 2.21). Time to first subsequent therapy (HR, 0.48) and time to treatment discontinuation or death (HR, 0.17) were significantly longer for OLA than for TPC.
Adverse events were consistent with the established safety profiles of OLA and chemotherapy. The most common grade 3 or higher adverse events were anemia (21.3%) with OLA and neutropenia (15.8%) and hand-foot syndrome (11.8%) with TPC.
However, median treatment durations were substantially and consistently longer for OLA than for TPC, and there were fewer treatment discontinuations because of toxicity for OLA than for TPC. At the time of data cutoff, 43 patients in the OLA group and 1 patient in the TPC cohort remained on treatment.
How these results influence practice
The results of the SOLO3 trial are clear: Treatment with OLA is a reasonable alternative to nonplatinum-containing chemotherapy for women with BRCA mutations and platinum-sensitive ovarian cancer. OLA is a “chemotherapy-free” option for these patients in the second- and later-line settings.
Less clear are the following:
- How many patients with BRCA mutations will not have already received a PARP inhibitor in the frontline maintenance setting in the future? SOLO3 required modification in the accrual target and endpoint because of challenges in patient recruitment from the entry of PARP inhibitors into routine clinical practice.
- Would OLA be superior to a carboplatin doublet rather than a nonplatinum single agent in patients with two prior relapses of platinum-sensitive ovarian cancer? Standard practice would be for patients in the second-line setting to receive a platinum doublet.
- Is extending the platinum-free interval a worthwhile objective, or would some patients prefer a finite interval of a platinum doublet over an indefinite period of treatment with OLA?
All phase 3 clinical trials have limitations since they require years to complete and the applicability of the results are challenged by intercurrent advances in treatment options and diagnostic tests.
However, overall, the results of SOLO3 are impressive and should influence clinical practice for the subset of relapsed ovarian cancer patients who would have qualified to participate in it. OLA represents an important treatment advance for a group of patients who are trying to string together remission after remission, with limited negative impact on quality of life.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
Perspective from the heartland: Cancer care and research during a public health crisis
I have no knowledge of, or experience with, managing a cancer patient during a pandemic. However, from the published and otherwise shared experience of others, we should not allow ourselves to underestimate the voracity of the coronavirus pandemic on our patients, communities, and health care systems.
Data from China suggest cancer patients infected with SARS-CoV-2 face a 3.5 times higher risk of mechanical ventilation, intensive care unit admission, or death, compared with infected patients without cancer (Lancet Oncol 2020;21:335-7).
Health care workers in Seattle have also shared their experiences battling coronavirus infections in cancer patients (J Natl Compr Canc Netw. 2020 Mar 20. doi: 10.6004/jnccn.2020.7560). Masumi Ueda, MD, of Seattle Cancer Care Alliance, and colleagues reviewed their decisions in multiple domains over a 7-week period, during which the state of Washington went from a single case of SARS-CoV-2 infection to nearly 650 cases and 40 deaths.
Making tough treatment decisions
Dr. Ueda and colleagues contrasted their customary resource-rich, innovation-oriented, cancer-combatting environment with their current circumstance, in which they must prioritize treatment for patients for whom the risk-reward balance has tilted substantially toward “risk.”
The authors noted that their most difficult decisions were those regarding delay of cancer treatment. They suggested that plans for potentially curative adjuvant therapy should likely proceed, but, for patients with metastatic disease, the equation is more nuanced.
In some cases, treatment should be delayed or interrupted with recognition of how that could result in worsening performance status and admission for symptom palliation, further stressing inpatient resources.
The authors suggested scenarios for prioritizing cancer surgery. For example, several months of systemic therapy (ideally, low-risk systemic therapy such as hormone therapy for breast or prostate cancer) and surgical delay may be worthwhile, without compromising patient care.
Patients with aggressive hematologic malignancy requiring urgent systemic treatment (potentially stem cell transplantation and cellular immunotherapies) should be treated promptly. However, even in those cases, opportunities should be sought to lessen immunosuppression and transition care as quickly as possible to the outpatient clinic, according to guidelines from the American Society of Transplantation and Cellular Therapy.
See one, do one, teach one
Rendering patient care during a pandemic would be unique for me. However, I, like all physicians, am familiar with feelings of inadequacy at times of professional challenge. On countless occasions, I have started my day or walked into a patient’s room wondering whether I will have the fortitude, knowledge, creativity, or help I need to get through that day or make that patient “better” by any definition of that word.
We all know the formula: “Work hard. Make evidence-based, personalized decisions for those who have entrusted their care to us. Learn from those encounters. Teach from our knowledge and experience – that is, ‘See one, do one, teach one.’ ”
The Seattle oncologists are living the lives of first responders and deserve our admiration for putting pen to paper so we can learn from their considerable, relevant experience.
Similar admiration is due to Giuseppe Curigliano, MD, of the European Institute of Oncology in Milan. In the ASCO Daily News, Dr. Curigliano described an epidemic that, within 3 weeks, overloaded the health care system across northern Italy.
Hospitalization was needed for over 60% of infected patients, and nearly 15% of those patients needed intensive care unit services for respiratory distress. The Italians centralized oncology care in specialized hubs, with spokes of institutions working in parallel to provide cancer-specific care in a COVID-free environment.
To build upon cancer-specific information from Italy and other areas hard-hit by COVID-19, more than 30 cancer centers have joined together to form the COVID-19 and Cancer Consortium. The consortium’s website hosts a survey designed to “capture details related to cancer patients presumed to have COVID-19.”
Calculating deaths and long-term consequences for cancer care delivery
It is proper that the authors from China, Italy, and Seattle did not focus attention on the case fatality rate from the COVID-19 pandemic among cancer patients. To say the least, it would be complicated to tally the direct mortality – either overall or in clinically important subsets of patients, including country-specific cohorts.
What we know from published reports is that, in Italy, cancer patients account for about 20% of deaths from coronavirus. In China, the case-fatality rate for patients with cancer was 5.6% (JAMA. 2020 Feb 24. doi: 10.1001/jama.2020.2648).
However, we know nothing about the indirect death toll from malignancy (without coronavirus infection) that was untreated or managed less than optimally because of personnel and physical resources that were diverted to COVID-19–associated cases.
Similarly, we cannot begin to estimate indirect consequences of the pandemic to oncology practices, such as accelerated burnout and posttraumatic stress disorder, as well as the long-range effects of economic turmoil on patients, health care workers, and provider organizations.
What happens to cancer trials?
From China, Italy, and Seattle, thus far, there is little information about how the pandemic will affect the vital clinical research endeavor. The Seattle physicians did say they plan to enroll patients on clinical trials only when the trial offers a high chance of benefiting the patient over standard therapy alone.
Fortunately, the National Institutes of Health and Food and Drug Administration have released guidance documents related to clinical trials.
The National Cancer Institute (NCI) has also released guidance documents (March 13 guidance; March 23 guidance) for patients on clinical trials supported by the NCI Cancer Therapy Evaluation Program (CTEP) and the NCI Community Oncology Research Program (NCORP).
CTEP and NCORP are making reasonable accommodations to suspend monitoring visits and audits, allow tele–follow-up visits for patients, and permit local physicians to provide care for patients on study. In addition, with appropriate procedural adherence and documentation, CTEP and NCORP will allow oral investigational medicines to be mailed directly to patients’ homes.
Planned NCI National Clinical Trials Network meetings will be conducted via remote access webinars, conference calls, and similar technology. These adjustments – and probably many more to come – are geared toward facilitating ongoing care to proceed safely and with minimal risk for patients currently receiving investigational therapies and for the sites and investigators engaged in those studies.
Each of us has probably faced a personal “defining professional moment,” when we had to utilize every skill in our arsenal and examine the motivations that led us to a career in oncology. However, it is clear from the forgoing clinical and research processes and guidelines that the COVID-19 pandemic is such a defining professional moment for each of us, in every community we serve.
Critical junctures like this cause more rapid behavior change and innovation than the slow-moving pace that characterizes our idealized preferences. As oncologists who embrace new data and behavioral change, we stand to learn processes that will facilitate more perfected systems of care than the one that preceded this unprecedented crisis, promote more efficient sharing of high-quality information, and improve the outcome for our future patients.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
I have no knowledge of, or experience with, managing a cancer patient during a pandemic. However, from the published and otherwise shared experience of others, we should not allow ourselves to underestimate the voracity of the coronavirus pandemic on our patients, communities, and health care systems.
Data from China suggest cancer patients infected with SARS-CoV-2 face a 3.5 times higher risk of mechanical ventilation, intensive care unit admission, or death, compared with infected patients without cancer (Lancet Oncol 2020;21:335-7).
Health care workers in Seattle have also shared their experiences battling coronavirus infections in cancer patients (J Natl Compr Canc Netw. 2020 Mar 20. doi: 10.6004/jnccn.2020.7560). Masumi Ueda, MD, of Seattle Cancer Care Alliance, and colleagues reviewed their decisions in multiple domains over a 7-week period, during which the state of Washington went from a single case of SARS-CoV-2 infection to nearly 650 cases and 40 deaths.
Making tough treatment decisions
Dr. Ueda and colleagues contrasted their customary resource-rich, innovation-oriented, cancer-combatting environment with their current circumstance, in which they must prioritize treatment for patients for whom the risk-reward balance has tilted substantially toward “risk.”
The authors noted that their most difficult decisions were those regarding delay of cancer treatment. They suggested that plans for potentially curative adjuvant therapy should likely proceed, but, for patients with metastatic disease, the equation is more nuanced.
In some cases, treatment should be delayed or interrupted with recognition of how that could result in worsening performance status and admission for symptom palliation, further stressing inpatient resources.
The authors suggested scenarios for prioritizing cancer surgery. For example, several months of systemic therapy (ideally, low-risk systemic therapy such as hormone therapy for breast or prostate cancer) and surgical delay may be worthwhile, without compromising patient care.
Patients with aggressive hematologic malignancy requiring urgent systemic treatment (potentially stem cell transplantation and cellular immunotherapies) should be treated promptly. However, even in those cases, opportunities should be sought to lessen immunosuppression and transition care as quickly as possible to the outpatient clinic, according to guidelines from the American Society of Transplantation and Cellular Therapy.
See one, do one, teach one
Rendering patient care during a pandemic would be unique for me. However, I, like all physicians, am familiar with feelings of inadequacy at times of professional challenge. On countless occasions, I have started my day or walked into a patient’s room wondering whether I will have the fortitude, knowledge, creativity, or help I need to get through that day or make that patient “better” by any definition of that word.
We all know the formula: “Work hard. Make evidence-based, personalized decisions for those who have entrusted their care to us. Learn from those encounters. Teach from our knowledge and experience – that is, ‘See one, do one, teach one.’ ”
The Seattle oncologists are living the lives of first responders and deserve our admiration for putting pen to paper so we can learn from their considerable, relevant experience.
Similar admiration is due to Giuseppe Curigliano, MD, of the European Institute of Oncology in Milan. In the ASCO Daily News, Dr. Curigliano described an epidemic that, within 3 weeks, overloaded the health care system across northern Italy.
Hospitalization was needed for over 60% of infected patients, and nearly 15% of those patients needed intensive care unit services for respiratory distress. The Italians centralized oncology care in specialized hubs, with spokes of institutions working in parallel to provide cancer-specific care in a COVID-free environment.
To build upon cancer-specific information from Italy and other areas hard-hit by COVID-19, more than 30 cancer centers have joined together to form the COVID-19 and Cancer Consortium. The consortium’s website hosts a survey designed to “capture details related to cancer patients presumed to have COVID-19.”
Calculating deaths and long-term consequences for cancer care delivery
It is proper that the authors from China, Italy, and Seattle did not focus attention on the case fatality rate from the COVID-19 pandemic among cancer patients. To say the least, it would be complicated to tally the direct mortality – either overall or in clinically important subsets of patients, including country-specific cohorts.
What we know from published reports is that, in Italy, cancer patients account for about 20% of deaths from coronavirus. In China, the case-fatality rate for patients with cancer was 5.6% (JAMA. 2020 Feb 24. doi: 10.1001/jama.2020.2648).
However, we know nothing about the indirect death toll from malignancy (without coronavirus infection) that was untreated or managed less than optimally because of personnel and physical resources that were diverted to COVID-19–associated cases.
Similarly, we cannot begin to estimate indirect consequences of the pandemic to oncology practices, such as accelerated burnout and posttraumatic stress disorder, as well as the long-range effects of economic turmoil on patients, health care workers, and provider organizations.
What happens to cancer trials?
From China, Italy, and Seattle, thus far, there is little information about how the pandemic will affect the vital clinical research endeavor. The Seattle physicians did say they plan to enroll patients on clinical trials only when the trial offers a high chance of benefiting the patient over standard therapy alone.
Fortunately, the National Institutes of Health and Food and Drug Administration have released guidance documents related to clinical trials.
The National Cancer Institute (NCI) has also released guidance documents (March 13 guidance; March 23 guidance) for patients on clinical trials supported by the NCI Cancer Therapy Evaluation Program (CTEP) and the NCI Community Oncology Research Program (NCORP).
CTEP and NCORP are making reasonable accommodations to suspend monitoring visits and audits, allow tele–follow-up visits for patients, and permit local physicians to provide care for patients on study. In addition, with appropriate procedural adherence and documentation, CTEP and NCORP will allow oral investigational medicines to be mailed directly to patients’ homes.
Planned NCI National Clinical Trials Network meetings will be conducted via remote access webinars, conference calls, and similar technology. These adjustments – and probably many more to come – are geared toward facilitating ongoing care to proceed safely and with minimal risk for patients currently receiving investigational therapies and for the sites and investigators engaged in those studies.
Each of us has probably faced a personal “defining professional moment,” when we had to utilize every skill in our arsenal and examine the motivations that led us to a career in oncology. However, it is clear from the forgoing clinical and research processes and guidelines that the COVID-19 pandemic is such a defining professional moment for each of us, in every community we serve.
Critical junctures like this cause more rapid behavior change and innovation than the slow-moving pace that characterizes our idealized preferences. As oncologists who embrace new data and behavioral change, we stand to learn processes that will facilitate more perfected systems of care than the one that preceded this unprecedented crisis, promote more efficient sharing of high-quality information, and improve the outcome for our future patients.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
I have no knowledge of, or experience with, managing a cancer patient during a pandemic. However, from the published and otherwise shared experience of others, we should not allow ourselves to underestimate the voracity of the coronavirus pandemic on our patients, communities, and health care systems.
Data from China suggest cancer patients infected with SARS-CoV-2 face a 3.5 times higher risk of mechanical ventilation, intensive care unit admission, or death, compared with infected patients without cancer (Lancet Oncol 2020;21:335-7).
Health care workers in Seattle have also shared their experiences battling coronavirus infections in cancer patients (J Natl Compr Canc Netw. 2020 Mar 20. doi: 10.6004/jnccn.2020.7560). Masumi Ueda, MD, of Seattle Cancer Care Alliance, and colleagues reviewed their decisions in multiple domains over a 7-week period, during which the state of Washington went from a single case of SARS-CoV-2 infection to nearly 650 cases and 40 deaths.
Making tough treatment decisions
Dr. Ueda and colleagues contrasted their customary resource-rich, innovation-oriented, cancer-combatting environment with their current circumstance, in which they must prioritize treatment for patients for whom the risk-reward balance has tilted substantially toward “risk.”
The authors noted that their most difficult decisions were those regarding delay of cancer treatment. They suggested that plans for potentially curative adjuvant therapy should likely proceed, but, for patients with metastatic disease, the equation is more nuanced.
In some cases, treatment should be delayed or interrupted with recognition of how that could result in worsening performance status and admission for symptom palliation, further stressing inpatient resources.
The authors suggested scenarios for prioritizing cancer surgery. For example, several months of systemic therapy (ideally, low-risk systemic therapy such as hormone therapy for breast or prostate cancer) and surgical delay may be worthwhile, without compromising patient care.
Patients with aggressive hematologic malignancy requiring urgent systemic treatment (potentially stem cell transplantation and cellular immunotherapies) should be treated promptly. However, even in those cases, opportunities should be sought to lessen immunosuppression and transition care as quickly as possible to the outpatient clinic, according to guidelines from the American Society of Transplantation and Cellular Therapy.
See one, do one, teach one
Rendering patient care during a pandemic would be unique for me. However, I, like all physicians, am familiar with feelings of inadequacy at times of professional challenge. On countless occasions, I have started my day or walked into a patient’s room wondering whether I will have the fortitude, knowledge, creativity, or help I need to get through that day or make that patient “better” by any definition of that word.
We all know the formula: “Work hard. Make evidence-based, personalized decisions for those who have entrusted their care to us. Learn from those encounters. Teach from our knowledge and experience – that is, ‘See one, do one, teach one.’ ”
The Seattle oncologists are living the lives of first responders and deserve our admiration for putting pen to paper so we can learn from their considerable, relevant experience.
Similar admiration is due to Giuseppe Curigliano, MD, of the European Institute of Oncology in Milan. In the ASCO Daily News, Dr. Curigliano described an epidemic that, within 3 weeks, overloaded the health care system across northern Italy.
Hospitalization was needed for over 60% of infected patients, and nearly 15% of those patients needed intensive care unit services for respiratory distress. The Italians centralized oncology care in specialized hubs, with spokes of institutions working in parallel to provide cancer-specific care in a COVID-free environment.
To build upon cancer-specific information from Italy and other areas hard-hit by COVID-19, more than 30 cancer centers have joined together to form the COVID-19 and Cancer Consortium. The consortium’s website hosts a survey designed to “capture details related to cancer patients presumed to have COVID-19.”
Calculating deaths and long-term consequences for cancer care delivery
It is proper that the authors from China, Italy, and Seattle did not focus attention on the case fatality rate from the COVID-19 pandemic among cancer patients. To say the least, it would be complicated to tally the direct mortality – either overall or in clinically important subsets of patients, including country-specific cohorts.
What we know from published reports is that, in Italy, cancer patients account for about 20% of deaths from coronavirus. In China, the case-fatality rate for patients with cancer was 5.6% (JAMA. 2020 Feb 24. doi: 10.1001/jama.2020.2648).
However, we know nothing about the indirect death toll from malignancy (without coronavirus infection) that was untreated or managed less than optimally because of personnel and physical resources that were diverted to COVID-19–associated cases.
Similarly, we cannot begin to estimate indirect consequences of the pandemic to oncology practices, such as accelerated burnout and posttraumatic stress disorder, as well as the long-range effects of economic turmoil on patients, health care workers, and provider organizations.
What happens to cancer trials?
From China, Italy, and Seattle, thus far, there is little information about how the pandemic will affect the vital clinical research endeavor. The Seattle physicians did say they plan to enroll patients on clinical trials only when the trial offers a high chance of benefiting the patient over standard therapy alone.
Fortunately, the National Institutes of Health and Food and Drug Administration have released guidance documents related to clinical trials.
The National Cancer Institute (NCI) has also released guidance documents (March 13 guidance; March 23 guidance) for patients on clinical trials supported by the NCI Cancer Therapy Evaluation Program (CTEP) and the NCI Community Oncology Research Program (NCORP).
CTEP and NCORP are making reasonable accommodations to suspend monitoring visits and audits, allow tele–follow-up visits for patients, and permit local physicians to provide care for patients on study. In addition, with appropriate procedural adherence and documentation, CTEP and NCORP will allow oral investigational medicines to be mailed directly to patients’ homes.
Planned NCI National Clinical Trials Network meetings will be conducted via remote access webinars, conference calls, and similar technology. These adjustments – and probably many more to come – are geared toward facilitating ongoing care to proceed safely and with minimal risk for patients currently receiving investigational therapies and for the sites and investigators engaged in those studies.
Each of us has probably faced a personal “defining professional moment,” when we had to utilize every skill in our arsenal and examine the motivations that led us to a career in oncology. However, it is clear from the forgoing clinical and research processes and guidelines that the COVID-19 pandemic is such a defining professional moment for each of us, in every community we serve.
Critical junctures like this cause more rapid behavior change and innovation than the slow-moving pace that characterizes our idealized preferences. As oncologists who embrace new data and behavioral change, we stand to learn processes that will facilitate more perfected systems of care than the one that preceded this unprecedented crisis, promote more efficient sharing of high-quality information, and improve the outcome for our future patients.
Dr. Lyss was an oncologist and researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.
Responsible use of breast cancer screening
In this edition of “Applying research to practice,” I examine a study suggesting that annual screening mammography does not reduce the risk of death from breast cancer in women aged 75 years and older. I also highlight a related editorial noting that we should optimize treatment as well as screening for breast cancer.
Regular screening mammography in women aged 50-69 years prevents 21.3 breast cancer deaths among 10,000 women over a 10-year time period (Ann Intern Med. 2016 Feb 16;164[4]:244-55). However, in the published screening trials, few participants were older than 70 years of age.
More than half of women above age 74 receive annual mammograms (Health, United States, 2018. www.cdc.gov/nchs/data/hus/hus18.pdf). And more than a third of breast cancer deaths occur in women aged 70 years or older (CA Cancer J Clin. 2016 Mar-Apr;66[2]:96-114).
Do older women benefit from annual mammography to the same extent as younger women? Is there a point at which benefit ends?
To answer these questions, Xabier García-Albéniz, MD, PhD, of Harvard Medical School in Boston, and colleagues studied 1,058,013 women enrolled in Medicare during 2000-2008 (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M18-1199).
The researchers examined data on patients aged 70-84 years who had a life expectancy of at least 10 years, at least one recent mammogram, and no history of breast cancer. The team emulated a prospective trial by examining deaths over an 8-year period for women aged 70 years and older who either continued or stopped screening mammography. The researchers conducted separate analyses for women aged 70-74 years and those aged 75-84 years.
Diagnoses of breast cancer were, not surprisingly, higher in the continued-screening group, but there were no major reductions in breast cancer–related deaths.
Among women aged 70-74 years, the estimated 8-year risk for breast cancer death was reduced for women who continued screening versus those who stopped it by one death per 1,000 women (hazard ratio, 0.78). Among women aged 75-84 years, the 8-year risk reduction was 0.07 deaths per 1,000 women (HR, 1.00).
The authors concluded that continuing mammographic screening past age 75 years resulted in no material difference in cancer-specific mortality over an 8-year time period, in comparison with stopping regular screening examinations.
Considering treatment as well as screening
For a variety of reasons (ethical, economic, methodologic), it is unreasonable to expect a randomized, clinical trial examining the value of mammography in older women. An informative alternative would be a well-designed, large-scale, population-based, observational study that takes into consideration potentially confounding variables of the binary strategies of continuing screening versus stopping it.
Although the 8-year risk of breast cancer in older women is not low among screened women – 5.5% in women aged 70-74 years and 5.8% in women aged 75-84 years – and mammography remains an effective screening tool, the effect of screening on breast cancer mortality appears to decline as women age.
In the editorial that accompanies the study by Dr. García-Albéniz and colleagues, Otis Brawley, MD, of Johns Hopkins University, Baltimore, highlighted the role of inadequate, ineffective, inconvenient, or poorly tolerated treatment in older women (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M20-0429).
Dr. Brawley illustrated that focusing too much on screening diverts attention from the major driver of cancer mortality in older women: suboptimal treatment. That certainly has been the case for the dramatic impact of improved lung cancer treatment on mortality, despite a statistically significant impact of screening on lung cancer mortality as well.
As with lung cancer screening, Dr. Brawley describes the goal of defining “personalized screening recommendations” in breast cancer, or screening that is targeted to the highest-risk women and those who stand a high chance of benefiting from treatment if they are diagnosed with breast cancer.
As our population ages and health care expenditures continue to rise, there can be little disagreement that responsible use of cancer diagnostics will be as vital as judicious application of treatment.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I examine a study suggesting that annual screening mammography does not reduce the risk of death from breast cancer in women aged 75 years and older. I also highlight a related editorial noting that we should optimize treatment as well as screening for breast cancer.
Regular screening mammography in women aged 50-69 years prevents 21.3 breast cancer deaths among 10,000 women over a 10-year time period (Ann Intern Med. 2016 Feb 16;164[4]:244-55). However, in the published screening trials, few participants were older than 70 years of age.
More than half of women above age 74 receive annual mammograms (Health, United States, 2018. www.cdc.gov/nchs/data/hus/hus18.pdf). And more than a third of breast cancer deaths occur in women aged 70 years or older (CA Cancer J Clin. 2016 Mar-Apr;66[2]:96-114).
Do older women benefit from annual mammography to the same extent as younger women? Is there a point at which benefit ends?
To answer these questions, Xabier García-Albéniz, MD, PhD, of Harvard Medical School in Boston, and colleagues studied 1,058,013 women enrolled in Medicare during 2000-2008 (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M18-1199).
The researchers examined data on patients aged 70-84 years who had a life expectancy of at least 10 years, at least one recent mammogram, and no history of breast cancer. The team emulated a prospective trial by examining deaths over an 8-year period for women aged 70 years and older who either continued or stopped screening mammography. The researchers conducted separate analyses for women aged 70-74 years and those aged 75-84 years.
Diagnoses of breast cancer were, not surprisingly, higher in the continued-screening group, but there were no major reductions in breast cancer–related deaths.
Among women aged 70-74 years, the estimated 8-year risk for breast cancer death was reduced for women who continued screening versus those who stopped it by one death per 1,000 women (hazard ratio, 0.78). Among women aged 75-84 years, the 8-year risk reduction was 0.07 deaths per 1,000 women (HR, 1.00).
The authors concluded that continuing mammographic screening past age 75 years resulted in no material difference in cancer-specific mortality over an 8-year time period, in comparison with stopping regular screening examinations.
Considering treatment as well as screening
For a variety of reasons (ethical, economic, methodologic), it is unreasonable to expect a randomized, clinical trial examining the value of mammography in older women. An informative alternative would be a well-designed, large-scale, population-based, observational study that takes into consideration potentially confounding variables of the binary strategies of continuing screening versus stopping it.
Although the 8-year risk of breast cancer in older women is not low among screened women – 5.5% in women aged 70-74 years and 5.8% in women aged 75-84 years – and mammography remains an effective screening tool, the effect of screening on breast cancer mortality appears to decline as women age.
In the editorial that accompanies the study by Dr. García-Albéniz and colleagues, Otis Brawley, MD, of Johns Hopkins University, Baltimore, highlighted the role of inadequate, ineffective, inconvenient, or poorly tolerated treatment in older women (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M20-0429).
Dr. Brawley illustrated that focusing too much on screening diverts attention from the major driver of cancer mortality in older women: suboptimal treatment. That certainly has been the case for the dramatic impact of improved lung cancer treatment on mortality, despite a statistically significant impact of screening on lung cancer mortality as well.
As with lung cancer screening, Dr. Brawley describes the goal of defining “personalized screening recommendations” in breast cancer, or screening that is targeted to the highest-risk women and those who stand a high chance of benefiting from treatment if they are diagnosed with breast cancer.
As our population ages and health care expenditures continue to rise, there can be little disagreement that responsible use of cancer diagnostics will be as vital as judicious application of treatment.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I examine a study suggesting that annual screening mammography does not reduce the risk of death from breast cancer in women aged 75 years and older. I also highlight a related editorial noting that we should optimize treatment as well as screening for breast cancer.
Regular screening mammography in women aged 50-69 years prevents 21.3 breast cancer deaths among 10,000 women over a 10-year time period (Ann Intern Med. 2016 Feb 16;164[4]:244-55). However, in the published screening trials, few participants were older than 70 years of age.
More than half of women above age 74 receive annual mammograms (Health, United States, 2018. www.cdc.gov/nchs/data/hus/hus18.pdf). And more than a third of breast cancer deaths occur in women aged 70 years or older (CA Cancer J Clin. 2016 Mar-Apr;66[2]:96-114).
Do older women benefit from annual mammography to the same extent as younger women? Is there a point at which benefit ends?
To answer these questions, Xabier García-Albéniz, MD, PhD, of Harvard Medical School in Boston, and colleagues studied 1,058,013 women enrolled in Medicare during 2000-2008 (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M18-1199).
The researchers examined data on patients aged 70-84 years who had a life expectancy of at least 10 years, at least one recent mammogram, and no history of breast cancer. The team emulated a prospective trial by examining deaths over an 8-year period for women aged 70 years and older who either continued or stopped screening mammography. The researchers conducted separate analyses for women aged 70-74 years and those aged 75-84 years.
Diagnoses of breast cancer were, not surprisingly, higher in the continued-screening group, but there were no major reductions in breast cancer–related deaths.
Among women aged 70-74 years, the estimated 8-year risk for breast cancer death was reduced for women who continued screening versus those who stopped it by one death per 1,000 women (hazard ratio, 0.78). Among women aged 75-84 years, the 8-year risk reduction was 0.07 deaths per 1,000 women (HR, 1.00).
The authors concluded that continuing mammographic screening past age 75 years resulted in no material difference in cancer-specific mortality over an 8-year time period, in comparison with stopping regular screening examinations.
Considering treatment as well as screening
For a variety of reasons (ethical, economic, methodologic), it is unreasonable to expect a randomized, clinical trial examining the value of mammography in older women. An informative alternative would be a well-designed, large-scale, population-based, observational study that takes into consideration potentially confounding variables of the binary strategies of continuing screening versus stopping it.
Although the 8-year risk of breast cancer in older women is not low among screened women – 5.5% in women aged 70-74 years and 5.8% in women aged 75-84 years – and mammography remains an effective screening tool, the effect of screening on breast cancer mortality appears to decline as women age.
In the editorial that accompanies the study by Dr. García-Albéniz and colleagues, Otis Brawley, MD, of Johns Hopkins University, Baltimore, highlighted the role of inadequate, ineffective, inconvenient, or poorly tolerated treatment in older women (Ann Intern Med. 2020 Feb 25. doi: 10.7326/M20-0429).
Dr. Brawley illustrated that focusing too much on screening diverts attention from the major driver of cancer mortality in older women: suboptimal treatment. That certainly has been the case for the dramatic impact of improved lung cancer treatment on mortality, despite a statistically significant impact of screening on lung cancer mortality as well.
As with lung cancer screening, Dr. Brawley describes the goal of defining “personalized screening recommendations” in breast cancer, or screening that is targeted to the highest-risk women and those who stand a high chance of benefiting from treatment if they are diagnosed with breast cancer.
As our population ages and health care expenditures continue to rise, there can be little disagreement that responsible use of cancer diagnostics will be as vital as judicious application of treatment.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
Real-world data are a wake-up call
In this edition of “Applying research to practice,” I highlight a study revealing real-world information about the clinical care of breast cancer patients with deleterious germline mutations.
While germline testing among breast cancer patients is becoming more commonplace, it isn’t clear how test results influence patient care. To gain some insight, Allison W. Kurian, MD, of Stanford (Calif.) University, and colleagues analyzed data on 20,568 women with stage 0-III breast cancer from the Surveillance, Epidemiology, and End Results (SEER) registries of Georgia and California (JAMA Oncol. 2020 Feb 6. doi: 10.1001/jamaoncol.2019.6400).
The researchers aimed to determine whether women with mutations in breast cancer–associated genes (BRCA1/2 or others) received guideline-concordant care to the same degree as women who lacked deleterious mutations. The authors evaluated guideline concordance with respect to three treatment modalities: surgery (bilateral vs. unilateral mastectomy in women who were eligible for unilateral surgery), radiotherapy after lumpectomy (for women aged less than 70 years with hormonally responsive, ErbB2-negative, stage I cancers), and chemotherapy (among women eligible for consideration of chemotherapy omission)
In alignment with guidelines, many clinicians correctly used genetic test results to guide surgical decisions. For example, 61.7% of women with BRCA mutations underwent bilateral mastectomy, compared with 24.3% who were mutation negative (odds ratio, 5.52). For other pathogenic variants (ATM, CDH1, CHEK2, NBN, NF1, PALB2, PTEN, and TP53), the rate of bilateral mastectomy was still elevated, albeit to a lesser degree (OR, 2.41).
In discord with guidelines, women with BRCA mutations were 78% less likely to receive radiotherapy after lumpectomy (OR, 0.22) and 76% more likely to receive chemotherapy for early-stage, hormone-positive disease (OR, 1.76), suggesting possible trends in under- and overtreatment, respectively. Chemotherapy utilization rates among mutation carriers and noncarriers became more similar after adjustment for clinical and demographic factors.
There are limits on the granularity of the SEER database, such that, if a patient had a mastectomy a year or more after lumpectomy in an effort to avoid radiotherapy, the database would not have reflected that. Clinical factors could have appropriately influenced chemotherapy receipt among patients with mutations, but those additional factors (including patient preference) would not be included in the SEER data.
The authors concluded that research should be conducted to confirm the results of this retrospective, population-based cohort analysis, in an effort to understand the decision-making process and consequences for long-term outcome.
How these findings should influence practice
With every new development, there are challenges – some expected, some unanticipated.
It is now feasible to obtain multigene panel testing reasonably inexpensively. There are concerns about undertesting of patients on the basis of family history alone. And some major professional organizations have endorsed routine gene panel testing for all breast cancer patients.
As a consequence of these factors, genetic test results are routinely available to clinicians who may lack formal training in clinical genetics. Whether these results influence the receipt of evidence-based clinical care is uncertain.
The information published by Dr. Kurian and colleagues is inherently limited by the methodology of a SEER database review. Among other limitations, as the authors comment:
- The genetic test results could have arrived after treatment decisions were made.
- Treatment delivered more than a year after diagnosis would not have been captured.
- There was selection of patients for genetic testing.
- There were few patients with particular germline mutations other than BRCA1/2 on whom to judge whether treatment was guideline concordant.
- The rationale for the treatment choices made by physicians and patients was not available.
- Impact of treatment choices on survival for carriers of deleterious mutations is uncertain.
Nonetheless, these data suggest a need to redouble efforts to educate patients, their family members, and health care professionals about evidence-based guidelines for care and the rationale for those recommendations.
Careful, prospective monitoring of any resultant differences in treatment outcome in patients treated with guideline-concordant and nonconcordant care is needed. When treatment choices appear to systematically deviate from published guidelines with no obvious rationale, it is a wake-up call for all of us.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I highlight a study revealing real-world information about the clinical care of breast cancer patients with deleterious germline mutations.
While germline testing among breast cancer patients is becoming more commonplace, it isn’t clear how test results influence patient care. To gain some insight, Allison W. Kurian, MD, of Stanford (Calif.) University, and colleagues analyzed data on 20,568 women with stage 0-III breast cancer from the Surveillance, Epidemiology, and End Results (SEER) registries of Georgia and California (JAMA Oncol. 2020 Feb 6. doi: 10.1001/jamaoncol.2019.6400).
The researchers aimed to determine whether women with mutations in breast cancer–associated genes (BRCA1/2 or others) received guideline-concordant care to the same degree as women who lacked deleterious mutations. The authors evaluated guideline concordance with respect to three treatment modalities: surgery (bilateral vs. unilateral mastectomy in women who were eligible for unilateral surgery), radiotherapy after lumpectomy (for women aged less than 70 years with hormonally responsive, ErbB2-negative, stage I cancers), and chemotherapy (among women eligible for consideration of chemotherapy omission)
In alignment with guidelines, many clinicians correctly used genetic test results to guide surgical decisions. For example, 61.7% of women with BRCA mutations underwent bilateral mastectomy, compared with 24.3% who were mutation negative (odds ratio, 5.52). For other pathogenic variants (ATM, CDH1, CHEK2, NBN, NF1, PALB2, PTEN, and TP53), the rate of bilateral mastectomy was still elevated, albeit to a lesser degree (OR, 2.41).
In discord with guidelines, women with BRCA mutations were 78% less likely to receive radiotherapy after lumpectomy (OR, 0.22) and 76% more likely to receive chemotherapy for early-stage, hormone-positive disease (OR, 1.76), suggesting possible trends in under- and overtreatment, respectively. Chemotherapy utilization rates among mutation carriers and noncarriers became more similar after adjustment for clinical and demographic factors.
There are limits on the granularity of the SEER database, such that, if a patient had a mastectomy a year or more after lumpectomy in an effort to avoid radiotherapy, the database would not have reflected that. Clinical factors could have appropriately influenced chemotherapy receipt among patients with mutations, but those additional factors (including patient preference) would not be included in the SEER data.
The authors concluded that research should be conducted to confirm the results of this retrospective, population-based cohort analysis, in an effort to understand the decision-making process and consequences for long-term outcome.
How these findings should influence practice
With every new development, there are challenges – some expected, some unanticipated.
It is now feasible to obtain multigene panel testing reasonably inexpensively. There are concerns about undertesting of patients on the basis of family history alone. And some major professional organizations have endorsed routine gene panel testing for all breast cancer patients.
As a consequence of these factors, genetic test results are routinely available to clinicians who may lack formal training in clinical genetics. Whether these results influence the receipt of evidence-based clinical care is uncertain.
The information published by Dr. Kurian and colleagues is inherently limited by the methodology of a SEER database review. Among other limitations, as the authors comment:
- The genetic test results could have arrived after treatment decisions were made.
- Treatment delivered more than a year after diagnosis would not have been captured.
- There was selection of patients for genetic testing.
- There were few patients with particular germline mutations other than BRCA1/2 on whom to judge whether treatment was guideline concordant.
- The rationale for the treatment choices made by physicians and patients was not available.
- Impact of treatment choices on survival for carriers of deleterious mutations is uncertain.
Nonetheless, these data suggest a need to redouble efforts to educate patients, their family members, and health care professionals about evidence-based guidelines for care and the rationale for those recommendations.
Careful, prospective monitoring of any resultant differences in treatment outcome in patients treated with guideline-concordant and nonconcordant care is needed. When treatment choices appear to systematically deviate from published guidelines with no obvious rationale, it is a wake-up call for all of us.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I highlight a study revealing real-world information about the clinical care of breast cancer patients with deleterious germline mutations.
While germline testing among breast cancer patients is becoming more commonplace, it isn’t clear how test results influence patient care. To gain some insight, Allison W. Kurian, MD, of Stanford (Calif.) University, and colleagues analyzed data on 20,568 women with stage 0-III breast cancer from the Surveillance, Epidemiology, and End Results (SEER) registries of Georgia and California (JAMA Oncol. 2020 Feb 6. doi: 10.1001/jamaoncol.2019.6400).
The researchers aimed to determine whether women with mutations in breast cancer–associated genes (BRCA1/2 or others) received guideline-concordant care to the same degree as women who lacked deleterious mutations. The authors evaluated guideline concordance with respect to three treatment modalities: surgery (bilateral vs. unilateral mastectomy in women who were eligible for unilateral surgery), radiotherapy after lumpectomy (for women aged less than 70 years with hormonally responsive, ErbB2-negative, stage I cancers), and chemotherapy (among women eligible for consideration of chemotherapy omission)
In alignment with guidelines, many clinicians correctly used genetic test results to guide surgical decisions. For example, 61.7% of women with BRCA mutations underwent bilateral mastectomy, compared with 24.3% who were mutation negative (odds ratio, 5.52). For other pathogenic variants (ATM, CDH1, CHEK2, NBN, NF1, PALB2, PTEN, and TP53), the rate of bilateral mastectomy was still elevated, albeit to a lesser degree (OR, 2.41).
In discord with guidelines, women with BRCA mutations were 78% less likely to receive radiotherapy after lumpectomy (OR, 0.22) and 76% more likely to receive chemotherapy for early-stage, hormone-positive disease (OR, 1.76), suggesting possible trends in under- and overtreatment, respectively. Chemotherapy utilization rates among mutation carriers and noncarriers became more similar after adjustment for clinical and demographic factors.
There are limits on the granularity of the SEER database, such that, if a patient had a mastectomy a year or more after lumpectomy in an effort to avoid radiotherapy, the database would not have reflected that. Clinical factors could have appropriately influenced chemotherapy receipt among patients with mutations, but those additional factors (including patient preference) would not be included in the SEER data.
The authors concluded that research should be conducted to confirm the results of this retrospective, population-based cohort analysis, in an effort to understand the decision-making process and consequences for long-term outcome.
How these findings should influence practice
With every new development, there are challenges – some expected, some unanticipated.
It is now feasible to obtain multigene panel testing reasonably inexpensively. There are concerns about undertesting of patients on the basis of family history alone. And some major professional organizations have endorsed routine gene panel testing for all breast cancer patients.
As a consequence of these factors, genetic test results are routinely available to clinicians who may lack formal training in clinical genetics. Whether these results influence the receipt of evidence-based clinical care is uncertain.
The information published by Dr. Kurian and colleagues is inherently limited by the methodology of a SEER database review. Among other limitations, as the authors comment:
- The genetic test results could have arrived after treatment decisions were made.
- Treatment delivered more than a year after diagnosis would not have been captured.
- There was selection of patients for genetic testing.
- There were few patients with particular germline mutations other than BRCA1/2 on whom to judge whether treatment was guideline concordant.
- The rationale for the treatment choices made by physicians and patients was not available.
- Impact of treatment choices on survival for carriers of deleterious mutations is uncertain.
Nonetheless, these data suggest a need to redouble efforts to educate patients, their family members, and health care professionals about evidence-based guidelines for care and the rationale for those recommendations.
Careful, prospective monitoring of any resultant differences in treatment outcome in patients treated with guideline-concordant and nonconcordant care is needed. When treatment choices appear to systematically deviate from published guidelines with no obvious rationale, it is a wake-up call for all of us.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
The possibilities of pembrolizumab plus chemo in breast cancer treatment
In this edition of “Applying research to practice,” I highlight I-SPY2 and other studies of pembrolizumab plus chemotherapy in breast cancer patients.
Pathologic complete response (pCR) rates up to 60% were reported for patients with high-risk, stage II/III breast cancer who received pembrolizumab plus standard neoadjuvant chemotherapy (NAC) in I-SPY2, an ongoing platform trial designed to screen multiple agents and pinpoint those with a high probability of success (JAMA Oncol. 2020 Feb 13. doi: 10.1001/jamaoncol.2019.6650).
The addition of pembrolizumab to NAC doubled pCR rates in all three biomarker signatures studied, including ERBB2 (HER2)-negative, hormone receptor (HR)-positive/ERBB2-negative, and triple-negative breast cancer (TNBC).
As a result, pembrolizumab “graduated” from I-SPY2, with a more than 99% predictive probability that the pembrolizumab-plus-NAC approach would be superior to NAC alone in a phase 3 trial. In the HR-positive/ERBB2-negative signature, pembrolizumab is the first agent to graduate among the 10 agents studied since I-SPY2 opened in 2010.
The control arm in I-SPY2 had 181 patients randomized to standard NAC (paclitaxel followed by doxorubicin plus cyclophosphamide). The pembrolizumab arm included 69 patients who received the same NAC regimen plus pembrolizumab, given concurrently with paclitaxel.
The estimated pCR rates in all ERBB2-negative patients were 44% in the pembrolizumab arm and 17% in the control arm. Among the 40 HR-positive/ERBB2-negative patients, the estimated pCR rates were 30% and 13%, respectively. In the 29 TNBC patients, the estimated pCR rates were 60% and 22%, respectively.
Extensive residual cancer burden was less often seen in the pembrolizumab-treated patients than in the comparison group. At a median follow-up of 2.8 years in the pembrolizumab arm and 3.5 years in the NAC arm, 3-year event-free survival was similar between the arms. However, the investigators cautioned against drawing conclusions from this exploratory analysis in a small number of patients. Testifying to the importance of the primary endpoint of pCR rate, patients who achieved pCR had excellent outcomes regardless of their assigned study arms.
Immune-related adverse events in the pembrolizumab-treated patients were generally grade 1 or 2 and were managed with dose interruption or corticosteroid therapy. Most commonly seen was thyroid dysfunction in 13% of patients, as in previously published reports. Adrenal insufficiency occurred more often than expected (8.7%), for unclear reasons, with five of the six reported cases occurring more than 30 days after the last dose of pembrolizumab.
The bigger picture: Putting I-SPY2 results into context
It is well known that responses to pembrolizumab monotherapy in patients with advanced, refractory breast cancer are infrequent. In contrast, in previously untreated patients with PD-L1 positive TNBC, pembrolizumab monotherapy produced a response rate of 21% in KEYNOTE-086 (Ann Oncol. 2019 Mar 1;30(3):405-11). This response rate is similar to that observed with standard chemotherapy, but responses with pembrolizumab were more durable.
In the phase 3 KEYNOTE-355 trial (NCT02819518), researchers are comparing pembrolizumab plus chemotherapy to placebo plus chemotherapy in patients with previously untreated, stage IV TNBC with high PD-L1 expression. Researchers saw a significant and clinically meaningful improvement in progression-free survival in the pembrolizumab arm, according to a recent announcement from Merck. These results lend credence to the I-SPY2 authors’ hypothesis that immune-targeted agents would show their greatest benefit in early-stage breast cancer patients.
In fact, results from I-SPY2 have been confirmed by results from the phase 3 KEYNOTE-522 trial, which were recently published (N Engl J Med 2020;382:810-21) and presented at the San Antonio Breast Cancer Symposium. I-SPY2 predicted that pembrolizumab would be superior to standard NAC in TNBC patients in a phase 3 trial, and it was.
In KEYNOTE-522, the pCR rate was significantly higher in early-stage TNBC patients who received pembrolizumab plus NAC than in early-stage TNBC patients who received placebo plus NAC. The pCR rate was 64.8% in the pembrolizumab-NAC arm and 51.2% in the placebo–NAC arm (estimated treatment difference, 13.6 percentage points; 95% CI, 5.4 to 21.8; P less than .001).
These results are exciting. Results from I-SPY2 and KEYNOTE-522 whet the appetite for results of KEYNOTE-756, an ongoing trial of pembrolizumab plus NAC in HR-positive/ERBB2-negative patients (NCT03725059). Hopefully, the efficacy and toxicity results of KEYNOTE-756 will be as exciting as the I-SPY2 results predict they will be. Among patients with early stage breast cancer whose tumor characteristics are adverse enough to require NAC, better regimens are needed to attain pCR, a validated surrogate for long-term freedom from recurrence.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I highlight I-SPY2 and other studies of pembrolizumab plus chemotherapy in breast cancer patients.
Pathologic complete response (pCR) rates up to 60% were reported for patients with high-risk, stage II/III breast cancer who received pembrolizumab plus standard neoadjuvant chemotherapy (NAC) in I-SPY2, an ongoing platform trial designed to screen multiple agents and pinpoint those with a high probability of success (JAMA Oncol. 2020 Feb 13. doi: 10.1001/jamaoncol.2019.6650).
The addition of pembrolizumab to NAC doubled pCR rates in all three biomarker signatures studied, including ERBB2 (HER2)-negative, hormone receptor (HR)-positive/ERBB2-negative, and triple-negative breast cancer (TNBC).
As a result, pembrolizumab “graduated” from I-SPY2, with a more than 99% predictive probability that the pembrolizumab-plus-NAC approach would be superior to NAC alone in a phase 3 trial. In the HR-positive/ERBB2-negative signature, pembrolizumab is the first agent to graduate among the 10 agents studied since I-SPY2 opened in 2010.
The control arm in I-SPY2 had 181 patients randomized to standard NAC (paclitaxel followed by doxorubicin plus cyclophosphamide). The pembrolizumab arm included 69 patients who received the same NAC regimen plus pembrolizumab, given concurrently with paclitaxel.
The estimated pCR rates in all ERBB2-negative patients were 44% in the pembrolizumab arm and 17% in the control arm. Among the 40 HR-positive/ERBB2-negative patients, the estimated pCR rates were 30% and 13%, respectively. In the 29 TNBC patients, the estimated pCR rates were 60% and 22%, respectively.
Extensive residual cancer burden was less often seen in the pembrolizumab-treated patients than in the comparison group. At a median follow-up of 2.8 years in the pembrolizumab arm and 3.5 years in the NAC arm, 3-year event-free survival was similar between the arms. However, the investigators cautioned against drawing conclusions from this exploratory analysis in a small number of patients. Testifying to the importance of the primary endpoint of pCR rate, patients who achieved pCR had excellent outcomes regardless of their assigned study arms.
Immune-related adverse events in the pembrolizumab-treated patients were generally grade 1 or 2 and were managed with dose interruption or corticosteroid therapy. Most commonly seen was thyroid dysfunction in 13% of patients, as in previously published reports. Adrenal insufficiency occurred more often than expected (8.7%), for unclear reasons, with five of the six reported cases occurring more than 30 days after the last dose of pembrolizumab.
The bigger picture: Putting I-SPY2 results into context
It is well known that responses to pembrolizumab monotherapy in patients with advanced, refractory breast cancer are infrequent. In contrast, in previously untreated patients with PD-L1 positive TNBC, pembrolizumab monotherapy produced a response rate of 21% in KEYNOTE-086 (Ann Oncol. 2019 Mar 1;30(3):405-11). This response rate is similar to that observed with standard chemotherapy, but responses with pembrolizumab were more durable.
In the phase 3 KEYNOTE-355 trial (NCT02819518), researchers are comparing pembrolizumab plus chemotherapy to placebo plus chemotherapy in patients with previously untreated, stage IV TNBC with high PD-L1 expression. Researchers saw a significant and clinically meaningful improvement in progression-free survival in the pembrolizumab arm, according to a recent announcement from Merck. These results lend credence to the I-SPY2 authors’ hypothesis that immune-targeted agents would show their greatest benefit in early-stage breast cancer patients.
In fact, results from I-SPY2 have been confirmed by results from the phase 3 KEYNOTE-522 trial, which were recently published (N Engl J Med 2020;382:810-21) and presented at the San Antonio Breast Cancer Symposium. I-SPY2 predicted that pembrolizumab would be superior to standard NAC in TNBC patients in a phase 3 trial, and it was.
In KEYNOTE-522, the pCR rate was significantly higher in early-stage TNBC patients who received pembrolizumab plus NAC than in early-stage TNBC patients who received placebo plus NAC. The pCR rate was 64.8% in the pembrolizumab-NAC arm and 51.2% in the placebo–NAC arm (estimated treatment difference, 13.6 percentage points; 95% CI, 5.4 to 21.8; P less than .001).
These results are exciting. Results from I-SPY2 and KEYNOTE-522 whet the appetite for results of KEYNOTE-756, an ongoing trial of pembrolizumab plus NAC in HR-positive/ERBB2-negative patients (NCT03725059). Hopefully, the efficacy and toxicity results of KEYNOTE-756 will be as exciting as the I-SPY2 results predict they will be. Among patients with early stage breast cancer whose tumor characteristics are adverse enough to require NAC, better regimens are needed to attain pCR, a validated surrogate for long-term freedom from recurrence.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
In this edition of “Applying research to practice,” I highlight I-SPY2 and other studies of pembrolizumab plus chemotherapy in breast cancer patients.
Pathologic complete response (pCR) rates up to 60% were reported for patients with high-risk, stage II/III breast cancer who received pembrolizumab plus standard neoadjuvant chemotherapy (NAC) in I-SPY2, an ongoing platform trial designed to screen multiple agents and pinpoint those with a high probability of success (JAMA Oncol. 2020 Feb 13. doi: 10.1001/jamaoncol.2019.6650).
The addition of pembrolizumab to NAC doubled pCR rates in all three biomarker signatures studied, including ERBB2 (HER2)-negative, hormone receptor (HR)-positive/ERBB2-negative, and triple-negative breast cancer (TNBC).
As a result, pembrolizumab “graduated” from I-SPY2, with a more than 99% predictive probability that the pembrolizumab-plus-NAC approach would be superior to NAC alone in a phase 3 trial. In the HR-positive/ERBB2-negative signature, pembrolizumab is the first agent to graduate among the 10 agents studied since I-SPY2 opened in 2010.
The control arm in I-SPY2 had 181 patients randomized to standard NAC (paclitaxel followed by doxorubicin plus cyclophosphamide). The pembrolizumab arm included 69 patients who received the same NAC regimen plus pembrolizumab, given concurrently with paclitaxel.
The estimated pCR rates in all ERBB2-negative patients were 44% in the pembrolizumab arm and 17% in the control arm. Among the 40 HR-positive/ERBB2-negative patients, the estimated pCR rates were 30% and 13%, respectively. In the 29 TNBC patients, the estimated pCR rates were 60% and 22%, respectively.
Extensive residual cancer burden was less often seen in the pembrolizumab-treated patients than in the comparison group. At a median follow-up of 2.8 years in the pembrolizumab arm and 3.5 years in the NAC arm, 3-year event-free survival was similar between the arms. However, the investigators cautioned against drawing conclusions from this exploratory analysis in a small number of patients. Testifying to the importance of the primary endpoint of pCR rate, patients who achieved pCR had excellent outcomes regardless of their assigned study arms.
Immune-related adverse events in the pembrolizumab-treated patients were generally grade 1 or 2 and were managed with dose interruption or corticosteroid therapy. Most commonly seen was thyroid dysfunction in 13% of patients, as in previously published reports. Adrenal insufficiency occurred more often than expected (8.7%), for unclear reasons, with five of the six reported cases occurring more than 30 days after the last dose of pembrolizumab.
The bigger picture: Putting I-SPY2 results into context
It is well known that responses to pembrolizumab monotherapy in patients with advanced, refractory breast cancer are infrequent. In contrast, in previously untreated patients with PD-L1 positive TNBC, pembrolizumab monotherapy produced a response rate of 21% in KEYNOTE-086 (Ann Oncol. 2019 Mar 1;30(3):405-11). This response rate is similar to that observed with standard chemotherapy, but responses with pembrolizumab were more durable.
In the phase 3 KEYNOTE-355 trial (NCT02819518), researchers are comparing pembrolizumab plus chemotherapy to placebo plus chemotherapy in patients with previously untreated, stage IV TNBC with high PD-L1 expression. Researchers saw a significant and clinically meaningful improvement in progression-free survival in the pembrolizumab arm, according to a recent announcement from Merck. These results lend credence to the I-SPY2 authors’ hypothesis that immune-targeted agents would show their greatest benefit in early-stage breast cancer patients.
In fact, results from I-SPY2 have been confirmed by results from the phase 3 KEYNOTE-522 trial, which were recently published (N Engl J Med 2020;382:810-21) and presented at the San Antonio Breast Cancer Symposium. I-SPY2 predicted that pembrolizumab would be superior to standard NAC in TNBC patients in a phase 3 trial, and it was.
In KEYNOTE-522, the pCR rate was significantly higher in early-stage TNBC patients who received pembrolizumab plus NAC than in early-stage TNBC patients who received placebo plus NAC. The pCR rate was 64.8% in the pembrolizumab-NAC arm and 51.2% in the placebo–NAC arm (estimated treatment difference, 13.6 percentage points; 95% CI, 5.4 to 21.8; P less than .001).
These results are exciting. Results from I-SPY2 and KEYNOTE-522 whet the appetite for results of KEYNOTE-756, an ongoing trial of pembrolizumab plus NAC in HR-positive/ERBB2-negative patients (NCT03725059). Hopefully, the efficacy and toxicity results of KEYNOTE-756 will be as exciting as the I-SPY2 results predict they will be. Among patients with early stage breast cancer whose tumor characteristics are adverse enough to require NAC, better regimens are needed to attain pCR, a validated surrogate for long-term freedom from recurrence.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations.
Guidelines for today and tomorrow
In this edition of “How I Will Treat My Next Patient,” I review “guidelines for today” and speculate about “guidelines for tomorrow,” highlighting recommendations from the American Society of Clinical Oncology about hereditary cancer testing in epithelial ovarian cancer (OC) and data that support a reexamination of the age at which screening for colorectal cancer (CRC) should begin.
ASCO guidelines on genetic testing in epithelial ovarian cancer
After reviewing 19 studies, including 6 meta-analyses; 11 randomized, controlled trials; and 2 observational studies, an ASCO panel recommended germline genetic testing for BRCA1, BRCA2, and other ovarian cancer susceptibility genes for all women with newly diagnosed epithelial OC, regardless of family history (J Clin Oncol. 2020 Jan 27. doi: 10.1200/JCO.19.02960).
For OC patients with a germline mutation, cascade testing of first- and second-degree relatives was strongly urged. For patients without a germline mutation, the guidelines recommended offering somatic tumor testing for BRCA1/2 pathogenic or likely pathogenic variants at disease recurrence or after initial therapy and for mismatch repair deficiency (MMRD) in patients with clear cell, endometrioid, or mucinous and potentially other histologic types of OC. The authors cautioned that the discussion of testing results should involve professionals with expertise in the surveillance and management of hereditary cancer syndromes.
The panel said the discovery of germline or somatic pathogenic or likely pathogenic BRCA1/2 variants should lead to considering treatment with Food and Drug Administration–approved poly (ADP-ribose) polymerase inhibitors, including niraparib, olaparib, and rucaparib. Identification of MMRD in a patient with recurrent OC should trigger consideration of treatment with pembrolizumab, consistent with its labeled indications, and surveillance for other malignancies.
The guidelines cautioned that, when patients have variants of uncertain significance on germline testing, “clinical features and family history should inform clinical decision making.” Similarly, the panel made no recommendation regarding testing for or making treatment decisions based on tests for homologous recombination deficiency.
How these results influence practice
Every oncologist recognizes that better understanding of cancer biology can guide personalized diagnostic, predictive, prognostic, and therapeutic strategies for patients and their family members.
It is estimated that approximately 25% of all OC is caused by a heritable genetic condition. Germline mutations in BRCA1 and BRCA2 are identified in 13%-15% of patients with OC, and somatic mutations are found in an additional 7%. Perhaps 6% of all ovarian/fallopian tube/peritoneal cancers are caused by mutations in genes other than BRCA1/2. For that reason, germline sequencing should be performed via multigene panels that assess BRCA1/2 and other relevant mutations.
MMDR has been found in 10%-12% of unselected epithelial OC, with increased representation in nonserous histologies. That frequency is high enough to justify testing for it routinely.
Unfortunately, only about 30% of women undergo genetic testing. Given the frequency of molecular abnormalities in OC, this is problematic in every conceivable domain of clinical care for patients and family members. ASCO’s comprehensive, educational guidelines provide a template for shared decision making and utilize resources that are available in almost all clinical settings. For those clinicians who have recommended genetic testing for all epithelial OC patients, these guidelines are practice reaffirming. For the rest of us, they are practice changing.
Colorectal cancer cases spike after start of routine screening
Instead of examining CRC incidence by the usual 5- or 10-year age ranges, a group of researchers looked at CRC incidence in 1-year intervals for adults aged 30-60 years in the SEER-18 registry from 2000 to 2015 (JAMA Network Open. 2020 Jan 31. doi: 10.1001/jamanetworkopen.2019.20407). The researchers focused their attention on the transition between age 49 and 50 years, which is when routine screening generally begins and case-finding based on symptoms and signs of CRC alone ideally ends.
The group’s hypothesis was that steep increases in CRC incidence between ages 49 and 50 would be consistent with a high, undetected preclinical case burden in patients aged younger than 50 years and that this “real-world” registry data could help estimate outcomes of screening at younger ages. The researchers found that CRC incidence increased by 46.1% in the transition period from age 49 to 50 years. A majority (93%) of these cases were invasive and, therefore, likely to be clinically relevant. The increase in cancer rates occurred across geographical regions, gender, and race, and likely reflected the impact of screening. The states with the steepest increases in CRC between ages 49 and 50 (Connecticut and Utah) were the states with the first and third highest CRC screening rates for individuals 50 years of age and older.
Stage stratification showed steep increases in incidence in the target age range for localized and regional CRC and for colon and rectal tumors. In the transition between age 49 and 50, the researchers found a significant increase in 5-year relative survival (6.9% absolute increase, 10% relative increase), suggesting that earlier screening had a survival impact, apart from the effects of treatment in cases diagnosed after symptoms occurred.
The authors concluded that their analysis of the transition from age 49 to 50 years provides registry-based data regarding CRC risk among individuals younger than 50, which can add to existing modeling studies to help inform guidelines about the age at which to initiate screening.
How these results influence practice
Early-onset CRC (EOCRC) incidence is increasing, with controversy regarding whether average-risk screening should begin before age 50 years. The justification for starting screening at age 50 is that there is a near doubling of incidence from patients aged 45-49 years (34 per 100,000) to those aged 50-54 years (60.2 per 100,000).
However, the increase in CRC incidence beyond age 50 may not be because rates are truly lower among younger individuals but rather because of uneven screening between the two populations. Doubling times for CRCs have been estimated to be perhaps as long as 1,000 days. Because many CRCs are asymptomatic, observed incidence rates of EOCRC in SEER registries do not reflect preclinical CRC case burdens in younger patients.
The current interrogation of SEER-18 data to identify preexisting CRC that was clinically silent in the 1-year interval between age 49 and 50 is highly supportive of a large undiagnosed number of EOCRC cases. In SEER-18, CRC rates increased 46.1% in this 1-year age transition, more than in earlier 1-year age transitions. With almost 93% of cases being invasive, these data suggest a high case burden of preclinical, undetected, clinically relevant EOCRC in younger patients that is not reflected in observed SEER incidence rates examining wider age group intervals.
The dual goals of screening for CRC are to prevent malignant neoplasms by the removal of precancerous polyps and improve cancer-specific survival. The data presented suggest that, by starting average-risk screening at age 50 years, we may be “missing the window.” The 6.9% absolute and 10.1% relative survival increase in the target transition period suggest the authors’ hypothesis is correct.
As in any real-world database survey, the analysis is limited by a lack of specific outcomes data, the inability to determine when the cancers developed, and how long they germinated. Because of those limitations and others, more detailed studies are needed to determine the ideal age at which to begin CRC screening.
Modeling studies incorporating the steep incidence inflection point at 49-50 years can be conducted to estimate the incidence rate increase at, for example, 45 years; the cost-benefit ratio; quality-adjusted life-years gained; and other important endpoints. However, this review of over 170,000 cases of CRC, with a data-completeness rate of over 98%, over the 15-year time frame when CRC screening became common, supports a fresh look at whether it is within our power to improve outcomes for EOCRC patients by using existing technology but applying it earlier.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I review “guidelines for today” and speculate about “guidelines for tomorrow,” highlighting recommendations from the American Society of Clinical Oncology about hereditary cancer testing in epithelial ovarian cancer (OC) and data that support a reexamination of the age at which screening for colorectal cancer (CRC) should begin.
ASCO guidelines on genetic testing in epithelial ovarian cancer
After reviewing 19 studies, including 6 meta-analyses; 11 randomized, controlled trials; and 2 observational studies, an ASCO panel recommended germline genetic testing for BRCA1, BRCA2, and other ovarian cancer susceptibility genes for all women with newly diagnosed epithelial OC, regardless of family history (J Clin Oncol. 2020 Jan 27. doi: 10.1200/JCO.19.02960).
For OC patients with a germline mutation, cascade testing of first- and second-degree relatives was strongly urged. For patients without a germline mutation, the guidelines recommended offering somatic tumor testing for BRCA1/2 pathogenic or likely pathogenic variants at disease recurrence or after initial therapy and for mismatch repair deficiency (MMRD) in patients with clear cell, endometrioid, or mucinous and potentially other histologic types of OC. The authors cautioned that the discussion of testing results should involve professionals with expertise in the surveillance and management of hereditary cancer syndromes.
The panel said the discovery of germline or somatic pathogenic or likely pathogenic BRCA1/2 variants should lead to considering treatment with Food and Drug Administration–approved poly (ADP-ribose) polymerase inhibitors, including niraparib, olaparib, and rucaparib. Identification of MMRD in a patient with recurrent OC should trigger consideration of treatment with pembrolizumab, consistent with its labeled indications, and surveillance for other malignancies.
The guidelines cautioned that, when patients have variants of uncertain significance on germline testing, “clinical features and family history should inform clinical decision making.” Similarly, the panel made no recommendation regarding testing for or making treatment decisions based on tests for homologous recombination deficiency.
How these results influence practice
Every oncologist recognizes that better understanding of cancer biology can guide personalized diagnostic, predictive, prognostic, and therapeutic strategies for patients and their family members.
It is estimated that approximately 25% of all OC is caused by a heritable genetic condition. Germline mutations in BRCA1 and BRCA2 are identified in 13%-15% of patients with OC, and somatic mutations are found in an additional 7%. Perhaps 6% of all ovarian/fallopian tube/peritoneal cancers are caused by mutations in genes other than BRCA1/2. For that reason, germline sequencing should be performed via multigene panels that assess BRCA1/2 and other relevant mutations.
MMDR has been found in 10%-12% of unselected epithelial OC, with increased representation in nonserous histologies. That frequency is high enough to justify testing for it routinely.
Unfortunately, only about 30% of women undergo genetic testing. Given the frequency of molecular abnormalities in OC, this is problematic in every conceivable domain of clinical care for patients and family members. ASCO’s comprehensive, educational guidelines provide a template for shared decision making and utilize resources that are available in almost all clinical settings. For those clinicians who have recommended genetic testing for all epithelial OC patients, these guidelines are practice reaffirming. For the rest of us, they are practice changing.
Colorectal cancer cases spike after start of routine screening
Instead of examining CRC incidence by the usual 5- or 10-year age ranges, a group of researchers looked at CRC incidence in 1-year intervals for adults aged 30-60 years in the SEER-18 registry from 2000 to 2015 (JAMA Network Open. 2020 Jan 31. doi: 10.1001/jamanetworkopen.2019.20407). The researchers focused their attention on the transition between age 49 and 50 years, which is when routine screening generally begins and case-finding based on symptoms and signs of CRC alone ideally ends.
The group’s hypothesis was that steep increases in CRC incidence between ages 49 and 50 would be consistent with a high, undetected preclinical case burden in patients aged younger than 50 years and that this “real-world” registry data could help estimate outcomes of screening at younger ages. The researchers found that CRC incidence increased by 46.1% in the transition period from age 49 to 50 years. A majority (93%) of these cases were invasive and, therefore, likely to be clinically relevant. The increase in cancer rates occurred across geographical regions, gender, and race, and likely reflected the impact of screening. The states with the steepest increases in CRC between ages 49 and 50 (Connecticut and Utah) were the states with the first and third highest CRC screening rates for individuals 50 years of age and older.
Stage stratification showed steep increases in incidence in the target age range for localized and regional CRC and for colon and rectal tumors. In the transition between age 49 and 50, the researchers found a significant increase in 5-year relative survival (6.9% absolute increase, 10% relative increase), suggesting that earlier screening had a survival impact, apart from the effects of treatment in cases diagnosed after symptoms occurred.
The authors concluded that their analysis of the transition from age 49 to 50 years provides registry-based data regarding CRC risk among individuals younger than 50, which can add to existing modeling studies to help inform guidelines about the age at which to initiate screening.
How these results influence practice
Early-onset CRC (EOCRC) incidence is increasing, with controversy regarding whether average-risk screening should begin before age 50 years. The justification for starting screening at age 50 is that there is a near doubling of incidence from patients aged 45-49 years (34 per 100,000) to those aged 50-54 years (60.2 per 100,000).
However, the increase in CRC incidence beyond age 50 may not be because rates are truly lower among younger individuals but rather because of uneven screening between the two populations. Doubling times for CRCs have been estimated to be perhaps as long as 1,000 days. Because many CRCs are asymptomatic, observed incidence rates of EOCRC in SEER registries do not reflect preclinical CRC case burdens in younger patients.
The current interrogation of SEER-18 data to identify preexisting CRC that was clinically silent in the 1-year interval between age 49 and 50 is highly supportive of a large undiagnosed number of EOCRC cases. In SEER-18, CRC rates increased 46.1% in this 1-year age transition, more than in earlier 1-year age transitions. With almost 93% of cases being invasive, these data suggest a high case burden of preclinical, undetected, clinically relevant EOCRC in younger patients that is not reflected in observed SEER incidence rates examining wider age group intervals.
The dual goals of screening for CRC are to prevent malignant neoplasms by the removal of precancerous polyps and improve cancer-specific survival. The data presented suggest that, by starting average-risk screening at age 50 years, we may be “missing the window.” The 6.9% absolute and 10.1% relative survival increase in the target transition period suggest the authors’ hypothesis is correct.
As in any real-world database survey, the analysis is limited by a lack of specific outcomes data, the inability to determine when the cancers developed, and how long they germinated. Because of those limitations and others, more detailed studies are needed to determine the ideal age at which to begin CRC screening.
Modeling studies incorporating the steep incidence inflection point at 49-50 years can be conducted to estimate the incidence rate increase at, for example, 45 years; the cost-benefit ratio; quality-adjusted life-years gained; and other important endpoints. However, this review of over 170,000 cases of CRC, with a data-completeness rate of over 98%, over the 15-year time frame when CRC screening became common, supports a fresh look at whether it is within our power to improve outcomes for EOCRC patients by using existing technology but applying it earlier.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I review “guidelines for today” and speculate about “guidelines for tomorrow,” highlighting recommendations from the American Society of Clinical Oncology about hereditary cancer testing in epithelial ovarian cancer (OC) and data that support a reexamination of the age at which screening for colorectal cancer (CRC) should begin.
ASCO guidelines on genetic testing in epithelial ovarian cancer
After reviewing 19 studies, including 6 meta-analyses; 11 randomized, controlled trials; and 2 observational studies, an ASCO panel recommended germline genetic testing for BRCA1, BRCA2, and other ovarian cancer susceptibility genes for all women with newly diagnosed epithelial OC, regardless of family history (J Clin Oncol. 2020 Jan 27. doi: 10.1200/JCO.19.02960).
For OC patients with a germline mutation, cascade testing of first- and second-degree relatives was strongly urged. For patients without a germline mutation, the guidelines recommended offering somatic tumor testing for BRCA1/2 pathogenic or likely pathogenic variants at disease recurrence or after initial therapy and for mismatch repair deficiency (MMRD) in patients with clear cell, endometrioid, or mucinous and potentially other histologic types of OC. The authors cautioned that the discussion of testing results should involve professionals with expertise in the surveillance and management of hereditary cancer syndromes.
The panel said the discovery of germline or somatic pathogenic or likely pathogenic BRCA1/2 variants should lead to considering treatment with Food and Drug Administration–approved poly (ADP-ribose) polymerase inhibitors, including niraparib, olaparib, and rucaparib. Identification of MMRD in a patient with recurrent OC should trigger consideration of treatment with pembrolizumab, consistent with its labeled indications, and surveillance for other malignancies.
The guidelines cautioned that, when patients have variants of uncertain significance on germline testing, “clinical features and family history should inform clinical decision making.” Similarly, the panel made no recommendation regarding testing for or making treatment decisions based on tests for homologous recombination deficiency.
How these results influence practice
Every oncologist recognizes that better understanding of cancer biology can guide personalized diagnostic, predictive, prognostic, and therapeutic strategies for patients and their family members.
It is estimated that approximately 25% of all OC is caused by a heritable genetic condition. Germline mutations in BRCA1 and BRCA2 are identified in 13%-15% of patients with OC, and somatic mutations are found in an additional 7%. Perhaps 6% of all ovarian/fallopian tube/peritoneal cancers are caused by mutations in genes other than BRCA1/2. For that reason, germline sequencing should be performed via multigene panels that assess BRCA1/2 and other relevant mutations.
MMDR has been found in 10%-12% of unselected epithelial OC, with increased representation in nonserous histologies. That frequency is high enough to justify testing for it routinely.
Unfortunately, only about 30% of women undergo genetic testing. Given the frequency of molecular abnormalities in OC, this is problematic in every conceivable domain of clinical care for patients and family members. ASCO’s comprehensive, educational guidelines provide a template for shared decision making and utilize resources that are available in almost all clinical settings. For those clinicians who have recommended genetic testing for all epithelial OC patients, these guidelines are practice reaffirming. For the rest of us, they are practice changing.
Colorectal cancer cases spike after start of routine screening
Instead of examining CRC incidence by the usual 5- or 10-year age ranges, a group of researchers looked at CRC incidence in 1-year intervals for adults aged 30-60 years in the SEER-18 registry from 2000 to 2015 (JAMA Network Open. 2020 Jan 31. doi: 10.1001/jamanetworkopen.2019.20407). The researchers focused their attention on the transition between age 49 and 50 years, which is when routine screening generally begins and case-finding based on symptoms and signs of CRC alone ideally ends.
The group’s hypothesis was that steep increases in CRC incidence between ages 49 and 50 would be consistent with a high, undetected preclinical case burden in patients aged younger than 50 years and that this “real-world” registry data could help estimate outcomes of screening at younger ages. The researchers found that CRC incidence increased by 46.1% in the transition period from age 49 to 50 years. A majority (93%) of these cases were invasive and, therefore, likely to be clinically relevant. The increase in cancer rates occurred across geographical regions, gender, and race, and likely reflected the impact of screening. The states with the steepest increases in CRC between ages 49 and 50 (Connecticut and Utah) were the states with the first and third highest CRC screening rates for individuals 50 years of age and older.
Stage stratification showed steep increases in incidence in the target age range for localized and regional CRC and for colon and rectal tumors. In the transition between age 49 and 50, the researchers found a significant increase in 5-year relative survival (6.9% absolute increase, 10% relative increase), suggesting that earlier screening had a survival impact, apart from the effects of treatment in cases diagnosed after symptoms occurred.
The authors concluded that their analysis of the transition from age 49 to 50 years provides registry-based data regarding CRC risk among individuals younger than 50, which can add to existing modeling studies to help inform guidelines about the age at which to initiate screening.
How these results influence practice
Early-onset CRC (EOCRC) incidence is increasing, with controversy regarding whether average-risk screening should begin before age 50 years. The justification for starting screening at age 50 is that there is a near doubling of incidence from patients aged 45-49 years (34 per 100,000) to those aged 50-54 years (60.2 per 100,000).
However, the increase in CRC incidence beyond age 50 may not be because rates are truly lower among younger individuals but rather because of uneven screening between the two populations. Doubling times for CRCs have been estimated to be perhaps as long as 1,000 days. Because many CRCs are asymptomatic, observed incidence rates of EOCRC in SEER registries do not reflect preclinical CRC case burdens in younger patients.
The current interrogation of SEER-18 data to identify preexisting CRC that was clinically silent in the 1-year interval between age 49 and 50 is highly supportive of a large undiagnosed number of EOCRC cases. In SEER-18, CRC rates increased 46.1% in this 1-year age transition, more than in earlier 1-year age transitions. With almost 93% of cases being invasive, these data suggest a high case burden of preclinical, undetected, clinically relevant EOCRC in younger patients that is not reflected in observed SEER incidence rates examining wider age group intervals.
The dual goals of screening for CRC are to prevent malignant neoplasms by the removal of precancerous polyps and improve cancer-specific survival. The data presented suggest that, by starting average-risk screening at age 50 years, we may be “missing the window.” The 6.9% absolute and 10.1% relative survival increase in the target transition period suggest the authors’ hypothesis is correct.
As in any real-world database survey, the analysis is limited by a lack of specific outcomes data, the inability to determine when the cancers developed, and how long they germinated. Because of those limitations and others, more detailed studies are needed to determine the ideal age at which to begin CRC screening.
Modeling studies incorporating the steep incidence inflection point at 49-50 years can be conducted to estimate the incidence rate increase at, for example, 45 years; the cost-benefit ratio; quality-adjusted life-years gained; and other important endpoints. However, this review of over 170,000 cases of CRC, with a data-completeness rate of over 98%, over the 15-year time frame when CRC screening became common, supports a fresh look at whether it is within our power to improve outcomes for EOCRC patients by using existing technology but applying it earlier.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
New tools could help predict complication risks in lung and breast cancer
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of new models for predicting risks of common, clinically important situations in general oncology practice: severe neutropenia in lung cancer patients and locoregional recurrence of breast cancer.
Predicting neutropenia
Accurate, lung cancer–specific prediction models would be useful to estimate risk of chemotherapy-induced neutropenia (CIN), especially febrile neutropenia (FN), since that particular toxicity is linked to infection, dose delays and dose reductions that can compromise treatment efficacy, and poor health-related quality of life. Lung cancer patients are often older adults, with advanced disease and comorbid conditions, so they are a particularly vulnerable population for CIN.
Xiaowen Cao of Duke University, Durham, N.C., and coinvestigators published a model for predicting risk of severe CIN in advanced lung cancer patients, based on 10 pretreatment variables (Lung Cancer. 2020 Jan 5. doi: 10.1016/j.lungcan.2020.01.004). They developed their model to overcome limitations of the previously published work of Gary H. Lyman, MD, and colleagues that is not specific to lung cancer and incorporated relative dose intensity as a predictor (Cancer. 2011;117:1917-27). Relative dose intensity is not determined until after a treatment course is completed.
The new prediction model was based on a lung cancer data set encompassing 11,352 patients from 67 phase 2-3 cooperative group studies conducted between 1991 and 2010. In this data set, the Lyman model had an area under the curve of 0.8772 in patients with small cell lung cancer, but an area under the curve of just 0.6787 in non–small cell lung cancer.
The derivation model was derived from about two-thirds of the patients, randomly selected. The validation set was conducted using the remaining third. The variables included were readily clinically available: age, gender, weight, body mass index, insurance status, disease stage, number of metastatic sites, chemotherapy agents used, number of chemotherapy agents, planned growth factor use, duration of planned therapy, pleural effusion, presence of symptoms, and performance status. Their model had an area under the curve of 0.8348 in the training set and 0.8234 in the testing set.
How these results influence practice
The risk of an initial episode of FN is highest during a patient’s initial cycle of chemotherapy, when most patients are receiving full-dose treatment, often without prophylactic measures. Guidelines from the National Comprehensive Cancer Network suggest the use of prophylactic growth factors in patients with more than a 20% risk of FN, and considering using prophylaxis in patients with 10%-20% risk of FN. Underestimating those risks and failure to take adequate precautions may be particularly important for patients with lung cancer who are generally older adults, with comorbid conditions.
The comprehensive risk model for neutropenic complications that was developed by Dr. Lyman and colleagues was based on a large, prospective cohort including nearly 3,800 patients. The model had a 90% sensitivity and 96% predictive value, but was not lung cancer specific and, in this latest study, did not perform as well in the 85% of lung cancer patients with non–small cell lung cancer. The Lyman data, however, was obtained in cancer patients treated with investigator-choice chemotherapy in community practices. It remains the National Comprehensive Cancer Network standard for evaluating FN risk in patients embarking on chemotherapy for advanced malignancies. That should remain the case, pending the additional validation testing of the new lung cancer–specific model at independent institutions, treating heterogeneous patients in real-world settings.
Locoregional recurrence
A retrospective cohort analysis of SWOG 8814, a phase 3 study of tamoxifen alone versus chemotherapy plus by tamoxifen in postmenopausal, node-positive, hormone receptor–positive breast cancer patients suggests that the 21-gene assay recurrence score (RS) can aid decisions about radiotherapy (RT).
Wendy A. Woodward, MD, PhD, and colleagues, analyzed patients who underwent mastectomy or breast-conserving surgery as their local therapy (JAMA Oncol. 2020 Jan 9. doi: 10.1001/jamaoncol.2019.5559). They found that patients with an intermediate or high RS – according to the 21-gene assay OncotypeDX – had more locoregional recurrences (LRR; breast, chest wall, axilla, internal mammary, supraclavicular or infraclavicular nodes).
There were 367 patients in SWOG 8814 who received tamoxifen alone or cyclophosphamide, doxorubicin, and fluorouracil followed by tamoxifen. LRR was observed in 5.8% of patients with a low RS (less than 18) and in 13.8% of patients with an intermediate or high RS (more than 18). The estimated 10-year cumulative LRR incidence rates were 9.7% and 16.5%, respectively (P = .02).
In the subset of patients with one to three positive nodes who had mastectomy without radiotherapy, the LRR was 1.5% for those with low RS and 11.1% for those with intermediate or high RS (P = .051). No difference by RS was found in the 10-year rates of LRR among patients with four or more involved nodes who received a mastectomy without RT (25.9% vs. 27.0%; P = .27).
In multivariate analysis, incorporating RS, type of surgery, and number of involved nodes, intermediate or high RS was a significant predictor of LRR, with a hazard ratio of 2.36 (P = .04). The investigators suggested that RS, when available, should be one of the factors considered in selecting patients for postmastectomy RT.
How these results influence practice
Selecting the node-positive, hormone receptor–positive, breast cancer patients who should receive postmastectomy RT is difficult and controversial. This is particularly true for those postmenopausal patients with fewer than four involved nodes, no lymphatic or vascular invasion, and no extracapsular spread of disease into the axillary fat. Limited information exists on the ability of genomic assays to identify LRR risk.
Eleftherios P. Mamounas, MD, and colleagues examined the results of NSABP B-28, a trial of chemotherapy plus tamoxifen (J Natl Cancer Inst. 2017;109[4]. doi:10.1093/jnci/djw259). Postmastectomy RT was not permitted. They found high RS correlated with greater LRR and low RS with decreased LRR among patients with one to three positive nodes. At first blush, the prospectively treated cohort of SWOG 8814 represents a uniformly treated cohort with long-term follow-up (median, 8.5 years) and extends in an independent analysis the findings of NSABP B-28.
However, as Dr. Woodward and colleagues point out, the current study has limitations. The use of RT was extracted retrospectively and may be underreported. More modern chemotherapy and RT may lower LRR from the risks observed in SWOG 8814. Finally, the modest numbers of LRR events precluded secondary analysis of RS as a continuous variable. This is important because the risk group cutoffs suggested by the authors are not aligned with those in the recently published TailorRx study or the ongoing RxPonder trial.
The TailorRT (Regional Radiotherapy in Biomarker Low Risk Node Positive Breast Cancer) study examines the safety of omitting RT among patients with low RS and one to three positive nodes. Until the TailorRT results are reported, the controversy regarding the role of postmastectomy RT in this group will continue for patients with low nodal tumor burden and less aggressive tumor features, including low RS.
An observed LRR risk of 11.1% in SWOG 8814 among patients with N1 disease and an RS above 18 suggest that genomic risk could be one of the factors that may justify postmastectomy RT in postmenopausal patients with node-positive, hormone receptor–positive breast cancer until additional data emerge from the contemporary trials.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of new models for predicting risks of common, clinically important situations in general oncology practice: severe neutropenia in lung cancer patients and locoregional recurrence of breast cancer.
Predicting neutropenia
Accurate, lung cancer–specific prediction models would be useful to estimate risk of chemotherapy-induced neutropenia (CIN), especially febrile neutropenia (FN), since that particular toxicity is linked to infection, dose delays and dose reductions that can compromise treatment efficacy, and poor health-related quality of life. Lung cancer patients are often older adults, with advanced disease and comorbid conditions, so they are a particularly vulnerable population for CIN.
Xiaowen Cao of Duke University, Durham, N.C., and coinvestigators published a model for predicting risk of severe CIN in advanced lung cancer patients, based on 10 pretreatment variables (Lung Cancer. 2020 Jan 5. doi: 10.1016/j.lungcan.2020.01.004). They developed their model to overcome limitations of the previously published work of Gary H. Lyman, MD, and colleagues that is not specific to lung cancer and incorporated relative dose intensity as a predictor (Cancer. 2011;117:1917-27). Relative dose intensity is not determined until after a treatment course is completed.
The new prediction model was based on a lung cancer data set encompassing 11,352 patients from 67 phase 2-3 cooperative group studies conducted between 1991 and 2010. In this data set, the Lyman model had an area under the curve of 0.8772 in patients with small cell lung cancer, but an area under the curve of just 0.6787 in non–small cell lung cancer.
The derivation model was derived from about two-thirds of the patients, randomly selected. The validation set was conducted using the remaining third. The variables included were readily clinically available: age, gender, weight, body mass index, insurance status, disease stage, number of metastatic sites, chemotherapy agents used, number of chemotherapy agents, planned growth factor use, duration of planned therapy, pleural effusion, presence of symptoms, and performance status. Their model had an area under the curve of 0.8348 in the training set and 0.8234 in the testing set.
How these results influence practice
The risk of an initial episode of FN is highest during a patient’s initial cycle of chemotherapy, when most patients are receiving full-dose treatment, often without prophylactic measures. Guidelines from the National Comprehensive Cancer Network suggest the use of prophylactic growth factors in patients with more than a 20% risk of FN, and considering using prophylaxis in patients with 10%-20% risk of FN. Underestimating those risks and failure to take adequate precautions may be particularly important for patients with lung cancer who are generally older adults, with comorbid conditions.
The comprehensive risk model for neutropenic complications that was developed by Dr. Lyman and colleagues was based on a large, prospective cohort including nearly 3,800 patients. The model had a 90% sensitivity and 96% predictive value, but was not lung cancer specific and, in this latest study, did not perform as well in the 85% of lung cancer patients with non–small cell lung cancer. The Lyman data, however, was obtained in cancer patients treated with investigator-choice chemotherapy in community practices. It remains the National Comprehensive Cancer Network standard for evaluating FN risk in patients embarking on chemotherapy for advanced malignancies. That should remain the case, pending the additional validation testing of the new lung cancer–specific model at independent institutions, treating heterogeneous patients in real-world settings.
Locoregional recurrence
A retrospective cohort analysis of SWOG 8814, a phase 3 study of tamoxifen alone versus chemotherapy plus by tamoxifen in postmenopausal, node-positive, hormone receptor–positive breast cancer patients suggests that the 21-gene assay recurrence score (RS) can aid decisions about radiotherapy (RT).
Wendy A. Woodward, MD, PhD, and colleagues, analyzed patients who underwent mastectomy or breast-conserving surgery as their local therapy (JAMA Oncol. 2020 Jan 9. doi: 10.1001/jamaoncol.2019.5559). They found that patients with an intermediate or high RS – according to the 21-gene assay OncotypeDX – had more locoregional recurrences (LRR; breast, chest wall, axilla, internal mammary, supraclavicular or infraclavicular nodes).
There were 367 patients in SWOG 8814 who received tamoxifen alone or cyclophosphamide, doxorubicin, and fluorouracil followed by tamoxifen. LRR was observed in 5.8% of patients with a low RS (less than 18) and in 13.8% of patients with an intermediate or high RS (more than 18). The estimated 10-year cumulative LRR incidence rates were 9.7% and 16.5%, respectively (P = .02).
In the subset of patients with one to three positive nodes who had mastectomy without radiotherapy, the LRR was 1.5% for those with low RS and 11.1% for those with intermediate or high RS (P = .051). No difference by RS was found in the 10-year rates of LRR among patients with four or more involved nodes who received a mastectomy without RT (25.9% vs. 27.0%; P = .27).
In multivariate analysis, incorporating RS, type of surgery, and number of involved nodes, intermediate or high RS was a significant predictor of LRR, with a hazard ratio of 2.36 (P = .04). The investigators suggested that RS, when available, should be one of the factors considered in selecting patients for postmastectomy RT.
How these results influence practice
Selecting the node-positive, hormone receptor–positive, breast cancer patients who should receive postmastectomy RT is difficult and controversial. This is particularly true for those postmenopausal patients with fewer than four involved nodes, no lymphatic or vascular invasion, and no extracapsular spread of disease into the axillary fat. Limited information exists on the ability of genomic assays to identify LRR risk.
Eleftherios P. Mamounas, MD, and colleagues examined the results of NSABP B-28, a trial of chemotherapy plus tamoxifen (J Natl Cancer Inst. 2017;109[4]. doi:10.1093/jnci/djw259). Postmastectomy RT was not permitted. They found high RS correlated with greater LRR and low RS with decreased LRR among patients with one to three positive nodes. At first blush, the prospectively treated cohort of SWOG 8814 represents a uniformly treated cohort with long-term follow-up (median, 8.5 years) and extends in an independent analysis the findings of NSABP B-28.
However, as Dr. Woodward and colleagues point out, the current study has limitations. The use of RT was extracted retrospectively and may be underreported. More modern chemotherapy and RT may lower LRR from the risks observed in SWOG 8814. Finally, the modest numbers of LRR events precluded secondary analysis of RS as a continuous variable. This is important because the risk group cutoffs suggested by the authors are not aligned with those in the recently published TailorRx study or the ongoing RxPonder trial.
The TailorRT (Regional Radiotherapy in Biomarker Low Risk Node Positive Breast Cancer) study examines the safety of omitting RT among patients with low RS and one to three positive nodes. Until the TailorRT results are reported, the controversy regarding the role of postmastectomy RT in this group will continue for patients with low nodal tumor burden and less aggressive tumor features, including low RS.
An observed LRR risk of 11.1% in SWOG 8814 among patients with N1 disease and an RS above 18 suggest that genomic risk could be one of the factors that may justify postmastectomy RT in postmenopausal patients with node-positive, hormone receptor–positive breast cancer until additional data emerge from the contemporary trials.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of new models for predicting risks of common, clinically important situations in general oncology practice: severe neutropenia in lung cancer patients and locoregional recurrence of breast cancer.
Predicting neutropenia
Accurate, lung cancer–specific prediction models would be useful to estimate risk of chemotherapy-induced neutropenia (CIN), especially febrile neutropenia (FN), since that particular toxicity is linked to infection, dose delays and dose reductions that can compromise treatment efficacy, and poor health-related quality of life. Lung cancer patients are often older adults, with advanced disease and comorbid conditions, so they are a particularly vulnerable population for CIN.
Xiaowen Cao of Duke University, Durham, N.C., and coinvestigators published a model for predicting risk of severe CIN in advanced lung cancer patients, based on 10 pretreatment variables (Lung Cancer. 2020 Jan 5. doi: 10.1016/j.lungcan.2020.01.004). They developed their model to overcome limitations of the previously published work of Gary H. Lyman, MD, and colleagues that is not specific to lung cancer and incorporated relative dose intensity as a predictor (Cancer. 2011;117:1917-27). Relative dose intensity is not determined until after a treatment course is completed.
The new prediction model was based on a lung cancer data set encompassing 11,352 patients from 67 phase 2-3 cooperative group studies conducted between 1991 and 2010. In this data set, the Lyman model had an area under the curve of 0.8772 in patients with small cell lung cancer, but an area under the curve of just 0.6787 in non–small cell lung cancer.
The derivation model was derived from about two-thirds of the patients, randomly selected. The validation set was conducted using the remaining third. The variables included were readily clinically available: age, gender, weight, body mass index, insurance status, disease stage, number of metastatic sites, chemotherapy agents used, number of chemotherapy agents, planned growth factor use, duration of planned therapy, pleural effusion, presence of symptoms, and performance status. Their model had an area under the curve of 0.8348 in the training set and 0.8234 in the testing set.
How these results influence practice
The risk of an initial episode of FN is highest during a patient’s initial cycle of chemotherapy, when most patients are receiving full-dose treatment, often without prophylactic measures. Guidelines from the National Comprehensive Cancer Network suggest the use of prophylactic growth factors in patients with more than a 20% risk of FN, and considering using prophylaxis in patients with 10%-20% risk of FN. Underestimating those risks and failure to take adequate precautions may be particularly important for patients with lung cancer who are generally older adults, with comorbid conditions.
The comprehensive risk model for neutropenic complications that was developed by Dr. Lyman and colleagues was based on a large, prospective cohort including nearly 3,800 patients. The model had a 90% sensitivity and 96% predictive value, but was not lung cancer specific and, in this latest study, did not perform as well in the 85% of lung cancer patients with non–small cell lung cancer. The Lyman data, however, was obtained in cancer patients treated with investigator-choice chemotherapy in community practices. It remains the National Comprehensive Cancer Network standard for evaluating FN risk in patients embarking on chemotherapy for advanced malignancies. That should remain the case, pending the additional validation testing of the new lung cancer–specific model at independent institutions, treating heterogeneous patients in real-world settings.
Locoregional recurrence
A retrospective cohort analysis of SWOG 8814, a phase 3 study of tamoxifen alone versus chemotherapy plus by tamoxifen in postmenopausal, node-positive, hormone receptor–positive breast cancer patients suggests that the 21-gene assay recurrence score (RS) can aid decisions about radiotherapy (RT).
Wendy A. Woodward, MD, PhD, and colleagues, analyzed patients who underwent mastectomy or breast-conserving surgery as their local therapy (JAMA Oncol. 2020 Jan 9. doi: 10.1001/jamaoncol.2019.5559). They found that patients with an intermediate or high RS – according to the 21-gene assay OncotypeDX – had more locoregional recurrences (LRR; breast, chest wall, axilla, internal mammary, supraclavicular or infraclavicular nodes).
There were 367 patients in SWOG 8814 who received tamoxifen alone or cyclophosphamide, doxorubicin, and fluorouracil followed by tamoxifen. LRR was observed in 5.8% of patients with a low RS (less than 18) and in 13.8% of patients with an intermediate or high RS (more than 18). The estimated 10-year cumulative LRR incidence rates were 9.7% and 16.5%, respectively (P = .02).
In the subset of patients with one to three positive nodes who had mastectomy without radiotherapy, the LRR was 1.5% for those with low RS and 11.1% for those with intermediate or high RS (P = .051). No difference by RS was found in the 10-year rates of LRR among patients with four or more involved nodes who received a mastectomy without RT (25.9% vs. 27.0%; P = .27).
In multivariate analysis, incorporating RS, type of surgery, and number of involved nodes, intermediate or high RS was a significant predictor of LRR, with a hazard ratio of 2.36 (P = .04). The investigators suggested that RS, when available, should be one of the factors considered in selecting patients for postmastectomy RT.
How these results influence practice
Selecting the node-positive, hormone receptor–positive, breast cancer patients who should receive postmastectomy RT is difficult and controversial. This is particularly true for those postmenopausal patients with fewer than four involved nodes, no lymphatic or vascular invasion, and no extracapsular spread of disease into the axillary fat. Limited information exists on the ability of genomic assays to identify LRR risk.
Eleftherios P. Mamounas, MD, and colleagues examined the results of NSABP B-28, a trial of chemotherapy plus tamoxifen (J Natl Cancer Inst. 2017;109[4]. doi:10.1093/jnci/djw259). Postmastectomy RT was not permitted. They found high RS correlated with greater LRR and low RS with decreased LRR among patients with one to three positive nodes. At first blush, the prospectively treated cohort of SWOG 8814 represents a uniformly treated cohort with long-term follow-up (median, 8.5 years) and extends in an independent analysis the findings of NSABP B-28.
However, as Dr. Woodward and colleagues point out, the current study has limitations. The use of RT was extracted retrospectively and may be underreported. More modern chemotherapy and RT may lower LRR from the risks observed in SWOG 8814. Finally, the modest numbers of LRR events precluded secondary analysis of RS as a continuous variable. This is important because the risk group cutoffs suggested by the authors are not aligned with those in the recently published TailorRx study or the ongoing RxPonder trial.
The TailorRT (Regional Radiotherapy in Biomarker Low Risk Node Positive Breast Cancer) study examines the safety of omitting RT among patients with low RS and one to three positive nodes. Until the TailorRT results are reported, the controversy regarding the role of postmastectomy RT in this group will continue for patients with low nodal tumor burden and less aggressive tumor features, including low RS.
An observed LRR risk of 11.1% in SWOG 8814 among patients with N1 disease and an RS above 18 suggest that genomic risk could be one of the factors that may justify postmastectomy RT in postmenopausal patients with node-positive, hormone receptor–positive breast cancer until additional data emerge from the contemporary trials.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
ctDNA and avapritinib in GI cancer management
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of circulating tumor DNA (ctDNA) monitoring in patients with a history of nonmetastatic colorectal cancer (CRC) and the approval of a new targeted agent for a subset of patients with gastrointestinal stromal tumors (GISTs). Taken together, this information may guide the management of selected patients with gastrointestinal malignancies, now and in the future.
ctDNA in colorectal cancer
Unfortunately, among patients with CRC with standard blood monitoring, multiple, incurable metastases are the predominant finding when relapse occurs. To improve upon the detection of potentially curable recurrent disease, researchers performed a cross-sectional, observational study of routine monitoring of ctDNA in patients with stage I-III CRC. The patients were also monitored with carcinoembryonic antigen (CEA) levels and imaging in accordance with guidelines from the National Comprehensive Cancer Network (Cancer. 2020 Jan 7. doi: 10.1002/cncr.32695). The investigators used the COLVERA assay of methylated CBAT1/IKZF1 for ctDNA monitoring and the LIAISON CEA plasma test.
Among the 50 patients with recurrence and 177 without recurrence of CRC who met all blood and imaging collection criteria, ctDNA testing offered a sensitivity of 68.1%, compared with 31.9% for CEA (P = .0002), with comparable specificity (97.9% vs. 96.4%; P = 1.000). Over about a 4-year time period, ctDNA detected an additional 18 patients (38%) with recurrence who did not have an elevated plasma CEA, whereas there was only 1 case (2.1%) with an elevated plasma CEA and negative ctDNA test.
Among recurrences considered amenable to surgery with curative intent, (n = 20), ctDNA was positive in 60% of patients, compared with only 20% for CEA (P = .010). Multivariate analysis indicated that ctDNA was an independent predictor of recurrence, whereas CEA was not.
The authors concluded that the methylated BCAT1/IKZF1 ctDNA test was superior to CEA monitoring after initial treatment for potentially curable CRC.
How these results influence practice
The current study adds to the body of work showing superior sensitivity of ctDNA monitoring in the detection of recurrence in patients with solid tumors. In May 2019, this column highlighted the work of Yuxuan Wang, MD, PhD, and colleagues in 58 patients with stages I-III CRC; over a similar duration of follow-up as in the current study (JAMA Oncol. 2019;5[8]:1118-23), Dr. Wang found that 10 of 13 recurrences (77%) were detected by monitoring ctDNA levels. CEA levels were detected 63% of recurrences.
The central, critical hypothesis of these efforts is that the earlier detection of metastatic disease will lead to improved survival for patients with CRC. As the authors state in their discussion, that hypothesis remains tantalizing but unproven.
Prospective, randomized trials like the recently opened COBRA trial (NRG-GI005), which tests ctDNA assay–directed therapy in patients with stage IIA CRC, deserve our enthusiastic support. Research with similar designs will establish the value of ctDNA monitoring as a biomarker for early intervention in patients with stage IV disease and for selecting patients after primary therapy who are unlikely to have been cured and who could potentially benefit from adjuvant therapy.
Monitoring ctDNA is a research tool that could be practice changing in the near future. It has shown enough promise to justify prospective, randomized trials to determine whether its superior sensitivity translates into improved survival.
Avapritinib for mutated GIST
The platelet-derived growth factor receptor–alpha (PDGFRA) gene encodes PDGFRA, a member of the type III tyrosine kinase receptor family, which includes the stem cell factor receptor, KIT (present in 95% of GISTs). PDGFRA and KIT have domains with specific roles in tyrosine kinase activation. Several PDGF isoforms bind and activate PDGFRA.
Among patients with GISTs, clinical responses to imatinib and other tyrosine kinase inhibitors (TKIs) correlate with tumor genotype. For example, many GISTs that progress within 6 months of TKI initiation lack mutations in KIT or PDGFRA or have a PDGFRA D842 mutation.
The Food and Drug Administration recently approved avapritinib (Ayvakit) for adults with unresectable or metastatic GISTs with a PDGFRA exon 18 mutation. The approval was based on the results of the NAVIGATOR trial, a single-arm, multicenter, open-label study of 43 patients with PDGFRA exon 18 mutations, including 38 patients with a PDGFRA D842V mutation.
The patients received 300 mg avapritinib once daily and 84% responded to avapritinib, though most were partial responses. Among patients with a D842V mutation, the response rate was also high – at 89% – but most responses were partial. At a median follow-up of 10.6 months, median response duration was not reached, but 61% of responses exceeded 6 months. Toxicity included asthenia, gastrointestinal and central nervous system side effects (including intracranial hemorrhage), hair color changes, lacrimation, and dizziness.
Avapritinib is the first drug approved specifically for patients with advanced or unresectable GISTs with a PDGFRA exon 18 mutation.
How these results influence practice
PDGFRA exon 18 mutations occur in 5%-7% of GISTs and are believed to stabilize the kinase activation loop, perhaps accounting for slow, steady growth of these tumors over a long time period and resistance to TKIs. The most frequent mutation results in an exon 18 D842V substitution (75% of all PDGFRA-mutated tumors). Although dramatic progress has been made in the treatment of patients with GIST since 2000, in patients with GISTs harboring PDGFRA exon 18 mutations, responses to treatment are rare and, when they occur, are more abbreviated.
With a high response rate and impressive response duration, avapritinib will be a valuable resource for oncologists treating this uncommon subset of patients with GISTs. The opportunity to convert unresectable tumors to potentially curable ones seems within reach.
The difficult dilemma will be to decide when treatment is needed and drug-related toxicity is justified in patients with slowly progressive tumors and few symptoms. As with so many decisions in oncology, until newer agents with fewer toxicities and higher complete response rates are developed, the proper time for any individual patient to embark on treatment with avapritinib will be found at the intersection of “precision medicine” and “clinical judgment.”
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of circulating tumor DNA (ctDNA) monitoring in patients with a history of nonmetastatic colorectal cancer (CRC) and the approval of a new targeted agent for a subset of patients with gastrointestinal stromal tumors (GISTs). Taken together, this information may guide the management of selected patients with gastrointestinal malignancies, now and in the future.
ctDNA in colorectal cancer
Unfortunately, among patients with CRC with standard blood monitoring, multiple, incurable metastases are the predominant finding when relapse occurs. To improve upon the detection of potentially curable recurrent disease, researchers performed a cross-sectional, observational study of routine monitoring of ctDNA in patients with stage I-III CRC. The patients were also monitored with carcinoembryonic antigen (CEA) levels and imaging in accordance with guidelines from the National Comprehensive Cancer Network (Cancer. 2020 Jan 7. doi: 10.1002/cncr.32695). The investigators used the COLVERA assay of methylated CBAT1/IKZF1 for ctDNA monitoring and the LIAISON CEA plasma test.
Among the 50 patients with recurrence and 177 without recurrence of CRC who met all blood and imaging collection criteria, ctDNA testing offered a sensitivity of 68.1%, compared with 31.9% for CEA (P = .0002), with comparable specificity (97.9% vs. 96.4%; P = 1.000). Over about a 4-year time period, ctDNA detected an additional 18 patients (38%) with recurrence who did not have an elevated plasma CEA, whereas there was only 1 case (2.1%) with an elevated plasma CEA and negative ctDNA test.
Among recurrences considered amenable to surgery with curative intent, (n = 20), ctDNA was positive in 60% of patients, compared with only 20% for CEA (P = .010). Multivariate analysis indicated that ctDNA was an independent predictor of recurrence, whereas CEA was not.
The authors concluded that the methylated BCAT1/IKZF1 ctDNA test was superior to CEA monitoring after initial treatment for potentially curable CRC.
How these results influence practice
The current study adds to the body of work showing superior sensitivity of ctDNA monitoring in the detection of recurrence in patients with solid tumors. In May 2019, this column highlighted the work of Yuxuan Wang, MD, PhD, and colleagues in 58 patients with stages I-III CRC; over a similar duration of follow-up as in the current study (JAMA Oncol. 2019;5[8]:1118-23), Dr. Wang found that 10 of 13 recurrences (77%) were detected by monitoring ctDNA levels. CEA levels were detected 63% of recurrences.
The central, critical hypothesis of these efforts is that the earlier detection of metastatic disease will lead to improved survival for patients with CRC. As the authors state in their discussion, that hypothesis remains tantalizing but unproven.
Prospective, randomized trials like the recently opened COBRA trial (NRG-GI005), which tests ctDNA assay–directed therapy in patients with stage IIA CRC, deserve our enthusiastic support. Research with similar designs will establish the value of ctDNA monitoring as a biomarker for early intervention in patients with stage IV disease and for selecting patients after primary therapy who are unlikely to have been cured and who could potentially benefit from adjuvant therapy.
Monitoring ctDNA is a research tool that could be practice changing in the near future. It has shown enough promise to justify prospective, randomized trials to determine whether its superior sensitivity translates into improved survival.
Avapritinib for mutated GIST
The platelet-derived growth factor receptor–alpha (PDGFRA) gene encodes PDGFRA, a member of the type III tyrosine kinase receptor family, which includes the stem cell factor receptor, KIT (present in 95% of GISTs). PDGFRA and KIT have domains with specific roles in tyrosine kinase activation. Several PDGF isoforms bind and activate PDGFRA.
Among patients with GISTs, clinical responses to imatinib and other tyrosine kinase inhibitors (TKIs) correlate with tumor genotype. For example, many GISTs that progress within 6 months of TKI initiation lack mutations in KIT or PDGFRA or have a PDGFRA D842 mutation.
The Food and Drug Administration recently approved avapritinib (Ayvakit) for adults with unresectable or metastatic GISTs with a PDGFRA exon 18 mutation. The approval was based on the results of the NAVIGATOR trial, a single-arm, multicenter, open-label study of 43 patients with PDGFRA exon 18 mutations, including 38 patients with a PDGFRA D842V mutation.
The patients received 300 mg avapritinib once daily and 84% responded to avapritinib, though most were partial responses. Among patients with a D842V mutation, the response rate was also high – at 89% – but most responses were partial. At a median follow-up of 10.6 months, median response duration was not reached, but 61% of responses exceeded 6 months. Toxicity included asthenia, gastrointestinal and central nervous system side effects (including intracranial hemorrhage), hair color changes, lacrimation, and dizziness.
Avapritinib is the first drug approved specifically for patients with advanced or unresectable GISTs with a PDGFRA exon 18 mutation.
How these results influence practice
PDGFRA exon 18 mutations occur in 5%-7% of GISTs and are believed to stabilize the kinase activation loop, perhaps accounting for slow, steady growth of these tumors over a long time period and resistance to TKIs. The most frequent mutation results in an exon 18 D842V substitution (75% of all PDGFRA-mutated tumors). Although dramatic progress has been made in the treatment of patients with GIST since 2000, in patients with GISTs harboring PDGFRA exon 18 mutations, responses to treatment are rare and, when they occur, are more abbreviated.
With a high response rate and impressive response duration, avapritinib will be a valuable resource for oncologists treating this uncommon subset of patients with GISTs. The opportunity to convert unresectable tumors to potentially curable ones seems within reach.
The difficult dilemma will be to decide when treatment is needed and drug-related toxicity is justified in patients with slowly progressive tumors and few symptoms. As with so many decisions in oncology, until newer agents with fewer toxicities and higher complete response rates are developed, the proper time for any individual patient to embark on treatment with avapritinib will be found at the intersection of “precision medicine” and “clinical judgment.”
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight the potential role of circulating tumor DNA (ctDNA) monitoring in patients with a history of nonmetastatic colorectal cancer (CRC) and the approval of a new targeted agent for a subset of patients with gastrointestinal stromal tumors (GISTs). Taken together, this information may guide the management of selected patients with gastrointestinal malignancies, now and in the future.
ctDNA in colorectal cancer
Unfortunately, among patients with CRC with standard blood monitoring, multiple, incurable metastases are the predominant finding when relapse occurs. To improve upon the detection of potentially curable recurrent disease, researchers performed a cross-sectional, observational study of routine monitoring of ctDNA in patients with stage I-III CRC. The patients were also monitored with carcinoembryonic antigen (CEA) levels and imaging in accordance with guidelines from the National Comprehensive Cancer Network (Cancer. 2020 Jan 7. doi: 10.1002/cncr.32695). The investigators used the COLVERA assay of methylated CBAT1/IKZF1 for ctDNA monitoring and the LIAISON CEA plasma test.
Among the 50 patients with recurrence and 177 without recurrence of CRC who met all blood and imaging collection criteria, ctDNA testing offered a sensitivity of 68.1%, compared with 31.9% for CEA (P = .0002), with comparable specificity (97.9% vs. 96.4%; P = 1.000). Over about a 4-year time period, ctDNA detected an additional 18 patients (38%) with recurrence who did not have an elevated plasma CEA, whereas there was only 1 case (2.1%) with an elevated plasma CEA and negative ctDNA test.
Among recurrences considered amenable to surgery with curative intent, (n = 20), ctDNA was positive in 60% of patients, compared with only 20% for CEA (P = .010). Multivariate analysis indicated that ctDNA was an independent predictor of recurrence, whereas CEA was not.
The authors concluded that the methylated BCAT1/IKZF1 ctDNA test was superior to CEA monitoring after initial treatment for potentially curable CRC.
How these results influence practice
The current study adds to the body of work showing superior sensitivity of ctDNA monitoring in the detection of recurrence in patients with solid tumors. In May 2019, this column highlighted the work of Yuxuan Wang, MD, PhD, and colleagues in 58 patients with stages I-III CRC; over a similar duration of follow-up as in the current study (JAMA Oncol. 2019;5[8]:1118-23), Dr. Wang found that 10 of 13 recurrences (77%) were detected by monitoring ctDNA levels. CEA levels were detected 63% of recurrences.
The central, critical hypothesis of these efforts is that the earlier detection of metastatic disease will lead to improved survival for patients with CRC. As the authors state in their discussion, that hypothesis remains tantalizing but unproven.
Prospective, randomized trials like the recently opened COBRA trial (NRG-GI005), which tests ctDNA assay–directed therapy in patients with stage IIA CRC, deserve our enthusiastic support. Research with similar designs will establish the value of ctDNA monitoring as a biomarker for early intervention in patients with stage IV disease and for selecting patients after primary therapy who are unlikely to have been cured and who could potentially benefit from adjuvant therapy.
Monitoring ctDNA is a research tool that could be practice changing in the near future. It has shown enough promise to justify prospective, randomized trials to determine whether its superior sensitivity translates into improved survival.
Avapritinib for mutated GIST
The platelet-derived growth factor receptor–alpha (PDGFRA) gene encodes PDGFRA, a member of the type III tyrosine kinase receptor family, which includes the stem cell factor receptor, KIT (present in 95% of GISTs). PDGFRA and KIT have domains with specific roles in tyrosine kinase activation. Several PDGF isoforms bind and activate PDGFRA.
Among patients with GISTs, clinical responses to imatinib and other tyrosine kinase inhibitors (TKIs) correlate with tumor genotype. For example, many GISTs that progress within 6 months of TKI initiation lack mutations in KIT or PDGFRA or have a PDGFRA D842 mutation.
The Food and Drug Administration recently approved avapritinib (Ayvakit) for adults with unresectable or metastatic GISTs with a PDGFRA exon 18 mutation. The approval was based on the results of the NAVIGATOR trial, a single-arm, multicenter, open-label study of 43 patients with PDGFRA exon 18 mutations, including 38 patients with a PDGFRA D842V mutation.
The patients received 300 mg avapritinib once daily and 84% responded to avapritinib, though most were partial responses. Among patients with a D842V mutation, the response rate was also high – at 89% – but most responses were partial. At a median follow-up of 10.6 months, median response duration was not reached, but 61% of responses exceeded 6 months. Toxicity included asthenia, gastrointestinal and central nervous system side effects (including intracranial hemorrhage), hair color changes, lacrimation, and dizziness.
Avapritinib is the first drug approved specifically for patients with advanced or unresectable GISTs with a PDGFRA exon 18 mutation.
How these results influence practice
PDGFRA exon 18 mutations occur in 5%-7% of GISTs and are believed to stabilize the kinase activation loop, perhaps accounting for slow, steady growth of these tumors over a long time period and resistance to TKIs. The most frequent mutation results in an exon 18 D842V substitution (75% of all PDGFRA-mutated tumors). Although dramatic progress has been made in the treatment of patients with GIST since 2000, in patients with GISTs harboring PDGFRA exon 18 mutations, responses to treatment are rare and, when they occur, are more abbreviated.
With a high response rate and impressive response duration, avapritinib will be a valuable resource for oncologists treating this uncommon subset of patients with GISTs. The opportunity to convert unresectable tumors to potentially curable ones seems within reach.
The difficult dilemma will be to decide when treatment is needed and drug-related toxicity is justified in patients with slowly progressive tumors and few symptoms. As with so many decisions in oncology, until newer agents with fewer toxicities and higher complete response rates are developed, the proper time for any individual patient to embark on treatment with avapritinib will be found at the intersection of “precision medicine” and “clinical judgment.”
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
New evidence informs discussions on FL treatment and breast screening
In this edition of “How I Will Treat My Next Patient,” I highlight two recent studies that reinforce prior diagnostic and treatment preferences among many oncologists and will certainly influence the discussions we all will have with our patients in the coming weeks and months.
Rituximab maintenance in follicular lymphoma
The largest trial addressing the role of maintenance CD20-targeted antibody therapy was the Primary Rituximab and Maintenance (PRIMA) phase 3, intergroup study in 1,018 advanced follicular lymphoma patients with an initial response to induction chemoimmunotherapy. The induction regimen could be either of three commonly used regimens plus rituximab. Importantly, none of the induction regimens included bendamustine.
Patients were randomized to 2 years of rituximab maintenance or observation. Prior interim analyses at 3 and 6 years showed improvements for the rituximab maintenance patients in progression-free survival and other secondary endpoints, but no improvement in overall survival. Emmanuel Bachy, MD, PhD, and colleagues published the final survival data after 9 years of follow-up, including a final safety analysis (J Clin Oncol. 2019 Nov 1;37[31]:2815-24).
Among the 607 patients consenting to extended follow-up, median progression-free survival was 10.5 years with rituximab maintenance, compared with 4.1 years for observation (hazard ratio, 0.61; P less than .001). There were more patients with progression-free survival at 3 years, complete response or unconfirmed complete response at 2 years, longer time to next antilymphoma treatment, and later time to next chemotherapy. But the 10-year overall survival was similar in the two groups, as were the quality of life ratings.
In all, 503 patients experienced disease progression. Overall survival after progression was shorter in the rituximab maintenance arm versus the observation arm. Among the approximately 4% of patients experiencing transformation from follicular lymphoma to a more aggressive histology, there was no difference observed in time to transformation. Results were independent of the induction regimen received, response to induction, Follicular Lymphoma International Prognostic Index score, and other clinical factors.
In the safety analysis, there were more grade 3-4 adverse events among the rituximab maintenance patients – primarily cytopenia and infections – more serious adverse events, and more adverse events overall. Fatal adverse events were low in both groups (1.6 vs. 0.6%).
How these results influence practice
Many years ago, a senior mentor of mine taught that “progression-free survival is an important endpoint since the quality of life for patients is generally superior before relapse than afterwards.”
With that mantra playing in my mind and with the reality that discussions about relapse and subsequent treatment are never easy, the results of PRIMA seem straightforward: Rituximab maintenance is beneficial, despite the absence of an overall survival benefit, in a disease with multiple options for subsequent therapy and a long natural history. In this 9-year, final analysis of PRIMA, rituximab maintenance seems to have achieved its goals of deepening responses with low (but not inconsequential) toxicity and delaying substantially those difficult conversations with patients about how to proceed after relapse.
The influence of the final analysis of PRIMA, however, may be more complicated that it would initially seem. Bendamustine-containing regimens were not employed, are commonly used now, and some studies with bendamustine have suggested higher nonrelapse mortality with rituximab maintenance.
Low-grade adverse events take on greater importance for patients on long-term therapy than they do for patients receiving abbreviated treatment, and with noncurative treatment, the higher adverse event profile with 2 years of rituximab maintenance needs to be taken seriously. Induction and maintenance regimens involving rituximab alone, lenalidomide, or obinutuzumab may be preferred for some patients. For all of those reasons, the discussion about rituximab with advanced follicular lymphoma patients remains a long one, demanding detailed descriptions of risks, benefits, limitations of the data, and multiple modern day alternatives to the treatments employed in PRIMA.
Supplemental MRI for dense breast tissue
In the DENSE trial, investigators assigned 40,373 women with extremely dense breast tissue and negative results on screening mammography to be offered either supplemental MRI or mammography screening every other year only. The women were 50-75 years old and were enrolled in the Dutch population-based digital mammography screening program. The primary outcome was the difference in the number of cancers that developed in the 2-year interval between mammograms (N Engl J Med 2019;381:2091-102).
The interval cancer rate was 2.5 per 1,000 screenings among 4,783 women in the “MRI-invitation” group (41% of whom did not actually agree to have an MRI), and 5 per 1,000 in the 32,312 women in the “mammography-only” group, a difference of 2.5 per 1,000 screenings (P less than .001).
Of the 20 interval cancers diagnosed in the MRI-invitation group, 4 were diagnosed in the 59% of women who had undergone MRI (0.8 per 1,000 screenings). The remaining 16 were diagnosed in those who had declined an MRI (4.9 per 1,000 screenings) – virtually identical to the rate of interval cancers in the group that was not invited to have an MRI. This speaks against nonrandom allocation of patients between the mammography-only and the supplemental MRI groups.
Although supplemental MRI was associated with a cancer-detection rate of 16.5 per 1,000 screenings, there was a false positive rate of 8.0% (79.8 per 1,000 screenings). Of the women who underwent a breast biopsy after MRI, 73.7% did not have cancer. Among the women who had MRI-detected cancers, in general, the malignancies diagnosed were smaller, more likely node negative, better differentiated, and more often hormone receptor–positive, compared with those in the mammography-only group.
At the next screening mammogram, the cancer detection rate was lower in the MRI-screened group (2 per 1,000 screenings) than in the MRI–“offered but declined” (7.1 per 1,000 screenings) or mammogram-only groups (6 per 1,000 screenings).
How these results influence practice
American physicians are obligated to inform women with dense breast tissue about the limitations (for them) of conventional mammography, an unquantified, but elevated, risk of breast cancer in women with dense breast tissue, and the fact that there are no universally accepted recommended subsequent steps those women should take.
One side benefit of the requirement to disclose information about breast density, however, is that disclosure can prompt discussions about breast cancer risk reduction – an important discussion with multiple possible health-maintaining interventions.
The DENSE trial is very important news, with potential for even more value in future years. It is finite (2 years of study, with only one MRI scan mandated), narrow (Vopara or BI-RADs breast density grade 4 only), and oligo-institutional (eight participating centers in the Netherlands). Still, it provides randomized, rigorously gathered and analyzed data where there were previously none. Importantly, it answers the question, “So what can I do now, doctor?”
I have some concerns about whether we, as a medical community, have the discipline to restrict application of supplemental MRI screening beyond the population that was studied in the Netherlands. Will we be able to restrain ourselves from ordering MRI scans in women with heterogeneously dense – but not extremely dense – breasts? Will we truly manage patients whose MRI scans had BI-RADs readings of less than 4 in the rigorous, but conservative, fashion employed in the trial? Would we miss fewer significant cancers and subject fewer patients to potential expense and harm with annual mammography, instead of the biennial screening performed in the Netherlands? Does tomo-synthesis improve the interpretation of screening mammograms so much that the interval cancer rate is a lot closer to the rate obtained with supplemental MRI scans? Is the expense of MRI scanning, with the resultant subsequent tests and procedures, justified by the reduction in detecting relatively favorable breast cancers that may not materially impact overall survival?
The DENSE study promises to be a “gift that keeps on giving” as the investigators continue to assess a number of factors including: the value of ongoing supplemental MRI scans (compared with the one-time-only screening reported here); whether there will be a reduction in the advanced cancers and subsequent mortality benefit; the extent of over diagnosis, costs, and impact on quality-of-life; and the applicability of artificial intelligence techniques to reduce false positive MRI results.
While the study may not be practice changing in the United States at the present time, it may be just that as subsequent analyses emerge.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight two recent studies that reinforce prior diagnostic and treatment preferences among many oncologists and will certainly influence the discussions we all will have with our patients in the coming weeks and months.
Rituximab maintenance in follicular lymphoma
The largest trial addressing the role of maintenance CD20-targeted antibody therapy was the Primary Rituximab and Maintenance (PRIMA) phase 3, intergroup study in 1,018 advanced follicular lymphoma patients with an initial response to induction chemoimmunotherapy. The induction regimen could be either of three commonly used regimens plus rituximab. Importantly, none of the induction regimens included bendamustine.
Patients were randomized to 2 years of rituximab maintenance or observation. Prior interim analyses at 3 and 6 years showed improvements for the rituximab maintenance patients in progression-free survival and other secondary endpoints, but no improvement in overall survival. Emmanuel Bachy, MD, PhD, and colleagues published the final survival data after 9 years of follow-up, including a final safety analysis (J Clin Oncol. 2019 Nov 1;37[31]:2815-24).
Among the 607 patients consenting to extended follow-up, median progression-free survival was 10.5 years with rituximab maintenance, compared with 4.1 years for observation (hazard ratio, 0.61; P less than .001). There were more patients with progression-free survival at 3 years, complete response or unconfirmed complete response at 2 years, longer time to next antilymphoma treatment, and later time to next chemotherapy. But the 10-year overall survival was similar in the two groups, as were the quality of life ratings.
In all, 503 patients experienced disease progression. Overall survival after progression was shorter in the rituximab maintenance arm versus the observation arm. Among the approximately 4% of patients experiencing transformation from follicular lymphoma to a more aggressive histology, there was no difference observed in time to transformation. Results were independent of the induction regimen received, response to induction, Follicular Lymphoma International Prognostic Index score, and other clinical factors.
In the safety analysis, there were more grade 3-4 adverse events among the rituximab maintenance patients – primarily cytopenia and infections – more serious adverse events, and more adverse events overall. Fatal adverse events were low in both groups (1.6 vs. 0.6%).
How these results influence practice
Many years ago, a senior mentor of mine taught that “progression-free survival is an important endpoint since the quality of life for patients is generally superior before relapse than afterwards.”
With that mantra playing in my mind and with the reality that discussions about relapse and subsequent treatment are never easy, the results of PRIMA seem straightforward: Rituximab maintenance is beneficial, despite the absence of an overall survival benefit, in a disease with multiple options for subsequent therapy and a long natural history. In this 9-year, final analysis of PRIMA, rituximab maintenance seems to have achieved its goals of deepening responses with low (but not inconsequential) toxicity and delaying substantially those difficult conversations with patients about how to proceed after relapse.
The influence of the final analysis of PRIMA, however, may be more complicated that it would initially seem. Bendamustine-containing regimens were not employed, are commonly used now, and some studies with bendamustine have suggested higher nonrelapse mortality with rituximab maintenance.
Low-grade adverse events take on greater importance for patients on long-term therapy than they do for patients receiving abbreviated treatment, and with noncurative treatment, the higher adverse event profile with 2 years of rituximab maintenance needs to be taken seriously. Induction and maintenance regimens involving rituximab alone, lenalidomide, or obinutuzumab may be preferred for some patients. For all of those reasons, the discussion about rituximab with advanced follicular lymphoma patients remains a long one, demanding detailed descriptions of risks, benefits, limitations of the data, and multiple modern day alternatives to the treatments employed in PRIMA.
Supplemental MRI for dense breast tissue
In the DENSE trial, investigators assigned 40,373 women with extremely dense breast tissue and negative results on screening mammography to be offered either supplemental MRI or mammography screening every other year only. The women were 50-75 years old and were enrolled in the Dutch population-based digital mammography screening program. The primary outcome was the difference in the number of cancers that developed in the 2-year interval between mammograms (N Engl J Med 2019;381:2091-102).
The interval cancer rate was 2.5 per 1,000 screenings among 4,783 women in the “MRI-invitation” group (41% of whom did not actually agree to have an MRI), and 5 per 1,000 in the 32,312 women in the “mammography-only” group, a difference of 2.5 per 1,000 screenings (P less than .001).
Of the 20 interval cancers diagnosed in the MRI-invitation group, 4 were diagnosed in the 59% of women who had undergone MRI (0.8 per 1,000 screenings). The remaining 16 were diagnosed in those who had declined an MRI (4.9 per 1,000 screenings) – virtually identical to the rate of interval cancers in the group that was not invited to have an MRI. This speaks against nonrandom allocation of patients between the mammography-only and the supplemental MRI groups.
Although supplemental MRI was associated with a cancer-detection rate of 16.5 per 1,000 screenings, there was a false positive rate of 8.0% (79.8 per 1,000 screenings). Of the women who underwent a breast biopsy after MRI, 73.7% did not have cancer. Among the women who had MRI-detected cancers, in general, the malignancies diagnosed were smaller, more likely node negative, better differentiated, and more often hormone receptor–positive, compared with those in the mammography-only group.
At the next screening mammogram, the cancer detection rate was lower in the MRI-screened group (2 per 1,000 screenings) than in the MRI–“offered but declined” (7.1 per 1,000 screenings) or mammogram-only groups (6 per 1,000 screenings).
How these results influence practice
American physicians are obligated to inform women with dense breast tissue about the limitations (for them) of conventional mammography, an unquantified, but elevated, risk of breast cancer in women with dense breast tissue, and the fact that there are no universally accepted recommended subsequent steps those women should take.
One side benefit of the requirement to disclose information about breast density, however, is that disclosure can prompt discussions about breast cancer risk reduction – an important discussion with multiple possible health-maintaining interventions.
The DENSE trial is very important news, with potential for even more value in future years. It is finite (2 years of study, with only one MRI scan mandated), narrow (Vopara or BI-RADs breast density grade 4 only), and oligo-institutional (eight participating centers in the Netherlands). Still, it provides randomized, rigorously gathered and analyzed data where there were previously none. Importantly, it answers the question, “So what can I do now, doctor?”
I have some concerns about whether we, as a medical community, have the discipline to restrict application of supplemental MRI screening beyond the population that was studied in the Netherlands. Will we be able to restrain ourselves from ordering MRI scans in women with heterogeneously dense – but not extremely dense – breasts? Will we truly manage patients whose MRI scans had BI-RADs readings of less than 4 in the rigorous, but conservative, fashion employed in the trial? Would we miss fewer significant cancers and subject fewer patients to potential expense and harm with annual mammography, instead of the biennial screening performed in the Netherlands? Does tomo-synthesis improve the interpretation of screening mammograms so much that the interval cancer rate is a lot closer to the rate obtained with supplemental MRI scans? Is the expense of MRI scanning, with the resultant subsequent tests and procedures, justified by the reduction in detecting relatively favorable breast cancers that may not materially impact overall survival?
The DENSE study promises to be a “gift that keeps on giving” as the investigators continue to assess a number of factors including: the value of ongoing supplemental MRI scans (compared with the one-time-only screening reported here); whether there will be a reduction in the advanced cancers and subsequent mortality benefit; the extent of over diagnosis, costs, and impact on quality-of-life; and the applicability of artificial intelligence techniques to reduce false positive MRI results.
While the study may not be practice changing in the United States at the present time, it may be just that as subsequent analyses emerge.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight two recent studies that reinforce prior diagnostic and treatment preferences among many oncologists and will certainly influence the discussions we all will have with our patients in the coming weeks and months.
Rituximab maintenance in follicular lymphoma
The largest trial addressing the role of maintenance CD20-targeted antibody therapy was the Primary Rituximab and Maintenance (PRIMA) phase 3, intergroup study in 1,018 advanced follicular lymphoma patients with an initial response to induction chemoimmunotherapy. The induction regimen could be either of three commonly used regimens plus rituximab. Importantly, none of the induction regimens included bendamustine.
Patients were randomized to 2 years of rituximab maintenance or observation. Prior interim analyses at 3 and 6 years showed improvements for the rituximab maintenance patients in progression-free survival and other secondary endpoints, but no improvement in overall survival. Emmanuel Bachy, MD, PhD, and colleagues published the final survival data after 9 years of follow-up, including a final safety analysis (J Clin Oncol. 2019 Nov 1;37[31]:2815-24).
Among the 607 patients consenting to extended follow-up, median progression-free survival was 10.5 years with rituximab maintenance, compared with 4.1 years for observation (hazard ratio, 0.61; P less than .001). There were more patients with progression-free survival at 3 years, complete response or unconfirmed complete response at 2 years, longer time to next antilymphoma treatment, and later time to next chemotherapy. But the 10-year overall survival was similar in the two groups, as were the quality of life ratings.
In all, 503 patients experienced disease progression. Overall survival after progression was shorter in the rituximab maintenance arm versus the observation arm. Among the approximately 4% of patients experiencing transformation from follicular lymphoma to a more aggressive histology, there was no difference observed in time to transformation. Results were independent of the induction regimen received, response to induction, Follicular Lymphoma International Prognostic Index score, and other clinical factors.
In the safety analysis, there were more grade 3-4 adverse events among the rituximab maintenance patients – primarily cytopenia and infections – more serious adverse events, and more adverse events overall. Fatal adverse events were low in both groups (1.6 vs. 0.6%).
How these results influence practice
Many years ago, a senior mentor of mine taught that “progression-free survival is an important endpoint since the quality of life for patients is generally superior before relapse than afterwards.”
With that mantra playing in my mind and with the reality that discussions about relapse and subsequent treatment are never easy, the results of PRIMA seem straightforward: Rituximab maintenance is beneficial, despite the absence of an overall survival benefit, in a disease with multiple options for subsequent therapy and a long natural history. In this 9-year, final analysis of PRIMA, rituximab maintenance seems to have achieved its goals of deepening responses with low (but not inconsequential) toxicity and delaying substantially those difficult conversations with patients about how to proceed after relapse.
The influence of the final analysis of PRIMA, however, may be more complicated that it would initially seem. Bendamustine-containing regimens were not employed, are commonly used now, and some studies with bendamustine have suggested higher nonrelapse mortality with rituximab maintenance.
Low-grade adverse events take on greater importance for patients on long-term therapy than they do for patients receiving abbreviated treatment, and with noncurative treatment, the higher adverse event profile with 2 years of rituximab maintenance needs to be taken seriously. Induction and maintenance regimens involving rituximab alone, lenalidomide, or obinutuzumab may be preferred for some patients. For all of those reasons, the discussion about rituximab with advanced follicular lymphoma patients remains a long one, demanding detailed descriptions of risks, benefits, limitations of the data, and multiple modern day alternatives to the treatments employed in PRIMA.
Supplemental MRI for dense breast tissue
In the DENSE trial, investigators assigned 40,373 women with extremely dense breast tissue and negative results on screening mammography to be offered either supplemental MRI or mammography screening every other year only. The women were 50-75 years old and were enrolled in the Dutch population-based digital mammography screening program. The primary outcome was the difference in the number of cancers that developed in the 2-year interval between mammograms (N Engl J Med 2019;381:2091-102).
The interval cancer rate was 2.5 per 1,000 screenings among 4,783 women in the “MRI-invitation” group (41% of whom did not actually agree to have an MRI), and 5 per 1,000 in the 32,312 women in the “mammography-only” group, a difference of 2.5 per 1,000 screenings (P less than .001).
Of the 20 interval cancers diagnosed in the MRI-invitation group, 4 were diagnosed in the 59% of women who had undergone MRI (0.8 per 1,000 screenings). The remaining 16 were diagnosed in those who had declined an MRI (4.9 per 1,000 screenings) – virtually identical to the rate of interval cancers in the group that was not invited to have an MRI. This speaks against nonrandom allocation of patients between the mammography-only and the supplemental MRI groups.
Although supplemental MRI was associated with a cancer-detection rate of 16.5 per 1,000 screenings, there was a false positive rate of 8.0% (79.8 per 1,000 screenings). Of the women who underwent a breast biopsy after MRI, 73.7% did not have cancer. Among the women who had MRI-detected cancers, in general, the malignancies diagnosed were smaller, more likely node negative, better differentiated, and more often hormone receptor–positive, compared with those in the mammography-only group.
At the next screening mammogram, the cancer detection rate was lower in the MRI-screened group (2 per 1,000 screenings) than in the MRI–“offered but declined” (7.1 per 1,000 screenings) or mammogram-only groups (6 per 1,000 screenings).
How these results influence practice
American physicians are obligated to inform women with dense breast tissue about the limitations (for them) of conventional mammography, an unquantified, but elevated, risk of breast cancer in women with dense breast tissue, and the fact that there are no universally accepted recommended subsequent steps those women should take.
One side benefit of the requirement to disclose information about breast density, however, is that disclosure can prompt discussions about breast cancer risk reduction – an important discussion with multiple possible health-maintaining interventions.
The DENSE trial is very important news, with potential for even more value in future years. It is finite (2 years of study, with only one MRI scan mandated), narrow (Vopara or BI-RADs breast density grade 4 only), and oligo-institutional (eight participating centers in the Netherlands). Still, it provides randomized, rigorously gathered and analyzed data where there were previously none. Importantly, it answers the question, “So what can I do now, doctor?”
I have some concerns about whether we, as a medical community, have the discipline to restrict application of supplemental MRI screening beyond the population that was studied in the Netherlands. Will we be able to restrain ourselves from ordering MRI scans in women with heterogeneously dense – but not extremely dense – breasts? Will we truly manage patients whose MRI scans had BI-RADs readings of less than 4 in the rigorous, but conservative, fashion employed in the trial? Would we miss fewer significant cancers and subject fewer patients to potential expense and harm with annual mammography, instead of the biennial screening performed in the Netherlands? Does tomo-synthesis improve the interpretation of screening mammograms so much that the interval cancer rate is a lot closer to the rate obtained with supplemental MRI scans? Is the expense of MRI scanning, with the resultant subsequent tests and procedures, justified by the reduction in detecting relatively favorable breast cancers that may not materially impact overall survival?
The DENSE study promises to be a “gift that keeps on giving” as the investigators continue to assess a number of factors including: the value of ongoing supplemental MRI scans (compared with the one-time-only screening reported here); whether there will be a reduction in the advanced cancers and subsequent mortality benefit; the extent of over diagnosis, costs, and impact on quality-of-life; and the applicability of artificial intelligence techniques to reduce false positive MRI results.
While the study may not be practice changing in the United States at the present time, it may be just that as subsequent analyses emerge.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
The clinical impact of new approvals in sickle cell, MCL
In this edition of “How I Will Treat My Next Patient,” I highlight two recent drug approvals by the Food and Drug Administration – crizanlizumab for sickle cell patients with painful crises and zanubrutinib for mantle cell lymphoma (MCL) patients in relapse.
Crizanlizumab
P-selectin is an adhesion molecule expressed on activated vascular endothelial cells and platelets. It is a key molecule in the initiation of leukocyte rolling on vessel walls and promotes firm attachment and extravasation to underlying tissues during inflammation. Up-regulation of P-selectin on endothelial cells and platelets contributes to the cell-cell interactions involved in the pathogenesis of sickle cell pain crises.
The SUSTAIN study was a multisite, placebo-controlled, randomized phase 2 trial of two different dosage levels of intravenous crizanlizumab (2.5 mg/kg or 5 mg/kg for 52 weeks), a humanized anti–P-selectin antibody, examining its effect on pain crises in patients with sickle cell disease. The primary endpoint was the annual rate of sickle cell pain crises, with a variety of clinically relevant secondary endpoints. The target population had 2-10 pain crises in the 12 months before enrollment. Patients on a stable dose of hydroxyurea for at least the most recent 3 months were allowed to enter, but if patients were not receiving hydroxyurea, it could not be initiated during the trial. Patients who were undergoing chronic red-cell transfusion therapy were excluded.
Among 198 enrolled patients, 35% did not complete the 52 weeks of treatment. Discontinuations were equally balanced among patients assigned to the high-dose, low-dose, and placebo cohorts. Adverse events associated with crizanlizumab included back pain, nausea, pyrexia, and arthralgia. Serious adverse events occurred in 55 patients, with 5 deaths, all of which were unrelated to treatment. Crizanlizumab did not augment hemolysis or bacterial infections.
In the efficacy analysis, patients receiving high-dose crizanlizumab had a median annual rate of 1.63 health care visits for sickle cell pain crises, compared with 2.98 visits for placebo patients (P = .01). In comparison with placebo, high-dose crizanlizumab also delayed the first pain crisis after starting treatment (4.1 months vs. 1.4 months), delayed the median time to a second pain crisis, and decreased the median number of pain crises annually.
More than twice as many high-dose crizanlizumab patients had no pain crisis episodes, compared with placebo patients. In general, differences were more striking in patients who were not taking hydroxyurea and who had non–hemoglobin SS disease. Differences in the primary endpoint between low-dose crizanlizumab and placebo were numerically, but not statistically, different.
How these results influence practice
It has been over 20 years since a new agent (hydroxyurea) was approved for sickle cell patients and, despite its use, sickle cell pain crises remain a frequent problem. Pain crises are associated with worse quality of life and increased risk of death. A promising advance is badly needed, especially in an era in which sensitivity to providers’ role in the opioid addiction crisis is highly scrutinized and may contribute to future undertreatment of pain episodes. This is especially true for patients from areas with high levels of opioid misuse.
The SUSTAIN trial was international, multi-institutional, placebo-controlled, and inclusive. These attributes enhance the likelihood that crizanlizumab will enhance patient care in routine practice. As an intravenous agent, monitoring adherence and toxicity are less challenging than with hydroxyurea. Despite these factors, however, there are some concerns. Crizanlizumab was not free of toxicity, quality of life via the Brief Pain Inventory used in the trial was not improved, and changes in the pain-severity and pain-interference domains were small. Treatment in SUSTAIN ensued for 52 weeks, so the emergence of late neutralizing antibodies and late toxicities with longer-term therapy will require careful postmarketing assessment.
These concerns notwithstanding, anyone who has cared for sickle cell patients would be excited about the potential benefits crizanlizumab could bring to patient care.
Zanubrutinib
The FDA has approved zanubrutinib for the treatment of MCL in adult patients who have received at least one prior therapy. The approval is based on the results of two studies in which overall response rate was the primary endpoint.
BGB-3111-206 (NCT03206970) was a phase 2, open-label, multicenter, single-arm trial of 86 patients with MCL who received at least one prior therapy. Zanubrutinib was given orally at 160 mg twice daily until disease progression or unacceptable toxicity. BGB-3111-AU-003 (NCT 02343120) was a phase 1/2, open-label, dose-escalation trial of B-cell malignancies, including 32 previously treated MCL patients treated with zanubrutinib at 160 mg twice daily or 320 mg once daily.
In the phase 2 trial, 18fluorodeoxyglucose (FDG)–PET scans were required and the ORR was 84% (95% confidence interval, 74%-91%), with a complete response rate of 59% (95% CI, 48%-70%) and a median response duration of 19.5 months (95% CI, 16.6% to not estimable). In the phase 1/2 dose-escalation trial, FDG-PET scans were not required and the ORR was 84% (95% CI, 67%-95%), with a complete response rate of 22% (95% CI, 9%-40%) and a median response duration of 18.5 months (95% CI, 12.6% to not estimable). In both trials, median follow-up on study was about 18 months.
The most common adverse reactions were cytopenias, upper respiratory tract infection, rash, bruising, diarrhea, and cough. The most common serious adverse reactions were pneumonia in 11% and hemorrhage in 5% of patients. Of 118 MCL patients, 8 stopped therapy because of an adverse event, most frequently pneumonia (3.4%).
How these results influence practice
Unfortunately, the therapy of recurrent MCL is noncurative, because of the rapid development of treatment resistance. There are multiple single-and multiagent chemotherapy regimens that may be tried, many incorporating immunotherapy options such as anti-CD20- or Bruton tyrosine kinase (BTK)–targeted agents. Given the limited efficacy of these agents, temporary nature of remissions, and paucity of data comparing these various treatment options, participation in clinical trials is encouraged whenever possible.
Outside of a clinical trial, zanubrutinib joins ibrutinib and acalabrutinib as approved single-agent BTK inhibitors for adult MCL patients in relapse. The impressive ORR and response duration reported for zanubrutinib are similar to the results achieved with the other agents, but the toxicity pattern may be slightly different.
As in the treatment of hormonally sensitive breast cancer, clinicians and patients benefit when they have multiple similar, equally efficacious oral agents with slightly different toxicity patterns so that quality of life can be improved and treatment duration maximized before treatment resistance develops and a more toxic and/or inconvenient therapy needs to be employed.
Whether zanubrutinib has benefits beyond those for MCL patients in relapse will depend on the results of confirmatory trials and patient-reported outcome data.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight two recent drug approvals by the Food and Drug Administration – crizanlizumab for sickle cell patients with painful crises and zanubrutinib for mantle cell lymphoma (MCL) patients in relapse.
Crizanlizumab
P-selectin is an adhesion molecule expressed on activated vascular endothelial cells and platelets. It is a key molecule in the initiation of leukocyte rolling on vessel walls and promotes firm attachment and extravasation to underlying tissues during inflammation. Up-regulation of P-selectin on endothelial cells and platelets contributes to the cell-cell interactions involved in the pathogenesis of sickle cell pain crises.
The SUSTAIN study was a multisite, placebo-controlled, randomized phase 2 trial of two different dosage levels of intravenous crizanlizumab (2.5 mg/kg or 5 mg/kg for 52 weeks), a humanized anti–P-selectin antibody, examining its effect on pain crises in patients with sickle cell disease. The primary endpoint was the annual rate of sickle cell pain crises, with a variety of clinically relevant secondary endpoints. The target population had 2-10 pain crises in the 12 months before enrollment. Patients on a stable dose of hydroxyurea for at least the most recent 3 months were allowed to enter, but if patients were not receiving hydroxyurea, it could not be initiated during the trial. Patients who were undergoing chronic red-cell transfusion therapy were excluded.
Among 198 enrolled patients, 35% did not complete the 52 weeks of treatment. Discontinuations were equally balanced among patients assigned to the high-dose, low-dose, and placebo cohorts. Adverse events associated with crizanlizumab included back pain, nausea, pyrexia, and arthralgia. Serious adverse events occurred in 55 patients, with 5 deaths, all of which were unrelated to treatment. Crizanlizumab did not augment hemolysis or bacterial infections.
In the efficacy analysis, patients receiving high-dose crizanlizumab had a median annual rate of 1.63 health care visits for sickle cell pain crises, compared with 2.98 visits for placebo patients (P = .01). In comparison with placebo, high-dose crizanlizumab also delayed the first pain crisis after starting treatment (4.1 months vs. 1.4 months), delayed the median time to a second pain crisis, and decreased the median number of pain crises annually.
More than twice as many high-dose crizanlizumab patients had no pain crisis episodes, compared with placebo patients. In general, differences were more striking in patients who were not taking hydroxyurea and who had non–hemoglobin SS disease. Differences in the primary endpoint between low-dose crizanlizumab and placebo were numerically, but not statistically, different.
How these results influence practice
It has been over 20 years since a new agent (hydroxyurea) was approved for sickle cell patients and, despite its use, sickle cell pain crises remain a frequent problem. Pain crises are associated with worse quality of life and increased risk of death. A promising advance is badly needed, especially in an era in which sensitivity to providers’ role in the opioid addiction crisis is highly scrutinized and may contribute to future undertreatment of pain episodes. This is especially true for patients from areas with high levels of opioid misuse.
The SUSTAIN trial was international, multi-institutional, placebo-controlled, and inclusive. These attributes enhance the likelihood that crizanlizumab will enhance patient care in routine practice. As an intravenous agent, monitoring adherence and toxicity are less challenging than with hydroxyurea. Despite these factors, however, there are some concerns. Crizanlizumab was not free of toxicity, quality of life via the Brief Pain Inventory used in the trial was not improved, and changes in the pain-severity and pain-interference domains were small. Treatment in SUSTAIN ensued for 52 weeks, so the emergence of late neutralizing antibodies and late toxicities with longer-term therapy will require careful postmarketing assessment.
These concerns notwithstanding, anyone who has cared for sickle cell patients would be excited about the potential benefits crizanlizumab could bring to patient care.
Zanubrutinib
The FDA has approved zanubrutinib for the treatment of MCL in adult patients who have received at least one prior therapy. The approval is based on the results of two studies in which overall response rate was the primary endpoint.
BGB-3111-206 (NCT03206970) was a phase 2, open-label, multicenter, single-arm trial of 86 patients with MCL who received at least one prior therapy. Zanubrutinib was given orally at 160 mg twice daily until disease progression or unacceptable toxicity. BGB-3111-AU-003 (NCT 02343120) was a phase 1/2, open-label, dose-escalation trial of B-cell malignancies, including 32 previously treated MCL patients treated with zanubrutinib at 160 mg twice daily or 320 mg once daily.
In the phase 2 trial, 18fluorodeoxyglucose (FDG)–PET scans were required and the ORR was 84% (95% confidence interval, 74%-91%), with a complete response rate of 59% (95% CI, 48%-70%) and a median response duration of 19.5 months (95% CI, 16.6% to not estimable). In the phase 1/2 dose-escalation trial, FDG-PET scans were not required and the ORR was 84% (95% CI, 67%-95%), with a complete response rate of 22% (95% CI, 9%-40%) and a median response duration of 18.5 months (95% CI, 12.6% to not estimable). In both trials, median follow-up on study was about 18 months.
The most common adverse reactions were cytopenias, upper respiratory tract infection, rash, bruising, diarrhea, and cough. The most common serious adverse reactions were pneumonia in 11% and hemorrhage in 5% of patients. Of 118 MCL patients, 8 stopped therapy because of an adverse event, most frequently pneumonia (3.4%).
How these results influence practice
Unfortunately, the therapy of recurrent MCL is noncurative, because of the rapid development of treatment resistance. There are multiple single-and multiagent chemotherapy regimens that may be tried, many incorporating immunotherapy options such as anti-CD20- or Bruton tyrosine kinase (BTK)–targeted agents. Given the limited efficacy of these agents, temporary nature of remissions, and paucity of data comparing these various treatment options, participation in clinical trials is encouraged whenever possible.
Outside of a clinical trial, zanubrutinib joins ibrutinib and acalabrutinib as approved single-agent BTK inhibitors for adult MCL patients in relapse. The impressive ORR and response duration reported for zanubrutinib are similar to the results achieved with the other agents, but the toxicity pattern may be slightly different.
As in the treatment of hormonally sensitive breast cancer, clinicians and patients benefit when they have multiple similar, equally efficacious oral agents with slightly different toxicity patterns so that quality of life can be improved and treatment duration maximized before treatment resistance develops and a more toxic and/or inconvenient therapy needs to be employed.
Whether zanubrutinib has benefits beyond those for MCL patients in relapse will depend on the results of confirmatory trials and patient-reported outcome data.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.
In this edition of “How I Will Treat My Next Patient,” I highlight two recent drug approvals by the Food and Drug Administration – crizanlizumab for sickle cell patients with painful crises and zanubrutinib for mantle cell lymphoma (MCL) patients in relapse.
Crizanlizumab
P-selectin is an adhesion molecule expressed on activated vascular endothelial cells and platelets. It is a key molecule in the initiation of leukocyte rolling on vessel walls and promotes firm attachment and extravasation to underlying tissues during inflammation. Up-regulation of P-selectin on endothelial cells and platelets contributes to the cell-cell interactions involved in the pathogenesis of sickle cell pain crises.
The SUSTAIN study was a multisite, placebo-controlled, randomized phase 2 trial of two different dosage levels of intravenous crizanlizumab (2.5 mg/kg or 5 mg/kg for 52 weeks), a humanized anti–P-selectin antibody, examining its effect on pain crises in patients with sickle cell disease. The primary endpoint was the annual rate of sickle cell pain crises, with a variety of clinically relevant secondary endpoints. The target population had 2-10 pain crises in the 12 months before enrollment. Patients on a stable dose of hydroxyurea for at least the most recent 3 months were allowed to enter, but if patients were not receiving hydroxyurea, it could not be initiated during the trial. Patients who were undergoing chronic red-cell transfusion therapy were excluded.
Among 198 enrolled patients, 35% did not complete the 52 weeks of treatment. Discontinuations were equally balanced among patients assigned to the high-dose, low-dose, and placebo cohorts. Adverse events associated with crizanlizumab included back pain, nausea, pyrexia, and arthralgia. Serious adverse events occurred in 55 patients, with 5 deaths, all of which were unrelated to treatment. Crizanlizumab did not augment hemolysis or bacterial infections.
In the efficacy analysis, patients receiving high-dose crizanlizumab had a median annual rate of 1.63 health care visits for sickle cell pain crises, compared with 2.98 visits for placebo patients (P = .01). In comparison with placebo, high-dose crizanlizumab also delayed the first pain crisis after starting treatment (4.1 months vs. 1.4 months), delayed the median time to a second pain crisis, and decreased the median number of pain crises annually.
More than twice as many high-dose crizanlizumab patients had no pain crisis episodes, compared with placebo patients. In general, differences were more striking in patients who were not taking hydroxyurea and who had non–hemoglobin SS disease. Differences in the primary endpoint between low-dose crizanlizumab and placebo were numerically, but not statistically, different.
How these results influence practice
It has been over 20 years since a new agent (hydroxyurea) was approved for sickle cell patients and, despite its use, sickle cell pain crises remain a frequent problem. Pain crises are associated with worse quality of life and increased risk of death. A promising advance is badly needed, especially in an era in which sensitivity to providers’ role in the opioid addiction crisis is highly scrutinized and may contribute to future undertreatment of pain episodes. This is especially true for patients from areas with high levels of opioid misuse.
The SUSTAIN trial was international, multi-institutional, placebo-controlled, and inclusive. These attributes enhance the likelihood that crizanlizumab will enhance patient care in routine practice. As an intravenous agent, monitoring adherence and toxicity are less challenging than with hydroxyurea. Despite these factors, however, there are some concerns. Crizanlizumab was not free of toxicity, quality of life via the Brief Pain Inventory used in the trial was not improved, and changes in the pain-severity and pain-interference domains were small. Treatment in SUSTAIN ensued for 52 weeks, so the emergence of late neutralizing antibodies and late toxicities with longer-term therapy will require careful postmarketing assessment.
These concerns notwithstanding, anyone who has cared for sickle cell patients would be excited about the potential benefits crizanlizumab could bring to patient care.
Zanubrutinib
The FDA has approved zanubrutinib for the treatment of MCL in adult patients who have received at least one prior therapy. The approval is based on the results of two studies in which overall response rate was the primary endpoint.
BGB-3111-206 (NCT03206970) was a phase 2, open-label, multicenter, single-arm trial of 86 patients with MCL who received at least one prior therapy. Zanubrutinib was given orally at 160 mg twice daily until disease progression or unacceptable toxicity. BGB-3111-AU-003 (NCT 02343120) was a phase 1/2, open-label, dose-escalation trial of B-cell malignancies, including 32 previously treated MCL patients treated with zanubrutinib at 160 mg twice daily or 320 mg once daily.
In the phase 2 trial, 18fluorodeoxyglucose (FDG)–PET scans were required and the ORR was 84% (95% confidence interval, 74%-91%), with a complete response rate of 59% (95% CI, 48%-70%) and a median response duration of 19.5 months (95% CI, 16.6% to not estimable). In the phase 1/2 dose-escalation trial, FDG-PET scans were not required and the ORR was 84% (95% CI, 67%-95%), with a complete response rate of 22% (95% CI, 9%-40%) and a median response duration of 18.5 months (95% CI, 12.6% to not estimable). In both trials, median follow-up on study was about 18 months.
The most common adverse reactions were cytopenias, upper respiratory tract infection, rash, bruising, diarrhea, and cough. The most common serious adverse reactions were pneumonia in 11% and hemorrhage in 5% of patients. Of 118 MCL patients, 8 stopped therapy because of an adverse event, most frequently pneumonia (3.4%).
How these results influence practice
Unfortunately, the therapy of recurrent MCL is noncurative, because of the rapid development of treatment resistance. There are multiple single-and multiagent chemotherapy regimens that may be tried, many incorporating immunotherapy options such as anti-CD20- or Bruton tyrosine kinase (BTK)–targeted agents. Given the limited efficacy of these agents, temporary nature of remissions, and paucity of data comparing these various treatment options, participation in clinical trials is encouraged whenever possible.
Outside of a clinical trial, zanubrutinib joins ibrutinib and acalabrutinib as approved single-agent BTK inhibitors for adult MCL patients in relapse. The impressive ORR and response duration reported for zanubrutinib are similar to the results achieved with the other agents, but the toxicity pattern may be slightly different.
As in the treatment of hormonally sensitive breast cancer, clinicians and patients benefit when they have multiple similar, equally efficacious oral agents with slightly different toxicity patterns so that quality of life can be improved and treatment duration maximized before treatment resistance develops and a more toxic and/or inconvenient therapy needs to be employed.
Whether zanubrutinib has benefits beyond those for MCL patients in relapse will depend on the results of confirmatory trials and patient-reported outcome data.
Dr. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis. His clinical and research interests are in the prevention, diagnosis, and treatment of breast and lung cancers and in expanding access to clinical trials to medically underserved populations.