Immuno-oncology

Background

The rise in the number of T-cell engager therapies highlights their importance in modern cancer treatment paradigms. Having recognized the need for, and complexities of, administering these innovative medications to our patients, our team assessed our institution’s capability to provide these therapies to our patients. We identified that our facility was wellequipped for implementation of T-cell engager therapy due to inpatient administration capabilities, an outpatient infusion center, on-hand supportive care medications (tocilizumab), and access to higher levels of care. Key players included medical oncologists, pharmacists, inpatient and infusion nurses, staff physicians, critical care practitioners, and care coordinators.

Clinical Practice Initiative

Barriers identified: education, toxicity concerns, formulary management, and logistics. To overcome these obstacles, comprehensive plans for procurement, hospital admission, monitoring, and training were developed as a facility-specific standard operating procedure (SOP). All available Tcell engager therapies were presented to the formulary committee and received local approval. Physician and pharmacist champions were registered for the associated risk evaluation and mitigation strategies (REMS) programs. Recorded webinars were done to provide education on REMS requirements, medication logistics, and adverse event management.

An admission plan was formulated to outline admission criteria, medication administration, and safety logistics. Order sets created by pharmacists, encompassed pre, post, and as needed medications for cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome. To facilitate safe discharge and meet REMS criteria, patients received wallet cards, dexamethasone and acetaminophen PRNs with detailed instructions for use, and direction for seeking emergency care with consideration of local tocilizumab availability.

Conclusions

Our SOP has enabled administration of six T-cell engager therapies for six diseases. The primary limitation for some of these agents is the need for inpatient monitoring at initiation, which may not be available at smaller centers. Facilities that lack these capabilities could utilize community care or partner with a neighboring Veterans Affairs medical center for initial administration, then transition back for continued treatment. Facilities that lack inpatient oncology nursing could administer the drug in the infusion center followed by admission for monitoring and toxicity management. Our implementation plan serves as a scalable model for improving veteran access to novel therapies.

Background

The rise in the number of T-cell engager therapies highlights their importance in modern cancer treatment paradigms. Having recognized the need for, and complexities of, administering these innovative medications to our patients, our team assessed our institution’s capability to provide these therapies to our patients. We identified that our facility was wellequipped for implementation of T-cell engager therapy due to inpatient administration capabilities, an outpatient infusion center, on-hand supportive care medications (tocilizumab), and access to higher levels of care. Key players included medical oncologists, pharmacists, inpatient and infusion nurses, staff physicians, critical care practitioners, and care coordinators.

Clinical Practice Initiative

Barriers identified: education, toxicity concerns, formulary management, and logistics. To overcome these obstacles, comprehensive plans for procurement, hospital admission, monitoring, and training were developed as a facility-specific standard operating procedure (SOP). All available Tcell engager therapies were presented to the formulary committee and received local approval. Physician and pharmacist champions were registered for the associated risk evaluation and mitigation strategies (REMS) programs. Recorded webinars were done to provide education on REMS requirements, medication logistics, and adverse event management.

An admission plan was formulated to outline admission criteria, medication administration, and safety logistics. Order sets created by pharmacists, encompassed pre, post, and as needed medications for cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome. To facilitate safe discharge and meet REMS criteria, patients received wallet cards, dexamethasone and acetaminophen PRNs with detailed instructions for use, and direction for seeking emergency care with consideration of local tocilizumab availability.

Conclusions

Our SOP has enabled administration of six T-cell engager therapies for six diseases. The primary limitation for some of these agents is the need for inpatient monitoring at initiation, which may not be available at smaller centers. Facilities that lack these capabilities could utilize community care or partner with a neighboring Veterans Affairs medical center for initial administration, then transition back for continued treatment. Facilities that lack inpatient oncology nursing could administer the drug in the infusion center followed by admission for monitoring and toxicity management. Our implementation plan serves as a scalable model for improving veteran access to novel therapies.

Background

The rise in the number of T-cell engager therapies highlights their importance in modern cancer treatment paradigms. Having recognized the need for, and complexities of, administering these innovative medications to our patients, our team assessed our institution’s capability to provide these therapies to our patients. We identified that our facility was wellequipped for implementation of T-cell engager therapy due to inpatient administration capabilities, an outpatient infusion center, on-hand supportive care medications (tocilizumab), and access to higher levels of care. Key players included medical oncologists, pharmacists, inpatient and infusion nurses, staff physicians, critical care practitioners, and care coordinators.

Clinical Practice Initiative

Barriers identified: education, toxicity concerns, formulary management, and logistics. To overcome these obstacles, comprehensive plans for procurement, hospital admission, monitoring, and training were developed as a facility-specific standard operating procedure (SOP). All available Tcell engager therapies were presented to the formulary committee and received local approval. Physician and pharmacist champions were registered for the associated risk evaluation and mitigation strategies (REMS) programs. Recorded webinars were done to provide education on REMS requirements, medication logistics, and adverse event management.

An admission plan was formulated to outline admission criteria, medication administration, and safety logistics. Order sets created by pharmacists, encompassed pre, post, and as needed medications for cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome. To facilitate safe discharge and meet REMS criteria, patients received wallet cards, dexamethasone and acetaminophen PRNs with detailed instructions for use, and direction for seeking emergency care with consideration of local tocilizumab availability.

Conclusions

Our SOP has enabled administration of six T-cell engager therapies for six diseases. The primary limitation for some of these agents is the need for inpatient monitoring at initiation, which may not be available at smaller centers. Facilities that lack these capabilities could utilize community care or partner with a neighboring Veterans Affairs medical center for initial administration, then transition back for continued treatment. Facilities that lack inpatient oncology nursing could administer the drug in the infusion center followed by admission for monitoring and toxicity management. Our implementation plan serves as a scalable model for improving veteran access to novel therapies.

Case Presentation

A 75-year-old man with chronic kidney disease, hypertension and diabetes mellitus presented with acute kidney injury (creatinine 5.2 from baseline 4.2) and a two-week history of increased urinary frequency. Labs revealed high anion gap metabolic acidosis, proteinuria, hematuria, pyuria, and acute on chronic anemia. He was diagnosed with kappa light chain nephropathy and multiple myeloma with 32% plasma cells on bone marrow biopsy. He began treatment with bortezomib, cyclophosphamide, and dexamethasone (Cy- BorD). Three days after cyclophosphamide and five days after bortezomib, the patient developed persistent hypotension with systolic BP in the 50s, unresponsive to fluids and Trendelenburg position. Due to end-stage renal disease with anuria, fluid resuscitation was limited. He required norepinephrine and was transferred to the ICU. Given instability, hemodialysis was deferred, and continuous renal replacement therapy was initiated. Shock evaluation included a CT abdomen showing enteritis versus ileus; however, infectious workup was negative. Cardiogenic shock was ruled out with a serial echocardiogram showing normal ejection fractions of 59-67% without significant valvular disease. The workup for adrenal insufficiency was negative. After the exclusion of other potential causes of shock, severe refractory hypotension was attributed to bortezomib toxicity.Hypotension is a known adverse effect of bortezomib. Orthostatic hypotension may occur in 8 to 9% of patients, and rarely, patients may experience heart failure, conduction disorders and arrhythmias, or cardiogenic shock. The pathologic mechanism of this toxicity is still poorly understood. Proposed mechanisms include direct endothelial toxicity as evidenced by thrombotic microangiopathy or impairment of sympathetic and parasympathetic nerve fibres. Most commonly, patients experience neurotoxicity, which may manifest as autonomic dysfunction or peripheral neuropathy. Cardiovascular complications are typically reversible. Our patient’s cardiac function remained within normal limits; therefore, his persistent hypotension was felt to be the result of direct toxicity from bortezomib rather than cardiogenic shock. Ultimately, blood pressure did improve, and vasopressors were discontinued. However, he continued to have orthostatic hypotension and continued to require supportive fludrocortisone, midodrine, and pyridostigmine. Goals of care have been discussed, and he wished to continue pursuing restorative care, with a plan for transition to carfilzomib versus daratumumab outpatient.

Case Presentation

A 75-year-old man with chronic kidney disease, hypertension and diabetes mellitus presented with acute kidney injury (creatinine 5.2 from baseline 4.2) and a two-week history of increased urinary frequency. Labs revealed high anion gap metabolic acidosis, proteinuria, hematuria, pyuria, and acute on chronic anemia. He was diagnosed with kappa light chain nephropathy and multiple myeloma with 32% plasma cells on bone marrow biopsy. He began treatment with bortezomib, cyclophosphamide, and dexamethasone (Cy- BorD). Three days after cyclophosphamide and five days after bortezomib, the patient developed persistent hypotension with systolic BP in the 50s, unresponsive to fluids and Trendelenburg position. Due to end-stage renal disease with anuria, fluid resuscitation was limited. He required norepinephrine and was transferred to the ICU. Given instability, hemodialysis was deferred, and continuous renal replacement therapy was initiated. Shock evaluation included a CT abdomen showing enteritis versus ileus; however, infectious workup was negative. Cardiogenic shock was ruled out with a serial echocardiogram showing normal ejection fractions of 59-67% without significant valvular disease. The workup for adrenal insufficiency was negative. After the exclusion of other potential causes of shock, severe refractory hypotension was attributed to bortezomib toxicity.Hypotension is a known adverse effect of bortezomib. Orthostatic hypotension may occur in 8 to 9% of patients, and rarely, patients may experience heart failure, conduction disorders and arrhythmias, or cardiogenic shock. The pathologic mechanism of this toxicity is still poorly understood. Proposed mechanisms include direct endothelial toxicity as evidenced by thrombotic microangiopathy or impairment of sympathetic and parasympathetic nerve fibres. Most commonly, patients experience neurotoxicity, which may manifest as autonomic dysfunction or peripheral neuropathy. Cardiovascular complications are typically reversible. Our patient’s cardiac function remained within normal limits; therefore, his persistent hypotension was felt to be the result of direct toxicity from bortezomib rather than cardiogenic shock. Ultimately, blood pressure did improve, and vasopressors were discontinued. However, he continued to have orthostatic hypotension and continued to require supportive fludrocortisone, midodrine, and pyridostigmine. Goals of care have been discussed, and he wished to continue pursuing restorative care, with a plan for transition to carfilzomib versus daratumumab outpatient.

Case Presentation

A 75-year-old man with chronic kidney disease, hypertension and diabetes mellitus presented with acute kidney injury (creatinine 5.2 from baseline 4.2) and a two-week history of increased urinary frequency. Labs revealed high anion gap metabolic acidosis, proteinuria, hematuria, pyuria, and acute on chronic anemia. He was diagnosed with kappa light chain nephropathy and multiple myeloma with 32% plasma cells on bone marrow biopsy. He began treatment with bortezomib, cyclophosphamide, and dexamethasone (Cy- BorD). Three days after cyclophosphamide and five days after bortezomib, the patient developed persistent hypotension with systolic BP in the 50s, unresponsive to fluids and Trendelenburg position. Due to end-stage renal disease with anuria, fluid resuscitation was limited. He required norepinephrine and was transferred to the ICU. Given instability, hemodialysis was deferred, and continuous renal replacement therapy was initiated. Shock evaluation included a CT abdomen showing enteritis versus ileus; however, infectious workup was negative. Cardiogenic shock was ruled out with a serial echocardiogram showing normal ejection fractions of 59-67% without significant valvular disease. The workup for adrenal insufficiency was negative. After the exclusion of other potential causes of shock, severe refractory hypotension was attributed to bortezomib toxicity.Hypotension is a known adverse effect of bortezomib. Orthostatic hypotension may occur in 8 to 9% of patients, and rarely, patients may experience heart failure, conduction disorders and arrhythmias, or cardiogenic shock. The pathologic mechanism of this toxicity is still poorly understood. Proposed mechanisms include direct endothelial toxicity as evidenced by thrombotic microangiopathy or impairment of sympathetic and parasympathetic nerve fibres. Most commonly, patients experience neurotoxicity, which may manifest as autonomic dysfunction or peripheral neuropathy. Cardiovascular complications are typically reversible. Our patient’s cardiac function remained within normal limits; therefore, his persistent hypotension was felt to be the result of direct toxicity from bortezomib rather than cardiogenic shock. Ultimately, blood pressure did improve, and vasopressors were discontinued. However, he continued to have orthostatic hypotension and continued to require supportive fludrocortisone, midodrine, and pyridostigmine. Goals of care have been discussed, and he wished to continue pursuing restorative care, with a plan for transition to carfilzomib versus daratumumab outpatient.

Background

Immune-related adverse events (irAEs) associated with checkpoint inhibitors can involve virtually any organ system. Optic neuritis is a rare but potentially reversible toxicity, with limited reports in the literature.

Case Presentation

A 57-year-old male with Stage IV poorly-differentiated neuroendocrine carcinoma presented with progressive bilateral vision loss following a near-complete response to four cycles of atezolizumab, carboplatin, and etoposide chemotherapy, and one cycle of maintenance atezolizumab. Symptoms began in the right eye and progressed to the left over 12 days. Neurological and ophthalmological evaluations included brain and orbital MRI, autoimmune panels, and infectious workup, all of which were unrevealing. The clinical picture remained consistent with isolated, immunemediated optic neuritis.

Discussion

High-dose intravenous methylprednisolone was initiated, resulting in gradual improvement and partial visual recovery by day four. An oral prednisone taper was prescribed for continued treatment. This is the second reported case of isolated optic neuritis associated with PD-L1 inhibitor therapy and the second with negative imaging findings. The rarity of this irAE and the absence of radiographic abnormalities may delay diagnosis and treatment.

Conclusions

Checkpoint-inhibitor-induced optic neuritis should be considered in patients with visual symptoms on immunotherapy, even in the setting of negative imaging. Early recognition and corticosteroid therapy are critical in preserving visual function.

Background

Immune-related adverse events (irAEs) associated with checkpoint inhibitors can involve virtually any organ system. Optic neuritis is a rare but potentially reversible toxicity, with limited reports in the literature.

Case Presentation

A 57-year-old male with Stage IV poorly-differentiated neuroendocrine carcinoma presented with progressive bilateral vision loss following a near-complete response to four cycles of atezolizumab, carboplatin, and etoposide chemotherapy, and one cycle of maintenance atezolizumab. Symptoms began in the right eye and progressed to the left over 12 days. Neurological and ophthalmological evaluations included brain and orbital MRI, autoimmune panels, and infectious workup, all of which were unrevealing. The clinical picture remained consistent with isolated, immunemediated optic neuritis.

Discussion

High-dose intravenous methylprednisolone was initiated, resulting in gradual improvement and partial visual recovery by day four. An oral prednisone taper was prescribed for continued treatment. This is the second reported case of isolated optic neuritis associated with PD-L1 inhibitor therapy and the second with negative imaging findings. The rarity of this irAE and the absence of radiographic abnormalities may delay diagnosis and treatment.

Conclusions

Checkpoint-inhibitor-induced optic neuritis should be considered in patients with visual symptoms on immunotherapy, even in the setting of negative imaging. Early recognition and corticosteroid therapy are critical in preserving visual function.

Background

Immune-related adverse events (irAEs) associated with checkpoint inhibitors can involve virtually any organ system. Optic neuritis is a rare but potentially reversible toxicity, with limited reports in the literature.

Case Presentation

A 57-year-old male with Stage IV poorly-differentiated neuroendocrine carcinoma presented with progressive bilateral vision loss following a near-complete response to four cycles of atezolizumab, carboplatin, and etoposide chemotherapy, and one cycle of maintenance atezolizumab. Symptoms began in the right eye and progressed to the left over 12 days. Neurological and ophthalmological evaluations included brain and orbital MRI, autoimmune panels, and infectious workup, all of which were unrevealing. The clinical picture remained consistent with isolated, immunemediated optic neuritis.

Discussion

High-dose intravenous methylprednisolone was initiated, resulting in gradual improvement and partial visual recovery by day four. An oral prednisone taper was prescribed for continued treatment. This is the second reported case of isolated optic neuritis associated with PD-L1 inhibitor therapy and the second with negative imaging findings. The rarity of this irAE and the absence of radiographic abnormalities may delay diagnosis and treatment.

Conclusions

Checkpoint-inhibitor-induced optic neuritis should be considered in patients with visual symptoms on immunotherapy, even in the setting of negative imaging. Early recognition and corticosteroid therapy are critical in preserving visual function.

Background

Immune checkpoint inhibitors, including pembrolizumab, are associated with a spectrum of immune-related adverse events (irAEs), including immune- mediated hepatitis. Typically, this toxicity manifests within the first 14 weeks of therapy. Delayed presentations beyond one year are exceedingly rare and pose diagnostic challenges.

Case Presentation

We report an elderly patient (over 90 years old) with stage IVa squamous cell carcinoma of the lung and high microsatellite instability (MSI) who had been receiving pembrolizumab since 2023. In 2024—13 months into therapy—he presented with subjective fevers, weakness, and altered mental status. Laboratory evaluation revealed cholestatic jaundice with AST 310 U/L, ALT 291 U/L, alkaline phosphatase 860 U/L, and total bilirubin 5.7 mg/dL. Infectious workup was negative. Imaging via MRCP showed multiple scattered hepatic cysts and a small pancreatic cyst, without biliary obstruction.

Further evaluation, including serologies for hepatitis B and C, CMV, HSV, autoimmune hepatitis panel, iron studies, and ceruloplasmin, was unremarkable except for mildly elevated alpha-1 antitrypsin. Scattered liver cysts were seen on an MRI. The overall findings were most consistent with immune-related hepatitis, as pembrolizumab is known to cause both hepatocellular and cholestatic patterns of liver injury.

The patient was started on high-dose prednisone, resulting in rapid clinical and biochemical improvement. Two weeks post-discharge, liver function tests (LFTs) had markedly improved (bilirubin 1.3, AST 19, ALT 40, ALP 193). Given the severity of transaminitis and hyperbilirubinemia (AST >8x ULN, bilirubin >3x ULN), pembrolizumab was permanently discontinued. LFTs normalized after completion of the steroid taper.

Conclusions

This case highlights a rare instance of delayed immune-related hepatitis occurring over a year after initiation of pembrolizumab, far beyond the typical window of onset. Clinicians should maintain a high index of suspicion for irAEs even in late stages of immunotherapy, particularly when common etiologies are excluded. Prompt recognition and corticosteroid treatment can lead to favorable outcomes, even in older patients.

Background

Immune checkpoint inhibitors, including pembrolizumab, are associated with a spectrum of immune-related adverse events (irAEs), including immune- mediated hepatitis. Typically, this toxicity manifests within the first 14 weeks of therapy. Delayed presentations beyond one year are exceedingly rare and pose diagnostic challenges.

Case Presentation

We report an elderly patient (over 90 years old) with stage IVa squamous cell carcinoma of the lung and high microsatellite instability (MSI) who had been receiving pembrolizumab since 2023. In 2024—13 months into therapy—he presented with subjective fevers, weakness, and altered mental status. Laboratory evaluation revealed cholestatic jaundice with AST 310 U/L, ALT 291 U/L, alkaline phosphatase 860 U/L, and total bilirubin 5.7 mg/dL. Infectious workup was negative. Imaging via MRCP showed multiple scattered hepatic cysts and a small pancreatic cyst, without biliary obstruction.

Further evaluation, including serologies for hepatitis B and C, CMV, HSV, autoimmune hepatitis panel, iron studies, and ceruloplasmin, was unremarkable except for mildly elevated alpha-1 antitrypsin. Scattered liver cysts were seen on an MRI. The overall findings were most consistent with immune-related hepatitis, as pembrolizumab is known to cause both hepatocellular and cholestatic patterns of liver injury.

The patient was started on high-dose prednisone, resulting in rapid clinical and biochemical improvement. Two weeks post-discharge, liver function tests (LFTs) had markedly improved (bilirubin 1.3, AST 19, ALT 40, ALP 193). Given the severity of transaminitis and hyperbilirubinemia (AST >8x ULN, bilirubin >3x ULN), pembrolizumab was permanently discontinued. LFTs normalized after completion of the steroid taper.

Conclusions

This case highlights a rare instance of delayed immune-related hepatitis occurring over a year after initiation of pembrolizumab, far beyond the typical window of onset. Clinicians should maintain a high index of suspicion for irAEs even in late stages of immunotherapy, particularly when common etiologies are excluded. Prompt recognition and corticosteroid treatment can lead to favorable outcomes, even in older patients.

Background

Immune checkpoint inhibitors, including pembrolizumab, are associated with a spectrum of immune-related adverse events (irAEs), including immune- mediated hepatitis. Typically, this toxicity manifests within the first 14 weeks of therapy. Delayed presentations beyond one year are exceedingly rare and pose diagnostic challenges.

Case Presentation

We report an elderly patient (over 90 years old) with stage IVa squamous cell carcinoma of the lung and high microsatellite instability (MSI) who had been receiving pembrolizumab since 2023. In 2024—13 months into therapy—he presented with subjective fevers, weakness, and altered mental status. Laboratory evaluation revealed cholestatic jaundice with AST 310 U/L, ALT 291 U/L, alkaline phosphatase 860 U/L, and total bilirubin 5.7 mg/dL. Infectious workup was negative. Imaging via MRCP showed multiple scattered hepatic cysts and a small pancreatic cyst, without biliary obstruction.

Further evaluation, including serologies for hepatitis B and C, CMV, HSV, autoimmune hepatitis panel, iron studies, and ceruloplasmin, was unremarkable except for mildly elevated alpha-1 antitrypsin. Scattered liver cysts were seen on an MRI. The overall findings were most consistent with immune-related hepatitis, as pembrolizumab is known to cause both hepatocellular and cholestatic patterns of liver injury.

The patient was started on high-dose prednisone, resulting in rapid clinical and biochemical improvement. Two weeks post-discharge, liver function tests (LFTs) had markedly improved (bilirubin 1.3, AST 19, ALT 40, ALP 193). Given the severity of transaminitis and hyperbilirubinemia (AST >8x ULN, bilirubin >3x ULN), pembrolizumab was permanently discontinued. LFTs normalized after completion of the steroid taper.

Conclusions

This case highlights a rare instance of delayed immune-related hepatitis occurring over a year after initiation of pembrolizumab, far beyond the typical window of onset. Clinicians should maintain a high index of suspicion for irAEs even in late stages of immunotherapy, particularly when common etiologies are excluded. Prompt recognition and corticosteroid treatment can lead to favorable outcomes, even in older patients.

Do patients with early-stage non–small cell lung cancer (NSCLC) benefit from continuing immunotherapy beyond surgery?

The short answer: Oncologists don’t know for sure.

Since October 2023, the US Food and Drug Administration (FDA) has approved three checkpoint inhibitors — pembrolizumab (Keytruda), durvalumab (Imfinzi), and most recently nivolumab (Opdivo) — alongside platinum-containing chemotherapy before surgery and as monotherapy after surgery to treat resectable NSCLC.

But the trials leading to each approval had a major design flaw. The studies failed to distinguish when patients with resectable NSCLC benefited from immunotherapy — before surgery, after surgery, or at both points.

That missing piece has left oncologists without definitive guidance on how best to treat their patients with resectable disease. 

Jamie E. Chaft, MD, a thoracic medical oncologist and attending physician at Memorial Sloan Kettering Cancer Center in New York City, was “surprised” that the FDA had approved the three immunotherapy combination regimens without this clarity. Clinicians are now left with studies that can’t evaluate the contribution of the neoadjuvant and adjuvant phases, she said.

But that may soon change.

In July, an FDA advisory committee met to discuss the pending approval of durvalumab.

During this July meeting, the FDA’s Oncologic Drugs Advisory Committee (ODAC) called out issues with AstraZeneca’s design of the trial, expressing concern that AstraZeneca had not followed the agency’s advice to compare patient outcomes with durvalumab in the neoadjuvant and adjuvant phases.

The ODAC panel ultimately voted unanimously in favor of requiring drug companies to demonstrate that patients need immunotherapy both before and after surgery in resectable NSCLC. Several panelists said this requirement should extend beyond NSCLC to other tumor types.

“We need to understand who needs what therapy when,” Daniel Spratt, MD, chairman of the FDA’s ODAC, told Medscape Medical News.

But even if the FDA does require drug companies to assess the benefit of immunotherapy pre- and post-surgery, will oncologists get the answers they need for their patients with resectable NSCLC? Or will the new costly trial design requirements dead-end progress in this space?

 

Treating Patients Without Clear Evidence

Despite the ODAC’s strong urging to require — not simply request — that drug companies show patients with resectable NSCLC benefit from immunotherapy in both the neoadjuvant and adjuvant settings, the advisory panel did not think durvalumab’s approval should be delayed until the neoadjuvant vs adjuvant question is answered.

A month later, in August, the FDA approved durvalumab for this indication.

Pembrolizumab (Keytruda, Merck) had already been approved 10 months earlier in the neoadjuvant and adjuvant settings in this setting. And most recently, in October, the FDA added nivolumab (Opdivo, Bristol Myers Squibb) to these approvals.

No trial, however, identified when patients benefited from the drug.

Without this understanding, patients may be taking immunotherapy unnecessarily, at significant expense and toxicity risk.

“Toxicities from immunotherapy can occur at any time after initiation,” said Joshua Eric Reuss, MD, a thoracic medical oncologist at Georgetown University’s Lombardi Comprehensive Cancer Center in Washington, DC. And these “risks definitely continue into the adjuvant period.”

So far, the available evidence does suggest that the neoadjuvant phase of immunotherapy confers the greatest benefit, while adjuvant immunotherapy — which can last a year or longer — may expose patients to more costs and toxicities, with no clear benefit.

A 2024 meta-analysis, which included four trials of neoadjuvant-adjuvant immunotherapy and one trial of neoadjuvant immunotherapy in resectable NSCLC, suggested that the addition of adjuvant immunotherapy did not improve event-free survival (hazard ratio [HR], 0.90; P = .59) or overall survival (HR, 1.18; P = .51) compared with neoadjuvant immunotherapy alone.

According to Spratt, “It’s very clear that the neoadjuvant phase is the more important of the two phases.” Given that, “we’re probably overtreating some patients,” said Spratt, also chairman of Radiation Oncology at University Hospitals Seidman Cancer Center and Case Western Reserve University in Cleveland.

Chaft agreed that “there’s very little data that we need the postoperative phase, and what data we do have is post hoc and limited.”

This evidence gap “has created considerable dilemmas” for oncologists and patients who are faced with “the challenge of deciding which therapeutic options or approach are best suited for each individual,” experts wrote in recent consensus recommendations from the International Association for the Study of Lung Cancer.

Clinicians may ultimately be left to make decisions about prescribing postoperative immunotherapy based on their experience and comfort level.

When Chaft’s patients have a pathologic complete response with immunotherapy and chemotherapy in the neoadjuvant phase, “I’m comfortable stopping because the data would suggest they’re almost certainly cured,” she said.

For patients who have viable disease after neoadjuvant therapy, continuing an immunotherapy postoperatively when it didn’t work preoperatively “is not going to make a difference,” Chaft explained. In these cases, Chaft would look to enroll them in a clinical trial evaluating a different regimen because of the risk for relapse.

With patients who did well preoperatively but still have tumor left at the time of surgery, she would discuss continuing the immunotherapy or participating in a trial, she said.

All the FDA-approved regimens are covered by insurance, said Chaft. Clinicians are most comfortable with pembrolizumab because it is the most widely used immunotherapy in advanced NSCLC, she said. But, she added, “there’s really no strong differentiating data between any of the studies; all the results look very comparable.”

When assessing whether a patient may benefit from immunotherapy after surgery, Reuss looks at a range of factors, including disease stage, histology, gene mutations, and pathologic response. Reuss also weighs patient preferences. A patient coming from another country might only want a neoadjuvant regimen, for instance, he said.

That “isn’t exactly the kind of the level one evidence that one likes to see when making treatment decisions,” said Reuss. “Without prospective data, all we can do is cross-trial comparisons and assessment of subgroups.”

If a new regimen comes along that improves outcomes or decision-making, “I think we would pivot to that in a heartbeat,” he said.

 

But Will FDA Follow ODAC’s Recommendation?

“ODAC has made their point clear,” said Chaft. “Our patients deserve to know that whatever added risk and cost they’re incurring is merited by a clinical outcome.”

Despite the ODAC’s recommendation, it’s not guaranteed that the FDA will follow it.

An FDA spokesperson did not confirm the agency’s decision on the matter but noted that the FDA is “incorporating the panel’s advice.”

Spratt thinks that, going forward, companies will be held to “a higher bar,” but it’s unclear what that bar will look like.

“Whether this is a mandate or a strong recommendation, I think industry is definitely paying attention,” Spratt said. Companies that do not follow the guidance may risk not having their drug approved, “unless it’s just an absolute huge slam dunk of a major benefit to patients.”

In fact, according to Chaft, drug makers seeking approvals of novel entities in this space “won’t have a choice” but to follow any new trial design requirements from the FDA.

Still, getting answers may be a challenge.

Drug companies with immunotherapies already on the market are unlikely to invest the resources to conduct trials comparing the neoadjuvant and adjuvant settings, said Chaft. “It will take too long and cost too much,” she said.

And it remains unclear whether drug companies will decide to stop pursuing novel agents if approvals will ultimately require more expensive and time-consuming trials.

According to Chaft, oncologists have been discussing protocols that could help fill the knowledge gaps. Such trials will be conducted by the National Cancer Institute’s Cooperative Groups, she noted. But it’s early days.

For the time being, with comparative data from phase 3 trials years away, oncologists will have to work with the limited evidence and individual patients in front of them.

Chaft disclosed ties with AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Genentech/Roche, Guardant Health, Janssen Pharmaceuticals, Eli Lilly, and Merck. Reuss disclosed ties with AstraZeneca, Arcus, AbbVie, Bristol Myers Squibb, CatalYm, Daiichi Sankyo, and Eli Lilly, and that Georgetown has received research funding from Genentech/Roche, Verastem, Nuvalent, LUNGevity Foundation, Exelixis, Arcus, and Revolution Medicines. Spratt disclosed ties with Astellas, AstraZeneca, Bayer, Boston Scientific, Janssen Pharmaceuticals, Novartis, and Pfizer.

A version of this article appeared on Medscape.com.

Do patients with early-stage non–small cell lung cancer (NSCLC) benefit from continuing immunotherapy beyond surgery?

The short answer: Oncologists don’t know for sure.

Since October 2023, the US Food and Drug Administration (FDA) has approved three checkpoint inhibitors — pembrolizumab (Keytruda), durvalumab (Imfinzi), and most recently nivolumab (Opdivo) — alongside platinum-containing chemotherapy before surgery and as monotherapy after surgery to treat resectable NSCLC.

But the trials leading to each approval had a major design flaw. The studies failed to distinguish when patients with resectable NSCLC benefited from immunotherapy — before surgery, after surgery, or at both points.

That missing piece has left oncologists without definitive guidance on how best to treat their patients with resectable disease. 

Jamie E. Chaft, MD, a thoracic medical oncologist and attending physician at Memorial Sloan Kettering Cancer Center in New York City, was “surprised” that the FDA had approved the three immunotherapy combination regimens without this clarity. Clinicians are now left with studies that can’t evaluate the contribution of the neoadjuvant and adjuvant phases, she said.

But that may soon change.

In July, an FDA advisory committee met to discuss the pending approval of durvalumab.

During this July meeting, the FDA’s Oncologic Drugs Advisory Committee (ODAC) called out issues with AstraZeneca’s design of the trial, expressing concern that AstraZeneca had not followed the agency’s advice to compare patient outcomes with durvalumab in the neoadjuvant and adjuvant phases.

The ODAC panel ultimately voted unanimously in favor of requiring drug companies to demonstrate that patients need immunotherapy both before and after surgery in resectable NSCLC. Several panelists said this requirement should extend beyond NSCLC to other tumor types.

“We need to understand who needs what therapy when,” Daniel Spratt, MD, chairman of the FDA’s ODAC, told Medscape Medical News.

But even if the FDA does require drug companies to assess the benefit of immunotherapy pre- and post-surgery, will oncologists get the answers they need for their patients with resectable NSCLC? Or will the new costly trial design requirements dead-end progress in this space?

 

Treating Patients Without Clear Evidence

Despite the ODAC’s strong urging to require — not simply request — that drug companies show patients with resectable NSCLC benefit from immunotherapy in both the neoadjuvant and adjuvant settings, the advisory panel did not think durvalumab’s approval should be delayed until the neoadjuvant vs adjuvant question is answered.

A month later, in August, the FDA approved durvalumab for this indication.

Pembrolizumab (Keytruda, Merck) had already been approved 10 months earlier in the neoadjuvant and adjuvant settings in this setting. And most recently, in October, the FDA added nivolumab (Opdivo, Bristol Myers Squibb) to these approvals.

No trial, however, identified when patients benefited from the drug.

Without this understanding, patients may be taking immunotherapy unnecessarily, at significant expense and toxicity risk.

“Toxicities from immunotherapy can occur at any time after initiation,” said Joshua Eric Reuss, MD, a thoracic medical oncologist at Georgetown University’s Lombardi Comprehensive Cancer Center in Washington, DC. And these “risks definitely continue into the adjuvant period.”

So far, the available evidence does suggest that the neoadjuvant phase of immunotherapy confers the greatest benefit, while adjuvant immunotherapy — which can last a year or longer — may expose patients to more costs and toxicities, with no clear benefit.

A 2024 meta-analysis, which included four trials of neoadjuvant-adjuvant immunotherapy and one trial of neoadjuvant immunotherapy in resectable NSCLC, suggested that the addition of adjuvant immunotherapy did not improve event-free survival (hazard ratio [HR], 0.90; P = .59) or overall survival (HR, 1.18; P = .51) compared with neoadjuvant immunotherapy alone.

According to Spratt, “It’s very clear that the neoadjuvant phase is the more important of the two phases.” Given that, “we’re probably overtreating some patients,” said Spratt, also chairman of Radiation Oncology at University Hospitals Seidman Cancer Center and Case Western Reserve University in Cleveland.

Chaft agreed that “there’s very little data that we need the postoperative phase, and what data we do have is post hoc and limited.”

This evidence gap “has created considerable dilemmas” for oncologists and patients who are faced with “the challenge of deciding which therapeutic options or approach are best suited for each individual,” experts wrote in recent consensus recommendations from the International Association for the Study of Lung Cancer.

Clinicians may ultimately be left to make decisions about prescribing postoperative immunotherapy based on their experience and comfort level.

When Chaft’s patients have a pathologic complete response with immunotherapy and chemotherapy in the neoadjuvant phase, “I’m comfortable stopping because the data would suggest they’re almost certainly cured,” she said.

For patients who have viable disease after neoadjuvant therapy, continuing an immunotherapy postoperatively when it didn’t work preoperatively “is not going to make a difference,” Chaft explained. In these cases, Chaft would look to enroll them in a clinical trial evaluating a different regimen because of the risk for relapse.

With patients who did well preoperatively but still have tumor left at the time of surgery, she would discuss continuing the immunotherapy or participating in a trial, she said.

All the FDA-approved regimens are covered by insurance, said Chaft. Clinicians are most comfortable with pembrolizumab because it is the most widely used immunotherapy in advanced NSCLC, she said. But, she added, “there’s really no strong differentiating data between any of the studies; all the results look very comparable.”

When assessing whether a patient may benefit from immunotherapy after surgery, Reuss looks at a range of factors, including disease stage, histology, gene mutations, and pathologic response. Reuss also weighs patient preferences. A patient coming from another country might only want a neoadjuvant regimen, for instance, he said.

That “isn’t exactly the kind of the level one evidence that one likes to see when making treatment decisions,” said Reuss. “Without prospective data, all we can do is cross-trial comparisons and assessment of subgroups.”

If a new regimen comes along that improves outcomes or decision-making, “I think we would pivot to that in a heartbeat,” he said.

 

But Will FDA Follow ODAC’s Recommendation?

“ODAC has made their point clear,” said Chaft. “Our patients deserve to know that whatever added risk and cost they’re incurring is merited by a clinical outcome.”

Despite the ODAC’s recommendation, it’s not guaranteed that the FDA will follow it.

An FDA spokesperson did not confirm the agency’s decision on the matter but noted that the FDA is “incorporating the panel’s advice.”

Spratt thinks that, going forward, companies will be held to “a higher bar,” but it’s unclear what that bar will look like.

“Whether this is a mandate or a strong recommendation, I think industry is definitely paying attention,” Spratt said. Companies that do not follow the guidance may risk not having their drug approved, “unless it’s just an absolute huge slam dunk of a major benefit to patients.”

In fact, according to Chaft, drug makers seeking approvals of novel entities in this space “won’t have a choice” but to follow any new trial design requirements from the FDA.

Still, getting answers may be a challenge.

Drug companies with immunotherapies already on the market are unlikely to invest the resources to conduct trials comparing the neoadjuvant and adjuvant settings, said Chaft. “It will take too long and cost too much,” she said.

And it remains unclear whether drug companies will decide to stop pursuing novel agents if approvals will ultimately require more expensive and time-consuming trials.

According to Chaft, oncologists have been discussing protocols that could help fill the knowledge gaps. Such trials will be conducted by the National Cancer Institute’s Cooperative Groups, she noted. But it’s early days.

For the time being, with comparative data from phase 3 trials years away, oncologists will have to work with the limited evidence and individual patients in front of them.

Chaft disclosed ties with AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Genentech/Roche, Guardant Health, Janssen Pharmaceuticals, Eli Lilly, and Merck. Reuss disclosed ties with AstraZeneca, Arcus, AbbVie, Bristol Myers Squibb, CatalYm, Daiichi Sankyo, and Eli Lilly, and that Georgetown has received research funding from Genentech/Roche, Verastem, Nuvalent, LUNGevity Foundation, Exelixis, Arcus, and Revolution Medicines. Spratt disclosed ties with Astellas, AstraZeneca, Bayer, Boston Scientific, Janssen Pharmaceuticals, Novartis, and Pfizer.

A version of this article appeared on Medscape.com.

Do patients with early-stage non–small cell lung cancer (NSCLC) benefit from continuing immunotherapy beyond surgery?

The short answer: Oncologists don’t know for sure.

Since October 2023, the US Food and Drug Administration (FDA) has approved three checkpoint inhibitors — pembrolizumab (Keytruda), durvalumab (Imfinzi), and most recently nivolumab (Opdivo) — alongside platinum-containing chemotherapy before surgery and as monotherapy after surgery to treat resectable NSCLC.

But the trials leading to each approval had a major design flaw. The studies failed to distinguish when patients with resectable NSCLC benefited from immunotherapy — before surgery, after surgery, or at both points.

That missing piece has left oncologists without definitive guidance on how best to treat their patients with resectable disease. 

Jamie E. Chaft, MD, a thoracic medical oncologist and attending physician at Memorial Sloan Kettering Cancer Center in New York City, was “surprised” that the FDA had approved the three immunotherapy combination regimens without this clarity. Clinicians are now left with studies that can’t evaluate the contribution of the neoadjuvant and adjuvant phases, she said.

But that may soon change.

In July, an FDA advisory committee met to discuss the pending approval of durvalumab.

During this July meeting, the FDA’s Oncologic Drugs Advisory Committee (ODAC) called out issues with AstraZeneca’s design of the trial, expressing concern that AstraZeneca had not followed the agency’s advice to compare patient outcomes with durvalumab in the neoadjuvant and adjuvant phases.

The ODAC panel ultimately voted unanimously in favor of requiring drug companies to demonstrate that patients need immunotherapy both before and after surgery in resectable NSCLC. Several panelists said this requirement should extend beyond NSCLC to other tumor types.

“We need to understand who needs what therapy when,” Daniel Spratt, MD, chairman of the FDA’s ODAC, told Medscape Medical News.

But even if the FDA does require drug companies to assess the benefit of immunotherapy pre- and post-surgery, will oncologists get the answers they need for their patients with resectable NSCLC? Or will the new costly trial design requirements dead-end progress in this space?

 

Treating Patients Without Clear Evidence

Despite the ODAC’s strong urging to require — not simply request — that drug companies show patients with resectable NSCLC benefit from immunotherapy in both the neoadjuvant and adjuvant settings, the advisory panel did not think durvalumab’s approval should be delayed until the neoadjuvant vs adjuvant question is answered.

A month later, in August, the FDA approved durvalumab for this indication.

Pembrolizumab (Keytruda, Merck) had already been approved 10 months earlier in the neoadjuvant and adjuvant settings in this setting. And most recently, in October, the FDA added nivolumab (Opdivo, Bristol Myers Squibb) to these approvals.

No trial, however, identified when patients benefited from the drug.

Without this understanding, patients may be taking immunotherapy unnecessarily, at significant expense and toxicity risk.

“Toxicities from immunotherapy can occur at any time after initiation,” said Joshua Eric Reuss, MD, a thoracic medical oncologist at Georgetown University’s Lombardi Comprehensive Cancer Center in Washington, DC. And these “risks definitely continue into the adjuvant period.”

So far, the available evidence does suggest that the neoadjuvant phase of immunotherapy confers the greatest benefit, while adjuvant immunotherapy — which can last a year or longer — may expose patients to more costs and toxicities, with no clear benefit.

A 2024 meta-analysis, which included four trials of neoadjuvant-adjuvant immunotherapy and one trial of neoadjuvant immunotherapy in resectable NSCLC, suggested that the addition of adjuvant immunotherapy did not improve event-free survival (hazard ratio [HR], 0.90; P = .59) or overall survival (HR, 1.18; P = .51) compared with neoadjuvant immunotherapy alone.

According to Spratt, “It’s very clear that the neoadjuvant phase is the more important of the two phases.” Given that, “we’re probably overtreating some patients,” said Spratt, also chairman of Radiation Oncology at University Hospitals Seidman Cancer Center and Case Western Reserve University in Cleveland.

Chaft agreed that “there’s very little data that we need the postoperative phase, and what data we do have is post hoc and limited.”

This evidence gap “has created considerable dilemmas” for oncologists and patients who are faced with “the challenge of deciding which therapeutic options or approach are best suited for each individual,” experts wrote in recent consensus recommendations from the International Association for the Study of Lung Cancer.

Clinicians may ultimately be left to make decisions about prescribing postoperative immunotherapy based on their experience and comfort level.

When Chaft’s patients have a pathologic complete response with immunotherapy and chemotherapy in the neoadjuvant phase, “I’m comfortable stopping because the data would suggest they’re almost certainly cured,” she said.

For patients who have viable disease after neoadjuvant therapy, continuing an immunotherapy postoperatively when it didn’t work preoperatively “is not going to make a difference,” Chaft explained. In these cases, Chaft would look to enroll them in a clinical trial evaluating a different regimen because of the risk for relapse.

With patients who did well preoperatively but still have tumor left at the time of surgery, she would discuss continuing the immunotherapy or participating in a trial, she said.

All the FDA-approved regimens are covered by insurance, said Chaft. Clinicians are most comfortable with pembrolizumab because it is the most widely used immunotherapy in advanced NSCLC, she said. But, she added, “there’s really no strong differentiating data between any of the studies; all the results look very comparable.”

When assessing whether a patient may benefit from immunotherapy after surgery, Reuss looks at a range of factors, including disease stage, histology, gene mutations, and pathologic response. Reuss also weighs patient preferences. A patient coming from another country might only want a neoadjuvant regimen, for instance, he said.

That “isn’t exactly the kind of the level one evidence that one likes to see when making treatment decisions,” said Reuss. “Without prospective data, all we can do is cross-trial comparisons and assessment of subgroups.”

If a new regimen comes along that improves outcomes or decision-making, “I think we would pivot to that in a heartbeat,” he said.

 

But Will FDA Follow ODAC’s Recommendation?

“ODAC has made their point clear,” said Chaft. “Our patients deserve to know that whatever added risk and cost they’re incurring is merited by a clinical outcome.”

Despite the ODAC’s recommendation, it’s not guaranteed that the FDA will follow it.

An FDA spokesperson did not confirm the agency’s decision on the matter but noted that the FDA is “incorporating the panel’s advice.”

Spratt thinks that, going forward, companies will be held to “a higher bar,” but it’s unclear what that bar will look like.

“Whether this is a mandate or a strong recommendation, I think industry is definitely paying attention,” Spratt said. Companies that do not follow the guidance may risk not having their drug approved, “unless it’s just an absolute huge slam dunk of a major benefit to patients.”

In fact, according to Chaft, drug makers seeking approvals of novel entities in this space “won’t have a choice” but to follow any new trial design requirements from the FDA.

Still, getting answers may be a challenge.

Drug companies with immunotherapies already on the market are unlikely to invest the resources to conduct trials comparing the neoadjuvant and adjuvant settings, said Chaft. “It will take too long and cost too much,” she said.

And it remains unclear whether drug companies will decide to stop pursuing novel agents if approvals will ultimately require more expensive and time-consuming trials.

According to Chaft, oncologists have been discussing protocols that could help fill the knowledge gaps. Such trials will be conducted by the National Cancer Institute’s Cooperative Groups, she noted. But it’s early days.

For the time being, with comparative data from phase 3 trials years away, oncologists will have to work with the limited evidence and individual patients in front of them.

Chaft disclosed ties with AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Genentech/Roche, Guardant Health, Janssen Pharmaceuticals, Eli Lilly, and Merck. Reuss disclosed ties with AstraZeneca, Arcus, AbbVie, Bristol Myers Squibb, CatalYm, Daiichi Sankyo, and Eli Lilly, and that Georgetown has received research funding from Genentech/Roche, Verastem, Nuvalent, LUNGevity Foundation, Exelixis, Arcus, and Revolution Medicines. Spratt disclosed ties with Astellas, AstraZeneca, Bayer, Boston Scientific, Janssen Pharmaceuticals, Novartis, and Pfizer.

A version of this article appeared on Medscape.com.

The question has been lingering for years in medical science circles. Since 2020, when the artificial intelligence (AI) model AlphaFold made it possible to predict protein structures, would the technology open the drug discovery floodgates?

Short answer: No. At least not yet.

The longer answer goes something like this:

A drug target (such as a mutation) is like a lock. The right drug (a protein designed to bind to the mutation, stopping its activity) is the key. But proteins are fidgety and flexible.

“They’re basically molecular springs,” said Gabriel Monteiro da Silva, PhD, a computational chemistry research scientist at Genesis Therapeutics. “Your key can bend and alter the shape of the lock, and if you don’t account for that, your key might fail.”

This is the protein problem in drug development. Another issue making this challenge so vexing is that proteins don’t act in isolation. Their interactions with other proteins, ribonucleic acid, and DNA can affect how they bind to molecules and the shapes they adopt.

Newer versions of AlphaFold, such as AlphaFold Multimer and AlphaFold 3 (the code for which was recently revealed for academic use), can predict many interactions among proteins and between proteins and other molecules. But these tools still have weak points scientists are trying to overcome or work around.

“Those kinds of dynamics and multiple conformations are still quite challenging for the AI models to predict,” said James Zou, PhD, associate professor of biomedical data science at Stanford University in California.

“We’re finding more and more that the only way we can make these structures useful for drug discovery is if we incorporate dynamics, if we incorporate more physics into the model,” said Monteiro da Silva.

Monteiro da Silva spent 3 years during his PhD at Brown University, Providence, Rhode Island, running physics-based simulations in the lab, trying to understand why proteins carrying certain mutations are drug resistant. His results showed how “the changing landscape of shapes that a protein can take” prevented the drug from binding.

It took him 3 years to model just four mutations.

AI can do better — and the struggle is fascinating. By developing models that build on the predictive power of AlphaFold, scientists are uncovering new details about protein activity — insights that can lead to new therapeutics and reveal why existing ones stop working — much faster than they could with traditional methods or AlphaFold alone.

 

New Windows into Protein Dynamics

By predicting protein structural details, AlphaFold models also made it possible to predict pockets where drugs could bind.

A notable step, “but that’s just the starting point,” said Pedro Beltrao, PhD, an associate professor at Institute of Molecular Systems Biology, ETH Zurich in Switzerland. “It’s still very difficult, given a pocket, to actually design the drug or figure out what the pocket binds.”

Going back to the lock-and-key analogy: While he was at Brown, with a team of researchers in the Rubenstein Group, Monteiro da Silva helped create a model to better understand how mutations affect “the shape and dynamics of the lock.” They manipulated the amino acid sequences of proteins, guiding their evolution. This enabled them to use AlphaFold to predict “protein ensembles” and how frequently those ensembles appear. Each ensemble represents the many different shapes a protein can take under given conditions.

“Essentially, it tries to find the most common shapes that a protein will take over an arbitrary amount of time,” Monteiro da Silva said. “If we can predict these ensembles at scale and fast, then we can screen many mutations that cause resistance and develop drugs that will not be affected by that resistance.”

To evaluate their method, the researchers focused on ABL1, a well-studied kinase that causes leukemia. ABL1 can be drugged – unless it carries or develops a mutation that causes drug resistance. Currently there are no drugs that work against proteins carrying those mutations, according to Monteiro da Silva. The researchers used their hybrid AI-meets-physics method to investigate how drugs bind to different ABL1 mutations, screening 100 mutations in just 1 month.

“It’s not going to be perfect for every one of them. But if we have 100 and we get 20 with good accuracy, that’s better than doing four over 3 years,” Monteiro da Silva said.

A forthcoming paper will make their model publicly available in “an easy-to-use graphical interface” that they hope clinicians and medicinal chemists will try out. It can also complement other AI-based tools that dig into protein dynamics, according to Monteiro da Silva.

 

Complementary Tools to Speed Up Discovery 

Another aspect of the protein problem is scale. One protein can interact with hundreds of other proteins, which in turn may interact with hundreds more, all of which comprise the human interactome.

Feixiong Cheng, PhD, helped build PIONEER, a deep learning model that predicts the three-dimensional (3D) structure of interactions between proteins across the interactome.

Most disease mutations disrupt specific interactions between proteins, making their affinity stronger or weaker, explained Cheng. To treat a disease without causing major side effects, scientists need a precise understanding of those interactions.

“From the drug discovery perspective, we cannot just focus on single proteins. We have to understand the protein environment, in particular how the protein interacts with other proteins,” said Cheng, director of Cleveland Clinic Genome Center, Cleveland.

PIONEER helps by blending AlphaFold’s protein structure predictions with next-generation sequencing, a type of genomic research that identifies mutations in the human genome. The model predicts the 3D structure of the places where proteins interact — the binding sites, or interfaces — across the interactome.

“We tell you not only that a binds b, but where on a and where on b the two proteins interact,” said Haiyuan Yu, PhD, director of the Center for Innovative Proteomics, Cornell University, and co-creator of PIONEER.

This can help scientists understand “why a mutation, protein, or even network is a good target for therapeutic discovery,” Cheng said.

The researchers validated PIONEER’s predictions in the lab, testing the impacts of roughly 3000 mutations on 7000 pairs of interacting proteins. Based on their findings, they plan to develop and test treatments for lung and endometrial cancer.

PIONEER can also help scientists home in on how a mutation causes a disease, such as by showing recurrent mutations.

“If you find cancer mutations hitting an interface again and again and again, it means that this is likely to be driving cancer progression,” said Beltrao.

Beltrao’s lab and others have looked for recurrent mutations by using AlphaFold Multimer and AlphaFold 3 to directly model protein interactions. It’s a much slower approach (Pioneer is more than 5000 faster than AlphaFold Multimer, according to Cheng). But it could allow scientists to model interfaces that are not shown by PIONEER.

“You will need many different things to try to come up with a structural modeling of the interactome, and all these will have limitations,” said Beltrao. “Their method is a very good step forward, and there’ll be other approaches that are complementary, to continue to add details.”

 

And It Wouldn’t be an AI Mission Without ChatGPT

Large language models, such as ChatGPT, are another way that scientists are adding details to protein structure predictions. Zou used GPT-4 to “fine tune” a protein language model, called evolutionary scale modeling (ESM-2), which predicts protein structures directly from a protein sequence.

First, they trained ChatGPT on thousands of papers and studies containing information about the functions, biophysical properties, and disease relevance of different mutations. Next, they used the trained model to “teach” ESM-2, boosting its ability “to predict which mutations are likely to have larger effects or smaller effects,” Zou said. The same could be done for a model like AlphaFold, according to Zou.

“They are quite complementary in that the large language model contains a lot more information about the functions and the biophysics of different mutations and proteins as captured in text,” he said, whereas “you can’t give AlphaFold a piece of paper.”

Exactly how AlphaFold makes its predictions is another mystery. “It will somehow learn protein dynamics phenomenologically,” said Monteiro da Silva. He and others are trying to understand how that happens, in hopes of creating even more accurate predictive models. But for the time being, AI-based methods still need assistance from physics.

“The dream is that we achieve a state where we rely on just the fast methods, and they’re accurate enough,” he said. “But we’re so far from that.”

A version of this article first appeared on Medscape.com.

The question has been lingering for years in medical science circles. Since 2020, when the artificial intelligence (AI) model AlphaFold made it possible to predict protein structures, would the technology open the drug discovery floodgates?

Short answer: No. At least not yet.

The longer answer goes something like this:

A drug target (such as a mutation) is like a lock. The right drug (a protein designed to bind to the mutation, stopping its activity) is the key. But proteins are fidgety and flexible.

“They’re basically molecular springs,” said Gabriel Monteiro da Silva, PhD, a computational chemistry research scientist at Genesis Therapeutics. “Your key can bend and alter the shape of the lock, and if you don’t account for that, your key might fail.”

This is the protein problem in drug development. Another issue making this challenge so vexing is that proteins don’t act in isolation. Their interactions with other proteins, ribonucleic acid, and DNA can affect how they bind to molecules and the shapes they adopt.

Newer versions of AlphaFold, such as AlphaFold Multimer and AlphaFold 3 (the code for which was recently revealed for academic use), can predict many interactions among proteins and between proteins and other molecules. But these tools still have weak points scientists are trying to overcome or work around.

“Those kinds of dynamics and multiple conformations are still quite challenging for the AI models to predict,” said James Zou, PhD, associate professor of biomedical data science at Stanford University in California.

“We’re finding more and more that the only way we can make these structures useful for drug discovery is if we incorporate dynamics, if we incorporate more physics into the model,” said Monteiro da Silva.

Monteiro da Silva spent 3 years during his PhD at Brown University, Providence, Rhode Island, running physics-based simulations in the lab, trying to understand why proteins carrying certain mutations are drug resistant. His results showed how “the changing landscape of shapes that a protein can take” prevented the drug from binding.

It took him 3 years to model just four mutations.

AI can do better — and the struggle is fascinating. By developing models that build on the predictive power of AlphaFold, scientists are uncovering new details about protein activity — insights that can lead to new therapeutics and reveal why existing ones stop working — much faster than they could with traditional methods or AlphaFold alone.

 

New Windows into Protein Dynamics

By predicting protein structural details, AlphaFold models also made it possible to predict pockets where drugs could bind.

A notable step, “but that’s just the starting point,” said Pedro Beltrao, PhD, an associate professor at Institute of Molecular Systems Biology, ETH Zurich in Switzerland. “It’s still very difficult, given a pocket, to actually design the drug or figure out what the pocket binds.”

Going back to the lock-and-key analogy: While he was at Brown, with a team of researchers in the Rubenstein Group, Monteiro da Silva helped create a model to better understand how mutations affect “the shape and dynamics of the lock.” They manipulated the amino acid sequences of proteins, guiding their evolution. This enabled them to use AlphaFold to predict “protein ensembles” and how frequently those ensembles appear. Each ensemble represents the many different shapes a protein can take under given conditions.

“Essentially, it tries to find the most common shapes that a protein will take over an arbitrary amount of time,” Monteiro da Silva said. “If we can predict these ensembles at scale and fast, then we can screen many mutations that cause resistance and develop drugs that will not be affected by that resistance.”

To evaluate their method, the researchers focused on ABL1, a well-studied kinase that causes leukemia. ABL1 can be drugged – unless it carries or develops a mutation that causes drug resistance. Currently there are no drugs that work against proteins carrying those mutations, according to Monteiro da Silva. The researchers used their hybrid AI-meets-physics method to investigate how drugs bind to different ABL1 mutations, screening 100 mutations in just 1 month.

“It’s not going to be perfect for every one of them. But if we have 100 and we get 20 with good accuracy, that’s better than doing four over 3 years,” Monteiro da Silva said.

A forthcoming paper will make their model publicly available in “an easy-to-use graphical interface” that they hope clinicians and medicinal chemists will try out. It can also complement other AI-based tools that dig into protein dynamics, according to Monteiro da Silva.

 

Complementary Tools to Speed Up Discovery 

Another aspect of the protein problem is scale. One protein can interact with hundreds of other proteins, which in turn may interact with hundreds more, all of which comprise the human interactome.

Feixiong Cheng, PhD, helped build PIONEER, a deep learning model that predicts the three-dimensional (3D) structure of interactions between proteins across the interactome.

Most disease mutations disrupt specific interactions between proteins, making their affinity stronger or weaker, explained Cheng. To treat a disease without causing major side effects, scientists need a precise understanding of those interactions.

“From the drug discovery perspective, we cannot just focus on single proteins. We have to understand the protein environment, in particular how the protein interacts with other proteins,” said Cheng, director of Cleveland Clinic Genome Center, Cleveland.

PIONEER helps by blending AlphaFold’s protein structure predictions with next-generation sequencing, a type of genomic research that identifies mutations in the human genome. The model predicts the 3D structure of the places where proteins interact — the binding sites, or interfaces — across the interactome.

“We tell you not only that a binds b, but where on a and where on b the two proteins interact,” said Haiyuan Yu, PhD, director of the Center for Innovative Proteomics, Cornell University, and co-creator of PIONEER.

This can help scientists understand “why a mutation, protein, or even network is a good target for therapeutic discovery,” Cheng said.

The researchers validated PIONEER’s predictions in the lab, testing the impacts of roughly 3000 mutations on 7000 pairs of interacting proteins. Based on their findings, they plan to develop and test treatments for lung and endometrial cancer.

PIONEER can also help scientists home in on how a mutation causes a disease, such as by showing recurrent mutations.

“If you find cancer mutations hitting an interface again and again and again, it means that this is likely to be driving cancer progression,” said Beltrao.

Beltrao’s lab and others have looked for recurrent mutations by using AlphaFold Multimer and AlphaFold 3 to directly model protein interactions. It’s a much slower approach (Pioneer is more than 5000 faster than AlphaFold Multimer, according to Cheng). But it could allow scientists to model interfaces that are not shown by PIONEER.

“You will need many different things to try to come up with a structural modeling of the interactome, and all these will have limitations,” said Beltrao. “Their method is a very good step forward, and there’ll be other approaches that are complementary, to continue to add details.”

 

And It Wouldn’t be an AI Mission Without ChatGPT

Large language models, such as ChatGPT, are another way that scientists are adding details to protein structure predictions. Zou used GPT-4 to “fine tune” a protein language model, called evolutionary scale modeling (ESM-2), which predicts protein structures directly from a protein sequence.

First, they trained ChatGPT on thousands of papers and studies containing information about the functions, biophysical properties, and disease relevance of different mutations. Next, they used the trained model to “teach” ESM-2, boosting its ability “to predict which mutations are likely to have larger effects or smaller effects,” Zou said. The same could be done for a model like AlphaFold, according to Zou.

“They are quite complementary in that the large language model contains a lot more information about the functions and the biophysics of different mutations and proteins as captured in text,” he said, whereas “you can’t give AlphaFold a piece of paper.”

Exactly how AlphaFold makes its predictions is another mystery. “It will somehow learn protein dynamics phenomenologically,” said Monteiro da Silva. He and others are trying to understand how that happens, in hopes of creating even more accurate predictive models. But for the time being, AI-based methods still need assistance from physics.

“The dream is that we achieve a state where we rely on just the fast methods, and they’re accurate enough,” he said. “But we’re so far from that.”

A version of this article first appeared on Medscape.com.

The question has been lingering for years in medical science circles. Since 2020, when the artificial intelligence (AI) model AlphaFold made it possible to predict protein structures, would the technology open the drug discovery floodgates?

Short answer: No. At least not yet.

The longer answer goes something like this:

A drug target (such as a mutation) is like a lock. The right drug (a protein designed to bind to the mutation, stopping its activity) is the key. But proteins are fidgety and flexible.

“They’re basically molecular springs,” said Gabriel Monteiro da Silva, PhD, a computational chemistry research scientist at Genesis Therapeutics. “Your key can bend and alter the shape of the lock, and if you don’t account for that, your key might fail.”

This is the protein problem in drug development. Another issue making this challenge so vexing is that proteins don’t act in isolation. Their interactions with other proteins, ribonucleic acid, and DNA can affect how they bind to molecules and the shapes they adopt.

Newer versions of AlphaFold, such as AlphaFold Multimer and AlphaFold 3 (the code for which was recently revealed for academic use), can predict many interactions among proteins and between proteins and other molecules. But these tools still have weak points scientists are trying to overcome or work around.

“Those kinds of dynamics and multiple conformations are still quite challenging for the AI models to predict,” said James Zou, PhD, associate professor of biomedical data science at Stanford University in California.

“We’re finding more and more that the only way we can make these structures useful for drug discovery is if we incorporate dynamics, if we incorporate more physics into the model,” said Monteiro da Silva.

Monteiro da Silva spent 3 years during his PhD at Brown University, Providence, Rhode Island, running physics-based simulations in the lab, trying to understand why proteins carrying certain mutations are drug resistant. His results showed how “the changing landscape of shapes that a protein can take” prevented the drug from binding.

It took him 3 years to model just four mutations.

AI can do better — and the struggle is fascinating. By developing models that build on the predictive power of AlphaFold, scientists are uncovering new details about protein activity — insights that can lead to new therapeutics and reveal why existing ones stop working — much faster than they could with traditional methods or AlphaFold alone.

 

New Windows into Protein Dynamics

By predicting protein structural details, AlphaFold models also made it possible to predict pockets where drugs could bind.

A notable step, “but that’s just the starting point,” said Pedro Beltrao, PhD, an associate professor at Institute of Molecular Systems Biology, ETH Zurich in Switzerland. “It’s still very difficult, given a pocket, to actually design the drug or figure out what the pocket binds.”

Going back to the lock-and-key analogy: While he was at Brown, with a team of researchers in the Rubenstein Group, Monteiro da Silva helped create a model to better understand how mutations affect “the shape and dynamics of the lock.” They manipulated the amino acid sequences of proteins, guiding their evolution. This enabled them to use AlphaFold to predict “protein ensembles” and how frequently those ensembles appear. Each ensemble represents the many different shapes a protein can take under given conditions.

“Essentially, it tries to find the most common shapes that a protein will take over an arbitrary amount of time,” Monteiro da Silva said. “If we can predict these ensembles at scale and fast, then we can screen many mutations that cause resistance and develop drugs that will not be affected by that resistance.”

To evaluate their method, the researchers focused on ABL1, a well-studied kinase that causes leukemia. ABL1 can be drugged – unless it carries or develops a mutation that causes drug resistance. Currently there are no drugs that work against proteins carrying those mutations, according to Monteiro da Silva. The researchers used their hybrid AI-meets-physics method to investigate how drugs bind to different ABL1 mutations, screening 100 mutations in just 1 month.

“It’s not going to be perfect for every one of them. But if we have 100 and we get 20 with good accuracy, that’s better than doing four over 3 years,” Monteiro da Silva said.

A forthcoming paper will make their model publicly available in “an easy-to-use graphical interface” that they hope clinicians and medicinal chemists will try out. It can also complement other AI-based tools that dig into protein dynamics, according to Monteiro da Silva.

 

Complementary Tools to Speed Up Discovery 

Another aspect of the protein problem is scale. One protein can interact with hundreds of other proteins, which in turn may interact with hundreds more, all of which comprise the human interactome.

Feixiong Cheng, PhD, helped build PIONEER, a deep learning model that predicts the three-dimensional (3D) structure of interactions between proteins across the interactome.

Most disease mutations disrupt specific interactions between proteins, making their affinity stronger or weaker, explained Cheng. To treat a disease without causing major side effects, scientists need a precise understanding of those interactions.

“From the drug discovery perspective, we cannot just focus on single proteins. We have to understand the protein environment, in particular how the protein interacts with other proteins,” said Cheng, director of Cleveland Clinic Genome Center, Cleveland.

PIONEER helps by blending AlphaFold’s protein structure predictions with next-generation sequencing, a type of genomic research that identifies mutations in the human genome. The model predicts the 3D structure of the places where proteins interact — the binding sites, or interfaces — across the interactome.

“We tell you not only that a binds b, but where on a and where on b the two proteins interact,” said Haiyuan Yu, PhD, director of the Center for Innovative Proteomics, Cornell University, and co-creator of PIONEER.

This can help scientists understand “why a mutation, protein, or even network is a good target for therapeutic discovery,” Cheng said.

The researchers validated PIONEER’s predictions in the lab, testing the impacts of roughly 3000 mutations on 7000 pairs of interacting proteins. Based on their findings, they plan to develop and test treatments for lung and endometrial cancer.

PIONEER can also help scientists home in on how a mutation causes a disease, such as by showing recurrent mutations.

“If you find cancer mutations hitting an interface again and again and again, it means that this is likely to be driving cancer progression,” said Beltrao.

Beltrao’s lab and others have looked for recurrent mutations by using AlphaFold Multimer and AlphaFold 3 to directly model protein interactions. It’s a much slower approach (Pioneer is more than 5000 faster than AlphaFold Multimer, according to Cheng). But it could allow scientists to model interfaces that are not shown by PIONEER.

“You will need many different things to try to come up with a structural modeling of the interactome, and all these will have limitations,” said Beltrao. “Their method is a very good step forward, and there’ll be other approaches that are complementary, to continue to add details.”

 

And It Wouldn’t be an AI Mission Without ChatGPT

Large language models, such as ChatGPT, are another way that scientists are adding details to protein structure predictions. Zou used GPT-4 to “fine tune” a protein language model, called evolutionary scale modeling (ESM-2), which predicts protein structures directly from a protein sequence.

First, they trained ChatGPT on thousands of papers and studies containing information about the functions, biophysical properties, and disease relevance of different mutations. Next, they used the trained model to “teach” ESM-2, boosting its ability “to predict which mutations are likely to have larger effects or smaller effects,” Zou said. The same could be done for a model like AlphaFold, according to Zou.

“They are quite complementary in that the large language model contains a lot more information about the functions and the biophysics of different mutations and proteins as captured in text,” he said, whereas “you can’t give AlphaFold a piece of paper.”

Exactly how AlphaFold makes its predictions is another mystery. “It will somehow learn protein dynamics phenomenologically,” said Monteiro da Silva. He and others are trying to understand how that happens, in hopes of creating even more accurate predictive models. But for the time being, AI-based methods still need assistance from physics.

“The dream is that we achieve a state where we rely on just the fast methods, and they’re accurate enough,” he said. “But we’re so far from that.”

A version of this article first appeared on Medscape.com.

TOPLINE:

For patients with early-stage ERBB2 (formerly HER2)–positive breast cancer, injections of increasing doses of autologous conventional type 1 dendritic cells (cDC1) combined with ERBB2-targeted antibodies demonstrate safety and effectiveness in enhancing immune response. The higher dose (100 million cells) shows enhanced immune effector recruitment and significant tumor regression before chemotherapy initiation.

METHODOLOGY:

• ERBB2-positive breast cancer survival has improved with anti-ERBB2 antibodies trastuzumab and pertuzumab, but for a pathologic complete response, chemotherapy remains necessary, which comes with significant toxic effects.
• A phase 1, nonrandomized clinical trial enrolled 12 patients with early-stage ERBB2-positive breast cancer in Tampa, Florida, from October 2021 to October 2022.
• Participants received intratumoral (IT) cDC1 injections weekly for 6 weeks at two dose levels (50 million cells for dose level 1 and 100 million cells for dose level 2), with six patients in each group.
• Starting from day 1 of the cDC1 injections, treatment included trastuzumab (8-mg/kg loading dose, then 6 mg/kg) and pertuzumab (840-mg loading dose, then 420 mg) administered intravenously every 3 weeks for six cycles, followed by paclitaxel (80 mg/m²) weekly for 12 weeks and surgery with lumpectomy or mastectomy.
• Primary outcomes measured safety and immune response of increasing doses of cDC1 combined with anti-ERBB2 antibodies before neoadjuvant chemotherapy; secondary outcomes assessed antitumor efficacy through breast MRI and residual cancer burden at surgery.

TAKEAWAY:

• IT delivery of ERBB2 cDC1 was safe and not associated with any dose-limiting toxic effects. The most frequent adverse events attributed to cDC1 were grade 1-2 chills (50%), fatigue (41.7%), headache (33%), and injection-site reactions (33%).
• Dose level 2 showed enhanced recruitment of adaptive CD3, CD4, and CD8 T cells and B cells within the tumor microenvironment (TME), along with increased innate gamma delta T cells and natural killer T cells.
• Breast MRI revealed nine objective responses, including six partial responses and three complete responses, with three cases of stable disease.
• Following surgery, 7 of 12 patients (58%) achieved a pathologic complete response, including all 3 hormone receptor–negative patients and 4 of the 9 hormone receptor–positive patients.

IN PRACTICE:

“Overall, the clinical data shown here demonstrate the effects of combining ERBB2 antibodies with IT [intratumoral] delivery of targeted cDC1 to enhance immune cell infiltration within the TME [tumor microenvironment] and subsequently induce tumor regression before chemotherapy,” wrote the authors, who noted they will be testing the higher dose for an ongoing phase 2 trial with an additional 41 patients.

SOURCE:

The study was led by Hyo S. Han, MD, of H. Lee Moffitt Cancer Center and Research Institute in Tampa, Florida. It was published online on December 5, 2024, in JAMA Oncology.

LIMITATIONS:

Because only two dose levels of cDC1 were tested, it remains unclear whether higher doses or different administration schedules could further enhance immune response. Additionally, the nonrandomized design prevents definitive conclusions about whether the clinical benefits were solely from the anti-ERBB2 antibodies. The small sample size also makes it difficult to determine if the pathologic complete responses were primarily due to the 12 weeks of trastuzumab/pertuzumab/paclitaxel treatment.

DISCLOSURES:

This study was funded by the Moffitt Breast Cancer Research Fund, Shula Fund, and Pennies in Action. Several authors reported research support and personal and consulting fees from US funding agencies and multiple pharmaceutical companies outside of the submitted work, as well as related intellectual property and patents.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication. A version of this article first appeared on Medscape.com.

TOPLINE:

For patients with early-stage ERBB2 (formerly HER2)–positive breast cancer, injections of increasing doses of autologous conventional type 1 dendritic cells (cDC1) combined with ERBB2-targeted antibodies demonstrate safety and effectiveness in enhancing immune response. The higher dose (100 million cells) shows enhanced immune effector recruitment and significant tumor regression before chemotherapy initiation.

METHODOLOGY:

• ERBB2-positive breast cancer survival has improved with anti-ERBB2 antibodies trastuzumab and pertuzumab, but for a pathologic complete response, chemotherapy remains necessary, which comes with significant toxic effects.
• A phase 1, nonrandomized clinical trial enrolled 12 patients with early-stage ERBB2-positive breast cancer in Tampa, Florida, from October 2021 to October 2022.
• Participants received intratumoral (IT) cDC1 injections weekly for 6 weeks at two dose levels (50 million cells for dose level 1 and 100 million cells for dose level 2), with six patients in each group.
• Starting from day 1 of the cDC1 injections, treatment included trastuzumab (8-mg/kg loading dose, then 6 mg/kg) and pertuzumab (840-mg loading dose, then 420 mg) administered intravenously every 3 weeks for six cycles, followed by paclitaxel (80 mg/m²) weekly for 12 weeks and surgery with lumpectomy or mastectomy.
• Primary outcomes measured safety and immune response of increasing doses of cDC1 combined with anti-ERBB2 antibodies before neoadjuvant chemotherapy; secondary outcomes assessed antitumor efficacy through breast MRI and residual cancer burden at surgery.

TAKEAWAY:

• IT delivery of ERBB2 cDC1 was safe and not associated with any dose-limiting toxic effects. The most frequent adverse events attributed to cDC1 were grade 1-2 chills (50%), fatigue (41.7%), headache (33%), and injection-site reactions (33%).
• Dose level 2 showed enhanced recruitment of adaptive CD3, CD4, and CD8 T cells and B cells within the tumor microenvironment (TME), along with increased innate gamma delta T cells and natural killer T cells.
• Breast MRI revealed nine objective responses, including six partial responses and three complete responses, with three cases of stable disease.
• Following surgery, 7 of 12 patients (58%) achieved a pathologic complete response, including all 3 hormone receptor–negative patients and 4 of the 9 hormone receptor–positive patients.

IN PRACTICE:

“Overall, the clinical data shown here demonstrate the effects of combining ERBB2 antibodies with IT [intratumoral] delivery of targeted cDC1 to enhance immune cell infiltration within the TME [tumor microenvironment] and subsequently induce tumor regression before chemotherapy,” wrote the authors, who noted they will be testing the higher dose for an ongoing phase 2 trial with an additional 41 patients.

SOURCE:

The study was led by Hyo S. Han, MD, of H. Lee Moffitt Cancer Center and Research Institute in Tampa, Florida. It was published online on December 5, 2024, in JAMA Oncology.

LIMITATIONS:

Because only two dose levels of cDC1 were tested, it remains unclear whether higher doses or different administration schedules could further enhance immune response. Additionally, the nonrandomized design prevents definitive conclusions about whether the clinical benefits were solely from the anti-ERBB2 antibodies. The small sample size also makes it difficult to determine if the pathologic complete responses were primarily due to the 12 weeks of trastuzumab/pertuzumab/paclitaxel treatment.

DISCLOSURES:

This study was funded by the Moffitt Breast Cancer Research Fund, Shula Fund, and Pennies in Action. Several authors reported research support and personal and consulting fees from US funding agencies and multiple pharmaceutical companies outside of the submitted work, as well as related intellectual property and patents.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication. A version of this article first appeared on Medscape.com.

TOPLINE:

For patients with early-stage ERBB2 (formerly HER2)–positive breast cancer, injections of increasing doses of autologous conventional type 1 dendritic cells (cDC1) combined with ERBB2-targeted antibodies demonstrate safety and effectiveness in enhancing immune response. The higher dose (100 million cells) shows enhanced immune effector recruitment and significant tumor regression before chemotherapy initiation.

METHODOLOGY:

• ERBB2-positive breast cancer survival has improved with anti-ERBB2 antibodies trastuzumab and pertuzumab, but for a pathologic complete response, chemotherapy remains necessary, which comes with significant toxic effects.
• A phase 1, nonrandomized clinical trial enrolled 12 patients with early-stage ERBB2-positive breast cancer in Tampa, Florida, from October 2021 to October 2022.
• Participants received intratumoral (IT) cDC1 injections weekly for 6 weeks at two dose levels (50 million cells for dose level 1 and 100 million cells for dose level 2), with six patients in each group.
• Starting from day 1 of the cDC1 injections, treatment included trastuzumab (8-mg/kg loading dose, then 6 mg/kg) and pertuzumab (840-mg loading dose, then 420 mg) administered intravenously every 3 weeks for six cycles, followed by paclitaxel (80 mg/m²) weekly for 12 weeks and surgery with lumpectomy or mastectomy.
• Primary outcomes measured safety and immune response of increasing doses of cDC1 combined with anti-ERBB2 antibodies before neoadjuvant chemotherapy; secondary outcomes assessed antitumor efficacy through breast MRI and residual cancer burden at surgery.

TAKEAWAY:

• IT delivery of ERBB2 cDC1 was safe and not associated with any dose-limiting toxic effects. The most frequent adverse events attributed to cDC1 were grade 1-2 chills (50%), fatigue (41.7%), headache (33%), and injection-site reactions (33%).
• Dose level 2 showed enhanced recruitment of adaptive CD3, CD4, and CD8 T cells and B cells within the tumor microenvironment (TME), along with increased innate gamma delta T cells and natural killer T cells.
• Breast MRI revealed nine objective responses, including six partial responses and three complete responses, with three cases of stable disease.
• Following surgery, 7 of 12 patients (58%) achieved a pathologic complete response, including all 3 hormone receptor–negative patients and 4 of the 9 hormone receptor–positive patients.

IN PRACTICE:

“Overall, the clinical data shown here demonstrate the effects of combining ERBB2 antibodies with IT [intratumoral] delivery of targeted cDC1 to enhance immune cell infiltration within the TME [tumor microenvironment] and subsequently induce tumor regression before chemotherapy,” wrote the authors, who noted they will be testing the higher dose for an ongoing phase 2 trial with an additional 41 patients.

SOURCE:

The study was led by Hyo S. Han, MD, of H. Lee Moffitt Cancer Center and Research Institute in Tampa, Florida. It was published online on December 5, 2024, in JAMA Oncology.

LIMITATIONS:

Because only two dose levels of cDC1 were tested, it remains unclear whether higher doses or different administration schedules could further enhance immune response. Additionally, the nonrandomized design prevents definitive conclusions about whether the clinical benefits were solely from the anti-ERBB2 antibodies. The small sample size also makes it difficult to determine if the pathologic complete responses were primarily due to the 12 weeks of trastuzumab/pertuzumab/paclitaxel treatment.

DISCLOSURES:

This study was funded by the Moffitt Breast Cancer Research Fund, Shula Fund, and Pennies in Action. Several authors reported research support and personal and consulting fees from US funding agencies and multiple pharmaceutical companies outside of the submitted work, as well as related intellectual property and patents.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication. A version of this article first appeared on Medscape.com.

Vaccines for treating and preventing cancer have long been considered a holy grail in oncology.

But aside from a few notable exceptions — including the human papillomavirus (HPV) vaccine, which has dramatically reduced the incidence of HPV-related cancers, and a Bacillus Calmette-Guerin vaccine, which helps prevent early-stage bladder cancer recurrence — most have failed to deliver.

Following a string of disappointments over the past decade, recent advances in the immunotherapy space are bringing renewed hope for progress.

In an American Association for Cancer Research (AACR) series earlier in 2024, Catherine J. Wu, MD, predicted big strides for cancer vaccines, especially for personalized vaccines that target patient-specific neoantigens — the proteins that form on cancer cells — as well as vaccines that can treat diverse tumor types.

“A focus on neoantigens that arise from driver mutations in different tumor types could allow us to make progress in creating off-the-shelf vaccines,” said Wu, the Lavine Family Chair of Preventative Cancer Therapies at Dana-Farber Cancer Institute and a professor of medicine at Harvard Medical School, both in Boston, Massachusetts.

A prime example is a personalized, messenger RNA (mRNA)–based vaccine designed to prevent melanoma recurrence. The mRNA-4157 vaccine encodes up to 34 different patient-specific neoantigens.

“This is one of the most exciting developments in modern cancer therapy,” said Lawrence Young, a virologist and professor of molecular oncology at the University of Warwick, Coventry, England, who commented on the investigational vaccine via the UK-based Science Media Centre.

Other promising options are on the horizon as well. In August, BioNTech announced a phase 1 global trial to study BNT116 — a vaccine to treat non–small cell lung cancer (NSCLC). BNT116, like mRNA-4157, targets specific antigens in the lung cancer cells.

“This technology is the next big phase of cancer treatment,” Siow Ming Lee, MD, a consultant medical oncologist at University College London Hospitals in England, which is leading the UK trial for the lung cancer and melanoma vaccines, told The Guardian. “We are now entering this very exciting new era of mRNA-based immunotherapy clinical trials to investigate the treatment of lung cancer.”

Still, these predictions have a familiar ring. While the prospects are exciting, delivering on them is another story. There are simply no guarantees these strategies will work as hoped.

 

Then: Where We Were

Cancer vaccine research began to ramp up in the 2000s, and in 2006, the first-generation HPV vaccine, Gardasil, was approved. Gardasil prevents infection from four strains of HPV that cause about 80% of cervical cancer cases.

In 2010, the Food and Drug Administration approved sipuleucel-T, the first therapeutic cancer vaccine, which improved overall survival in patients with hormone-refractory prostate cancer.

Researchers predicted this approval would “pave the way for developing innovative, next generation of vaccines with enhanced antitumor potency.”

In a 2015 AACR research forecast report, Drew Pardoll, MD, PhD, co-director of the Cancer Immunology and Hematopoiesis Program at Johns Hopkins University, Baltimore, Maryland, said that “we can expect to see encouraging results from studies using cancer vaccines.”

Despite the excitement surrounding cancer vaccines alongside a few successes, the next decade brought a longer string of late-phase disappointments.

In 2016, the phase 3 ACT IV trial of a therapeutic vaccine to treat glioblastoma multiforme (CDX-110) was terminated after it failed to demonstrate improved survival.

In 2017, a phase 3 trial of the therapeutic pancreatic cancer vaccine, GVAX, was stopped early for lack of efficacy.

That year, an attenuated Listeria monocytogenes vaccine to treat pancreatic cancer and mesothelioma also failed to come to fruition. In late 2017, concerns over listeria infections prompted Aduro Biotech to cancel its listeria-based cancer treatment program.

In 2018, a phase 3 trial of belagenpumatucel-L, a therapeutic NSCLC vaccine, failed to demonstrate a significant improvement in survival and further study was discontinued.

And in 2019, a vaccine targeting MAGE-A3, a cancer-testis antigen present in multiple tumor types, failed to meet endpoints for improved survival in a phase 3 trial, leading to discontinuation of the vaccine program.

But these disappointments and failures are normal parts of medical research and drug development and have allowed for incremental advances that helped fuel renewed interest and hope for cancer vaccines, when the timing was right, explained vaccine pioneer Larry W. Kwak, MD, PhD, deputy director of the Comprehensive Cancer Center at City of Hope, Duarte, California.

When it comes to vaccine progress, timing makes a difference. In 2011, Kwak and colleagues published promising phase 3 trial results on a personalized vaccine. The vaccine was a patient-specific tumor-derived antigen for patients with follicular lymphoma in their first remission following chemotherapy. Patients who received the vaccine demonstrated significantly longer disease-free survival.

But, at the time, personalized vaccines faced strong headwinds due, largely, to high costs, and commercial interest failed to materialize. “That’s been the major hurdle for a long time,” said Kwak.

Now, however, interest has returned alongside advances in technology and research. The big shift has been the emergence of lower-cost rapid-production mRNA and DNA platforms and a better understanding of how vaccines and potent immune stimulants, like checkpoint inhibitors, can work together to improve outcomes, he explained.

“The timing wasn’t right” back then, Kwak noted. “Now, it’s a different environment and a different time.”

 

A Turning Point?

Indeed, a decade later, cancer vaccine development appears to be headed in a more promising direction.

Among key cancer vaccines to watch is the mRNA-4157 vaccine, developed by Merck and Moderna, designed to prevent melanoma recurrence. In a recent phase 2 study, patients receiving the mRNA-4157 vaccine alongside pembrolizumab had nearly half the risk for melanoma recurrence or death at 3 years compared with those receiving pembrolizumab alone. Investigators are now evaluating the vaccine in a global phase 3 study in patients with high-risk, stage IIB to IV melanoma following surgery.

Another one to watch is the BNT116 NSCLC vaccine from BioNTech. This vaccine presents the immune system with NSCLC tumor markers to encourage the body to fight cancer cells expressing those markers while ignoring healthy cells. BioNTech also launched a global clinical trial for its vaccine this year.

Other notables include a pancreatic cancer mRNA vaccine, which has shown promising early results in a small trial of 16 patients. Of 16 patients who received the vaccine alongside chemotherapy and after surgery and immunotherapy, 8 responded. Of these eight, six remained recurrence free at 3 years. Investigators noted that the vaccine appeared to stimulate a durable T-cell response in patients who responded.

Kwak has also continued his work on lymphoma vaccines. In August, his team published promising first-in-human data on the use of personalized neoantigen vaccines as an early intervention in untreated patients with lymphoplasmacytic lymphoma. Among nine asymptomatic patients who received the vaccine, all achieved stable disease or better, with no dose-limiting toxicities. One patient had a minor response, and the median time to progression was greater than 72 months.

“The current setting is more for advanced disease,” Kwak explained. “It’s a tougher task, but combined with checkpoint blockade, it may be potent enough to work.” 

Still, caution is important. Despite early promise, it’s too soon to tell which, if any, of these investigational vaccines will pan out in the long run. Like investigational drugs, cancer vaccines may show big promising initially but then fail in larger trials.

One key to success, according to Kwak, is to design trials so that even negative results will inform next steps.

But, he noted, failures in large clinical trials will “put a chilling effect on cancer vaccine research again.”

“That’s what keeps me up at night,” he said. “We know the science is fundamentally sound and we have seen glimpses over decades of research that cancer vaccines can work, so it’s really just a matter of tweaking things to optimize trial design.”

Companies tend to design trials to test if a vaccine works or not, without trying to understand why, he said.

“What we need to do is design those so that we can learn from negative results,” he said. That’s what he and his colleagues attempted to do in their recent trial. “We didn’t just look at clinical results; we’re interrogating the actual tumor environment to understand what worked and didn’t and how to tweak that for the next trial.”

Kwak and his colleagues found, for instance, that the vaccine had a greater effect on B cell–derived tumor cells than on cells of plasma origin, so “the most rational design for the next iteration is to combine the vaccine with agents that work directly against plasma cells,” he explained.

As for what’s next, Kwak said: “We’re just focused on trying to do good science and understand. We’ve seen glimpses of success. That’s where we are.”

A version of this article first appeared on Medscape.com.

Vaccines for treating and preventing cancer have long been considered a holy grail in oncology.

But aside from a few notable exceptions — including the human papillomavirus (HPV) vaccine, which has dramatically reduced the incidence of HPV-related cancers, and a Bacillus Calmette-Guerin vaccine, which helps prevent early-stage bladder cancer recurrence — most have failed to deliver.

Following a string of disappointments over the past decade, recent advances in the immunotherapy space are bringing renewed hope for progress.

In an American Association for Cancer Research (AACR) series earlier in 2024, Catherine J. Wu, MD, predicted big strides for cancer vaccines, especially for personalized vaccines that target patient-specific neoantigens — the proteins that form on cancer cells — as well as vaccines that can treat diverse tumor types.

“A focus on neoantigens that arise from driver mutations in different tumor types could allow us to make progress in creating off-the-shelf vaccines,” said Wu, the Lavine Family Chair of Preventative Cancer Therapies at Dana-Farber Cancer Institute and a professor of medicine at Harvard Medical School, both in Boston, Massachusetts.

A prime example is a personalized, messenger RNA (mRNA)–based vaccine designed to prevent melanoma recurrence. The mRNA-4157 vaccine encodes up to 34 different patient-specific neoantigens.

“This is one of the most exciting developments in modern cancer therapy,” said Lawrence Young, a virologist and professor of molecular oncology at the University of Warwick, Coventry, England, who commented on the investigational vaccine via the UK-based Science Media Centre.

Other promising options are on the horizon as well. In August, BioNTech announced a phase 1 global trial to study BNT116 — a vaccine to treat non–small cell lung cancer (NSCLC). BNT116, like mRNA-4157, targets specific antigens in the lung cancer cells.

“This technology is the next big phase of cancer treatment,” Siow Ming Lee, MD, a consultant medical oncologist at University College London Hospitals in England, which is leading the UK trial for the lung cancer and melanoma vaccines, told The Guardian. “We are now entering this very exciting new era of mRNA-based immunotherapy clinical trials to investigate the treatment of lung cancer.”

Still, these predictions have a familiar ring. While the prospects are exciting, delivering on them is another story. There are simply no guarantees these strategies will work as hoped.

 

Then: Where We Were

Cancer vaccine research began to ramp up in the 2000s, and in 2006, the first-generation HPV vaccine, Gardasil, was approved. Gardasil prevents infection from four strains of HPV that cause about 80% of cervical cancer cases.

In 2010, the Food and Drug Administration approved sipuleucel-T, the first therapeutic cancer vaccine, which improved overall survival in patients with hormone-refractory prostate cancer.

Researchers predicted this approval would “pave the way for developing innovative, next generation of vaccines with enhanced antitumor potency.”

In a 2015 AACR research forecast report, Drew Pardoll, MD, PhD, co-director of the Cancer Immunology and Hematopoiesis Program at Johns Hopkins University, Baltimore, Maryland, said that “we can expect to see encouraging results from studies using cancer vaccines.”

Despite the excitement surrounding cancer vaccines alongside a few successes, the next decade brought a longer string of late-phase disappointments.

In 2016, the phase 3 ACT IV trial of a therapeutic vaccine to treat glioblastoma multiforme (CDX-110) was terminated after it failed to demonstrate improved survival.

In 2017, a phase 3 trial of the therapeutic pancreatic cancer vaccine, GVAX, was stopped early for lack of efficacy.

That year, an attenuated Listeria monocytogenes vaccine to treat pancreatic cancer and mesothelioma also failed to come to fruition. In late 2017, concerns over listeria infections prompted Aduro Biotech to cancel its listeria-based cancer treatment program.

In 2018, a phase 3 trial of belagenpumatucel-L, a therapeutic NSCLC vaccine, failed to demonstrate a significant improvement in survival and further study was discontinued.

And in 2019, a vaccine targeting MAGE-A3, a cancer-testis antigen present in multiple tumor types, failed to meet endpoints for improved survival in a phase 3 trial, leading to discontinuation of the vaccine program.

But these disappointments and failures are normal parts of medical research and drug development and have allowed for incremental advances that helped fuel renewed interest and hope for cancer vaccines, when the timing was right, explained vaccine pioneer Larry W. Kwak, MD, PhD, deputy director of the Comprehensive Cancer Center at City of Hope, Duarte, California.

When it comes to vaccine progress, timing makes a difference. In 2011, Kwak and colleagues published promising phase 3 trial results on a personalized vaccine. The vaccine was a patient-specific tumor-derived antigen for patients with follicular lymphoma in their first remission following chemotherapy. Patients who received the vaccine demonstrated significantly longer disease-free survival.

But, at the time, personalized vaccines faced strong headwinds due, largely, to high costs, and commercial interest failed to materialize. “That’s been the major hurdle for a long time,” said Kwak.

Now, however, interest has returned alongside advances in technology and research. The big shift has been the emergence of lower-cost rapid-production mRNA and DNA platforms and a better understanding of how vaccines and potent immune stimulants, like checkpoint inhibitors, can work together to improve outcomes, he explained.

“The timing wasn’t right” back then, Kwak noted. “Now, it’s a different environment and a different time.”

 

A Turning Point?

Indeed, a decade later, cancer vaccine development appears to be headed in a more promising direction.

Among key cancer vaccines to watch is the mRNA-4157 vaccine, developed by Merck and Moderna, designed to prevent melanoma recurrence. In a recent phase 2 study, patients receiving the mRNA-4157 vaccine alongside pembrolizumab had nearly half the risk for melanoma recurrence or death at 3 years compared with those receiving pembrolizumab alone. Investigators are now evaluating the vaccine in a global phase 3 study in patients with high-risk, stage IIB to IV melanoma following surgery.

Another one to watch is the BNT116 NSCLC vaccine from BioNTech. This vaccine presents the immune system with NSCLC tumor markers to encourage the body to fight cancer cells expressing those markers while ignoring healthy cells. BioNTech also launched a global clinical trial for its vaccine this year.

Other notables include a pancreatic cancer mRNA vaccine, which has shown promising early results in a small trial of 16 patients. Of 16 patients who received the vaccine alongside chemotherapy and after surgery and immunotherapy, 8 responded. Of these eight, six remained recurrence free at 3 years. Investigators noted that the vaccine appeared to stimulate a durable T-cell response in patients who responded.

Kwak has also continued his work on lymphoma vaccines. In August, his team published promising first-in-human data on the use of personalized neoantigen vaccines as an early intervention in untreated patients with lymphoplasmacytic lymphoma. Among nine asymptomatic patients who received the vaccine, all achieved stable disease or better, with no dose-limiting toxicities. One patient had a minor response, and the median time to progression was greater than 72 months.

“The current setting is more for advanced disease,” Kwak explained. “It’s a tougher task, but combined with checkpoint blockade, it may be potent enough to work.” 

Still, caution is important. Despite early promise, it’s too soon to tell which, if any, of these investigational vaccines will pan out in the long run. Like investigational drugs, cancer vaccines may show big promising initially but then fail in larger trials.

One key to success, according to Kwak, is to design trials so that even negative results will inform next steps.

But, he noted, failures in large clinical trials will “put a chilling effect on cancer vaccine research again.”

“That’s what keeps me up at night,” he said. “We know the science is fundamentally sound and we have seen glimpses over decades of research that cancer vaccines can work, so it’s really just a matter of tweaking things to optimize trial design.”

Companies tend to design trials to test if a vaccine works or not, without trying to understand why, he said.

“What we need to do is design those so that we can learn from negative results,” he said. That’s what he and his colleagues attempted to do in their recent trial. “We didn’t just look at clinical results; we’re interrogating the actual tumor environment to understand what worked and didn’t and how to tweak that for the next trial.”

Kwak and his colleagues found, for instance, that the vaccine had a greater effect on B cell–derived tumor cells than on cells of plasma origin, so “the most rational design for the next iteration is to combine the vaccine with agents that work directly against plasma cells,” he explained.

As for what’s next, Kwak said: “We’re just focused on trying to do good science and understand. We’ve seen glimpses of success. That’s where we are.”

A version of this article first appeared on Medscape.com.

Vaccines for treating and preventing cancer have long been considered a holy grail in oncology.

But aside from a few notable exceptions — including the human papillomavirus (HPV) vaccine, which has dramatically reduced the incidence of HPV-related cancers, and a Bacillus Calmette-Guerin vaccine, which helps prevent early-stage bladder cancer recurrence — most have failed to deliver.

Following a string of disappointments over the past decade, recent advances in the immunotherapy space are bringing renewed hope for progress.

In an American Association for Cancer Research (AACR) series earlier in 2024, Catherine J. Wu, MD, predicted big strides for cancer vaccines, especially for personalized vaccines that target patient-specific neoantigens — the proteins that form on cancer cells — as well as vaccines that can treat diverse tumor types.

“A focus on neoantigens that arise from driver mutations in different tumor types could allow us to make progress in creating off-the-shelf vaccines,” said Wu, the Lavine Family Chair of Preventative Cancer Therapies at Dana-Farber Cancer Institute and a professor of medicine at Harvard Medical School, both in Boston, Massachusetts.

A prime example is a personalized, messenger RNA (mRNA)–based vaccine designed to prevent melanoma recurrence. The mRNA-4157 vaccine encodes up to 34 different patient-specific neoantigens.

“This is one of the most exciting developments in modern cancer therapy,” said Lawrence Young, a virologist and professor of molecular oncology at the University of Warwick, Coventry, England, who commented on the investigational vaccine via the UK-based Science Media Centre.

Other promising options are on the horizon as well. In August, BioNTech announced a phase 1 global trial to study BNT116 — a vaccine to treat non–small cell lung cancer (NSCLC). BNT116, like mRNA-4157, targets specific antigens in the lung cancer cells.

“This technology is the next big phase of cancer treatment,” Siow Ming Lee, MD, a consultant medical oncologist at University College London Hospitals in England, which is leading the UK trial for the lung cancer and melanoma vaccines, told The Guardian. “We are now entering this very exciting new era of mRNA-based immunotherapy clinical trials to investigate the treatment of lung cancer.”

Still, these predictions have a familiar ring. While the prospects are exciting, delivering on them is another story. There are simply no guarantees these strategies will work as hoped.

 

Then: Where We Were

Cancer vaccine research began to ramp up in the 2000s, and in 2006, the first-generation HPV vaccine, Gardasil, was approved. Gardasil prevents infection from four strains of HPV that cause about 80% of cervical cancer cases.

In 2010, the Food and Drug Administration approved sipuleucel-T, the first therapeutic cancer vaccine, which improved overall survival in patients with hormone-refractory prostate cancer.

Researchers predicted this approval would “pave the way for developing innovative, next generation of vaccines with enhanced antitumor potency.”

In a 2015 AACR research forecast report, Drew Pardoll, MD, PhD, co-director of the Cancer Immunology and Hematopoiesis Program at Johns Hopkins University, Baltimore, Maryland, said that “we can expect to see encouraging results from studies using cancer vaccines.”

Despite the excitement surrounding cancer vaccines alongside a few successes, the next decade brought a longer string of late-phase disappointments.

In 2016, the phase 3 ACT IV trial of a therapeutic vaccine to treat glioblastoma multiforme (CDX-110) was terminated after it failed to demonstrate improved survival.

In 2017, a phase 3 trial of the therapeutic pancreatic cancer vaccine, GVAX, was stopped early for lack of efficacy.

That year, an attenuated Listeria monocytogenes vaccine to treat pancreatic cancer and mesothelioma also failed to come to fruition. In late 2017, concerns over listeria infections prompted Aduro Biotech to cancel its listeria-based cancer treatment program.

In 2018, a phase 3 trial of belagenpumatucel-L, a therapeutic NSCLC vaccine, failed to demonstrate a significant improvement in survival and further study was discontinued.

And in 2019, a vaccine targeting MAGE-A3, a cancer-testis antigen present in multiple tumor types, failed to meet endpoints for improved survival in a phase 3 trial, leading to discontinuation of the vaccine program.

But these disappointments and failures are normal parts of medical research and drug development and have allowed for incremental advances that helped fuel renewed interest and hope for cancer vaccines, when the timing was right, explained vaccine pioneer Larry W. Kwak, MD, PhD, deputy director of the Comprehensive Cancer Center at City of Hope, Duarte, California.

When it comes to vaccine progress, timing makes a difference. In 2011, Kwak and colleagues published promising phase 3 trial results on a personalized vaccine. The vaccine was a patient-specific tumor-derived antigen for patients with follicular lymphoma in their first remission following chemotherapy. Patients who received the vaccine demonstrated significantly longer disease-free survival.

But, at the time, personalized vaccines faced strong headwinds due, largely, to high costs, and commercial interest failed to materialize. “That’s been the major hurdle for a long time,” said Kwak.

Now, however, interest has returned alongside advances in technology and research. The big shift has been the emergence of lower-cost rapid-production mRNA and DNA platforms and a better understanding of how vaccines and potent immune stimulants, like checkpoint inhibitors, can work together to improve outcomes, he explained.

“The timing wasn’t right” back then, Kwak noted. “Now, it’s a different environment and a different time.”

 

A Turning Point?

Indeed, a decade later, cancer vaccine development appears to be headed in a more promising direction.

Among key cancer vaccines to watch is the mRNA-4157 vaccine, developed by Merck and Moderna, designed to prevent melanoma recurrence. In a recent phase 2 study, patients receiving the mRNA-4157 vaccine alongside pembrolizumab had nearly half the risk for melanoma recurrence or death at 3 years compared with those receiving pembrolizumab alone. Investigators are now evaluating the vaccine in a global phase 3 study in patients with high-risk, stage IIB to IV melanoma following surgery.

Another one to watch is the BNT116 NSCLC vaccine from BioNTech. This vaccine presents the immune system with NSCLC tumor markers to encourage the body to fight cancer cells expressing those markers while ignoring healthy cells. BioNTech also launched a global clinical trial for its vaccine this year.

Other notables include a pancreatic cancer mRNA vaccine, which has shown promising early results in a small trial of 16 patients. Of 16 patients who received the vaccine alongside chemotherapy and after surgery and immunotherapy, 8 responded. Of these eight, six remained recurrence free at 3 years. Investigators noted that the vaccine appeared to stimulate a durable T-cell response in patients who responded.

Kwak has also continued his work on lymphoma vaccines. In August, his team published promising first-in-human data on the use of personalized neoantigen vaccines as an early intervention in untreated patients with lymphoplasmacytic lymphoma. Among nine asymptomatic patients who received the vaccine, all achieved stable disease or better, with no dose-limiting toxicities. One patient had a minor response, and the median time to progression was greater than 72 months.

“The current setting is more for advanced disease,” Kwak explained. “It’s a tougher task, but combined with checkpoint blockade, it may be potent enough to work.” 

Still, caution is important. Despite early promise, it’s too soon to tell which, if any, of these investigational vaccines will pan out in the long run. Like investigational drugs, cancer vaccines may show big promising initially but then fail in larger trials.

One key to success, according to Kwak, is to design trials so that even negative results will inform next steps.

But, he noted, failures in large clinical trials will “put a chilling effect on cancer vaccine research again.”

“That’s what keeps me up at night,” he said. “We know the science is fundamentally sound and we have seen glimpses over decades of research that cancer vaccines can work, so it’s really just a matter of tweaking things to optimize trial design.”

Companies tend to design trials to test if a vaccine works or not, without trying to understand why, he said.

“What we need to do is design those so that we can learn from negative results,” he said. That’s what he and his colleagues attempted to do in their recent trial. “We didn’t just look at clinical results; we’re interrogating the actual tumor environment to understand what worked and didn’t and how to tweak that for the next trial.”

Kwak and his colleagues found, for instance, that the vaccine had a greater effect on B cell–derived tumor cells than on cells of plasma origin, so “the most rational design for the next iteration is to combine the vaccine with agents that work directly against plasma cells,” he explained.

As for what’s next, Kwak said: “We’re just focused on trying to do good science and understand. We’ve seen glimpses of success. That’s where we are.”

A version of this article first appeared on Medscape.com.