Federal Practitioner

Infections were the most common adverse events associated with the use of microwave energy devices for hyperhidrosis over a 9-year period, an analysis of reports submitted to the FDA Manufacturer and User Facility Device Experience (MAUDE) database showed.

While microwave energy devices (MEDs) are used to treat hyperhidrosis, the largest MED clinical trial included only 101 patients, Samantha Jo Albucker and Shari Lipner, MD, PhD, wrote in a research letter reporting the results.

For the study, published online in the Journal of the American Academy of Dermatology, Ms. Albucker, a student at Tulane University, New Orleans, and Dr. Lipner, associate professor of clinical dermatology at Weill Cornell Medicine, New York, searched the MAUDE database between Feb. 28, 2013, and Dec. 29, 2022, for adverse events (AEs) involving MEDs for hyperhidrosis treatment. Of the 502 medical device reports identified over the study period, the axilla was the most frequent injury site in 50.4% of cases. The three most common complications were infections (45.4%); neurological symptoms including neuropathy, nerve damage, and numbness (21.7%); and burns/ulcerations/erosions (19.1%).

In other findings, 2.4% of patients required hospitalization, most often because of infection (83.3%), followed by burn and coma (8.3% each). The average symptom onset was 2 months postprocedure, and the most common treatment was antibiotics in 62.2% of cases, followed by incision and drainage/aspiration in 21.7% of cases.

A codiagnosis of hidradenitis suppurativa (HS) was reported in 5.4% of all medical device reports. The researchers noted that in a published randomized clinical trial of eight HS patients undergoing MED treatment to assess the effect on HS symptoms, the treatment showed no clinical advantage. In addition, they referred to two case reports describing new-onset HS after MED treatment for hyperhidrosis.

“Therefore, we recommend questioning patients about HS history and examining for HS clinical findings before performing MED for hyperhidrosis,” they wrote, adding that the data, “taken together, suggests that avoidance of MED treatment of hyperhidrosis in HS patients is prudent and alternative treatments may be prescribed.”

The researchers acknowledged certain limitations of their analysis, including uncompleted medical device reports, patient reporting, and unverified causes of adverse events. “Large multicenter studies are needed to corroborate our results,” they concluded.

Adam Friedman, MD, professor and chair of dermatology at George Washington University, Washington, who was asked to comment on the study, said that primary idiopathic hyperhidrosis is a common medical condition that is often overlooked as a legitimate concern, and causes a quality-of-life burden. “Even with the striking numbers in the millions, there are limited treatment options available for axillary let alone other forms of primary hyperhidrosis,” said Dr. Friedman, who was not involved with the study.

“Therefore, for the short treatment list we have, it is important to have some predictive power with respect to clinical impact to provide realistic expectations as well as potential adverse events to ensure best practices and meaningful patient guidance. In this research letter, our colleagues highlight complications that can ensue from microwave therapy for hyperhidrosis and the frequency of said adverse events. Knowing these data is half the battle, and I for one would not have assumed infection was number one on the list of adverse events.”

Ms. Albucker had no relevant conflicts of interest to disclose. Dr. Lipner disclosed that she has served as a consultant for Ortho Dermatologics, Hoth Therapeutics, BelleTorus Corporation, and Moberg Pharmaceuticals.

Dr. Friedman disclosed that he is a consultant and/or advisory board member for Medscape/SanovaWorks, Oakstone Institute, L’Oréal, La Roche Posay, Galderma, Aveeno, Ortho Dermatologic, Microcures, Pfizer, Novartis, Lilly, Hoth Therapeutics, Zylo Therapeutics, BMS, Vial, Janssen, Novocure, Dermavant, Regeneron/Sanofi, and Incyte. He has also received grants from Pfizer, the Dermatology Foundation, Lilly, Janssen, Incyte, and Galderma.

Infections were the most common adverse events associated with the use of microwave energy devices for hyperhidrosis over a 9-year period, an analysis of reports submitted to the FDA Manufacturer and User Facility Device Experience (MAUDE) database showed.

While microwave energy devices (MEDs) are used to treat hyperhidrosis, the largest MED clinical trial included only 101 patients, Samantha Jo Albucker and Shari Lipner, MD, PhD, wrote in a research letter reporting the results.

For the study, published online in the Journal of the American Academy of Dermatology, Ms. Albucker, a student at Tulane University, New Orleans, and Dr. Lipner, associate professor of clinical dermatology at Weill Cornell Medicine, New York, searched the MAUDE database between Feb. 28, 2013, and Dec. 29, 2022, for adverse events (AEs) involving MEDs for hyperhidrosis treatment. Of the 502 medical device reports identified over the study period, the axilla was the most frequent injury site in 50.4% of cases. The three most common complications were infections (45.4%); neurological symptoms including neuropathy, nerve damage, and numbness (21.7%); and burns/ulcerations/erosions (19.1%).

In other findings, 2.4% of patients required hospitalization, most often because of infection (83.3%), followed by burn and coma (8.3% each). The average symptom onset was 2 months postprocedure, and the most common treatment was antibiotics in 62.2% of cases, followed by incision and drainage/aspiration in 21.7% of cases.

A codiagnosis of hidradenitis suppurativa (HS) was reported in 5.4% of all medical device reports. The researchers noted that in a published randomized clinical trial of eight HS patients undergoing MED treatment to assess the effect on HS symptoms, the treatment showed no clinical advantage. In addition, they referred to two case reports describing new-onset HS after MED treatment for hyperhidrosis.

“Therefore, we recommend questioning patients about HS history and examining for HS clinical findings before performing MED for hyperhidrosis,” they wrote, adding that the data, “taken together, suggests that avoidance of MED treatment of hyperhidrosis in HS patients is prudent and alternative treatments may be prescribed.”

The researchers acknowledged certain limitations of their analysis, including uncompleted medical device reports, patient reporting, and unverified causes of adverse events. “Large multicenter studies are needed to corroborate our results,” they concluded.

Adam Friedman, MD, professor and chair of dermatology at George Washington University, Washington, who was asked to comment on the study, said that primary idiopathic hyperhidrosis is a common medical condition that is often overlooked as a legitimate concern, and causes a quality-of-life burden. “Even with the striking numbers in the millions, there are limited treatment options available for axillary let alone other forms of primary hyperhidrosis,” said Dr. Friedman, who was not involved with the study.

“Therefore, for the short treatment list we have, it is important to have some predictive power with respect to clinical impact to provide realistic expectations as well as potential adverse events to ensure best practices and meaningful patient guidance. In this research letter, our colleagues highlight complications that can ensue from microwave therapy for hyperhidrosis and the frequency of said adverse events. Knowing these data is half the battle, and I for one would not have assumed infection was number one on the list of adverse events.”

Ms. Albucker had no relevant conflicts of interest to disclose. Dr. Lipner disclosed that she has served as a consultant for Ortho Dermatologics, Hoth Therapeutics, BelleTorus Corporation, and Moberg Pharmaceuticals.

Dr. Friedman disclosed that he is a consultant and/or advisory board member for Medscape/SanovaWorks, Oakstone Institute, L’Oréal, La Roche Posay, Galderma, Aveeno, Ortho Dermatologic, Microcures, Pfizer, Novartis, Lilly, Hoth Therapeutics, Zylo Therapeutics, BMS, Vial, Janssen, Novocure, Dermavant, Regeneron/Sanofi, and Incyte. He has also received grants from Pfizer, the Dermatology Foundation, Lilly, Janssen, Incyte, and Galderma.

Infections were the most common adverse events associated with the use of microwave energy devices for hyperhidrosis over a 9-year period, an analysis of reports submitted to the FDA Manufacturer and User Facility Device Experience (MAUDE) database showed.

While microwave energy devices (MEDs) are used to treat hyperhidrosis, the largest MED clinical trial included only 101 patients, Samantha Jo Albucker and Shari Lipner, MD, PhD, wrote in a research letter reporting the results.

For the study, published online in the Journal of the American Academy of Dermatology, Ms. Albucker, a student at Tulane University, New Orleans, and Dr. Lipner, associate professor of clinical dermatology at Weill Cornell Medicine, New York, searched the MAUDE database between Feb. 28, 2013, and Dec. 29, 2022, for adverse events (AEs) involving MEDs for hyperhidrosis treatment. Of the 502 medical device reports identified over the study period, the axilla was the most frequent injury site in 50.4% of cases. The three most common complications were infections (45.4%); neurological symptoms including neuropathy, nerve damage, and numbness (21.7%); and burns/ulcerations/erosions (19.1%).

In other findings, 2.4% of patients required hospitalization, most often because of infection (83.3%), followed by burn and coma (8.3% each). The average symptom onset was 2 months postprocedure, and the most common treatment was antibiotics in 62.2% of cases, followed by incision and drainage/aspiration in 21.7% of cases.

A codiagnosis of hidradenitis suppurativa (HS) was reported in 5.4% of all medical device reports. The researchers noted that in a published randomized clinical trial of eight HS patients undergoing MED treatment to assess the effect on HS symptoms, the treatment showed no clinical advantage. In addition, they referred to two case reports describing new-onset HS after MED treatment for hyperhidrosis.

“Therefore, we recommend questioning patients about HS history and examining for HS clinical findings before performing MED for hyperhidrosis,” they wrote, adding that the data, “taken together, suggests that avoidance of MED treatment of hyperhidrosis in HS patients is prudent and alternative treatments may be prescribed.”

The researchers acknowledged certain limitations of their analysis, including uncompleted medical device reports, patient reporting, and unverified causes of adverse events. “Large multicenter studies are needed to corroborate our results,” they concluded.

Adam Friedman, MD, professor and chair of dermatology at George Washington University, Washington, who was asked to comment on the study, said that primary idiopathic hyperhidrosis is a common medical condition that is often overlooked as a legitimate concern, and causes a quality-of-life burden. “Even with the striking numbers in the millions, there are limited treatment options available for axillary let alone other forms of primary hyperhidrosis,” said Dr. Friedman, who was not involved with the study.

“Therefore, for the short treatment list we have, it is important to have some predictive power with respect to clinical impact to provide realistic expectations as well as potential adverse events to ensure best practices and meaningful patient guidance. In this research letter, our colleagues highlight complications that can ensue from microwave therapy for hyperhidrosis and the frequency of said adverse events. Knowing these data is half the battle, and I for one would not have assumed infection was number one on the list of adverse events.”

Ms. Albucker had no relevant conflicts of interest to disclose. Dr. Lipner disclosed that she has served as a consultant for Ortho Dermatologics, Hoth Therapeutics, BelleTorus Corporation, and Moberg Pharmaceuticals.

Dr. Friedman disclosed that he is a consultant and/or advisory board member for Medscape/SanovaWorks, Oakstone Institute, L’Oréal, La Roche Posay, Galderma, Aveeno, Ortho Dermatologic, Microcures, Pfizer, Novartis, Lilly, Hoth Therapeutics, Zylo Therapeutics, BMS, Vial, Janssen, Novocure, Dermavant, Regeneron/Sanofi, and Incyte. He has also received grants from Pfizer, the Dermatology Foundation, Lilly, Janssen, Incyte, and Galderma.

Medscape’s Editor-in-Chief Eric Topol, MD, referred to continual noninvasive, cuffless, accurate blood pressure devices as “a holy grail in sensor technology.”

He personally tested a cuff-calibrated, over-the-counter device available in Europe that claims to monitor daily blood pressure changes and produce data that can help physicians titrate medications.

Dr. Topol does not believe that it is ready for prime time. Yes, cuffless devices are easy to use, and generate lots of data. But are those data accurate?

Many experts say not yet, even as the market continues to grow and more devices are introduced and highlighted at high-profile consumer events.
 

Burned before

Limitations of cuffed devices are well known, including errors related to cuff size, patient positioning, patient habits or behaviors (for example, caffeine/nicotine use, acute meal digestion, full bladder, very recent physical activity) and clinicians’ failure to take accurate measurements.

But are the currently available cuffless devices acceptable substitutes?

Like many clinicians, Timothy B. Plante, MD, MHS, assistant professor at the University of Vermont Medical Center thrombosis & hemostasis program in Burlington, is very excited about cuffless technology. However, “we’ve been burned by it before,” he said in an interview.

Dr. Plante’s 2016 validation study of an instant blood pressure smartphone app found that its measurements were “highly inaccurate,” with such low sensitivity that more than three-quarters of individuals with hypertensive blood levels would be falsely reassured that their blood pressure was in the normal range.

His team’s 2023 review of the current landscape, which includes more sophisticated devices, concluded that accuracy remains an issue: “Unfortunately, the pace of regulation of these devices has failed to match the speed of innovation and direct availability to patient consumers. There is an urgent need to develop a consensus on standards by which cuffless BP devices can be tested for accuracy.”
 

Devices, indications differ

Cuffless devices estimate blood pressure indirectly. Most operate based on pulse wave analysis and pulse arrival time (PWA-PAT), explained Ramakrishna Mukkamala, PhD, in a commentary. Dr. Mukkamala is a professor in the departments of bioengineering and anesthesiology and perioperative medicine at the University of Pittsburgh.

PWA involves measuring a peripheral arterial waveform using an optical sensor such as the green lights on the back of a wrist-worn device, or a ‘force sensor’ such as a finger cuff or pressing on a smartphone. Certain features are extracted from the waveform using machine learning and calibrated to blood pressure values.

PAT techniques work together with PWA; they record the ECG and extract features from that signal as well as the arterial waveform for calibration to blood pressure values.

The algorithm used to generate the BP numbers comprises a proprietary baseline model that may include demographics and other patient characteristics. A cuff measurement is often part of the baseline model because most cuffless devices require periodic (typically weekly or monthly) calibration using a cuffed device.

Cuffless devices that require cuff calibration compare the estimate they get to the cuff-calibrated number. In this scenario, the cuffless device may come up with the same blood pressure numbers simply because the baseline model – which is made up of thousands of data points relevant to the patient – has not changed.

This has led some experts to question whether PWA-PAT cuffless device readings actually add anything to the baseline model.

They don’t, according to Microsoft Research in what Dr. Mukkamala and coauthors referred to (in a review published in Hypertension) as “a complex article describing perhaps the most important and highest resource project to date (Aurora Project) on assessing the accuracy of PWA and PWA devices.”

The Microsoft article was written for bioengineers. The review in Hypertension explains the project for clinicians, and concludes that, “Cuffless BP devices based on PWA and PWA-PAT, which are similar to some regulatory-cleared devices, were of no additional value in measuring auscultatory or 24-hour ambulatory cuff BP when compared with a baseline model in which BP was predicted without an actual measurement.”
 

IEEE and FDA validation

Despite these concerns, several cuffless devices using PWA and PAT have been cleared by the Food and Drug Administration.

Validating cuffless devices is no simple matter. The Institute of Electrical and Electronics Engineers published a validation protocol for cuffless blood pressure devices in 2014 that was amended in 2019 to include a requirement to evaluate performance in different positions and in the presence of motion with varying degrees of noise artifact.

However, Daichi Shimbo, MD, codirector of the Columbia Hypertension Center in New York and vice chair of the American Heart Association Statement on blood pressure monitoring, and colleagues point out limitations, even in the updated standard. These include not requiring evaluation for drift over time; lack of specific dynamic testing protocols for stressors such as exercise or environmental temperatures; and an unsuitable reference standard (oscillometric cuff-based devices) during movement.

Dr. Shimbo said in an interview that, although he is excited about them, “these cuffless devices are not aligned with regulatory bodies. If a device gives someone a wrong blood pressure, they might be diagnosed with hypertension when they don’t have it or might miss the fact that they’re hypertensive because they get a normal blood pressure reading. If there’s no yardstick by which you say these devices are good, what are we really doing – helping, or causing a problem?”

“The specifics of how a device estimates blood pressure can determine what testing is needed to ensure that it is providing accurate performance in the intended conditions of use,” Jeremy Kahn, an FDA press officer, said in an interview. “For example, for cuffless devices that are calibrated initially with a cuff-based blood pressure device, the cuffless device needs to specify the period over which it can provide accurate readings and have testing to demonstrate that it provides accurate results over that period of use.”

The FDA said its testing is different from what the Microsoft Aurora Project used in their study.

“The intent of that testing, as the agency understands it, is to evaluate whether the device is providing useful input based on the current physiology of the patient rather than relying on predetermined values based on calibration or patient attributes. We evaluate this clinically in two separate tests: an induced change in blood pressure test and tracking of natural blood pressure changes with longer term device use,” Mr. Kahn explained.

Analyzing a device’s performance on individuals who have had natural changes in blood pressure as compared to a calibration value or initial reading “can also help discern if the device is using physiological data from the patient to determine their blood pressure accurately,” he said.

Experts interviewed for this article who remain skeptical about cuffless BP monitoring question whether the numbers that appear during the induced blood pressure change, and with the natural blood pressure changes that may occur over time, accurately reflect a patient’s blood pressure.

“The FDA doesn’t approve these devices; they clear them,” Dr. Shimbo pointed out. “Clearing them means they can be sold to the general public in the U.S. It’s not a strong statement that they’re accurate.”
 

Moving toward validation, standards

Ultimately, cuffless BP monitors may require more than one validation protocol and standard, depending on their technology, how and where they will be used, and by whom.

And as Dr. Plante and colleagues write, “Importantly, validation should be performed in diverse and special populations, including pregnant women and individuals across a range of heart rates, skin tones, wrist sizes, common arrhythmias, and beta-blocker use.”

Organizations that might be expected to help move validation and standards forward have mostly remained silent. The American Medical Association’s US Blood Pressure Validated Device Listing website includes only cuffed devices, as does the website of the international scientific nonprofit STRIDE BP.

The European Society of Hypertension 2022 consensus statement on cuffless devices concluded that, until there is an internationally accepted accuracy standard and the devices have been tested in healthy people and those with suspected or diagnosed hypertension, “cuffless BP devices should not be used for the evaluation or management of hypertension in clinical practice.”

This month, ESH published recommendations for “specific, clinically meaningful, and pragmatic validation procedures for different types of intermittent cuffless devices” that will be presented at their upcoming annual meeting June 26.

Updated protocols from IEEE “are coming out soon,” according to Dr. Shimbo. The FDA says currently cleared devices won’t need to revalidate according to new standards unless the sponsor makes significant modifications in software algorithms, device hardware, or targeted patient populations.
 

Device makers take the initiative

In the face of conflicting reports on accuracy and lack of a robust standard, some device makers are publishing their own tests or encouraging validation by potential customers.

For example, institutions that are considering using the Biobeat cuffless blood pressure monitor watch “usually start with small pilots with our devices to do internal validation,” Lior Ben Shettrit, the company’s vice president of business development, said in an interview. “Only after they complete the internal validation are they willing to move forward to full implementation.”

Cardiologist Dean Nachman, MD, is leading validation studies of the Biobeat device at the Hadassah Ein Kerem Medical Center in Jerusalem. For the first validation, the team recruited 1,057 volunteers who did a single blood pressure measurement with the cuffless device and with a cuffed device.

“We found 96.3% agreement in identifying hypertension and an interclass correlation coefficient of 0.99 and 0.97 for systolic and diastolic measurements, respectively,” he said. “Then we took it to the next level and compared the device to ambulatory 24-hour blood pressure monitoring and found comparable measurements.”

The investigators are not done yet. “We need data from thousands of patients, with subgroups, to not have any concerns,” he says. “Right now, we are using the device as a general monitor – as an EKG plus heart rate plus oxygen saturation level monitor – and as a blood pressure monitor for 24-hour blood pressure monitoring.”

The developers of the Aktiia device, which is the one Dr. Topol tested, take a different perspective. “When somebody introduces a new technology that is disrupting something that has been in place for over 100 years, there will always be some grumblings, ruffling of feathers, people saying it’s not ready, it’s not ready, it’s not ready,” Aktiia’s chief medical officer Jay Shah, MD, noted.

“But a lot of those comments are coming from the isolation of an ivory tower,” he said.

Aktiia cofounder and chief technology officer Josep Solà said that “no device is probably as accurate as if you have an invasive catheter,” adding that “we engage patients to look at their blood pressure day by day. … If each individual measurement of each of those patient is slightly less accurate than a cuff, who cares? We have 40 measurements per day on each patient. The accuracy and precision of each of those is good.”

Researchers from the George Institute for Global Health recently compared the Aktiia device to conventional ambulatory monitoring in 41 patients and found that “it did not accurately track night-time BP decline and results suggested it was unable to track medication-induced BP changes.”

“In the context of 24/7 monitoring of hypertensive patients,” Mr. Solà said, “whatever you do, if it’s better than a sham device or a baseline model and you track the blood pressure changes, it’s a hundred times much better than doing nothing.”

Dr. Nachman and Dr. Plante reported no relevant financial relationships. Dr. Shimbo reported that he received funding from NIH and has consulted for Abbott Vascular, Edward Lifesciences, Medtronic, and Tryton Medical.

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

Medscape’s Editor-in-Chief Eric Topol, MD, referred to continual noninvasive, cuffless, accurate blood pressure devices as “a holy grail in sensor technology.”

He personally tested a cuff-calibrated, over-the-counter device available in Europe that claims to monitor daily blood pressure changes and produce data that can help physicians titrate medications.

Dr. Topol does not believe that it is ready for prime time. Yes, cuffless devices are easy to use, and generate lots of data. But are those data accurate?

Many experts say not yet, even as the market continues to grow and more devices are introduced and highlighted at high-profile consumer events.
 

Burned before

Limitations of cuffed devices are well known, including errors related to cuff size, patient positioning, patient habits or behaviors (for example, caffeine/nicotine use, acute meal digestion, full bladder, very recent physical activity) and clinicians’ failure to take accurate measurements.

But are the currently available cuffless devices acceptable substitutes?

Like many clinicians, Timothy B. Plante, MD, MHS, assistant professor at the University of Vermont Medical Center thrombosis & hemostasis program in Burlington, is very excited about cuffless technology. However, “we’ve been burned by it before,” he said in an interview.

Dr. Plante’s 2016 validation study of an instant blood pressure smartphone app found that its measurements were “highly inaccurate,” with such low sensitivity that more than three-quarters of individuals with hypertensive blood levels would be falsely reassured that their blood pressure was in the normal range.

His team’s 2023 review of the current landscape, which includes more sophisticated devices, concluded that accuracy remains an issue: “Unfortunately, the pace of regulation of these devices has failed to match the speed of innovation and direct availability to patient consumers. There is an urgent need to develop a consensus on standards by which cuffless BP devices can be tested for accuracy.”
 

Devices, indications differ

Cuffless devices estimate blood pressure indirectly. Most operate based on pulse wave analysis and pulse arrival time (PWA-PAT), explained Ramakrishna Mukkamala, PhD, in a commentary. Dr. Mukkamala is a professor in the departments of bioengineering and anesthesiology and perioperative medicine at the University of Pittsburgh.

PWA involves measuring a peripheral arterial waveform using an optical sensor such as the green lights on the back of a wrist-worn device, or a ‘force sensor’ such as a finger cuff or pressing on a smartphone. Certain features are extracted from the waveform using machine learning and calibrated to blood pressure values.

PAT techniques work together with PWA; they record the ECG and extract features from that signal as well as the arterial waveform for calibration to blood pressure values.

The algorithm used to generate the BP numbers comprises a proprietary baseline model that may include demographics and other patient characteristics. A cuff measurement is often part of the baseline model because most cuffless devices require periodic (typically weekly or monthly) calibration using a cuffed device.

Cuffless devices that require cuff calibration compare the estimate they get to the cuff-calibrated number. In this scenario, the cuffless device may come up with the same blood pressure numbers simply because the baseline model – which is made up of thousands of data points relevant to the patient – has not changed.

This has led some experts to question whether PWA-PAT cuffless device readings actually add anything to the baseline model.

They don’t, according to Microsoft Research in what Dr. Mukkamala and coauthors referred to (in a review published in Hypertension) as “a complex article describing perhaps the most important and highest resource project to date (Aurora Project) on assessing the accuracy of PWA and PWA devices.”

The Microsoft article was written for bioengineers. The review in Hypertension explains the project for clinicians, and concludes that, “Cuffless BP devices based on PWA and PWA-PAT, which are similar to some regulatory-cleared devices, were of no additional value in measuring auscultatory or 24-hour ambulatory cuff BP when compared with a baseline model in which BP was predicted without an actual measurement.”
 

IEEE and FDA validation

Despite these concerns, several cuffless devices using PWA and PAT have been cleared by the Food and Drug Administration.

Validating cuffless devices is no simple matter. The Institute of Electrical and Electronics Engineers published a validation protocol for cuffless blood pressure devices in 2014 that was amended in 2019 to include a requirement to evaluate performance in different positions and in the presence of motion with varying degrees of noise artifact.

However, Daichi Shimbo, MD, codirector of the Columbia Hypertension Center in New York and vice chair of the American Heart Association Statement on blood pressure monitoring, and colleagues point out limitations, even in the updated standard. These include not requiring evaluation for drift over time; lack of specific dynamic testing protocols for stressors such as exercise or environmental temperatures; and an unsuitable reference standard (oscillometric cuff-based devices) during movement.

Dr. Shimbo said in an interview that, although he is excited about them, “these cuffless devices are not aligned with regulatory bodies. If a device gives someone a wrong blood pressure, they might be diagnosed with hypertension when they don’t have it or might miss the fact that they’re hypertensive because they get a normal blood pressure reading. If there’s no yardstick by which you say these devices are good, what are we really doing – helping, or causing a problem?”

“The specifics of how a device estimates blood pressure can determine what testing is needed to ensure that it is providing accurate performance in the intended conditions of use,” Jeremy Kahn, an FDA press officer, said in an interview. “For example, for cuffless devices that are calibrated initially with a cuff-based blood pressure device, the cuffless device needs to specify the period over which it can provide accurate readings and have testing to demonstrate that it provides accurate results over that period of use.”

The FDA said its testing is different from what the Microsoft Aurora Project used in their study.

“The intent of that testing, as the agency understands it, is to evaluate whether the device is providing useful input based on the current physiology of the patient rather than relying on predetermined values based on calibration or patient attributes. We evaluate this clinically in two separate tests: an induced change in blood pressure test and tracking of natural blood pressure changes with longer term device use,” Mr. Kahn explained.

Analyzing a device’s performance on individuals who have had natural changes in blood pressure as compared to a calibration value or initial reading “can also help discern if the device is using physiological data from the patient to determine their blood pressure accurately,” he said.

Experts interviewed for this article who remain skeptical about cuffless BP monitoring question whether the numbers that appear during the induced blood pressure change, and with the natural blood pressure changes that may occur over time, accurately reflect a patient’s blood pressure.

“The FDA doesn’t approve these devices; they clear them,” Dr. Shimbo pointed out. “Clearing them means they can be sold to the general public in the U.S. It’s not a strong statement that they’re accurate.”
 

Moving toward validation, standards

Ultimately, cuffless BP monitors may require more than one validation protocol and standard, depending on their technology, how and where they will be used, and by whom.

And as Dr. Plante and colleagues write, “Importantly, validation should be performed in diverse and special populations, including pregnant women and individuals across a range of heart rates, skin tones, wrist sizes, common arrhythmias, and beta-blocker use.”

Organizations that might be expected to help move validation and standards forward have mostly remained silent. The American Medical Association’s US Blood Pressure Validated Device Listing website includes only cuffed devices, as does the website of the international scientific nonprofit STRIDE BP.

The European Society of Hypertension 2022 consensus statement on cuffless devices concluded that, until there is an internationally accepted accuracy standard and the devices have been tested in healthy people and those with suspected or diagnosed hypertension, “cuffless BP devices should not be used for the evaluation or management of hypertension in clinical practice.”

This month, ESH published recommendations for “specific, clinically meaningful, and pragmatic validation procedures for different types of intermittent cuffless devices” that will be presented at their upcoming annual meeting June 26.

Updated protocols from IEEE “are coming out soon,” according to Dr. Shimbo. The FDA says currently cleared devices won’t need to revalidate according to new standards unless the sponsor makes significant modifications in software algorithms, device hardware, or targeted patient populations.
 

Device makers take the initiative

In the face of conflicting reports on accuracy and lack of a robust standard, some device makers are publishing their own tests or encouraging validation by potential customers.

For example, institutions that are considering using the Biobeat cuffless blood pressure monitor watch “usually start with small pilots with our devices to do internal validation,” Lior Ben Shettrit, the company’s vice president of business development, said in an interview. “Only after they complete the internal validation are they willing to move forward to full implementation.”

Cardiologist Dean Nachman, MD, is leading validation studies of the Biobeat device at the Hadassah Ein Kerem Medical Center in Jerusalem. For the first validation, the team recruited 1,057 volunteers who did a single blood pressure measurement with the cuffless device and with a cuffed device.

“We found 96.3% agreement in identifying hypertension and an interclass correlation coefficient of 0.99 and 0.97 for systolic and diastolic measurements, respectively,” he said. “Then we took it to the next level and compared the device to ambulatory 24-hour blood pressure monitoring and found comparable measurements.”

The investigators are not done yet. “We need data from thousands of patients, with subgroups, to not have any concerns,” he says. “Right now, we are using the device as a general monitor – as an EKG plus heart rate plus oxygen saturation level monitor – and as a blood pressure monitor for 24-hour blood pressure monitoring.”

The developers of the Aktiia device, which is the one Dr. Topol tested, take a different perspective. “When somebody introduces a new technology that is disrupting something that has been in place for over 100 years, there will always be some grumblings, ruffling of feathers, people saying it’s not ready, it’s not ready, it’s not ready,” Aktiia’s chief medical officer Jay Shah, MD, noted.

“But a lot of those comments are coming from the isolation of an ivory tower,” he said.

Aktiia cofounder and chief technology officer Josep Solà said that “no device is probably as accurate as if you have an invasive catheter,” adding that “we engage patients to look at their blood pressure day by day. … If each individual measurement of each of those patient is slightly less accurate than a cuff, who cares? We have 40 measurements per day on each patient. The accuracy and precision of each of those is good.”

Researchers from the George Institute for Global Health recently compared the Aktiia device to conventional ambulatory monitoring in 41 patients and found that “it did not accurately track night-time BP decline and results suggested it was unable to track medication-induced BP changes.”

“In the context of 24/7 monitoring of hypertensive patients,” Mr. Solà said, “whatever you do, if it’s better than a sham device or a baseline model and you track the blood pressure changes, it’s a hundred times much better than doing nothing.”

Dr. Nachman and Dr. Plante reported no relevant financial relationships. Dr. Shimbo reported that he received funding from NIH and has consulted for Abbott Vascular, Edward Lifesciences, Medtronic, and Tryton Medical.

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

Medscape’s Editor-in-Chief Eric Topol, MD, referred to continual noninvasive, cuffless, accurate blood pressure devices as “a holy grail in sensor technology.”

He personally tested a cuff-calibrated, over-the-counter device available in Europe that claims to monitor daily blood pressure changes and produce data that can help physicians titrate medications.

Dr. Topol does not believe that it is ready for prime time. Yes, cuffless devices are easy to use, and generate lots of data. But are those data accurate?

Many experts say not yet, even as the market continues to grow and more devices are introduced and highlighted at high-profile consumer events.
 

Burned before

Limitations of cuffed devices are well known, including errors related to cuff size, patient positioning, patient habits or behaviors (for example, caffeine/nicotine use, acute meal digestion, full bladder, very recent physical activity) and clinicians’ failure to take accurate measurements.

But are the currently available cuffless devices acceptable substitutes?

Like many clinicians, Timothy B. Plante, MD, MHS, assistant professor at the University of Vermont Medical Center thrombosis & hemostasis program in Burlington, is very excited about cuffless technology. However, “we’ve been burned by it before,” he said in an interview.

Dr. Plante’s 2016 validation study of an instant blood pressure smartphone app found that its measurements were “highly inaccurate,” with such low sensitivity that more than three-quarters of individuals with hypertensive blood levels would be falsely reassured that their blood pressure was in the normal range.

His team’s 2023 review of the current landscape, which includes more sophisticated devices, concluded that accuracy remains an issue: “Unfortunately, the pace of regulation of these devices has failed to match the speed of innovation and direct availability to patient consumers. There is an urgent need to develop a consensus on standards by which cuffless BP devices can be tested for accuracy.”
 

Devices, indications differ

Cuffless devices estimate blood pressure indirectly. Most operate based on pulse wave analysis and pulse arrival time (PWA-PAT), explained Ramakrishna Mukkamala, PhD, in a commentary. Dr. Mukkamala is a professor in the departments of bioengineering and anesthesiology and perioperative medicine at the University of Pittsburgh.

PWA involves measuring a peripheral arterial waveform using an optical sensor such as the green lights on the back of a wrist-worn device, or a ‘force sensor’ such as a finger cuff or pressing on a smartphone. Certain features are extracted from the waveform using machine learning and calibrated to blood pressure values.

PAT techniques work together with PWA; they record the ECG and extract features from that signal as well as the arterial waveform for calibration to blood pressure values.

The algorithm used to generate the BP numbers comprises a proprietary baseline model that may include demographics and other patient characteristics. A cuff measurement is often part of the baseline model because most cuffless devices require periodic (typically weekly or monthly) calibration using a cuffed device.

Cuffless devices that require cuff calibration compare the estimate they get to the cuff-calibrated number. In this scenario, the cuffless device may come up with the same blood pressure numbers simply because the baseline model – which is made up of thousands of data points relevant to the patient – has not changed.

This has led some experts to question whether PWA-PAT cuffless device readings actually add anything to the baseline model.

They don’t, according to Microsoft Research in what Dr. Mukkamala and coauthors referred to (in a review published in Hypertension) as “a complex article describing perhaps the most important and highest resource project to date (Aurora Project) on assessing the accuracy of PWA and PWA devices.”

The Microsoft article was written for bioengineers. The review in Hypertension explains the project for clinicians, and concludes that, “Cuffless BP devices based on PWA and PWA-PAT, which are similar to some regulatory-cleared devices, were of no additional value in measuring auscultatory or 24-hour ambulatory cuff BP when compared with a baseline model in which BP was predicted without an actual measurement.”
 

IEEE and FDA validation

Despite these concerns, several cuffless devices using PWA and PAT have been cleared by the Food and Drug Administration.

Validating cuffless devices is no simple matter. The Institute of Electrical and Electronics Engineers published a validation protocol for cuffless blood pressure devices in 2014 that was amended in 2019 to include a requirement to evaluate performance in different positions and in the presence of motion with varying degrees of noise artifact.

However, Daichi Shimbo, MD, codirector of the Columbia Hypertension Center in New York and vice chair of the American Heart Association Statement on blood pressure monitoring, and colleagues point out limitations, even in the updated standard. These include not requiring evaluation for drift over time; lack of specific dynamic testing protocols for stressors such as exercise or environmental temperatures; and an unsuitable reference standard (oscillometric cuff-based devices) during movement.

Dr. Shimbo said in an interview that, although he is excited about them, “these cuffless devices are not aligned with regulatory bodies. If a device gives someone a wrong blood pressure, they might be diagnosed with hypertension when they don’t have it or might miss the fact that they’re hypertensive because they get a normal blood pressure reading. If there’s no yardstick by which you say these devices are good, what are we really doing – helping, or causing a problem?”

“The specifics of how a device estimates blood pressure can determine what testing is needed to ensure that it is providing accurate performance in the intended conditions of use,” Jeremy Kahn, an FDA press officer, said in an interview. “For example, for cuffless devices that are calibrated initially with a cuff-based blood pressure device, the cuffless device needs to specify the period over which it can provide accurate readings and have testing to demonstrate that it provides accurate results over that period of use.”

The FDA said its testing is different from what the Microsoft Aurora Project used in their study.

“The intent of that testing, as the agency understands it, is to evaluate whether the device is providing useful input based on the current physiology of the patient rather than relying on predetermined values based on calibration or patient attributes. We evaluate this clinically in two separate tests: an induced change in blood pressure test and tracking of natural blood pressure changes with longer term device use,” Mr. Kahn explained.

Analyzing a device’s performance on individuals who have had natural changes in blood pressure as compared to a calibration value or initial reading “can also help discern if the device is using physiological data from the patient to determine their blood pressure accurately,” he said.

Experts interviewed for this article who remain skeptical about cuffless BP monitoring question whether the numbers that appear during the induced blood pressure change, and with the natural blood pressure changes that may occur over time, accurately reflect a patient’s blood pressure.

“The FDA doesn’t approve these devices; they clear them,” Dr. Shimbo pointed out. “Clearing them means they can be sold to the general public in the U.S. It’s not a strong statement that they’re accurate.”
 

Moving toward validation, standards

Ultimately, cuffless BP monitors may require more than one validation protocol and standard, depending on their technology, how and where they will be used, and by whom.

And as Dr. Plante and colleagues write, “Importantly, validation should be performed in diverse and special populations, including pregnant women and individuals across a range of heart rates, skin tones, wrist sizes, common arrhythmias, and beta-blocker use.”

Organizations that might be expected to help move validation and standards forward have mostly remained silent. The American Medical Association’s US Blood Pressure Validated Device Listing website includes only cuffed devices, as does the website of the international scientific nonprofit STRIDE BP.

The European Society of Hypertension 2022 consensus statement on cuffless devices concluded that, until there is an internationally accepted accuracy standard and the devices have been tested in healthy people and those with suspected or diagnosed hypertension, “cuffless BP devices should not be used for the evaluation or management of hypertension in clinical practice.”

This month, ESH published recommendations for “specific, clinically meaningful, and pragmatic validation procedures for different types of intermittent cuffless devices” that will be presented at their upcoming annual meeting June 26.

Updated protocols from IEEE “are coming out soon,” according to Dr. Shimbo. The FDA says currently cleared devices won’t need to revalidate according to new standards unless the sponsor makes significant modifications in software algorithms, device hardware, or targeted patient populations.
 

Device makers take the initiative

In the face of conflicting reports on accuracy and lack of a robust standard, some device makers are publishing their own tests or encouraging validation by potential customers.

For example, institutions that are considering using the Biobeat cuffless blood pressure monitor watch “usually start with small pilots with our devices to do internal validation,” Lior Ben Shettrit, the company’s vice president of business development, said in an interview. “Only after they complete the internal validation are they willing to move forward to full implementation.”

Cardiologist Dean Nachman, MD, is leading validation studies of the Biobeat device at the Hadassah Ein Kerem Medical Center in Jerusalem. For the first validation, the team recruited 1,057 volunteers who did a single blood pressure measurement with the cuffless device and with a cuffed device.

“We found 96.3% agreement in identifying hypertension and an interclass correlation coefficient of 0.99 and 0.97 for systolic and diastolic measurements, respectively,” he said. “Then we took it to the next level and compared the device to ambulatory 24-hour blood pressure monitoring and found comparable measurements.”

The investigators are not done yet. “We need data from thousands of patients, with subgroups, to not have any concerns,” he says. “Right now, we are using the device as a general monitor – as an EKG plus heart rate plus oxygen saturation level monitor – and as a blood pressure monitor for 24-hour blood pressure monitoring.”

The developers of the Aktiia device, which is the one Dr. Topol tested, take a different perspective. “When somebody introduces a new technology that is disrupting something that has been in place for over 100 years, there will always be some grumblings, ruffling of feathers, people saying it’s not ready, it’s not ready, it’s not ready,” Aktiia’s chief medical officer Jay Shah, MD, noted.

“But a lot of those comments are coming from the isolation of an ivory tower,” he said.

Aktiia cofounder and chief technology officer Josep Solà said that “no device is probably as accurate as if you have an invasive catheter,” adding that “we engage patients to look at their blood pressure day by day. … If each individual measurement of each of those patient is slightly less accurate than a cuff, who cares? We have 40 measurements per day on each patient. The accuracy and precision of each of those is good.”

Researchers from the George Institute for Global Health recently compared the Aktiia device to conventional ambulatory monitoring in 41 patients and found that “it did not accurately track night-time BP decline and results suggested it was unable to track medication-induced BP changes.”

“In the context of 24/7 monitoring of hypertensive patients,” Mr. Solà said, “whatever you do, if it’s better than a sham device or a baseline model and you track the blood pressure changes, it’s a hundred times much better than doing nothing.”

Dr. Nachman and Dr. Plante reported no relevant financial relationships. Dr. Shimbo reported that he received funding from NIH and has consulted for Abbott Vascular, Edward Lifesciences, Medtronic, and Tryton Medical.

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

Patients with “treatment-resistant depression” may be taking other medications with side effects that interfere with antidepressant efficacy.

Investigators studied over 800 patients who were taking antidepressants for major depressive disorder (MDD) and found that close to two-thirds were taking at least one nonpsychiatric medication with potential depressive symptom side effects (PDSS), more than 30% were taking two or more such medications, and 20% at least three such medications.

These medications, which included antihypertensive medications and corticosteroids, among others, were associated with higher odds of moderate-to-severe depressive symptoms, compared with medications without PDSS.

“When evaluating the reasons for inadequate response to treatment for depression, clinicians should consider whether their patient is also receiving a nonpsychiatric medication with a potential for depressive symptom side effects,” study investigator Mark Olfson, MD, MPH, Elizabeth K. Dollard professor of psychiatry, medicine, and law and professor of epidemiology, Columbia University Irving Medical Center, New York, said in an interview.

Columbia University

Previous research limited

“In earlier research, we found that people who were taking medications with a potential to cause depressive symptom side effects were at increased risk of depression, especially those adults who were taking more than one of these medications,” said Dr. Olfson.

This finding led Dr. Olfson and his team to “wonder whether the risks of depressive symptoms associated with these medications extended to people who were being actively treated with antidepressants for depression.”

To investigate, they turned to the National Health and Nutrition Examination Survey (NHANES) – a nationally representative cross-sectional survey of the United States general population.

The study was based on the 2013-2014, 2015-2016, and 2017-2018 waves and included 885 adults who reported using antidepressant medications for greater than or equal to 6 weeks for depression and whose depression could be ascertained.

Prescription medications with PDSS were identified through Micromedex, whose accuracy is “established” and primarily based on the U.S. Food and Drug Administration’s labeled side effects.

Nonantidepressant psychiatric medications and medications for Alzheimer’s disease or substance use disorders were not included in the analysis.

Antidepressant-treated MDD was defined as taking an antidepressant for MDD for greater than or equal to 6 weeks. Depressive symptoms were ascertained using the Patient Health Questionnaire-9 (PHQ-9) with a score of less than 5 representing no/minimal depressive symptoms and a score of greater than or equal to 10 indicating moderate/severe symptoms.

Other variables included self-reported sex, age, race/ethnicity, income, education, health insurance, and common chronic medical conditions such as hypertension, arthritis, lung disease, diabetes mellitus, thyroid disease, cancer, heart disease, liver disease, stroke, and congestive heart failure.
 

Recovery interrupted

Of the patients in the study treated with antidepressants, most were female, greater than or equal to 50 years, non-Hispanic White, and with a college education (70.55, 62.0%, 81.7%, and 69.4%, respectively).

Selective serotonin reuptake inhibitors were used by 67.9% of participants with MDD. Most had been on the same antidepressant medication for a “long time,” the authors report, with 79.2% and 67.8% taking them for greater than 1 year and greater than 2 years, respectively.

Despite the large number of patients on antidepressants, only 43.0% scored in the no/minimal symptoms range, based on the PHQ-9, while 28.4% scored in the moderate/severe range.

Most patients (85%) took at least one medication for medical conditions, with the majority medications with PDSS: 66.7% took at least one medication with PDSS, 37.3% took at least two, 21.6% took at least three, 10.7% took at least four, and 4.9% took at least five.

Almost 75% were using greater than or equal to 1 medication without PDSS, and about 50% were using greater than 1.

The number of medications with PDSS was significantly associated with lower odds of no/minimal depressive symptoms (AOR, 0.75 [95% CI, 0.64-0.87]; P < .001) and higher odds of moderate/severe symptoms (AOR, 1.14 [1.004-1.29]; P = .044).

“The predicted probability of no/minimal symptoms in those taking 5 medications with PDSS was less than half the predicted probability in those taking no medications with PDSS (0.23 vs. 0.52),” the authors report.

Conversely, the predicted probability of moderate/severe symptoms was ~50% higher in individuals taking 5 versus 0 medications with PDSS (0.36 vs. 0.24).

No corresponding associations were found for medications without PDSS.

The results were even stronger when the researchers repeated their adjusted regression analyses to focus on the 10 individual medications most associated with the severity of depressive symptoms. These were omeprazole, gabapentin, meloxicam, tramadol, ranitidine, baclofen, oxycodone, tizanidine, propranolol, and morphine, with an AOR of 0.42 [0.30-0.60] for no/minimal symptoms and 1.68 [1.24-2.27] for moderate/severe symptoms.

“Many widely prescribed medications, from antihypertensives, such as atenolol and metoprolol to corticosteroids, such as dexamethasone and triamcinolone, are associated with depression side effects,” said Dr. Olfson.

“These medications could interfere with recovery from depression. When available, consideration should be given to selecting a substitute with lower risk for depressive symptoms,” he said.
 

Role in treatment-resistant depression

In a comment, Dima Qato, PharmD, MPH, PhD, Hygeia Centennial chair and associate professor, University of Southern California School of Pharmacy, Los Angeles, said the study “is an important reminder that the use of medications with depressive symptoms side effects is increasingly common and may contribute to delays in responsiveness or worsen depressive symptoms among individuals being treated for depression.”

University of Southern California

Dr. Qato, who is also the director of the Program on Medicines and Public Health, USC School of Pharmacy, and was not involved with the study, recommended that clinicians “consider the role of medications with depression side effects when evaluating patients with treatment-resistant depression.”

The study was not supported by any funding agency. Dr. Olfson and coauthors have disclosed no relevant financial relationships. Dr. Qato is a consultant for the Public Citizen Health Research Group.

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

Patients with “treatment-resistant depression” may be taking other medications with side effects that interfere with antidepressant efficacy.

Investigators studied over 800 patients who were taking antidepressants for major depressive disorder (MDD) and found that close to two-thirds were taking at least one nonpsychiatric medication with potential depressive symptom side effects (PDSS), more than 30% were taking two or more such medications, and 20% at least three such medications.

These medications, which included antihypertensive medications and corticosteroids, among others, were associated with higher odds of moderate-to-severe depressive symptoms, compared with medications without PDSS.

“When evaluating the reasons for inadequate response to treatment for depression, clinicians should consider whether their patient is also receiving a nonpsychiatric medication with a potential for depressive symptom side effects,” study investigator Mark Olfson, MD, MPH, Elizabeth K. Dollard professor of psychiatry, medicine, and law and professor of epidemiology, Columbia University Irving Medical Center, New York, said in an interview.

Columbia University

Previous research limited

“In earlier research, we found that people who were taking medications with a potential to cause depressive symptom side effects were at increased risk of depression, especially those adults who were taking more than one of these medications,” said Dr. Olfson.

This finding led Dr. Olfson and his team to “wonder whether the risks of depressive symptoms associated with these medications extended to people who were being actively treated with antidepressants for depression.”

To investigate, they turned to the National Health and Nutrition Examination Survey (NHANES) – a nationally representative cross-sectional survey of the United States general population.

The study was based on the 2013-2014, 2015-2016, and 2017-2018 waves and included 885 adults who reported using antidepressant medications for greater than or equal to 6 weeks for depression and whose depression could be ascertained.

Prescription medications with PDSS were identified through Micromedex, whose accuracy is “established” and primarily based on the U.S. Food and Drug Administration’s labeled side effects.

Nonantidepressant psychiatric medications and medications for Alzheimer’s disease or substance use disorders were not included in the analysis.

Antidepressant-treated MDD was defined as taking an antidepressant for MDD for greater than or equal to 6 weeks. Depressive symptoms were ascertained using the Patient Health Questionnaire-9 (PHQ-9) with a score of less than 5 representing no/minimal depressive symptoms and a score of greater than or equal to 10 indicating moderate/severe symptoms.

Other variables included self-reported sex, age, race/ethnicity, income, education, health insurance, and common chronic medical conditions such as hypertension, arthritis, lung disease, diabetes mellitus, thyroid disease, cancer, heart disease, liver disease, stroke, and congestive heart failure.
 

Recovery interrupted

Of the patients in the study treated with antidepressants, most were female, greater than or equal to 50 years, non-Hispanic White, and with a college education (70.55, 62.0%, 81.7%, and 69.4%, respectively).

Selective serotonin reuptake inhibitors were used by 67.9% of participants with MDD. Most had been on the same antidepressant medication for a “long time,” the authors report, with 79.2% and 67.8% taking them for greater than 1 year and greater than 2 years, respectively.

Despite the large number of patients on antidepressants, only 43.0% scored in the no/minimal symptoms range, based on the PHQ-9, while 28.4% scored in the moderate/severe range.

Most patients (85%) took at least one medication for medical conditions, with the majority medications with PDSS: 66.7% took at least one medication with PDSS, 37.3% took at least two, 21.6% took at least three, 10.7% took at least four, and 4.9% took at least five.

Almost 75% were using greater than or equal to 1 medication without PDSS, and about 50% were using greater than 1.

The number of medications with PDSS was significantly associated with lower odds of no/minimal depressive symptoms (AOR, 0.75 [95% CI, 0.64-0.87]; P < .001) and higher odds of moderate/severe symptoms (AOR, 1.14 [1.004-1.29]; P = .044).

“The predicted probability of no/minimal symptoms in those taking 5 medications with PDSS was less than half the predicted probability in those taking no medications with PDSS (0.23 vs. 0.52),” the authors report.

Conversely, the predicted probability of moderate/severe symptoms was ~50% higher in individuals taking 5 versus 0 medications with PDSS (0.36 vs. 0.24).

No corresponding associations were found for medications without PDSS.

The results were even stronger when the researchers repeated their adjusted regression analyses to focus on the 10 individual medications most associated with the severity of depressive symptoms. These were omeprazole, gabapentin, meloxicam, tramadol, ranitidine, baclofen, oxycodone, tizanidine, propranolol, and morphine, with an AOR of 0.42 [0.30-0.60] for no/minimal symptoms and 1.68 [1.24-2.27] for moderate/severe symptoms.

“Many widely prescribed medications, from antihypertensives, such as atenolol and metoprolol to corticosteroids, such as dexamethasone and triamcinolone, are associated with depression side effects,” said Dr. Olfson.

“These medications could interfere with recovery from depression. When available, consideration should be given to selecting a substitute with lower risk for depressive symptoms,” he said.
 

Role in treatment-resistant depression

In a comment, Dima Qato, PharmD, MPH, PhD, Hygeia Centennial chair and associate professor, University of Southern California School of Pharmacy, Los Angeles, said the study “is an important reminder that the use of medications with depressive symptoms side effects is increasingly common and may contribute to delays in responsiveness or worsen depressive symptoms among individuals being treated for depression.”

University of Southern California

Dr. Qato, who is also the director of the Program on Medicines and Public Health, USC School of Pharmacy, and was not involved with the study, recommended that clinicians “consider the role of medications with depression side effects when evaluating patients with treatment-resistant depression.”

The study was not supported by any funding agency. Dr. Olfson and coauthors have disclosed no relevant financial relationships. Dr. Qato is a consultant for the Public Citizen Health Research Group.

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

Patients with “treatment-resistant depression” may be taking other medications with side effects that interfere with antidepressant efficacy.

Investigators studied over 800 patients who were taking antidepressants for major depressive disorder (MDD) and found that close to two-thirds were taking at least one nonpsychiatric medication with potential depressive symptom side effects (PDSS), more than 30% were taking two or more such medications, and 20% at least three such medications.

These medications, which included antihypertensive medications and corticosteroids, among others, were associated with higher odds of moderate-to-severe depressive symptoms, compared with medications without PDSS.

“When evaluating the reasons for inadequate response to treatment for depression, clinicians should consider whether their patient is also receiving a nonpsychiatric medication with a potential for depressive symptom side effects,” study investigator Mark Olfson, MD, MPH, Elizabeth K. Dollard professor of psychiatry, medicine, and law and professor of epidemiology, Columbia University Irving Medical Center, New York, said in an interview.

Columbia University

Previous research limited

“In earlier research, we found that people who were taking medications with a potential to cause depressive symptom side effects were at increased risk of depression, especially those adults who were taking more than one of these medications,” said Dr. Olfson.

This finding led Dr. Olfson and his team to “wonder whether the risks of depressive symptoms associated with these medications extended to people who were being actively treated with antidepressants for depression.”

To investigate, they turned to the National Health and Nutrition Examination Survey (NHANES) – a nationally representative cross-sectional survey of the United States general population.

The study was based on the 2013-2014, 2015-2016, and 2017-2018 waves and included 885 adults who reported using antidepressant medications for greater than or equal to 6 weeks for depression and whose depression could be ascertained.

Prescription medications with PDSS were identified through Micromedex, whose accuracy is “established” and primarily based on the U.S. Food and Drug Administration’s labeled side effects.

Nonantidepressant psychiatric medications and medications for Alzheimer’s disease or substance use disorders were not included in the analysis.

Antidepressant-treated MDD was defined as taking an antidepressant for MDD for greater than or equal to 6 weeks. Depressive symptoms were ascertained using the Patient Health Questionnaire-9 (PHQ-9) with a score of less than 5 representing no/minimal depressive symptoms and a score of greater than or equal to 10 indicating moderate/severe symptoms.

Other variables included self-reported sex, age, race/ethnicity, income, education, health insurance, and common chronic medical conditions such as hypertension, arthritis, lung disease, diabetes mellitus, thyroid disease, cancer, heart disease, liver disease, stroke, and congestive heart failure.
 

Recovery interrupted

Of the patients in the study treated with antidepressants, most were female, greater than or equal to 50 years, non-Hispanic White, and with a college education (70.55, 62.0%, 81.7%, and 69.4%, respectively).

Selective serotonin reuptake inhibitors were used by 67.9% of participants with MDD. Most had been on the same antidepressant medication for a “long time,” the authors report, with 79.2% and 67.8% taking them for greater than 1 year and greater than 2 years, respectively.

Despite the large number of patients on antidepressants, only 43.0% scored in the no/minimal symptoms range, based on the PHQ-9, while 28.4% scored in the moderate/severe range.

Most patients (85%) took at least one medication for medical conditions, with the majority medications with PDSS: 66.7% took at least one medication with PDSS, 37.3% took at least two, 21.6% took at least three, 10.7% took at least four, and 4.9% took at least five.

Almost 75% were using greater than or equal to 1 medication without PDSS, and about 50% were using greater than 1.

The number of medications with PDSS was significantly associated with lower odds of no/minimal depressive symptoms (AOR, 0.75 [95% CI, 0.64-0.87]; P < .001) and higher odds of moderate/severe symptoms (AOR, 1.14 [1.004-1.29]; P = .044).

“The predicted probability of no/minimal symptoms in those taking 5 medications with PDSS was less than half the predicted probability in those taking no medications with PDSS (0.23 vs. 0.52),” the authors report.

Conversely, the predicted probability of moderate/severe symptoms was ~50% higher in individuals taking 5 versus 0 medications with PDSS (0.36 vs. 0.24).

No corresponding associations were found for medications without PDSS.

The results were even stronger when the researchers repeated their adjusted regression analyses to focus on the 10 individual medications most associated with the severity of depressive symptoms. These were omeprazole, gabapentin, meloxicam, tramadol, ranitidine, baclofen, oxycodone, tizanidine, propranolol, and morphine, with an AOR of 0.42 [0.30-0.60] for no/minimal symptoms and 1.68 [1.24-2.27] for moderate/severe symptoms.

“Many widely prescribed medications, from antihypertensives, such as atenolol and metoprolol to corticosteroids, such as dexamethasone and triamcinolone, are associated with depression side effects,” said Dr. Olfson.

“These medications could interfere with recovery from depression. When available, consideration should be given to selecting a substitute with lower risk for depressive symptoms,” he said.
 

Role in treatment-resistant depression

In a comment, Dima Qato, PharmD, MPH, PhD, Hygeia Centennial chair and associate professor, University of Southern California School of Pharmacy, Los Angeles, said the study “is an important reminder that the use of medications with depressive symptoms side effects is increasingly common and may contribute to delays in responsiveness or worsen depressive symptoms among individuals being treated for depression.”

University of Southern California

Dr. Qato, who is also the director of the Program on Medicines and Public Health, USC School of Pharmacy, and was not involved with the study, recommended that clinicians “consider the role of medications with depression side effects when evaluating patients with treatment-resistant depression.”

The study was not supported by any funding agency. Dr. Olfson and coauthors have disclosed no relevant financial relationships. Dr. Qato is a consultant for the Public Citizen Health Research Group.

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

TOPLINE

Patients who received zuranolone 50 mg/d demonstrated significantly greater improvement in depressive symptoms than those who received placebo, with a rapid onset of effect.

METHODOLOGY

The Food and Drug Administration has accepted filing of a new drug application for zuranolone, a neuroactive steroid that targets g-aminobutyric acid type A receptors (GABAAR), for the treatment of major depressive disorder (MDD) and postpartum depression.

The study included 543 mostly White female patients with MDD. The mean age of the patients was 40 years. Participants were randomly assigned to receive oral zuranolone 50 mg or placebo once daily for 14 days.

About 30% of patients were taking an antidepressant.

The primary endpoint was change in Hamilton Depression Rating Scale (HAM-D) score at day 15.
 

TAKEAWAY

The zuranolone group showed significantly greater improvement in depressive symptoms at 15 days compared with the placebo group (least square mean [LSM] change on HAM-D, –14.1, vs. –12.3; P = .01; Cohen’s d = 0.23).

Improvements were observed on day 3, the earliest assessment, and were sustained at all subsequent visits during the treatment and follow-up period (through day 42).

Results favored zuranolone regardless of the use of antidepressant therapies.

Patients with anxiety who received the active drug experienced improvement in anxiety symptoms compared to the patients who received placebo.

The drug was well tolerated, and there were no new safety findings. The most common treatment-emergent adverse events were somnolence and headache. There was no weight gain, sexual dysfunction, withdrawal symptoms, or increased suicidal ideation or behavior.
 

IN PRACTICE

The study adds to evidence suggesting zuranolone is a promising novel therapy for treating MDD, the authors noted.

STUDY DETAILS

The study was conducted by Anita H. Clayton, MD, department of psychiatry and neurobehavioral sciences, University of Virginia, Charlottesville, and colleagues. It was published online May 3 in The American Journal of Psychiatry.

LIMITATIONS

The study was short term, and the patient population was severely depressed at study entry, which may limit application to those with mild or moderate symptoms. There was a robust placebo response, possibly partly due to the COVID-19 pandemic, when there was an increase in depressive symptoms in the U.S. population, and so frequent in-person visits may have led to an improvement in symptoms even if the patient was receiving placebo.

DISCLOSURES

The study was funded by Sage Therapeutics and Biogen.

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

TOPLINE

Patients who received zuranolone 50 mg/d demonstrated significantly greater improvement in depressive symptoms than those who received placebo, with a rapid onset of effect.

METHODOLOGY

The Food and Drug Administration has accepted filing of a new drug application for zuranolone, a neuroactive steroid that targets g-aminobutyric acid type A receptors (GABAAR), for the treatment of major depressive disorder (MDD) and postpartum depression.

The study included 543 mostly White female patients with MDD. The mean age of the patients was 40 years. Participants were randomly assigned to receive oral zuranolone 50 mg or placebo once daily for 14 days.

About 30% of patients were taking an antidepressant.

The primary endpoint was change in Hamilton Depression Rating Scale (HAM-D) score at day 15.
 

TAKEAWAY

The zuranolone group showed significantly greater improvement in depressive symptoms at 15 days compared with the placebo group (least square mean [LSM] change on HAM-D, –14.1, vs. –12.3; P = .01; Cohen’s d = 0.23).

Improvements were observed on day 3, the earliest assessment, and were sustained at all subsequent visits during the treatment and follow-up period (through day 42).

Results favored zuranolone regardless of the use of antidepressant therapies.

Patients with anxiety who received the active drug experienced improvement in anxiety symptoms compared to the patients who received placebo.

The drug was well tolerated, and there were no new safety findings. The most common treatment-emergent adverse events were somnolence and headache. There was no weight gain, sexual dysfunction, withdrawal symptoms, or increased suicidal ideation or behavior.
 

IN PRACTICE

The study adds to evidence suggesting zuranolone is a promising novel therapy for treating MDD, the authors noted.

STUDY DETAILS

The study was conducted by Anita H. Clayton, MD, department of psychiatry and neurobehavioral sciences, University of Virginia, Charlottesville, and colleagues. It was published online May 3 in The American Journal of Psychiatry.

LIMITATIONS

The study was short term, and the patient population was severely depressed at study entry, which may limit application to those with mild or moderate symptoms. There was a robust placebo response, possibly partly due to the COVID-19 pandemic, when there was an increase in depressive symptoms in the U.S. population, and so frequent in-person visits may have led to an improvement in symptoms even if the patient was receiving placebo.

DISCLOSURES

The study was funded by Sage Therapeutics and Biogen.

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

TOPLINE

Patients who received zuranolone 50 mg/d demonstrated significantly greater improvement in depressive symptoms than those who received placebo, with a rapid onset of effect.

METHODOLOGY

The Food and Drug Administration has accepted filing of a new drug application for zuranolone, a neuroactive steroid that targets g-aminobutyric acid type A receptors (GABAAR), for the treatment of major depressive disorder (MDD) and postpartum depression.

The study included 543 mostly White female patients with MDD. The mean age of the patients was 40 years. Participants were randomly assigned to receive oral zuranolone 50 mg or placebo once daily for 14 days.

About 30% of patients were taking an antidepressant.

The primary endpoint was change in Hamilton Depression Rating Scale (HAM-D) score at day 15.
 

TAKEAWAY

The zuranolone group showed significantly greater improvement in depressive symptoms at 15 days compared with the placebo group (least square mean [LSM] change on HAM-D, –14.1, vs. –12.3; P = .01; Cohen’s d = 0.23).

Improvements were observed on day 3, the earliest assessment, and were sustained at all subsequent visits during the treatment and follow-up period (through day 42).

Results favored zuranolone regardless of the use of antidepressant therapies.

Patients with anxiety who received the active drug experienced improvement in anxiety symptoms compared to the patients who received placebo.

The drug was well tolerated, and there were no new safety findings. The most common treatment-emergent adverse events were somnolence and headache. There was no weight gain, sexual dysfunction, withdrawal symptoms, or increased suicidal ideation or behavior.
 

IN PRACTICE

The study adds to evidence suggesting zuranolone is a promising novel therapy for treating MDD, the authors noted.

STUDY DETAILS

The study was conducted by Anita H. Clayton, MD, department of psychiatry and neurobehavioral sciences, University of Virginia, Charlottesville, and colleagues. It was published online May 3 in The American Journal of Psychiatry.

LIMITATIONS

The study was short term, and the patient population was severely depressed at study entry, which may limit application to those with mild or moderate symptoms. There was a robust placebo response, possibly partly due to the COVID-19 pandemic, when there was an increase in depressive symptoms in the U.S. population, and so frequent in-person visits may have led to an improvement in symptoms even if the patient was receiving placebo.

DISCLOSURES

The study was funded by Sage Therapeutics and Biogen.

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

SEOUL, SOUTH KOREA – B cell–depleting therapies in patients with lupus nephritis have a higher likelihood of complete response if B cells are almost completely depleted, and strategies for achieving more complete B-cell depletion continue to be tested, according to evidence presented by Richard A. Furie, MD, at an international congress on systemic lupus erythematosus (SLE).

“If you go back about 20 years ago or so, when we designed the LUNAR and EXPLORER trials, we were scared to death of rituximab [Rituxan and biosimilars], about what would happen when you deplete B cells,” said Dr. Furie, chief of the division of rheumatology at Northwell Health in New York.

Sara Freeman/MDedge News

The LUNAR trial, which compared rituximab with placebo in patients with lupus nephritis, did not show a statistically significant difference in renal outcomes at 1 year. However, a post hoc analysis done several years later told a different story. It looked at patients who achieved complete peripheral depletion of B cells, defined as zero cells per microliter in peripheral blood. “You can see about a fourfold increase in complete response rates in those who were complete B-cell depleters at 1 year,” Dr. Furie told the conference.

It therefore raises the question of how to achieve greater B-cell depletion rates in patients. Dr. Furie said one strategy might be to first mobilize memory B cells and neutralize B cell–activating factor using belimumab (Benlysta), and then treat with rituximab to eliminate B cells. This strategy of sequential belimumab-rituximab treatment has been taken in several clinical trials.
 

More potent B-cell depletion with obinutuzumab

Another approach is to choose more potent B cell–depleting therapies, such as obinutuzumab (Gazyva), which is an anti-CD20 monoclonal antibody that was approved in 2013 for the treatment of chronic lymphocytic leukemia.

The NOBILITY trial compared obinutuzumab with placebo in 125 patients with lupus nephritis who were on background treatment with mycophenolate and corticosteroids. At 1 year, significantly more patients achieved B-cell thresholds either below 5 cells per microliter or even zero cells per microliter than had been seen previously with rituximab.

That also translated into clinical benefit, Dr. Furie said. By week 76, half the patients who had sustained depletion of B cells below 0.4 cells per microliter had a complete response, compared with 35% of those who still had detectable B cells and 18% of the placebo group. Treatment with obinutuzumab did not show any link to higher rates of serious adverse events, serious infections, or deaths.

“I think we’re all pretty much convinced more is better, without introducing safety issues,” Dr. Furie said in an interview.

Bianca Nogrady/MDedge News

Joan Merrill, MD, professor of medicine at the University of Oklahoma Health Sciences Center, Oklahoma City, said the data did suggest that renal outcomes were better with more complete depletion, but raised the question of whether this might increase the risk of infections or infectious severity.

Dr. Furie noted that complete response not only required improvement in proteinuria, complement levels, and anti–double-stranded DNA antibodies, but also in serum creatinine, “because maintenance of eGFR [estimated glomerular filtration rate] is the name of the game with lupus nephritis.”

However, he also pointed out that there may be a ceiling for response rates in patients with lupus nephritis when using stricter endpoints for serum creatinine. The NOBILITY trial required patients to achieve a serum creatinine that did not increase by more than 15% from baseline. But when researchers did an analysis that instead only required patients to achieve a reduction in proteinuria and maintain normal creatinine, the complete response rate in complete B-cell depleters increased to 72%, compared with 50% in partial depleters and 37% in the placebo group.
 

Newer strategies for greater B-cell depletion

A third strategy for achieving greater B-cell depletion is bispecific T-cell engagers, or BiTEs. “I called it a ‘frenemy,’ where it’s taking the activated T cell and introducing it to the B cell, and it can kill it via direct T-cell killing,” Dr. Furie said in an interview. Mosunetuzumab (Lunsumio) is one example, and is currently in a phase 1 clinical trial of patients with SLE.

And the fourth strategy, which has proved so successful in lymphoma, is chimeric antigen receptor T-cell therapy (CAR T). Dr. Furie cited the recent publication of data from a CAR T clinical trial in five patients with refractory SLE. He said the data were impressive but the question for this treatment approach will be which patients are most likely to benefit and whether CAR T will experience the same ceiling effect because of pre-existing kidney damage.

“We won’t be seeing 100% response rates,” he said. “What we’ll be seeing, as a maximum, might be about 70%.” The big question for B-cell depletion in lupus was therefore how best to achieve it. “Is the future a potent monoclonal antibody, or is it in fact CAR T?”

Dr. Merrill said the analyses from B-cell depletion trials, showing greater response rates among more complete depleters, highlighted the importance of a personalized approach to treating lupus.

“One size fits all is never optimal in any disease, but it will prove a nonstarter in lupus, where we ought to be trying to find the optimal treatment regimen for each patient guided by biomarkers,” she said in an interview.

Dr. Furie reported having financial relationships with Genentech/Roche, which manufactures obinutuzumab and rituximab, as well as GlaxoSmithKline, Kezar Life Sciences, Kyverna Therapeutics, and Takeda. Dr. Merrill reported consulting for and receiving research support from a range of pharmaceutical companies including Genentech/Roche, GlaxoSmithKline, Pfizer, Janssen, Bristol-Myers Squibb, AbbVie, and AstraZeneca.

SEOUL, SOUTH KOREA – B cell–depleting therapies in patients with lupus nephritis have a higher likelihood of complete response if B cells are almost completely depleted, and strategies for achieving more complete B-cell depletion continue to be tested, according to evidence presented by Richard A. Furie, MD, at an international congress on systemic lupus erythematosus (SLE).

“If you go back about 20 years ago or so, when we designed the LUNAR and EXPLORER trials, we were scared to death of rituximab [Rituxan and biosimilars], about what would happen when you deplete B cells,” said Dr. Furie, chief of the division of rheumatology at Northwell Health in New York.

Sara Freeman/MDedge News

The LUNAR trial, which compared rituximab with placebo in patients with lupus nephritis, did not show a statistically significant difference in renal outcomes at 1 year. However, a post hoc analysis done several years later told a different story. It looked at patients who achieved complete peripheral depletion of B cells, defined as zero cells per microliter in peripheral blood. “You can see about a fourfold increase in complete response rates in those who were complete B-cell depleters at 1 year,” Dr. Furie told the conference.

It therefore raises the question of how to achieve greater B-cell depletion rates in patients. Dr. Furie said one strategy might be to first mobilize memory B cells and neutralize B cell–activating factor using belimumab (Benlysta), and then treat with rituximab to eliminate B cells. This strategy of sequential belimumab-rituximab treatment has been taken in several clinical trials.
 

More potent B-cell depletion with obinutuzumab

Another approach is to choose more potent B cell–depleting therapies, such as obinutuzumab (Gazyva), which is an anti-CD20 monoclonal antibody that was approved in 2013 for the treatment of chronic lymphocytic leukemia.

The NOBILITY trial compared obinutuzumab with placebo in 125 patients with lupus nephritis who were on background treatment with mycophenolate and corticosteroids. At 1 year, significantly more patients achieved B-cell thresholds either below 5 cells per microliter or even zero cells per microliter than had been seen previously with rituximab.

That also translated into clinical benefit, Dr. Furie said. By week 76, half the patients who had sustained depletion of B cells below 0.4 cells per microliter had a complete response, compared with 35% of those who still had detectable B cells and 18% of the placebo group. Treatment with obinutuzumab did not show any link to higher rates of serious adverse events, serious infections, or deaths.

“I think we’re all pretty much convinced more is better, without introducing safety issues,” Dr. Furie said in an interview.

Bianca Nogrady/MDedge News

Joan Merrill, MD, professor of medicine at the University of Oklahoma Health Sciences Center, Oklahoma City, said the data did suggest that renal outcomes were better with more complete depletion, but raised the question of whether this might increase the risk of infections or infectious severity.

Dr. Furie noted that complete response not only required improvement in proteinuria, complement levels, and anti–double-stranded DNA antibodies, but also in serum creatinine, “because maintenance of eGFR [estimated glomerular filtration rate] is the name of the game with lupus nephritis.”

However, he also pointed out that there may be a ceiling for response rates in patients with lupus nephritis when using stricter endpoints for serum creatinine. The NOBILITY trial required patients to achieve a serum creatinine that did not increase by more than 15% from baseline. But when researchers did an analysis that instead only required patients to achieve a reduction in proteinuria and maintain normal creatinine, the complete response rate in complete B-cell depleters increased to 72%, compared with 50% in partial depleters and 37% in the placebo group.
 

Newer strategies for greater B-cell depletion

A third strategy for achieving greater B-cell depletion is bispecific T-cell engagers, or BiTEs. “I called it a ‘frenemy,’ where it’s taking the activated T cell and introducing it to the B cell, and it can kill it via direct T-cell killing,” Dr. Furie said in an interview. Mosunetuzumab (Lunsumio) is one example, and is currently in a phase 1 clinical trial of patients with SLE.

And the fourth strategy, which has proved so successful in lymphoma, is chimeric antigen receptor T-cell therapy (CAR T). Dr. Furie cited the recent publication of data from a CAR T clinical trial in five patients with refractory SLE. He said the data were impressive but the question for this treatment approach will be which patients are most likely to benefit and whether CAR T will experience the same ceiling effect because of pre-existing kidney damage.

“We won’t be seeing 100% response rates,” he said. “What we’ll be seeing, as a maximum, might be about 70%.” The big question for B-cell depletion in lupus was therefore how best to achieve it. “Is the future a potent monoclonal antibody, or is it in fact CAR T?”

Dr. Merrill said the analyses from B-cell depletion trials, showing greater response rates among more complete depleters, highlighted the importance of a personalized approach to treating lupus.

“One size fits all is never optimal in any disease, but it will prove a nonstarter in lupus, where we ought to be trying to find the optimal treatment regimen for each patient guided by biomarkers,” she said in an interview.

Dr. Furie reported having financial relationships with Genentech/Roche, which manufactures obinutuzumab and rituximab, as well as GlaxoSmithKline, Kezar Life Sciences, Kyverna Therapeutics, and Takeda. Dr. Merrill reported consulting for and receiving research support from a range of pharmaceutical companies including Genentech/Roche, GlaxoSmithKline, Pfizer, Janssen, Bristol-Myers Squibb, AbbVie, and AstraZeneca.

SEOUL, SOUTH KOREA – B cell–depleting therapies in patients with lupus nephritis have a higher likelihood of complete response if B cells are almost completely depleted, and strategies for achieving more complete B-cell depletion continue to be tested, according to evidence presented by Richard A. Furie, MD, at an international congress on systemic lupus erythematosus (SLE).

“If you go back about 20 years ago or so, when we designed the LUNAR and EXPLORER trials, we were scared to death of rituximab [Rituxan and biosimilars], about what would happen when you deplete B cells,” said Dr. Furie, chief of the division of rheumatology at Northwell Health in New York.

Sara Freeman/MDedge News

The LUNAR trial, which compared rituximab with placebo in patients with lupus nephritis, did not show a statistically significant difference in renal outcomes at 1 year. However, a post hoc analysis done several years later told a different story. It looked at patients who achieved complete peripheral depletion of B cells, defined as zero cells per microliter in peripheral blood. “You can see about a fourfold increase in complete response rates in those who were complete B-cell depleters at 1 year,” Dr. Furie told the conference.

It therefore raises the question of how to achieve greater B-cell depletion rates in patients. Dr. Furie said one strategy might be to first mobilize memory B cells and neutralize B cell–activating factor using belimumab (Benlysta), and then treat with rituximab to eliminate B cells. This strategy of sequential belimumab-rituximab treatment has been taken in several clinical trials.
 

More potent B-cell depletion with obinutuzumab

Another approach is to choose more potent B cell–depleting therapies, such as obinutuzumab (Gazyva), which is an anti-CD20 monoclonal antibody that was approved in 2013 for the treatment of chronic lymphocytic leukemia.

The NOBILITY trial compared obinutuzumab with placebo in 125 patients with lupus nephritis who were on background treatment with mycophenolate and corticosteroids. At 1 year, significantly more patients achieved B-cell thresholds either below 5 cells per microliter or even zero cells per microliter than had been seen previously with rituximab.

That also translated into clinical benefit, Dr. Furie said. By week 76, half the patients who had sustained depletion of B cells below 0.4 cells per microliter had a complete response, compared with 35% of those who still had detectable B cells and 18% of the placebo group. Treatment with obinutuzumab did not show any link to higher rates of serious adverse events, serious infections, or deaths.

“I think we’re all pretty much convinced more is better, without introducing safety issues,” Dr. Furie said in an interview.

Bianca Nogrady/MDedge News

Joan Merrill, MD, professor of medicine at the University of Oklahoma Health Sciences Center, Oklahoma City, said the data did suggest that renal outcomes were better with more complete depletion, but raised the question of whether this might increase the risk of infections or infectious severity.

Dr. Furie noted that complete response not only required improvement in proteinuria, complement levels, and anti–double-stranded DNA antibodies, but also in serum creatinine, “because maintenance of eGFR [estimated glomerular filtration rate] is the name of the game with lupus nephritis.”

However, he also pointed out that there may be a ceiling for response rates in patients with lupus nephritis when using stricter endpoints for serum creatinine. The NOBILITY trial required patients to achieve a serum creatinine that did not increase by more than 15% from baseline. But when researchers did an analysis that instead only required patients to achieve a reduction in proteinuria and maintain normal creatinine, the complete response rate in complete B-cell depleters increased to 72%, compared with 50% in partial depleters and 37% in the placebo group.
 

Newer strategies for greater B-cell depletion

A third strategy for achieving greater B-cell depletion is bispecific T-cell engagers, or BiTEs. “I called it a ‘frenemy,’ where it’s taking the activated T cell and introducing it to the B cell, and it can kill it via direct T-cell killing,” Dr. Furie said in an interview. Mosunetuzumab (Lunsumio) is one example, and is currently in a phase 1 clinical trial of patients with SLE.

And the fourth strategy, which has proved so successful in lymphoma, is chimeric antigen receptor T-cell therapy (CAR T). Dr. Furie cited the recent publication of data from a CAR T clinical trial in five patients with refractory SLE. He said the data were impressive but the question for this treatment approach will be which patients are most likely to benefit and whether CAR T will experience the same ceiling effect because of pre-existing kidney damage.

“We won’t be seeing 100% response rates,” he said. “What we’ll be seeing, as a maximum, might be about 70%.” The big question for B-cell depletion in lupus was therefore how best to achieve it. “Is the future a potent monoclonal antibody, or is it in fact CAR T?”

Dr. Merrill said the analyses from B-cell depletion trials, showing greater response rates among more complete depleters, highlighted the importance of a personalized approach to treating lupus.

“One size fits all is never optimal in any disease, but it will prove a nonstarter in lupus, where we ought to be trying to find the optimal treatment regimen for each patient guided by biomarkers,” she said in an interview.

Dr. Furie reported having financial relationships with Genentech/Roche, which manufactures obinutuzumab and rituximab, as well as GlaxoSmithKline, Kezar Life Sciences, Kyverna Therapeutics, and Takeda. Dr. Merrill reported consulting for and receiving research support from a range of pharmaceutical companies including Genentech/Roche, GlaxoSmithKline, Pfizer, Janssen, Bristol-Myers Squibb, AbbVie, and AstraZeneca.

Triple-negative breast cancer (TNBC) is characterized by the absence of hormonal receptors and human epidermal growth factor receptor 2 (HER2) expression. Historically, treatments targeting HER2 were found to be ineffective in patients with TNBC and known HER2-zero status. Researchers more recently identified a new type of TNBC involving patients with low expression of HER2. Patients with this type of breast cancer, now referred to as HER2-low, have immunohistochemical (IHC) analysis scores of 1+ or 2+ and negative in situ hybridization (ISH) stain.

In a new study, patients with TNBC who initially tested as having HER2-zero status were later found to have HER2-low status following repeated biopsies. These HER2-low results were of great clinical significance for this patient population, said Yael Bar, MD, PhD, during her presentation of the research, at the annual meeting of the American Society of Clinical Oncology (ASCO).

Previously, the DESTINY-Breast04 trial demonstrated that the antibody-drug conjugate (ADC) trastuzumab deruxtecan (T-DXd) improved progression-free survival (PFS) and overall survival (OS) for patients with HER2-low metastatic breast cancer. “As a result [of the DESTINY-Breast04 findings], T-DXd is now approved for HER2-low but not HER2-zero triple-negative metastatic breast cancer."

“While HER2-low is detected in about 30%-50% of patients with triple-negative breast cancer, several studies have shown that HER2 status is heterogeneous and also dynamic over time, said Dr. Bar, who is an international research fellow in the breast cancer group at Mass General Cancer Center, Boston.

In the new study, Dr. Bar and her co-authors retrospectively identified 512 TNBC patients from 2000 to 2022 from an institutional database. They included core, surgical, or metastatic biopsies. Participants had a mean age of 52 years, with 54% over age 50. They were 83% White, 7% African American, 5% Asian, 3% Hispanic, and 2% other. Stage II was most common at diagnosis at 48%, followed by stage 1 (28%), stage 3 (14%), and stage IV (8%).

Most patients had undergone one (38%) or two (45%) biopsies, while 9% underwent three biopsies, 6% underwent four biopsies, and 2% underwent five or more.

Among all 512 patients in the study, 60% had a HER2-low result on their first biopsy. As of the second biopsy, 73% had at least one HER2-low result, with 13% of the first HER2-low results occurring at the second biopsy. As of the third biopsy, 81% had a HER2-low result, with 9% occurring for the first time. At the fourth biopsy, 86% had a positive result, with 8% occurring for the first time. All patients with five or more biopsies had at least one HER2-low result and none were first-time results.

At the second biopsy, a HER2-low result was detected for 32% of patients for the first time. At the third biopsy, a new HER2-low result was detected in 33%, and at the fourth biopsy, a new HER-2 result was detected in 38%.

The researchers matched early and metastatic biopsies in 71 patients, and 44% had changed status: 68% of those with a status change went HER2-low to HER2-zero, 26% from HER2-zero to HER2-low, and 6% from HER2-low to HER2-positive. Among 50 patients with matched metastatic biopsies, 33% had a change in status, with 63% going from HER2-zero to HER2-low, 31% from HER2-low to HER2-zero, and 6% from HER2-low to HER2-positive.

“We showed here that repeat biopsies can identify new HER2-low results for patients who were previously ineligible for T-DXd; and therefore, we think that a repeat biopsy could be considered if feasible and safe. Also, if a repeat biopsy is performed for any reason, but mainly upon metastatic recurrence, receptors should be retested,” said Dr. Bar.

After Dr. Bar’s presentation, Barbara Pistilli, MD served as a discussant. She noted the increased HER2-low results over successive biopsies. “However, here the question is, are these results related to the changes in the analytical methods over the past 20 years or the changes in the guidelines in terms of definition of HER2 status, or are they more related to a true evolution of HER2 status with the evolution of the disease?” she said during her presentation. Dr. Pistilli is chair of the breast disease committee at Gustave Roussy in Villejuif, France.

She also said that HER2 expression can vary even between different parts of the same tumor and called for alternative methods to following HER2 expression. “I don’t think that we can follow our patients with multiple biopsies over the disease evolution, so we have to find other tools, such as target-positive [circulating tumor cells], or antibody-radiolabeled PET scan in order to better follow the intermetastasis target heterogeneity over time, and finally define what is the optimal ADC sequential strategy for each patient,” said Dr. Pistilli.

Comoderator Michael Danso, MD, also weighed in when asked for comment. 

“It was an important trial to show that serial biopsies potentially allow more patients to receive trastuzumab deruxtecan,” said Dr. Danso, who is the research director at Virginia Oncology Associates, Norfolk. However, he pointed out the concerns of a statistician who had spoken up during the question-and-answer session who said that the positive results could simply be the consequence of repeated testing. “If you do a test often enough, statistically you’re going to get a difference in outcome. That was an important point made. Also, if you’re going to get 100% of patients who are eventually going to [develop HER2-low status], the question is, can you just treat everybody with trastuzumab deruxtecan and not do these sequential biopsies? Obviously that is subject to cost; it’s subject to toxicity as well, so you probably want documentation that there is a HER2-low result,” said Dr. Danso.

Dr. Bar has no relevant financial disclosures. Dr. Pistilli has consulted for or advised AstraZeneca, Daiichi Sankyo/UCB Japan, Myriad Genetics, Novartis, PIERRE FABRE, and Puma Biotechnology. She has received research funding through her institution from AstraZeneca, Daiichi Sankyo, Gilead Sciences, Merus, Pfizer, and Puma Biotechnology. She has received travel or accommodation expenses from AstraZeneca, Daiichi Sankyo Europe, MSD Oncology, Novartis, Pfizer, and Pierre Fabre. Dr. Danso has received honoraria from Amgen and has consulted or advised Immunomedics, Novartis, Pfizer, and Seagen.

*This story was updated on 6/13/2023.

Triple-negative breast cancer (TNBC) is characterized by the absence of hormonal receptors and human epidermal growth factor receptor 2 (HER2) expression. Historically, treatments targeting HER2 were found to be ineffective in patients with TNBC and known HER2-zero status. Researchers more recently identified a new type of TNBC involving patients with low expression of HER2. Patients with this type of breast cancer, now referred to as HER2-low, have immunohistochemical (IHC) analysis scores of 1+ or 2+ and negative in situ hybridization (ISH) stain.

In a new study, patients with TNBC who initially tested as having HER2-zero status were later found to have HER2-low status following repeated biopsies. These HER2-low results were of great clinical significance for this patient population, said Yael Bar, MD, PhD, during her presentation of the research, at the annual meeting of the American Society of Clinical Oncology (ASCO).

Previously, the DESTINY-Breast04 trial demonstrated that the antibody-drug conjugate (ADC) trastuzumab deruxtecan (T-DXd) improved progression-free survival (PFS) and overall survival (OS) for patients with HER2-low metastatic breast cancer. “As a result [of the DESTINY-Breast04 findings], T-DXd is now approved for HER2-low but not HER2-zero triple-negative metastatic breast cancer."

“While HER2-low is detected in about 30%-50% of patients with triple-negative breast cancer, several studies have shown that HER2 status is heterogeneous and also dynamic over time, said Dr. Bar, who is an international research fellow in the breast cancer group at Mass General Cancer Center, Boston.

In the new study, Dr. Bar and her co-authors retrospectively identified 512 TNBC patients from 2000 to 2022 from an institutional database. They included core, surgical, or metastatic biopsies. Participants had a mean age of 52 years, with 54% over age 50. They were 83% White, 7% African American, 5% Asian, 3% Hispanic, and 2% other. Stage II was most common at diagnosis at 48%, followed by stage 1 (28%), stage 3 (14%), and stage IV (8%).

Most patients had undergone one (38%) or two (45%) biopsies, while 9% underwent three biopsies, 6% underwent four biopsies, and 2% underwent five or more.

Among all 512 patients in the study, 60% had a HER2-low result on their first biopsy. As of the second biopsy, 73% had at least one HER2-low result, with 13% of the first HER2-low results occurring at the second biopsy. As of the third biopsy, 81% had a HER2-low result, with 9% occurring for the first time. At the fourth biopsy, 86% had a positive result, with 8% occurring for the first time. All patients with five or more biopsies had at least one HER2-low result and none were first-time results.

At the second biopsy, a HER2-low result was detected for 32% of patients for the first time. At the third biopsy, a new HER2-low result was detected in 33%, and at the fourth biopsy, a new HER-2 result was detected in 38%.

The researchers matched early and metastatic biopsies in 71 patients, and 44% had changed status: 68% of those with a status change went HER2-low to HER2-zero, 26% from HER2-zero to HER2-low, and 6% from HER2-low to HER2-positive. Among 50 patients with matched metastatic biopsies, 33% had a change in status, with 63% going from HER2-zero to HER2-low, 31% from HER2-low to HER2-zero, and 6% from HER2-low to HER2-positive.

“We showed here that repeat biopsies can identify new HER2-low results for patients who were previously ineligible for T-DXd; and therefore, we think that a repeat biopsy could be considered if feasible and safe. Also, if a repeat biopsy is performed for any reason, but mainly upon metastatic recurrence, receptors should be retested,” said Dr. Bar.

After Dr. Bar’s presentation, Barbara Pistilli, MD served as a discussant. She noted the increased HER2-low results over successive biopsies. “However, here the question is, are these results related to the changes in the analytical methods over the past 20 years or the changes in the guidelines in terms of definition of HER2 status, or are they more related to a true evolution of HER2 status with the evolution of the disease?” she said during her presentation. Dr. Pistilli is chair of the breast disease committee at Gustave Roussy in Villejuif, France.

She also said that HER2 expression can vary even between different parts of the same tumor and called for alternative methods to following HER2 expression. “I don’t think that we can follow our patients with multiple biopsies over the disease evolution, so we have to find other tools, such as target-positive [circulating tumor cells], or antibody-radiolabeled PET scan in order to better follow the intermetastasis target heterogeneity over time, and finally define what is the optimal ADC sequential strategy for each patient,” said Dr. Pistilli.

Comoderator Michael Danso, MD, also weighed in when asked for comment. 

“It was an important trial to show that serial biopsies potentially allow more patients to receive trastuzumab deruxtecan,” said Dr. Danso, who is the research director at Virginia Oncology Associates, Norfolk. However, he pointed out the concerns of a statistician who had spoken up during the question-and-answer session who said that the positive results could simply be the consequence of repeated testing. “If you do a test often enough, statistically you’re going to get a difference in outcome. That was an important point made. Also, if you’re going to get 100% of patients who are eventually going to [develop HER2-low status], the question is, can you just treat everybody with trastuzumab deruxtecan and not do these sequential biopsies? Obviously that is subject to cost; it’s subject to toxicity as well, so you probably want documentation that there is a HER2-low result,” said Dr. Danso.

Dr. Bar has no relevant financial disclosures. Dr. Pistilli has consulted for or advised AstraZeneca, Daiichi Sankyo/UCB Japan, Myriad Genetics, Novartis, PIERRE FABRE, and Puma Biotechnology. She has received research funding through her institution from AstraZeneca, Daiichi Sankyo, Gilead Sciences, Merus, Pfizer, and Puma Biotechnology. She has received travel or accommodation expenses from AstraZeneca, Daiichi Sankyo Europe, MSD Oncology, Novartis, Pfizer, and Pierre Fabre. Dr. Danso has received honoraria from Amgen and has consulted or advised Immunomedics, Novartis, Pfizer, and Seagen.

*This story was updated on 6/13/2023.

Triple-negative breast cancer (TNBC) is characterized by the absence of hormonal receptors and human epidermal growth factor receptor 2 (HER2) expression. Historically, treatments targeting HER2 were found to be ineffective in patients with TNBC and known HER2-zero status. Researchers more recently identified a new type of TNBC involving patients with low expression of HER2. Patients with this type of breast cancer, now referred to as HER2-low, have immunohistochemical (IHC) analysis scores of 1+ or 2+ and negative in situ hybridization (ISH) stain.

In a new study, patients with TNBC who initially tested as having HER2-zero status were later found to have HER2-low status following repeated biopsies. These HER2-low results were of great clinical significance for this patient population, said Yael Bar, MD, PhD, during her presentation of the research, at the annual meeting of the American Society of Clinical Oncology (ASCO).

Previously, the DESTINY-Breast04 trial demonstrated that the antibody-drug conjugate (ADC) trastuzumab deruxtecan (T-DXd) improved progression-free survival (PFS) and overall survival (OS) for patients with HER2-low metastatic breast cancer. “As a result [of the DESTINY-Breast04 findings], T-DXd is now approved for HER2-low but not HER2-zero triple-negative metastatic breast cancer."

“While HER2-low is detected in about 30%-50% of patients with triple-negative breast cancer, several studies have shown that HER2 status is heterogeneous and also dynamic over time, said Dr. Bar, who is an international research fellow in the breast cancer group at Mass General Cancer Center, Boston.

In the new study, Dr. Bar and her co-authors retrospectively identified 512 TNBC patients from 2000 to 2022 from an institutional database. They included core, surgical, or metastatic biopsies. Participants had a mean age of 52 years, with 54% over age 50. They were 83% White, 7% African American, 5% Asian, 3% Hispanic, and 2% other. Stage II was most common at diagnosis at 48%, followed by stage 1 (28%), stage 3 (14%), and stage IV (8%).

Most patients had undergone one (38%) or two (45%) biopsies, while 9% underwent three biopsies, 6% underwent four biopsies, and 2% underwent five or more.

Among all 512 patients in the study, 60% had a HER2-low result on their first biopsy. As of the second biopsy, 73% had at least one HER2-low result, with 13% of the first HER2-low results occurring at the second biopsy. As of the third biopsy, 81% had a HER2-low result, with 9% occurring for the first time. At the fourth biopsy, 86% had a positive result, with 8% occurring for the first time. All patients with five or more biopsies had at least one HER2-low result and none were first-time results.

At the second biopsy, a HER2-low result was detected for 32% of patients for the first time. At the third biopsy, a new HER2-low result was detected in 33%, and at the fourth biopsy, a new HER-2 result was detected in 38%.

The researchers matched early and metastatic biopsies in 71 patients, and 44% had changed status: 68% of those with a status change went HER2-low to HER2-zero, 26% from HER2-zero to HER2-low, and 6% from HER2-low to HER2-positive. Among 50 patients with matched metastatic biopsies, 33% had a change in status, with 63% going from HER2-zero to HER2-low, 31% from HER2-low to HER2-zero, and 6% from HER2-low to HER2-positive.

“We showed here that repeat biopsies can identify new HER2-low results for patients who were previously ineligible for T-DXd; and therefore, we think that a repeat biopsy could be considered if feasible and safe. Also, if a repeat biopsy is performed for any reason, but mainly upon metastatic recurrence, receptors should be retested,” said Dr. Bar.

After Dr. Bar’s presentation, Barbara Pistilli, MD served as a discussant. She noted the increased HER2-low results over successive biopsies. “However, here the question is, are these results related to the changes in the analytical methods over the past 20 years or the changes in the guidelines in terms of definition of HER2 status, or are they more related to a true evolution of HER2 status with the evolution of the disease?” she said during her presentation. Dr. Pistilli is chair of the breast disease committee at Gustave Roussy in Villejuif, France.

She also said that HER2 expression can vary even between different parts of the same tumor and called for alternative methods to following HER2 expression. “I don’t think that we can follow our patients with multiple biopsies over the disease evolution, so we have to find other tools, such as target-positive [circulating tumor cells], or antibody-radiolabeled PET scan in order to better follow the intermetastasis target heterogeneity over time, and finally define what is the optimal ADC sequential strategy for each patient,” said Dr. Pistilli.

Comoderator Michael Danso, MD, also weighed in when asked for comment. 

“It was an important trial to show that serial biopsies potentially allow more patients to receive trastuzumab deruxtecan,” said Dr. Danso, who is the research director at Virginia Oncology Associates, Norfolk. However, he pointed out the concerns of a statistician who had spoken up during the question-and-answer session who said that the positive results could simply be the consequence of repeated testing. “If you do a test often enough, statistically you’re going to get a difference in outcome. That was an important point made. Also, if you’re going to get 100% of patients who are eventually going to [develop HER2-low status], the question is, can you just treat everybody with trastuzumab deruxtecan and not do these sequential biopsies? Obviously that is subject to cost; it’s subject to toxicity as well, so you probably want documentation that there is a HER2-low result,” said Dr. Danso.

Dr. Bar has no relevant financial disclosures. Dr. Pistilli has consulted for or advised AstraZeneca, Daiichi Sankyo/UCB Japan, Myriad Genetics, Novartis, PIERRE FABRE, and Puma Biotechnology. She has received research funding through her institution from AstraZeneca, Daiichi Sankyo, Gilead Sciences, Merus, Pfizer, and Puma Biotechnology. She has received travel or accommodation expenses from AstraZeneca, Daiichi Sankyo Europe, MSD Oncology, Novartis, Pfizer, and Pierre Fabre. Dr. Danso has received honoraria from Amgen and has consulted or advised Immunomedics, Novartis, Pfizer, and Seagen.

*This story was updated on 6/13/2023.

The development of [artificial intelligence] is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other.

Bill Gates ¹

As the world emerges from the pandemic and the health care system faces new challenges, technology has become an increasingly important tool for health care professionals (HCPs). One such technology is the large language model (LLM), which has the potential to revolutionize the health care industry. ChatGPT, a popular LLM developed by OpenAI, has gained particular attention in the medical community for its ability to pass the United States Medical Licensing Exam.² This article will explore the benefits and potential pitfalls of using LLMs like ChatGPT in medicine and health care.

Benefits

HCP burnout is a serious issue that can lead to lower productivity, increased medical errors, and decreased patient satisfaction.³ LLMs can alleviate some administrative burdens on HCPs, allowing them to focus on patient care. By assisting with billing, coding, insurance claims, and organizing schedules, LLMs like ChatGPT can free up time for HCPs to focus on what they do best: providing quality patient care.⁴ ChatGPT also can assist with diagnoses by providing accurate and reliable information based on a vast amount of clinical data. By learning the relationships between different medical conditions, symptoms, and treatment options, ChatGPT can provide an appropriate differential diagnosis (Figure 1).

Imaging medical specialists like radiologists, pathologists, dermatologists, and others can benefit from combining computer vision diagnostics with ChatGPT report creation abilities to streamline the diagnostic workflow and improve diagnostic accuracy (Figure 2).

AI Bill of Rights

In order to protect the American public, the White House Office of Science and Technology Policy (OSTP) has released a blueprint for an AI Bill of Rights that emphasizes 5 principles to protect the public from the harmful effects of AI models, including safe and effective systems; algorithmic discrimination protection; data privacy; notice and explanation; and human alternatives, considerations, and fallback (Figure 3).¹⁷

One of the biggest challenges with LLMs like ChatGPT is the prevalence of inaccurate information or so-called hallucinations.¹⁶ These inaccuracies stem from the inability of LLMs to distinguish between real and fake information. To prevent hallucinations, researchers have proposed several methods, including training models on more diverse data, using adversarial training methods, and human-in-the-loop approaches.²¹ In addition, medicine-specific models like GatorTron, medPaLM, and Almanac were developed, increasing the accuracy of factual results.^22-24 Unfortunately, only the GatorTron model is available to the public through the NVIDIA developers’ program.²⁵

Despite these shortcomings, the future of LLMs in health care is promising. Although these models will not replace HCPs, they can help reduce the unnecessary burden on them, prevent burnout, and enable HCPs and patients spend more time together. Establishing an official hospital AI oversight governing body that would promote best practices could ensure the trustworthy implementation of these new technologies.²⁶

Conclusions

The use of ChatGPT and other LLMs in health care has the potential to revolutionize the industry. By assisting HCPs with administrative tasks, improving the accuracy and reliability of diagnoses, and engaging patients, ChatGPT can help health care organizations provide better care to their patients. While LLMs are not a substitute for human interaction and personalized care, they can augment the work of HCPs, making health care more accessible and efficient. As the health care industry continues to evolve, it will be exciting to see how ChatGPT and other LLMs are used to improve patient outcomes and quality of care. In addition, AI technologies like ChatGPT offer enormous potential in medical education and research. To ensure that the benefits outweigh the risks, developing trustworthy AI health care products and establishing oversight governing bodies to ensure their implementation is essential. By doing so, we can help HCPs focus on what matters most, providing high-quality care to patients.

Acknowledgments

This material is the result of work supported by resources and the use of facilities at the James A. Haley Veterans’ Hospital.

The development of [artificial intelligence] is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other.

Bill Gates ¹

As the world emerges from the pandemic and the health care system faces new challenges, technology has become an increasingly important tool for health care professionals (HCPs). One such technology is the large language model (LLM), which has the potential to revolutionize the health care industry. ChatGPT, a popular LLM developed by OpenAI, has gained particular attention in the medical community for its ability to pass the United States Medical Licensing Exam.² This article will explore the benefits and potential pitfalls of using LLMs like ChatGPT in medicine and health care.

Benefits

HCP burnout is a serious issue that can lead to lower productivity, increased medical errors, and decreased patient satisfaction.³ LLMs can alleviate some administrative burdens on HCPs, allowing them to focus on patient care. By assisting with billing, coding, insurance claims, and organizing schedules, LLMs like ChatGPT can free up time for HCPs to focus on what they do best: providing quality patient care.⁴ ChatGPT also can assist with diagnoses by providing accurate and reliable information based on a vast amount of clinical data. By learning the relationships between different medical conditions, symptoms, and treatment options, ChatGPT can provide an appropriate differential diagnosis (Figure 1).

Imaging medical specialists like radiologists, pathologists, dermatologists, and others can benefit from combining computer vision diagnostics with ChatGPT report creation abilities to streamline the diagnostic workflow and improve diagnostic accuracy (Figure 2).

AI Bill of Rights

In order to protect the American public, the White House Office of Science and Technology Policy (OSTP) has released a blueprint for an AI Bill of Rights that emphasizes 5 principles to protect the public from the harmful effects of AI models, including safe and effective systems; algorithmic discrimination protection; data privacy; notice and explanation; and human alternatives, considerations, and fallback (Figure 3).¹⁷

One of the biggest challenges with LLMs like ChatGPT is the prevalence of inaccurate information or so-called hallucinations.¹⁶ These inaccuracies stem from the inability of LLMs to distinguish between real and fake information. To prevent hallucinations, researchers have proposed several methods, including training models on more diverse data, using adversarial training methods, and human-in-the-loop approaches.²¹ In addition, medicine-specific models like GatorTron, medPaLM, and Almanac were developed, increasing the accuracy of factual results.^22-24 Unfortunately, only the GatorTron model is available to the public through the NVIDIA developers’ program.²⁵

Despite these shortcomings, the future of LLMs in health care is promising. Although these models will not replace HCPs, they can help reduce the unnecessary burden on them, prevent burnout, and enable HCPs and patients spend more time together. Establishing an official hospital AI oversight governing body that would promote best practices could ensure the trustworthy implementation of these new technologies.²⁶

Conclusions

The use of ChatGPT and other LLMs in health care has the potential to revolutionize the industry. By assisting HCPs with administrative tasks, improving the accuracy and reliability of diagnoses, and engaging patients, ChatGPT can help health care organizations provide better care to their patients. While LLMs are not a substitute for human interaction and personalized care, they can augment the work of HCPs, making health care more accessible and efficient. As the health care industry continues to evolve, it will be exciting to see how ChatGPT and other LLMs are used to improve patient outcomes and quality of care. In addition, AI technologies like ChatGPT offer enormous potential in medical education and research. To ensure that the benefits outweigh the risks, developing trustworthy AI health care products and establishing oversight governing bodies to ensure their implementation is essential. By doing so, we can help HCPs focus on what matters most, providing high-quality care to patients.

Acknowledgments

This material is the result of work supported by resources and the use of facilities at the James A. Haley Veterans’ Hospital.

The development of [artificial intelligence] is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other.

Bill Gates ¹

As the world emerges from the pandemic and the health care system faces new challenges, technology has become an increasingly important tool for health care professionals (HCPs). One such technology is the large language model (LLM), which has the potential to revolutionize the health care industry. ChatGPT, a popular LLM developed by OpenAI, has gained particular attention in the medical community for its ability to pass the United States Medical Licensing Exam.² This article will explore the benefits and potential pitfalls of using LLMs like ChatGPT in medicine and health care.

Benefits

HCP burnout is a serious issue that can lead to lower productivity, increased medical errors, and decreased patient satisfaction.³ LLMs can alleviate some administrative burdens on HCPs, allowing them to focus on patient care. By assisting with billing, coding, insurance claims, and organizing schedules, LLMs like ChatGPT can free up time for HCPs to focus on what they do best: providing quality patient care.⁴ ChatGPT also can assist with diagnoses by providing accurate and reliable information based on a vast amount of clinical data. By learning the relationships between different medical conditions, symptoms, and treatment options, ChatGPT can provide an appropriate differential diagnosis (Figure 1).

Imaging medical specialists like radiologists, pathologists, dermatologists, and others can benefit from combining computer vision diagnostics with ChatGPT report creation abilities to streamline the diagnostic workflow and improve diagnostic accuracy (Figure 2).

AI Bill of Rights

In order to protect the American public, the White House Office of Science and Technology Policy (OSTP) has released a blueprint for an AI Bill of Rights that emphasizes 5 principles to protect the public from the harmful effects of AI models, including safe and effective systems; algorithmic discrimination protection; data privacy; notice and explanation; and human alternatives, considerations, and fallback (Figure 3).¹⁷

One of the biggest challenges with LLMs like ChatGPT is the prevalence of inaccurate information or so-called hallucinations.¹⁶ These inaccuracies stem from the inability of LLMs to distinguish between real and fake information. To prevent hallucinations, researchers have proposed several methods, including training models on more diverse data, using adversarial training methods, and human-in-the-loop approaches.²¹ In addition, medicine-specific models like GatorTron, medPaLM, and Almanac were developed, increasing the accuracy of factual results.^22-24 Unfortunately, only the GatorTron model is available to the public through the NVIDIA developers’ program.²⁵

Despite these shortcomings, the future of LLMs in health care is promising. Although these models will not replace HCPs, they can help reduce the unnecessary burden on them, prevent burnout, and enable HCPs and patients spend more time together. Establishing an official hospital AI oversight governing body that would promote best practices could ensure the trustworthy implementation of these new technologies.²⁶

Conclusions

The use of ChatGPT and other LLMs in health care has the potential to revolutionize the industry. By assisting HCPs with administrative tasks, improving the accuracy and reliability of diagnoses, and engaging patients, ChatGPT can help health care organizations provide better care to their patients. While LLMs are not a substitute for human interaction and personalized care, they can augment the work of HCPs, making health care more accessible and efficient. As the health care industry continues to evolve, it will be exciting to see how ChatGPT and other LLMs are used to improve patient outcomes and quality of care. In addition, AI technologies like ChatGPT offer enormous potential in medical education and research. To ensure that the benefits outweigh the risks, developing trustworthy AI health care products and establishing oversight governing bodies to ensure their implementation is essential. By doing so, we can help HCPs focus on what matters most, providing high-quality care to patients.

Acknowledgments

This material is the result of work supported by resources and the use of facilities at the James A. Haley Veterans’ Hospital.

Unlike many of you, maybe even most of you, I can recall when my office records were handwritten, some would say scribbled, on pieces of paper. They were decipherable by a select few. Some veteran assistants never mastered the skill. Pages were sometimes lavishly illustrated with drawings of body parts, often because I couldn’t remember or spell the correct anatomic term. When I needed to send a referring letter to another provider I typed it myself because dictating never quite suited my personality.

When I joined a small primary care group, the computer-savvy lead physician and a programmer developed our own homegrown EHR. It relied on scanning documents, as so many of us still generated handwritten notes. Even the most vociferous Luddites among us loved the system from day 2.

However, for a variety of reasons, some defensible some just plain bad, our beloved system needed to be replaced after 7 years. We then invested in an off-the-shelf EHR system that promised more capabilities. We were told there would be a learning curve but the plateau would come quickly and we would enjoy our new electronic assistant.

You’ve lived the rest of the story. The learning curve was steep and long and the plateau was a time gobbler. I was probably the most efficient provider in the group, and after 6 months I was leaving the office an hour later than I had been and was seeing the same number of patients. Most of my coworkers were staying and/or working on the computer at home for an extra 2 hours. This change could be easily documented by speaking with our spouses and children. I understand from my colleagues who have stayed in the business that over the ensuing decade and a half since my first experience with the EHR, its insatiable appetite for a clinician’s time has not abated.

The authors of a recent article in Annals of Family Medicine offer up some advice on how this tragic situation might be brought under control. First, the investigators point out that the phenomenon of after-hours EHR work, sometimes referred to as WOW (work outside of work), has not gone unnoticed by health system administrators and vendors who develop and sell the EHRs. However, analyzing the voluminous data necessary is not any easy task and for the most part has resulted in metrics that cannot be easily applied over a variety of practice scenarios. Many health care organizations, even large ones, have simply given up and rely on the WOW data and recommendations provided by the vendors, obviously lending the situation a faint odor of conflict of interest.

The bottom line is that after a couple of decades, there is no well defined way of quantifying the time-gobbling effect of an EHR system. It would seem to me just asking the spouses and significant others of the clinicians would be sufficient. But, authors of the paper have more specific recommendations. First, they suggest that time working on the computer outside of scheduled time with patients should be separated from any other calculation of EHR usage. They encourage vendors and time-management researchers to develop standardized and validated methods for measuring active EHR use. And, finally they recommend that all EHR work done outside of time scheduled with patients be attributed to WOW. They feel that clearly labeling it work outside of work offers health care organizations a better chance of developing policies that will address the scourge of burnout.

This, unfortunately, is another tragic example of how clinicians have lost control of our work environments. The fact that 20 years have passed and there is still no standardized method for determining how much time we spend on the computer is more evidence we need to raise our voices.
 

Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.” Other than a Littman stethoscope he accepted as a first-year medical student in 1966, Dr. Wilkoff reports having nothing to disclose. Email him at pdnews@mdedge.com.

Unlike many of you, maybe even most of you, I can recall when my office records were handwritten, some would say scribbled, on pieces of paper. They were decipherable by a select few. Some veteran assistants never mastered the skill. Pages were sometimes lavishly illustrated with drawings of body parts, often because I couldn’t remember or spell the correct anatomic term. When I needed to send a referring letter to another provider I typed it myself because dictating never quite suited my personality.

When I joined a small primary care group, the computer-savvy lead physician and a programmer developed our own homegrown EHR. It relied on scanning documents, as so many of us still generated handwritten notes. Even the most vociferous Luddites among us loved the system from day 2.

However, for a variety of reasons, some defensible some just plain bad, our beloved system needed to be replaced after 7 years. We then invested in an off-the-shelf EHR system that promised more capabilities. We were told there would be a learning curve but the plateau would come quickly and we would enjoy our new electronic assistant.

You’ve lived the rest of the story. The learning curve was steep and long and the plateau was a time gobbler. I was probably the most efficient provider in the group, and after 6 months I was leaving the office an hour later than I had been and was seeing the same number of patients. Most of my coworkers were staying and/or working on the computer at home for an extra 2 hours. This change could be easily documented by speaking with our spouses and children. I understand from my colleagues who have stayed in the business that over the ensuing decade and a half since my first experience with the EHR, its insatiable appetite for a clinician’s time has not abated.

The authors of a recent article in Annals of Family Medicine offer up some advice on how this tragic situation might be brought under control. First, the investigators point out that the phenomenon of after-hours EHR work, sometimes referred to as WOW (work outside of work), has not gone unnoticed by health system administrators and vendors who develop and sell the EHRs. However, analyzing the voluminous data necessary is not any easy task and for the most part has resulted in metrics that cannot be easily applied over a variety of practice scenarios. Many health care organizations, even large ones, have simply given up and rely on the WOW data and recommendations provided by the vendors, obviously lending the situation a faint odor of conflict of interest.

The bottom line is that after a couple of decades, there is no well defined way of quantifying the time-gobbling effect of an EHR system. It would seem to me just asking the spouses and significant others of the clinicians would be sufficient. But, authors of the paper have more specific recommendations. First, they suggest that time working on the computer outside of scheduled time with patients should be separated from any other calculation of EHR usage. They encourage vendors and time-management researchers to develop standardized and validated methods for measuring active EHR use. And, finally they recommend that all EHR work done outside of time scheduled with patients be attributed to WOW. They feel that clearly labeling it work outside of work offers health care organizations a better chance of developing policies that will address the scourge of burnout.

This, unfortunately, is another tragic example of how clinicians have lost control of our work environments. The fact that 20 years have passed and there is still no standardized method for determining how much time we spend on the computer is more evidence we need to raise our voices.
 

Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.” Other than a Littman stethoscope he accepted as a first-year medical student in 1966, Dr. Wilkoff reports having nothing to disclose. Email him at pdnews@mdedge.com.

Unlike many of you, maybe even most of you, I can recall when my office records were handwritten, some would say scribbled, on pieces of paper. They were decipherable by a select few. Some veteran assistants never mastered the skill. Pages were sometimes lavishly illustrated with drawings of body parts, often because I couldn’t remember or spell the correct anatomic term. When I needed to send a referring letter to another provider I typed it myself because dictating never quite suited my personality.

When I joined a small primary care group, the computer-savvy lead physician and a programmer developed our own homegrown EHR. It relied on scanning documents, as so many of us still generated handwritten notes. Even the most vociferous Luddites among us loved the system from day 2.

However, for a variety of reasons, some defensible some just plain bad, our beloved system needed to be replaced after 7 years. We then invested in an off-the-shelf EHR system that promised more capabilities. We were told there would be a learning curve but the plateau would come quickly and we would enjoy our new electronic assistant.

You’ve lived the rest of the story. The learning curve was steep and long and the plateau was a time gobbler. I was probably the most efficient provider in the group, and after 6 months I was leaving the office an hour later than I had been and was seeing the same number of patients. Most of my coworkers were staying and/or working on the computer at home for an extra 2 hours. This change could be easily documented by speaking with our spouses and children. I understand from my colleagues who have stayed in the business that over the ensuing decade and a half since my first experience with the EHR, its insatiable appetite for a clinician’s time has not abated.

The authors of a recent article in Annals of Family Medicine offer up some advice on how this tragic situation might be brought under control. First, the investigators point out that the phenomenon of after-hours EHR work, sometimes referred to as WOW (work outside of work), has not gone unnoticed by health system administrators and vendors who develop and sell the EHRs. However, analyzing the voluminous data necessary is not any easy task and for the most part has resulted in metrics that cannot be easily applied over a variety of practice scenarios. Many health care organizations, even large ones, have simply given up and rely on the WOW data and recommendations provided by the vendors, obviously lending the situation a faint odor of conflict of interest.

The bottom line is that after a couple of decades, there is no well defined way of quantifying the time-gobbling effect of an EHR system. It would seem to me just asking the spouses and significant others of the clinicians would be sufficient. But, authors of the paper have more specific recommendations. First, they suggest that time working on the computer outside of scheduled time with patients should be separated from any other calculation of EHR usage. They encourage vendors and time-management researchers to develop standardized and validated methods for measuring active EHR use. And, finally they recommend that all EHR work done outside of time scheduled with patients be attributed to WOW. They feel that clearly labeling it work outside of work offers health care organizations a better chance of developing policies that will address the scourge of burnout.

This, unfortunately, is another tragic example of how clinicians have lost control of our work environments. The fact that 20 years have passed and there is still no standardized method for determining how much time we spend on the computer is more evidence we need to raise our voices.
 

Dr. Wilkoff practiced primary care pediatrics in Brunswick, Maine, for nearly 40 years. He has authored several books on behavioral pediatrics, including “How to Say No to Your Toddler.” Other than a Littman stethoscope he accepted as a first-year medical student in 1966, Dr. Wilkoff reports having nothing to disclose. Email him at pdnews@mdedge.com.