A new tool developed by researchers at Dana-Farber Cancer Institute and Brigham and Women's Hospital outperforms conventional methods for predicting whether certain medications are likely to lead to frailty in older patients with blood cancers who are taking multiple medications, a study being published online indicates.
The tool, called the Geriatric Oncology Potentially Inappropriate Medications (GO-PIM) scale, is based on a list of supportive medications that the National Comprehensive Cancer Network (NCCN) has identified as "of concern" for patients with cancer. For each such medication a patient is taking, the tool adds a point.
"A large portion of research for older adults identifies patients as ‘having polypharmacy’ – taking multiple medications at the same time, leading to a higher risk of adverse events – based on a cutoff of five or more medications; but they don’t specify which medications,” explains co-lead author Tammy Hshieh, MD, MPH, of Dana-Farber and Brigham and Women’s Hospital. Unfortunately, 50% of Americans 75 and older meet this definition for polypharmacy, making it challenging for busy oncology teams to adequately review chronic medications to prevent potential problems
“We were interested in the NCCN list of medications of concern because, in contrast to other lists, it was developed by geriatricians and oncologists to be specific for older adults with cancer,” said Hshieh “We found the GO-PIM scale carried the strongest association with frailty and can be used by oncology teams to help decrease inappropriate medication usage. The GO-PIM scale can be an effective way to potentially improve patients’ overall health.”
GO-PIM was developed jointly by the Older Adult Hematologic Malignancy Program at Dana-Farber and the Division of Aging at Brigham and Women's. A report on the scale's utility was published online by the Journal of the National Comprehensive Cancer Network.
Tammy T. Hshieh, MD, MPH
In the study, researchers conducted physical exams of 785 patients over the age of 74 who had blood cancer and were ineligible for a stem cell transplant. They found that each additional medication increased patients' relative odds of being pre-frail or frail by 8%, while each additional medication on the GO-PIM scale increased their relative odds by 65%. Compared to robust patients, frail and pre-frail patients were more likely to be taking benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), or corticosteroids. (Frailty was defined by combining two approaches, one of which is based on gait speed, grip strength, self-reported exhaustion, and low physical activity, and one based on the presence of multiple health deficits including comorbidity, cognition, and function.)
“The health risks from the medications on the NCCN list – including falls and fatigue – often outweigh their benefits, especially in older adults whose ability to tolerate side effects are reduced,” says co-lead author Clark DuMontier, MD, MPH, of Brigham and Women’s. “Our GO-PIM tool makes it easier for oncology teams to identify these medications and consider deprescribing them. This scale can also be easily converted into an automated tool that identifies high-risk medications within an electronic health record.”
The senior author of the study is Gregory Abel, MD, MPH, of Dana-Farber. Co-authors include Richard M. Stone, MD, Robert Soiffer, MD, and Jane A. Driver, MD, MPH, of Dana-Farber; Timothy Jaung, MS, Nupur E. Bahl, and Chelsea E. Hawley, PharmD, MPH, of New England Geriatric Research Education and Clinical Center, Bedford, Mass.; and Lee Mozessohn, MD, of Sunnybrook Health Services Centre, Toronto.
Research reported in this publication was supported by Dana-Farber Murphy Family Fund for Hematologic Malignancies (T.T. Hshieh, N.E. Bahl, G. Abel); Dana-Farber/Harvard Cancer Center SPORE in Multiple Myeloma (NCI of the NIH: P50 CA100707; C. DuMontier); and Boston Claude D. Pepper Older Americans Independence Center, National Institute on Aging of the NIH under award number P30-AG013679 (C. DuMontier).
Keyword: Blood Cancer
