Identifying older patients with advanced cancer who have a high risk for hospitalization and death could be a game changer. A new analysis suggests that goal is within reach.
Using an unsupervised machine-learning algorithm, researchers found that patients’ baseline symptom severity was associated with their risk of unplanned hospitalization and death after initiating treatment.
Compared with patients in the lowest symptom severity group, those in the moderate-severity cluster had a 36% greater likelihood of being hospitalized and a 31% greater risk of death, whereas those in the high-severity cluster had a twofold higher risk for death.
Overall, the findings suggest that moderate-to-severe baseline symptom severity is independently associated with all-cause mortality in this patient population, the authors, led by Huiwen Xu, paxil cure insomnia PhD, MHA, of the University of Texas Medical Branch, Galveston, concluded.
Assessing “symptom burden prior to treatment initiation contributes clinically useful prognostic information for treatment decision-making among older adults,” Xu and colleagues said.
The findings were published online last month in JAMA Network Open.
Patient-reported outcomes are increasingly being used to identify and manage patients’ symptom burden, an effort that can improve a patient’s quality of life and reduce emergency department visits.
Given the wide range of possible symptoms, machine learning may be a useful tool to guide risk stratification and help healthcare professionals identify patients at high risk for hospitalization and death before starting treatment.
This analysis included data from patients who completed the National Cancer Institute Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) before initiating their new treatment regimen. The algorithm clustered patients based on similar baseline symptom severities. The severity of symptoms was scored from 0 to 4 — which corresponded to none, mild, moderate, severe, and very severe symptoms — and the total severity score was calculated as the sum of 24 items (overall score range, 0-96).
Data from 706 older adults were included; patients’ mean age was 77.2 years, more than half were male, and almost 90% were non-Hispanic White. About 35% of patients had gastrointestinal cancer, and 25% had lung cancer.
The algorithm classified 310 patients in the low-severity group (43.9%), 295 in the moderate-severity group (41.8%), and 101 (14.3%) in the high-severity cluster. The mean scores were 6.3 in the low group, 16.6 in the moderate group, and 29.8 in the high group.
Compared with patients in the low-severity cluster, those in the moderate-severity cluster were significantly more likely to experience hospitalization (adjusted risk ratio [RR], 1.36, P = .046), but not those in the high-severity cluster (adjusted RR, 1.44; 95% CI, 0.99 – 2.10; P = .05). Patients in all groups had similar risks for toxic effects.
But compared with patients in the low-severity cluster, those in both the moderate- and high-severity clusters had a significantly higher risk of death (HR, 1.31; P = .04 and HR, 2.00; P < .001, respectively), after controlling for sociodemographic variables, clinical factors, study group, and practice site.
The inclusion of patient-centered assessment tools in clinical practice “can assist clinicians in treatment decision-making and supportive care recommendations,” the authors conclude. “Moreover, machine learning algorithms could identify patients with advanced cancer who are at a higher risk of poor treatment tolerability and short-term mortality,” but additional study in a more diverse patient population and involving other treatments than chemotherapy is needed.
In an invited commentary, Carolyn J. Presley, MD, stated that predicting treatment tolerability among older adults with advanced cancer is “the holy grail.”
“Estimating tolerability of cancer treatments is a continued unmet need among older adults with advanced cancer,” writes Presley, from The Ohio State University Comprehensive Cancer Center, Columbus. “Uncertain tolerability among older adults can lead to both undertreatment and overtreatment after a new cancer diagnosis. Understanding who will experience treatment-related side effects is incredibly important, as the number of new cancer drugs is rapidly increasing.”
This study was supported by grants from the National Cancer Institute and the National Institute on Aging. Xu, the study co-authors, and editorialist Presley report no relevant financial relationships.
JAMA Netw Open. Published online March 22, 2023. Full text, Editorial
Sharon Worcester, MA, is an award-winning medical journalist based in Birmingham, Alabama, writing for Medscape, MDedge and other affiliate sites. She currently covers oncology, but she has also written on a variety of other medical specialties and healthcare topics. She can be reached at [email protected] or on Twitter: @SW_MedReporter
For more from Medscape Oncology, join us on Twitter and Facebook
Source: Read Full Article