Comparison between Vulnerable Elders Survey-13 and Comprehensive Geriatric Assessment in Predicting Clinical Outcomes in Aggressive Lymphomas

Background: Aggressive non-Hodgkin lymphoma (aNHL) is common in older adults. Comprehensive geriatric assessment (CGA) has emerged as a tool for assessing older adults with cancer but can be challenging to implement in aNHL due to logistical and time constraints from the need for urgent therapy. VES-13 is a brief 13-question self-administered survey that has shown promise in identifying vulnerable older adults with cancer and is easily completed in less than five minutes; however, no data have compared the performance of the VES-13 compared to CGA in identifying vulnerable patients with aNHL. Thus, we sought to compare the performance of the VES-13 and CGA for predicting clinical outcomes in older adults with aNHL.

Methods: We conducted a longitudinal prospective study of adults with aggressive aNHL at two academic centers and their community affiliates. Eligible patients were age ≥65 years at time of initiation of chemoimmunotherapy. At baseline an oncologist and research assistant administered the Fondazione Italiana Linfomi (FIL) CGA for all patients and categorized patients as frail or not frail. Patients also completed the VES-13 at baseline, with a score of 3+ denoting vulnerability. We assessed the association of VES-13 and CGA with death within 1-year post therapy initiation using Fisher’s exact test due to a small number of events. We obtained the number of unplanned hospitalizations within 6 months of treatment start and rates of chemotherapy dose reduction, grade 3+ nonhematologic toxicity, and death or disease progression within 1-year post therapy initiation using the electronic health record. We compared receiver operating curves (ROC) for the outcomes of interest between VES-13 and CGA. We also compared the ROC characteristics between VES-13 and CGA added to high age-adjusted International Prognostic Index (aaIPI) score for death or progression within 1 year post therapy initiation.

Results: 105 patients were enrolled between 9/2020 and 1/2023. The median age was 73 years (range: 65-99, 41.9% >=75) and most were male (64.8%). The majority had Diffuse Large B-Cell Lymphoma (70.5%), aaIPI of 2-3 (53.8%), and ECOG performance status of 0-1 (82.9%). 40.0%, 9.5%, and 50.5% were fit, unfit, and frail on CGA, respectively, and 45.7% of patients were vulnerable on the VES-13. The most common treatment was R-CHOP (60.0%) followed by R-mini-CHOP (13.3%). Vulnerable patients on VES-13 were more likely to receive dose-attenuated chemotherapy (33.3% vs 10.5%, p=0.007). Vulnerability on VES-13 (16.7% vs 0%, p=0.001) and frail on CGA (15.1% vs 0%, p=0.006) were both significantly associated with death within 1 year of therapy initiation. For CGA frail patients, 45.3% had an unplanned hospitalization, 38.5% had grade 3+ toxicity, 26.4% had a dose reduction, and 41.5% had disease progression/death. For VES-13 vulnerable patients, 47.9% had an unplanned hospitalization, 40.5% had grade 3+ toxicity, 31.3% had a dose reduction, and 41.7% had death/disease progression. Both VES-13 and CGA models, respectively, were comparable on ROC area under the curve (AUC) characteristics for unplanned hospitalization (0.59 vs 0.58, p=0.79), grade 3+ nonhematologic toxicity (0.60 vs 0.59, p=0.87), dose reduction (0.61 vs 0.55, p=0.50), and death/disease progression within 1 year of therapy initiation (0.62 vs 0.62, p=0.98). In models for death/disease progression within 1 year post therapy initiation, the AUC for high aaIPI, high aaIPI and frailty on CGA, and high aaIPI and vulnerable on VES-13 were 0.72, 0.78, and 0.76, respectively.

Conclusion: The VES-13 and FIL CGA both identify older adults with high rates of unplanned hospitalization, grade 3+ toxicity, dose reductions, and disease progression/death. The VES-13 was similar to FIL CGA in associations with these important outcomes but may be a more practical instrument for vulnerability assessment, highlighting its potential use in clinical practice as a tool to identify aNHL patients at higher risk of adverse events and those in need of formal geriatric assessment.