Geriatric Assessment Metrics Are Associated with Hospital Length of Stay in Pre-Bone Marrow Transplant Myeloma Patients

A geriatric assessment (GA) is a global approach to improve healthy aging, wherein occult problems are assessed and intervened upon using a multidisciplinary method. A GA is feasible and can predict chemotherapy-induced toxicity and overall survival in cancer patients. Biomarkers of aging are also being explored as objective and reproducible measures of health and fitness. p16^INK4a (p16) is a marker of cellular senescence that rises exponentially with chronologic age and is influenced by factors such as physical activity, smoking, and solid tumor chemotherapy. Here, we investigated the relationship of both the GA and molecular (i.e. p16) metrics in pre-bone marrow transplant (BMT) multiple myeloma (MM) patients. We selected this group for our studies as BMT patients are a vulnerable cohort in which transplant eligibility is subjective and age related. BMT patients are also at high risk for adverse events and treatment toxicity. In this preliminary analysis, we explored the predictive value of GA metrics and p16 with inpatient length of stay (LOS) during autologous BMT.

Methods: We performed a pilot prospective cohort study on 55 MM patients during their pre-transplant evaluation. MM patients >18 years completed GA assessments related to physical function, distress, comorbidities, social support, and cognition. Patients completed surveys using the Brief Fatigue Inventory (BFI) (scale 1-10; moderate fatigue 4-6, severe fatigue 7+); Hospital Anxiety and Depression (HADS) (borderline case 8-10, definite case 11+); medical outcome study-social support survey (MOS-SSS) (scale 0-100, higher scores indicated greater support), Human Activity Profile (HAP) maximum activity score (MAS) and HAP-adjusted activity score (AAS), a 94-item questionnaire ranking tasks according to energy use validated in the BMT population, with higher scores indicating higher activity (Herzberg BBMT 2010). Objective measures of physical activity were measured using the Short Physical Performance Battery (SPPB) (range 0-12; impairment <9) and cognition was evaluated using the Modified Mini Mental Status exam (3MS). At the pre-transplant evaluation, p16 mRNA was measured in peripheral blood T-cells using established laboratory techniques (Liu Aging Cell 2009). The association between GA metrics and p16 were evaluated using Spearman's correlation coefficient. Univariable generalized linear models were fit to model LOS as a function of GA metrics or p16.

Results: The median patient age was 61 (range 42-76). Most patients exhibited early stage disease (ISS Stage 1 53%) with minimal comorbidities (HCT-CI median 1; range 0-8) and a median of 2 prior lines of treatment (range 1-11). Pre-transplant Karnofsky Performance Status (KPS) was reported as 70% (n=10), 80% (n=10), 90% (n=15) and 100% (n=12). 7 patients did not proceed with BMT, 1 inpatient for BMT. Patients reported moderate fatigue by BFI (median 4.3; range 0-9.8), with minimal anxiety or depression as measured by the HADS. Self-reported physical activity by HAP-MAS was 73 (range 30-94) and HAP-AAS was 64 (range 20-94). Patients reported high levels of social support (median 86.7%; range 18.2-100) by MOS-SSS. Objective measures of physical function were also high as measured by the SPPB (median 10; range 4-12) and no cognitive impairment was identified by the 3MS. p16 expression was adjusted for age and did not correlate with GA tools including BFI, HADS, HAP-AAS, HAP-MAS, MOS-SSS, SPPB or 3MS. The median length LOS during transplant was 16 days (range 12-36). Univariate analysis revealed that SPPB score was significantly associated with LOS, where each one unit increase in physical performance corresponded to an average LOS decrease of 0.63 days (p=0.04). Self-reported activity by HAP-AAS also correlated with LOS (p=0.05). LOS was not influenced by p16, age, KPS or HCT-CI. Age and HCT-CI had no relationship with SPPB scores, but KPS did (p=0.03727).

Conclusions: A comprehensive GA can be used to identify factors that contribute to BMT outcomes. Physical function appears to be most predictive of hospital LOS as measured by SPPB or a detailed self-report of physical function. Baseline p16 levels had no relationship with GA metrics in this selected population. A standardized approach for determining patient fitness including SPPB and HAP-AAS assessments may improve treatment tolerance, reduce hospital LOS, and decrease the risk for adverse outcomes in BMT populations.