Real-World Prospective Evaluation of Different Geriatric Assessment Tools in Unselected Elderly Patients with Symptomatic Myeloma

The diagnosis of MM that requires therapy in elderly individuals is increasing. The management of such patients is challenging due to several factors, besides disease characteristics and age that affect outcome. Geriatric assessment (GA) is a multidimensional diagnostic approach that collects data on the medical, psychosocial and functional capabilities and limitations of elderly patients to develop treatment and care decisions and improve the use of health care resources. The IMWG has proposed a simplified GA ("frailty score") based on 3 tools (Katz Activity of Daily Living (ADL), the Lawton Instrumental Activity of Daily Living (IADL) and the Charlson Comorbidity Index (CCI)) as a measure of frailty (Palumbo et al, Blood 2015). "Frailty score" was developed on patients who participated in clinical trials, thus, may have a selection bias. In our current "Real-World" study, we prospectively evaluated consecutive patients >65 years, irrespective of their participation in clinical trials and physical condition, in order to evaluate several different GA tools and comorbidity indices, along with standard disease related prognostic factors. The following tools were used: G8 geriatric assessment screening tool (G8-GAS), VES-13, GDS, Katz ADL, Lawton IADL, MMSE, KPS (%), ECOG PS, number of falls in the past 1 & 6 months, lower-extremity function and disability in elderly tool, nutritional assessment tools (DETERMINE and Mini Nutritional Assessment), social support score, cognition evaluation tools (MMSE), Geriatric Depression Scale and comorbidity indices (CCI, CIRS-G, ACE-27 tool).

Since January 2012, 120 consecutive patients >65 years were diagnosed with symptomatic MM in our center (Department of Clinical Therapeutics, University of Athens) and had a GA. The median age of patients with a GA was 76 years (range 66-92); 55% were males; 26% had ISS-1, 24% ISS-2 and 50% ISS-3. In 100 patients cytogenetics were available: 19% had high risk cytogenetics (del17p or t(4;14)). Median eGFR was 60 ml/min/1.73 m² and 22% had eGFR<30 ml/min/1.73 m². Treatment was based on IMiDs in 47% of patients (thalidomide in 13% and lenalidomide in 34%) and proteasome inhibitors (mainly bortezomib) in 53%. At least PR was achieved by 78% of evaluable patients. Median follow up was 20 months and 2-year overall survival (OS) was 71%. Age was associated with OS and the risk of early death (<3 months): the respective HRs for death for ages ≤70 vs 71-80 vs >80 years was 1, 1.5 and 3 and early death rates were 3%, 8% and 20%, respectively. ISS was associated with OS (p=0.004) but the presence of high risk cytogenetics was not (2-year OS 75% vs 68%, p=0.714). There was no significant difference in the OS according to different types of primary therapy (p=0.593).

Per IMWG "frailty score", 29% were fit, 17% intermediately fit and 54% frail; the respective 2-year OS was 77%, 81% and 62%. The differences in the allocation of patients in frailty categories compared to the original IMWG cohort (39%, 31% & 30%) is probably due to the fact that our patients were unselected. In univariate analysis several different GA tools showed prognostic significance: number of falls in the past 6 months (0 vs ≥1, p=0.002), lower extremity function (score <9 vs ≥9, p=0.014), mini nutritional assessment (score <11 vs ≥11, p=0.014), G8-GAS (score <12 vs ≥12, p<0.001), KPS <50% (p<0.001), ECOG PS >2 (p=0.04) and MMSE (score ≥6 vs <6, p=0.024). There was an association of early death with KPS ≤50% (p=0.003), ECOG PS >2 (p=0.05), Geriatric depression score (p=0.018) and G8-GAS score (p=0.015). IMWG "frailty score" was not associated with early death. In multivariate analysis, which included ISS and age, number of falls in the past 6 months (0 vs ≥1, HR: 4.7, p=0.007) and score <12 in the G8-GAS tool (HR: 4.7, p=0.004) were independent factors for survival. Addition of cytogenetics did not change the multivariate model. We also evaluated IMWG "frailty score" in a multivariate analysis, which included ISS stage, and we did not find statistical significance for OS.

In conclusion, in elderly myeloma patients, G8-GAS provides prognostic information related to the risk of early death and overall survival, independently from disease characteristics and the treatment type. IMWG "frailty score" provides a simple tool which may be useful for patients fit to participate in clinical trials but in unselected, "Real-World", patients may have reduced prognostic performance.