Compass: a prospective study comparing clinical evaluation with different geriatric screening methods in newly diagnosed elderly multiple myeloma patients

Background Multiple myeloma (MM) affects primarily older patients. Frail older MM patients benefit from a less intensive therapeutic approach to prevent severe drug-related toxicities. To diagnose frailty, several geriatric risk scores have been developed. However, comprehensive geriatric assessment (cGA) is not widely adopted in daily practice and many physicians still prefer to rely on their clinical skills to judge if a patient is fit or frail. Methods The Compass trial is a prospective study in newly diagnosed MM patients ≥ 70 y comparing frailty assessed by clinical assessment (CA) and geriatric screening (GA) at diagnosis and after 3 months of treatment. Initial geriatric screening was performed by the G8 score followed by cGA if the G8 score was ≤14/17, and by IMWG score calculation. CA was performed by the treating physician and G8 independently by a trained health care worker. Results Between 04/2017 and 10/2019 200 patients were enrolled from 20 Belgian hematological sites. 74% of patients were ≥ 75y at enrollment. By CA 43% of patients were scored as frail whereas 69% had a geriatric risk profile by independently performed G8. Compared with patients who scored fit both by CA and G8 (fit-fit), patients who scored fit by CA but frail by G8 (fit-frail) were older (p=0,002), had reduced nutritional status (p<0,001), more recent weight loss (p<0,001), received more polypharmacy (p<0,001) and considered their own health status as poorer (p<0,001). In this fit-frail cohort, 23% of physicians modified the upfront treatment regimen based on the information from the G8 compared to 8% in the fit-fit group (p=0.04). During treatment, additional dose reductions were performed in 45% of the fit-frail vs 33% of the fit-fit cohort (p=ns). In all patients with a G8≤ 14/17, full cGA was performed. CA fit but G8 frail patients were more independent on ADL (63% vs 30,3%; p <0,001), iADL (56,5% vs 26,3%; p <0.001) and had less cognitive impairment (8,4% vs 30,7%; p=0,004) compared with patients scored frail by both CA and G8. The patient subgroup fit by CA but frail by G8 score were primarily categorized into intermediate fitness (31%) and frail (57%) subgroups of the IMWG score. After 3 months of treatment, the majority of evaluable patients remained in the same category (fit or frail) by CA and by G8 (respectively 82% and 80%) and only a small proportion of frail patients were reclassified as fit by CA (5%) and by G8 (6%) reinforcing that frailty status at diagnosis is not driven by myeloma-related symptoms. Conclusion Treatment optimization in elderly MM patients requires a frailty adapted approach. We demonstrated that clinical judgment alone underestimates the geriatric risk profile in 25% of newly diagnosed elderly MM patients requiring more reactive modifications in the treatment regimen. Based on our study results minimal geriatric screening tools like G8 are recommended in all elderly MM patients before treatment initiation. Multiple myeloma (MM) affects primarily older patients. Frail older MM patients benefit from a less intensive therapeutic approach to prevent severe drug-related toxicities. To diagnose frailty, several geriatric risk scores have been developed. However, comprehensive geriatric assessment (cGA) is not widely adopted in daily practice and many physicians still prefer to rely on their clinical skills to judge if a patient is fit or frail. The Compass trial is a prospective study in newly diagnosed MM patients ≥ 70 y comparing frailty assessed by clinical assessment (CA) and geriatric screening (GA) at diagnosis and after 3 months of treatment. Initial geriatric screening was performed by the G8 score followed by cGA if the G8 score was ≤14/17, and by IMWG score calculation. CA was performed by the treating physician and G8 independently by a trained health care worker. Between 04/2017 and 10/2019 200 patients were enrolled from 20 Belgian hematological sites. 74% of patients were ≥ 75y at enrollment. By CA 43% of patients were scored as frail whereas 69% had a geriatric risk profile by independently performed G8. Compared with patients who scored fit both by CA and G8 (fit-fit), patients who scored fit by CA but frail by G8 (fit-frail) were older (p=0,002), had reduced nutritional status (p<0,001), more recent weight loss (p<0,001), received more polypharmacy (p<0,001) and considered their own health status as poorer (p<0,001). In this fit-frail cohort, 23% of physicians modified the upfront treatment regimen based on the information from the G8 compared to 8% in the fit-fit group (p=0.04). During treatment, additional dose reductions were performed in 45% of the fit-frail vs 33% of the fit-fit cohort (p=ns). In all patients with a G8≤ 14/17, full cGA was performed. CA fit but G8 frail patients were more independent on ADL (63% vs 30,3%; p <0,001), iADL (56,5% vs 26,3%; p <0.001) and had less cognitive impairment (8,4% vs 30,7%; p=0,004) compared with patients scored frail by both CA and G8. The patient subgroup fit by CA but frail by G8 score were primarily categorized into intermediate fitness (31%) and frail (57%) subgroups of the IMWG score. After 3 months of treatment, the majority of evaluable patients remained in the same category (fit or frail) by CA and by G8 (respectively 82% and 80%) and only a small proportion of frail patients were reclassified as fit by CA (5%) and by G8 (6%) reinforcing that frailty status at diagnosis is not driven by myeloma-related symptoms. Treatment optimization in elderly MM patients requires a frailty adapted approach. We demonstrated that clinical judgment alone underestimates the geriatric risk profile in 25% of newly diagnosed elderly MM patients requiring more reactive modifications in the treatment regimen. Based on our study results minimal geriatric screening tools like G8 are recommended in all elderly MM patients before treatment initiation.