Objective: To assess the predictive value of gait speed for early death in older outpatients with cancer.
Design: Prospective bicentric observational cohort study.
Setting: The Physical Frailty in Elder Cancer patients (PF-EC) study (France).
Participants: One hundred and ninety outpatients with cancer during the first 6 months of follow up in the PF-EC study.
Measurements: The association between usual gait speed over 4 m alone (GS) or included in the short physical performance battery (SPPB) and overall survival within 6 months following a comprehensive geriatric assessment (CGA). A Cox proportional-hazard regression model was performed in non-survivors for clinical factors from the CGA, along with c reactive protein (CRP). Two models were created to assess GS alone and from inclusion in the SPPB.
Results: The mean age was 80.6 years, and 50.5% of the participants were men. Death occurred in 11% (n=22) of the participants within the 6 month follow up period. Of these participants, 98% had solid cancers, and 33% had a metastatic disease. A GS < 0.8 m/s (HR=5.6, 95%CI=1.6-19.7, p=0.007), a SPPB < 9 (HR=5.8, 95%CI=1.6-20.9, p=0.007) and a CRP of 50 mg/l or greater (p<0.0001) were significantly associated with early death in the two multivariate analyses. Cancer site and extension were not significantly associated with early death.
Conclusion: Walking tests are associated with early death within the 6 month follow up period after a CGA independent of cancer site and cancer extension. GS alone < 0.8 m/s is at least as efficacious as the SPPB in predicting this outcome. GS alone could be used routinely as a marker of early death to adapt oncologic therapeutics. Further studies are needed to validate these preliminary data.