Personalized risk assessment of frailty in long-term cancer survivors

Due to the growth of the cancer survivor population, strategies to facilitate efficient delivery of survivorship care are critical to reduce the risk of adverse events associated with frailty. The objective of this study was to develop a risk stratification tool to identify long-term survivors at the highest risk of becoming frail 5-10 years after cancer diagnosis.

We used the Surveillance, Epidemiology, and End Results (SEER) dataset linked with Medicare data to identify patients with stage I-III breast, prostate, colon, or rectal cancers who lived at least five years from diagnosis and were not severely frail at year five post-diagnosis. Frailty was assessed using the claims-based Kim Frailty Index (FI) categorized by recommended thresholds. Restricted mean survival time (RMST) regression was used to identify clinical and demographic characteristics associated with frailty progression, defined as an increased category of the FI. Significant predictors were used to create clinical prediction rules and stratify survivors into low, intermediate, and high-risk groups.

A total of 87,229 five-year survivors were included. At five years from diagnosis (time zero), 22 % of patients not frail at cancer diagnosis had new onset frailty and were mildly or moderately frail; at 10 years from diagnosis, 61 % had developed new or worsening frailty. Advanced age, comorbidities (RMST ratios ranging from 0.67 [95 % CI 0.65-0.70] to 0.80 [95 % CI 0.77-0.84], baseline moderate frailty at cancer diagnosis (RMST ratios ranging from 0.79 [95 % CI 0.76-0.83] to 0.86 [95 % CI 0.83-0.90]) and at five years post-diagnosis (RMST ratios ranging from 0.63 [95 % CI 0.62-0.64] to 0.71 [95 % CI 0.69-0.73]), living in a high poverty area (RMST ratios ranging from 0.91 [95 % CI 0.87-0.94] to 0.96, [95 % CI 0.93-0.99], and systemic treatments four to five years post-diagnosis (RMST ratios ranging from 0.77 [95 % CI: 0.70-0.84] to 0.86, [95 % CI: 0.84-0.89] were associated with less average time without frailty.

Age, comorbidities, prior frailty, and late treatment were associated with frailty in older breast, prostate, colon, and rectal cancer survivors. This risk stratification model can be used by clinicians to assess cancer and age-related risk of frailty and facilitate timely intervention.