Cancer Treatment Among the Elderly in the United States: Complexities of Aging and Comorbidities

Abstract

Background: As the life expectancy of Americans becomes longer, the number of individuals who are diagnosed with cancer and other comorbidities increases rapidly. Management of these patients can become increasingly complicated as physicians administer multiple combinations of drugs and therapies. However, the complexities of how age and comorbidities affect receipt of cancer treatment are not well understood. Aim: To explore the association between age, comorbidities, and subsequent cancer treatment in the elderly diagnosed with the four most common types of cancer in the United States. Methods: We used SEER-Medicare data, which covers 28% of the U.S. population, to explore the association between age, comorbidities, and receipt of cancer treatment within 6 months of diagnosis in 727,136 individuals over 65 years old and diagnosed with breast, colorectal, lung, and prostate cancer from 1992-2011. Comorbidity burden was measured using the Charlson Comorbidity Index (CCI) and analyzed as four quantities (Q1: lowest CCI score to Q4: highest CCI score). Poisson regression models were used to assess the associations between comorbidities and cancer treatment, and whether age modified this relationship. Results: Cancer treatment proportion declined rapidly with age for all cancers while median CCI scores increased with age among breast, colorectal, prostate cancer patients and appeared stable among lung cancer patients. For example, individuals aged 76-99 had higher CCI scores ( P < 0.001) and were less likely to be treated (69.8% vs. 81.1% of those age 66-75 year; P < 0.001). After adjustment for potential confounders, we found that high CCI scores (Q3-Q4) were associated with substantially lower cancer treatment rates compared to low CCI score (Q1) in all cancer patients aged 76-99. Regarding individuals aged 66-75, high CCI scores (Q3-Q4) were not associated with lower colorectal cancer treatment rates, and only the highest CCI score group (Q4) was associated with a modest reduction in breast and prostate cancer treatment rates compared to low CCI score (Q1) (PR [95% CI]: 0.97 [0.95-0.99] and 0.91 [0.88-0.93], respectively). Additional multivariable analysis showed that older patients (aged 76-99) with low CCI score (Q1-Q2) had equal or lower treatment rates compared to younger patients (aged 66-75) with the highest CCI scores (Q4). Conclusion: Our findings suggested that among those aged 66-75 years, comorbidities are less likely to influence the receipt of treatment when compared to individuals aged 76-99. The potential harms and benefits of treatment given these age by comorbidity interactions are not clear, but using curative interventions that only have a modest benefit in a highly comorbid aging population could potentially decrease patients' quality of life.