What is the relative burden of oncologic and nononcologic mortality among long-term survivors of cancer in the US?

By Susha Cheriyedath, M.Sc.Jul 17 2023Reviewed by Sophia Coveney

In a recent study in the journal JAMA Network Open, a team of researchers provided quantitative risk estimates for both cancer-specific and non-cancer-specific outcomes by studying long-term survivors of breast, prostate, colon, and rectal cancers.

Background

Cancer is a leading cause of death worldwide. However, due to advancements in early detection, treatment, and cancer outcomes, many cancer survivors in the United States are living longer, increasing their likelihood of dying from non-cancer-related conditions.

Risk-stratified care models are crucial in allocating appropriate care between oncologists and primary care physicians. This approach, supported by various organizations, emphasizes coordination and addresses survivors' unique needs.

About the study

The present study used data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program, focusing on patients diagnosed with breast, prostate, or colorectal cancer between 2003 and 2014. Only patients who received definitive treatment and survived at least five years after diagnosis were included.

Various factors were examined, including demographic information, tumor characteristics, and treatment details. The outcomes of interest were defined based on SEER cause-of-death classification variables.

Statistical analysis involved using the least absolute shrinkage and selection operator (LASSO) method to select factors associated with cancer-specific and non-cancer-specific mortality separately.

Accelerated failure time (AFT) models were employed to account for violations of the proportional hazards assumption. Risk groups for oncologic mortality were established based on known risk factors. Cumulative incidence function curves were generated to visualize mortality rates by risk group. 

Study results

The study included a cohort of 627,702 patients diagnosed with breast, prostate, colon, or rectal cancer between 2003 and 2014. The majority were women (69.3%), and the mean age was 61.1 years.

The cohort consisted of 364,230 breast cancer patients, 118,839 prostate cancer patients, 104,488 colon cancer patients, and 40,145 rectal cancer patients. All patients had stage I, II, or III disease and received definitive treatment through surgery and/or radiotherapy.

Researchers reported that the majority of cancer patients died due to non-cancer-related causes. In the breast cancer cohort, two-thirds of the patients died from causes unrelated to their primary cancer.

Among these non-cancer-specific deaths, heart disease was the leading cause, accounting for 24.0% of the total deaths. Alzheimer’s disease contributed to 7.1% of the non-cancer-specific deaths, followed by cerebrovascular diseases at 6.6% and chronic obstructive pulmonary disease (COPD) at 6.5%.

Similarly, in the prostate cancer cohort, a significant majority of deaths (77.9%) were non-cancer-specific. Among these deaths, heart disease was the primary cause, responsible for 24.5% of the non-cancer-related deaths. COPD accounted for 6.1%, cerebrovascular diseases for 4.8%, and Alzheimer’s disease for 3.5% of the total deaths in this cohort.

These analyses revealed that non-cancer-related causes were the leading cause of death among the patients. For all cancer cohorts, heart disease was the primary cause of death, followed by Alzheimer’s disease, COPD, and cerebrovascular disease.

Factors associated with non-cancer-specific and cancer-specific mortality varied across cancer types. In breast cancer patients, higher stage and grade were associated with reduced median survival time for cancer-specific mortality.

For prostate cancer patients, higher prostate-specific antigen (PSA) levels and a Gleason score of 8 or higher were associated with decreased median survival time for cancer-specific mortality. Stage III colon and rectal cancer patients also had reduced median survival time for cancer-specific mortality compared to stage I patients.

Cumulative mortality analysis by risk groups demonstrated substantial differences between cancer-specific and non-cancer-specific mortality. In the low-risk group of breast cancer patients (age 65 or older with stage I disease), the incidence of non-breast cancer-specific mortality was almost seven times higher than breast cancer-specific mortality.

Similarly, in the low-risk group of prostate cancer patients (age 65 or older with a Gleason score of 6 or lower), the incidence of non-prostate cancer-specific mortality was almost nine times higher than prostate cancer-specific mortality.

Among the low-risk groups of colon and rectal cancer patients, the cumulative incidence of non-cancer-specific mortality was seven and three times higher than cancer-specific mortality, respectively.

Conclusions

To summarize, this cohort study categorized cancer patients into low, intermediate, and high-risk groups based on cancer-specific prognostic factors. The study revealed significant variations in the risk of cancer-specific and non-cancer-specific mortality among these risk groups, emphasizing the importance of personalized, risk-stratified care.

However, the study had some limitations, including limited clinical and treatment information, lack of data on comorbidities, and inability to assess treatment patterns and access to care. Disparities in treatment and access to quality healthcare, particularly among racial and ethnic minority groups, may also impact the outcomes.

Overall, the findings contribute to understanding the risk profiles of long-term cancer survivors and the importance of comprehensive care beyond the initial cancer diagnosis.

They highlight the need for improved coordination between oncologists and primary care physicians to eliminate unnecessary oncologic follow-up. Future studies should incorporate additional follow-up data on treatment and disease recurrence.