Non-cancer causes of death following cervical cancer diagnosis: a SEER database study

To quantify non-cancer mortality after cervical cancer diagnosis and examine how these risks vary over time, by disease stage, race, and registry-recorded treatment patterns.

We analyzed 65,852 women diagnosed with cervical cancer in Surveillance, Epidemiology, and End Results-17 (2000-2021). Standardized mortality ratios (SMRs) compared observed deaths with age-, race-, sex-, and calendar-period-specific mortality in the U.S. general population. Fine-Gray models retained every death other than the target cause as a competing event.

Of 24,857 deaths, 19,975 (80.4%) were cancer-related and 4,882 (19.6%) were non-cancer-related. The overall non-cancer SMR was 1.65 (95% confidence interval [CI]=1.60-1.70). After International Federation of Gynecology and Obstetrics stage- and race-exact propensity score matching (4,506 pairs; maximum absolute standardized mean difference, 0.079), registry-inferred combined radiotherapy plus chemotherapy without surgery was not associated with overall non-cancer death (subdistribution hazard ratio [sHR]=1.01; 95% CI=0.88-1.16); the infectious-disease (sHR=1.72; 95% CI=0.75-3.95) and septicemia estimates (sHR=1.60; 95% CI=0.69-3.72; 23 events) were imprecise. Among 31,617 patients surviving beyond 5 years, the conditional non-cancer SMR remained elevated (1.44; 95% CI=1.38-1.49).

Non-cancer mortality remains elevated among cervical cancer survivors, including beyond the 5-year landmark. Treatment-group contrasts are strongly affected by disease stage, treatment-selection mechanisms, and unmeasured clinical factors; they should therefore be interpreted as hypothesis-generating rather than causal. Survivorship care should integrate cardiovascular, infectious, renal, respiratory, and metabolic risk management.