Abstract
To determine the association of chemotherapy delay with overall survival (OS) and investigate predictors of delay among a population-representative American ovarian cancer cohort.
An observational retrospective cohort analysis of women with ovarian cancer who received National Comprehensive Cancer Network guideline-consistent care was performed with the 1998-2011 National Cancer Data Base. Chemotherapy delay was defined as initiation of multiagent chemotherapy >28days from primary debulking surgery. Associations of patient and disease characteristics with chemotherapy delay were tested with multivariate logistic regression. Survival analyses for women diagnosed from 2003 to 2006 approximated a 21-daycycle intravenous platinum-taxane chemotherapy cohort. Overall survival was estimated by Kaplan-Meier analyses and Cox proportional-hazards regressions, with sensitivity analyses using matched cohorts.
58.1% (26,149/45,001) of women experienced chemotherapy delay. Race, insurance status, cancer center type, and community median income were significantly associated with chemotherapy delay (P<0.001). Odds for chemotherapy delay were higher for older or sicker women, women with endometrioid or mucinous histology, lower stage or grade disease, and uninsured or low-income women (P<0.05). Chemotherapy delay >35days from surgery was associated with a 7% (95% confidence interval, 2-13%) increased hazard of death (P=0.01). Relative hazard of death was lowest between 25 and 29days after surgery but was not significantly different within the longer two-week interval from 21 to 35days.
A survival benefit may be achieved by consistently starting chemotherapy between 21 and 35days from primary debulking surgery. Women at higher risk for chemotherapy delay may be targeted for close follow-up.
