Predictors of Interstitial Brachytherapy Utilization in Locally Advanced Cervical Cancer and Impact on Overall Survival

Abstract

<h3>Purpose/Objective(s)</h3> The standard of care for locally advanced cervical cancer (LACC) is multimodality therapy with chemotherapy, external beam radiation therapy (EBRT) and brachytherapy (BT) boost. The addition of BT improves overall survival (OS); interstitial BT has been prospectively shown to improve target and organ at risk dose, especially for larger tumors. Prior research suggests declining utilization of BT particularly in patients without insurance, underrepresented minorities, and in low volume cancer centers, which may widen cancer disparities. Here we utilize the National Cancer Database (NCDB) to examine patterns and predictors of BT use, particularly interstitial BT, in patients with LACC and its impact on OS. <h3>Materials/Methods</h3> The NCDB was used to identify patients with LACC (stage IIb-IVa) diagnosed between 2004-2018 and treated with radiation that included a component of BT. A multivariable model to predict receipt of interstitial vs intracavitary BT was created including age, race, comorbidity score (CCI), stage, receipt of chemotherapy, urban-rural continuum, insurance, facility type, facility region, facility quartile of cases reported and year of diagnosis. For survival analysis, patients who received interstitial BT were propensity score matched 1:1 to intracavitary patients using the same variables and nearest neighbor matching without replacement. Log rank survival analysis was performed for the matched cohort. <h3>Results</h3> A total of 9,591 patients were included with median age of 55 years. Overall, 14% of patients who received BT as part of their radiation treatment received interstitial BT. Ninety-three percent of patients received chemotherapy. On multivariable logistic regression analysis, higher stage (stage IIIa: OR 2.04 [95% CI 1.64 to 2.54], stage IIIb: OR 1.38 [1.22-1.57], stage IVa: OR 2.09 [1.09-4.00]), treatment at an academic center (OR 1.31 [1.12-1.54]), treatment at a higher volume center (highest quartile: OR 1.38 [1.12-1.70]) and later year of diagnosis (OR 1.03 per year [1.02-1.05]) were associated with increased interstitial BT utilization. Conversely, uninsured status (OR 0.772 [0.62-0.97]) or Medicare (OR 0.78 [0.65-0.94]) were associated with decreased interstitial BT utilization. There was no significant association between age at diagnosis, race, or CCI and patterns of interstitial BT use. After propensity score matching there was no overall survival difference between patients who received interstitial vs those treated with intracavitary BT (p=0.471). <h3>Conclusion</h3> Patients with more locally advanced cervical cancer and those treated at a high volume or academic center were more likely to receive interstitial BT. This pattern likely reflects appropriate intensification of therapy for larger tumors at centers with the expertise to perform interstitial implants. Utilization of interstitial implant did not translate into an OS benefit likely due to selection of cases for interstitial implant that have poor initial response to chemoRT.