Do Air Gaps With Image-Guided Vaginal Cuff Brachytherapy Impact Failure Rates in High Intermediate Risk FIGO Stage I Endometrial Cancer Patients?

Abstract

In patients with intermediate to high intermediate risk (HIR) FIGO stage 1 endometrial cancer, adjuvant vaginal cuff brachytherapy (VCBT) reduces the risk of vaginal cuff failure (VCF). Image-guided brachytherapy (IGBT) utilizing computed tomography (CT) scans helps confirm applicator placement but may show air gaps at the applicator surface. While multiple studies have quantified air gaps and their dosimetric impact, the objective of this study is to assess their impact on the rate of VCF after IGBT. We performed a retrospective review of patients treated with adjuvant image-guided VCBT alone from 2009-2016 for HIR FIGO stage I endometrial cancer. Patients were defined as HIR if they met the criteria outlined in GOG 99 or PORTEC-2. VCBT was 21 Gy in 3 fractions (85% of patients) or 30 Gy in 5 fractions (15%). Air gaps present at the applicator surface on the first post-insertion CT were contoured. VCFs were captured by review of imaging, pathology, and/or clinical notes. Vaginal cuff failure free survival (VCFFS) was measured from the start of VCBT to VCF. Patients without VCF were censored at last follow up or death. The Kaplan-Meier method was used to estimate actuarial event-time probabilities for VCFFS. Predictors of VCFFS were analyzed using log rank tests between groups and with Cox regression for continuous variables. A total of 234 patients were identified. Median age was 66 years (Interquartile range [IQR] 60-72 years). Most patients (80%) had FIGO grade 2 or 3 carcinoma; 49% were stage FIGO IB, and 54% had LVSI. A total of 85% were HIR by GOG 99 criteria and 68% by PORTEC-2 criteria. 25% had non-endometrioid histology. Air gaps were present on the first post-insertion CT scan in 82% of patients. They were present laterally in 74% of patients, anteriorly in 35%, at the apex in 30%, and posteriorly in 18%. One hundred eleven patients (47%) had air gaps in multiple locations. The median number of air gaps was 2 (IQR 1-3). The median depth of the largest air gap was 2.7 mm (IQR 2.1-3.4 mm), and the median cumulative volume of air gaps was less than 0.1 cm3 (range < 0.1 – 0.7 cm3). At a median follow-up of 56 months (IQR 41 – 69), 12 patients (5%) experienced VCF, of which 4 had an isolated VCF and 8 had synchronous pelvic and/or distant failure. Five-year any VCFFS and isolated VCFFS were 96% (95% CI 93-98%) and 98% (95% CI 96% - 100%), respectively. Mean air gap volume was no different between patients with or without VCF (0.03 vs 0.06 cm3, p = 0.35). On univariate analysis, no factors, including the presence, number, maximum depth, or cumulative volume of air gaps, were predictive for VCFFS. In this population of HIR FIGO stage I endometrial cancer patients treated with adjuvant image-guided VCBT, VCFFS remained high (96% at 5 years) despite the majority having small volume air gaps present on CT scan, suggesting that small air gaps < 0.7 cm3 do not increase the risk for VCF.