Dose-volume effects in pathologic lymph nodes in locally advanced cervical cancer

In cervical cancer patients, dose-volume relationships have been demonstrated for tumor and organs-at-risk, but not for pathologic nodes. The nodal control probability (NCP) according to dose/volume parameters was investigated.Patients with node-positive cervical cancer treated curatively with external beam radiotherapy (EBRT) and image-guided brachytherapy (IGABT) were identified. Nodal doses during EBRT, IGABT and boost were converted to 2-Gy equivalent (α/β = 10 Gy) and summed. Pathologic nodes were followed individually from diagnosis to relapse. Statistical analyses comprised log-rank tests (univariate analyses), Cox proportional model (factors with p ≤ 0.1 in univariate) and Probit analyses.A total of 108 patients with 254 unresected pathological nodes were identified. The mean nodal volume at diagnosis was 3.4 ± 5.8 cm3. The mean total nodal EQD2 doses were 55.3 ± 5.6 Gy. Concurrent chemotherapy was given in 96%. With a median follow-up of 33.5 months, 20 patients (18.5%) experienced relapse in nodes considered pathologic at diagnosis. Overall nodal recurrence rate was 9.1% (23/254). On univariate analyses, nodal volume (threshold: 3 cm3, p < .0001) and lymph node dose (≥57.5 Gyα/β10, p = .039) were significant for nodal control. The use of simultaneous boost was borderline for significance (p = .07). On multivariate analysis, volume (HR = 8.2, 4.0-16.6, p < .0001) and dose (HR = 2, 1.05-3.9, p = .034) remained independent factors. Probit analysis combining dose and volume showed significant relationships with NCP, with increasing gap between the curves with higher nodal volumes.A nodal dose-volume effect on NCP is demonstrated for the first time, with increasing NCP benefit of additional doses to higher-volume nodes.