Geometric Feature Extraction and Spatial Relationship Analysis of the Hrctv Relative to Organs at Risk In Cervical Cancer Brachytherapy with Intracavitary and Intracavitary–Interstitial Applicators
Abstract
Purpose
There is little objective data to guide clinician decisions regarding use of intracavitary (IC) versus combined intracavitary–interstitial (IC/IS) applicators during cervical cancer brachytherapy. Data from retroEMBRACE indicate HRCTV volumes ≥30 cc predict benefit. We hypothesized that distance-based geometric features of the HRCTV and its spatial relationships with surrounding OARs in cervical cancer patients treated with IC or combined IC/IS high-dose-rate brachytherapy provide additional evidence for increased probability of benefiting from IC/IS applicators beyond HRCTV volume alone
Methods
Complete post-implant first-fraction DICOM datasets from 83 consecutive cervical cancer patients treated with IC or combined IC/IS brachytherapy were extracted from the Oncentra-Brachy TPS. An in-house Python-based geometric analysis code computed slice-by-slice (1) Euclidean distances between the uterine tandem axis and the most distant HRCTV boundary (dHRCTV), and (2) the shortest Euclidean distance between the tandem and the rectum (dREC), bladder (dBL), sigmoid colon (dSIG), small bowel (dSMB), and large bowel (dLB) on corresponding axial slices. HRCTV volumes were also collected. Slice-by-slice ratios between dHRCTV and each OAR (RREC, RBL, RSIG, RSMB, RLB) were calculated, and the maximum ratio (RMAX) and corresponding OAR were identified. Statistical analyses were performed using SAS-9.4.
Results
Mann–Whitney and two-sample t tests demonstrated statistically significant differences between IC and IC/IS patients for all ratios (RREC, RBL, RSIG, RSMB, RLB, and RMAX; p1, and the Bladder to be statistically significant the closest OAR to HRCTV. No statistically significant difference in HRCTV volume was observed between IC and IC/IS patients (p=0.35).
Conclusion
We parameterized and quantified distance-based geometric features and characterized spatial relationships between the HRCTV and surrounding OARs, distinguishing patients treated with IC from those treated with combined IC/IS brachytherapy. These geometric measures may better predict benefit from IC/IS applicators than HRCTV volume alone.