3804 - HR-CTV Volume as a Predictor of Organ at Risk Dose Constraint Achievement in Image Guided Adaptive Brachytherapy for Cervical Cancer

Purpose/Objective(s):

To evaluate whether high risk clinical target volume (HRCTV) predicts the need to add interstitial needles to achieve predefined organ at risk dose constraints in patients with cervical cancer treated with external beam radiation (EBRT) and brachytherapy, and to explore volumetric cutoffs using receiver operating characteristic (ROC) analysis.

Materials/Methods:

We retrospectively reviewed 82 patients with primary cervical cancer treated with definitive pelvic EBRT and high-dose-rate brachytherapy at a single institution. Brachytherapy was delivered using either intracavitary (IC) alone or combined intracavitary/interstitial (ICIS) techniques. All patients received concurrent chemotherapy with EBRT followed by brachytherapy.

For original IC cases (n = 41), treatment plans were evaluated to determine whether all OAR dose constraints were achieved. For ICIS cases, plans were re-optimized after virtually removing interstitial needles, adjusting dwell times to achieve the lowest feasible OAR D2cc values while maintaining HRCTV D90 coverage, thus generating virtual IC cohort.

The prescribed dose was 600 cGy per fraction for five fractions. Dose constraints per fraction were: D2cc bladder < 475 cGy, D2cc rectum < 340 cGy, D2cc sigmoid < 390–400 cGy, and D2cc small bowel < 390–400 cGy. Plans meeting all constraints were classified as dosimetrically acceptable. HRCTV volumes were collected for each patient. The primary endpoint was plan acceptability where all constraints met. ROC analysis was performed in SPSS using HRCTV volume as the test variable and plan acceptability as the state variable to derive the area under the curve (AUC), 95% confidence intervals, and corresponding sensitivity and specificity at varying HRCTV volume thresholds.

Results:

Of 82 patients, 18 (22%) achieved dosimetrically acceptable plans meeting all OAR constraints. Mean HR-CTV volume was 38.3 cc (SD 23.9 cc, range 11.2–212.3 cc). HRCTV volume demonstrated- limited discriminative ability for predicting plan acceptability, with an AUC of 0.58 (95% CI 0.44–0.72, p = 0.30). ROC coordinate analysis identified an exploratory cutoff of 35 cc, which yielded sensitivity of 53% and specificity of 67% for identifying- acceptable plans. A lower cut-off of 30 cc increased sensitivity to 67% at the expense of reduced specificity (50%), illustrating the trade-off between target coverage and OAR sparing. No single HR-CTV volume threshold provided strong discrimination between acceptable and unacceptable plans.

Conclusion:

The discriminative ability of any single HRCTV volume threshold was limited, indicating that HRCTV size alone is insufficient to determine the need for interstitial needle placement to meet OAR constraints without compromising target coverage.