Intrapatient Dosimetric Variability Between Two Adjustable Brachytherapy Applicators for Cervical Cancer: Tandem/Ovoid vs. Tandem/Split-Ring

Abstract

The choice of intracavitary applicator for high-dose rate brachytherapy for cervical cancer has dosimetric implications for target coverage as well as toxicity. Prior analyses have compared rigid (tandem/ring) and adjustable applicators, however there is a deficiency of data clarifying the differences and indications among tandem/ovoid (TO) and the increasingly-used tandem/split-ring (TSR) applicator. To our knowledge, our institution is the first to analyze intrapatient spatial and dosimetric variability between adjustable TO and TSR applicators. Between 2010 and 2017, 108 patients with cervical cancer were treated with a brachytherapy boost (median 28 Gy in 4 fractions) with two insertions under general anesthesia one week apart, after pelvic external beam radiotherapy to 45-50.4 Gy. Sixteen of these patients were treated with both TO and TSR applicators. CT-based high-risk CTVs were delineated according to GEC-ESTRO guidelines. Doses to ICRU points A, B, P, organs at risk (OAR) including bladder, rectum, vagina, and D2cc of OAR were compared. Spatial characteristics of the 100% isodose distribution and physical geometry of the applicators were analyzed as well (craniocaudal, anterior-posterior, and horizontal left-right dimensions). Target coverage was evaluated by comparing the CTV volume, D90 (Gy) and V100 (%) between insertions. Paired two-tailed T-tests were used, and significance was assessed at p<0.05. Across the 32 insertions studied, there were no statistically significant differences in CTV volume, D90, V100, or dose to points A, B, P, rectum and bladder between the two applicators. TO applicators had a significantly lower dose delivered to the left and right vaginal mucosae than TSR applicators (left mean 8.1 vs. 13.3 Gy, p<0.01; right mean 7.67 vs. 12.5 Gy, p<0.01). TO applicators also yielded a larger maximum craniocaudal dimension of the 100% isodose distribution (mean 8.2 vs. 6.8cm, p<0.01). TSR applicators, however, yielded a wider horizontal max dimension of the 100% isodose distribution (mean 7.8 vs. 6.7cm, p<0.01) and a correspondingly wider horizontal colpostat separation (mean 5.4 vs. 3.7 cm, p<0.01). Because this analysis evaluates both TO and TSR applicators in the same patients, it provides the unique advantage of contrasting dosimetry and target coverage while controlling for tumor and pelvic anatomy. Our study demonstrates that TO applicators consistently yield lower vaginal mucosal doses compared to TSR, likely reflective of the generally larger ovoid caps, without significantly different dose to the rectum or bladder. Although tumor coverage was equivalent, differences in dose distribution persist, such as greater horizontal width with TSR and improved craniocaudal coverage with TO, which may guide selection between the two devices.