GSOR5 Presentation Time: 9:20 AM

Purpose

To compare differences in normal tissue toxicity between conventional gauze vs. balloon intravaginal packing in women receiving brachytherapy for cervical cancer.

Materials and Methods

We retrospectively identified women with 2009 FIGO stage IB1-IVB cervical cancer who received definitive EBRT with cisplatin and T&O or T&R brachytherapy between 2010-2016 at our institution. The primary endpoints were rates of grade ≥2 vaginal, GU, and GI toxicity. Dose metrics were compared via the Mann-Whitney U test. Toxicity rates were estimated via the Kaplan-Meier method and compared via log-rank test, and factors associated with toxicity were identified with Cox regression analysis.

Results

96 patients received brachytherapy with a median follow-up of 39 months. 32 patients received intravaginal gauze packing, while 64 patients received intravaginal balloon packing. Two balloon leakages occurred and were replaced with gauze packing for that fraction. Gauze packing was associated with significantly higher rectal D2cc vs. balloon packing (median 66 Gy gauze vs. 55 Gy balloon, p<0.01), but not with vaginal or bladder D2cc. As shown in the figure, on Kaplan-Meier analysis, gauze packing was also associated with higher rates of grade ≥2 vaginal toxicity vs. balloon packing (3-year rate 30.1% vs. 20.0%, p<0.05), GU (32.2% vs. 11.3%, p<0.05), and GI (42.3% vs. 18.4%, p<0.05) toxicity. Grade ≥3 vaginal (3-year rate 11.9% vs. 5.5%, p<0.05) and GI (21.0% vs. 4.8%, p<0.05) toxicity was also higher with gauze vs. balloon packing. On multivariate Cox regression analysis, gauze packing remained significantly associated with grade ≥2 vaginal toxicity (HR 5.15, 95% CI 1.41-18.85, p<0.05), but not with GU or GI toxicity.

Conclusions

Compared to conventional gauze packing, intravaginal packing balloons were associated with reduced rectal D2cc and late GI toxicity. Decreased rates of late vaginal and GU toxicity were also observed, though corresponding D2cc metrics were not significantly reduced.