A Systematic Study on Local Failure Events Post Chemoradiation Therapy for Cervical Cancer: Understanding the Impact of Baseline Lateral Anatomic Compartment Involvement

Abstract

Purpose: To dissect the local failure (LF) events, including specific rates by anatomic compartments, after definitive chemoradiation therapy for locally advanced cervical cancer.

Methods and materials: Records of all consecutive women with locally advanced cervical cancer treated with definitive chemoradiation therapy and image guided adaptive brachytherapy were reviewed, focusing on the local disease status. No patient received external beam radiation therapy parametrial boost. Incidence estimations, timing analysis (true persistence [TP]) versus new recurrence [NR]), calculations of LF rates by anatomic compartments (at the time of LF detection), and test of association of compartment-specific LF with baseline (at diagnosis) involvement (adjusted odds ratio [aOR]), were performed.

Results: Among the 225 patients included (75% International Federation of Gynecology and Obstetrics [FIGO] III-IV; 32% intracavitary/interstitial brachytherapy use; median follow-up 46.2 months), 34 LFs occurred (24 TP; 10 NR), involving 92 anatomic compartments. Isolated LF was the first relapse in 55.9% (19/34), with no events after 3 years, resulting in 6.2% 3-year/5-year overall NR and 16.3% 3-year/5-year overall LF (NR + TP) rates. TP had worse survival compared to NR (2 year overall survival: 19.1% vs 72.9%, P = .04). The most involved compartments were the cervix (79.4%), parametrium (PMT) (61.8%), uterine corpus (33.3%), bladder (14.7%), vagina (11.8%), and mesorectum (2.9%). Recurrence followed the cardinal ligament (CL), uterosacral ligament, and pubocervical ligament directions in 58.8%, 35.3%, and 23.5% of cases, respectively. Only 2 intact PMT at diagnosis developed ipsilateral LF (1.1%, 2/181, CL only), whereas baseline ipsilateral PMT involvement with hydronephrosis was associated with ipsilateral LF (aOR 22.2, P < .01, rate: 21.3%, 10/47, involving CL ± pubocervical ligament ± uterosacral ligament), but not with contralateral LF (aOR 0.4, P = .28).

Conclusions: In the nonoperative setting, the lateral PMT is the second most involved anatomic compartment by LF (>60%) after the cervix (80%). An intact baseline PMT has a low frequency (1%) and limited extension (CL only) of ipsilateral failure. In contrast, baseline hydronephrosis is strongly associated with extensive ipsilateral PMT relapse, requiring side-specific treatment intensification.