Comprehensive analysis of patient outcome after local recurrence of locally advanced cervical cancer treated with concomitant chemoradiation and image-guided adaptive brachytherapy

Introduction: Since dose escalation allowed by image-guided adaptive brachytherapy (IGABT) in locally advanced cervical cancer (LACC), local relapses have become a rare event. Only scarce data are available on the outcome of patients experiencing a local relapse after IGABT. Methods: Between 2004 and 2016, all consecutive patients treated at Gustave Roussy Institute for LACC and receiving concomitant chemoradiation and IGABT were analysed. Clinical and treatment-related prognostic factors for survival after local relapse were searched, in order to potentially identify patients requiring salvage treatment. Results: Two hundred and fifty-nine patients were treated during this period. With a median follow-up of 4.1 years, 10.8% (n = 28) had a local relapse. Among these patients, 53.6% had synchronous lymph nodes or distant metastatic relapse and only 13 patients (5% of all patients) had isolated local relapse. After local relapse, median survival was 47 months and three patients were alive at last follow-up. Only three patients with local relapse could receive salvage surgery (10.7%). Metastases occurrence and pelvic wall involvement were the main contraindications (67.9%) for salvage surgery. Among the three patients treated with surgery, two are still alive at last follow-up without significant complication. Improved survival was observed among the two patients who could have surgery (p =.02). Local progression led to serious symptoms in 75% of patients. Only the time interval between brachytherapy and relapse (<1 year) was prognostic for 2-year overall survival (p =.005). Conclusion: Salvage surgery is feasible in a very low number of highly selected patients with local relapse following IGABT. Local failure is a major cause of severe local symptoms, confirming that every effort should be done to achieve long-term local control through dose escalation.