Risk Factors for Local Failure Following Chemoradiation and Magnetic Resonance Image-Guided Brachytherapy in Locally Advanced Cervical Cancer: Results From the EMBRACE-I Study.

Purpose: To report clinical and treatment characteristics, remission and failure patterns, and risk factors for local failure (LF) from the EMBRACE-I study.
Materials and Methods: EMBRACE-I was a prospective, observational, multicenter cohort study on magnetic resonance imaging-based image-guided adaptive brachytherapy (MR-IGABT) in locally advanced cervical cancer. Treatment consisted of external beam radiotherapy, concurrent chemotherapy, and MR-IGABT. LF was defined as progressive or recurrent disease in the cervix, uterus, parametria, pelvic wall, or vagina. Competing risk analysis was used to estimate local tumor control (LC) and Cox proportional regression models for multivariable analysis and dose-response analysis.
Results: One thousand three hundred eighteen patients with a median follow-up of 52 months were available for this analysis. Eighty-one patients had persistent disease 3 months after end of treatment. Of those, 60 patients achieved LC at 6-9 months without further treatment, whereas 21 patients had progressive disease. In addition, 77 patients developed a local recurrence after complete remission comprising a total number of 98 LFs. LFs were located inside the MR-IGABT target volumes in 90% of patients with LF. In multivariable analysis, histology, minimal dose to 90% of high-risk clinical target volume (CTV _HR ), maximum tumor dimension, CTV _HR > 45 cm ³ , overall treatment time, tumor necrosis on magnetic resonance imaging at diagnosis, uterine corpus infiltration at diagnosis and at MR-IGABT, and mesorectal infiltration at MR-IGABT had significant impact on LF. Dose-response analysis showed that a minimal dose to 90% of 85 Gy to the CTV _HR led to 95% (95% CI, 94 to 97) LC 3 years postintervention for squamous cell in comparison to 86% (95% CI, 81 to 90) for adeno/adenosquamous carcinoma histology.
Conclusion: The present study demonstrates the safety and validity of the GYN GEC-ESTRO/ICRU-89 target concept and provides large-scale evidence for dose prescription and new risk factors for LF in MR-IGABT in locally advanced cervical cancer.
Competing Interests: Maximilian P. SchmidHonoraria: ElektaResearch Funding: Elekta (Inst), Varian Medical Systems (Inst)Patents, Royalties, Other Intellectual Property: Patent: US20170120072 Umesh MahantshettyConsulting or Advisory Role: AstraZenecaResearch Funding: Varian Medical Systems (Inst) Kari TanderupResearch Funding: Varian Medical Systems (Inst), Elekta (Inst) Peter HoskinResearch Funding: Varian Medical Systems (Inst), Astellas Pharma (Inst), Bayer (Inst), Roche (Inst), Pfizer (Inst), Elekta (Inst), Bristol Meyers (Inst) Bradley R. PietersConsulting or Advisory Role: BD Bard (Inst)Research Funding: ElektaTravel, Accommodations, Expenses: BD BardOther Relationship: Elekta (Inst) Remi NoutConsulting or Advisory Role: Merck KGaA (Inst)Research Funding: Varian Medical Systems (Inst), Elekta (Inst), Accuray (Inst) Nicole NesvacilHonoraria: ElektaConsulting or Advisory Role: ElektaTravel, Accommodations, Expenses: Elekta Richard PötterResearch Funding: Elekta (Inst), Varian Medical Systems (Inst) Christian KirisitsHonoraria: ElektaResearch Funding: Elekta (Inst), Varian Medical (Inst)Patents, Royalties, Other Intellectual Property: Patent WO2015181632A1 on Methods and Systems for Brachytherapy Planning Based on Imaging DataNo other potential conflicts of interest were reported.