Results of the Gynecology Oncology Group (GOG) phase III study presented in a plenary session of 2017 ASTRO Annual Meeting (24-27 September, San Diego, USA) did not demonstrate a superiority of vaginal cuff brachytherapy and chemotherapy (VCB/C) to pelvic external beam radiation therapy (PXRT) in women with high-risk, early endometrial cancer. Acute and late toxicity and pelvic and para-aortic nodal failure were more frequent in the VCB/C arm. Both arms appeared to be well tolerated with high completion rates.
The primary study objective was to determine if VCB/C could increase recurrence-free survival (RFS) compared to PXRT. Secondary objectives included comparisons of overall survival (OS), patterns of failure, and frequency/severity of adverse events between the treatment arms.
All patients in the study were required to undergo hysterectomy. Eligible patients had stage I endometrioid histology with GOG 99-based high intermediate risk criteria (based on age, tumour grade, depth of invasion, and presence of lymphovascular space invasion), stage II, or stage I-II serous or clear cell tumours. Central pathology review was performed.
Patients assigned to PXRT were treated with standard 4-field or Intensity-Modulated Radiation Therapy (IMRT) techniques to a mean dose of 45 Gy over 5 weeks. Additional VCB was optional for patients with serous/clear cell tumours or stage II disease. Patients assigned to VCB/C received HDR or LDR brachytherapy followed by paclitaxel 175 mg/m2 plus carboplatin AUC 6 for a total of 3 cycles.
A total of 601 patients were accrued; PXRT was assigned to 301 (18 did not receive study treatment) and VCB/C to 300 (9 did not receive study treatment).
The median age was 63 years, 74% had stage I disease, and 89% underwent lymphadenectomy. Histology included 71% with endometrioid type, 15% serous, and 5% clear cell. Nearly all patients completed the prescribed therapy.
In the PXRT arm, IMRT was used in 36%, and vaginal cuff brachytherapy boost was added in approximately 35%.
Acute toxicity was more common and more severe with VCB/C. Grade 3 or higher adverse events were reported in 32 patients on the PXRT arm versus 187 patients on the VCB/C arm. Grade 3 or higher late effects were seen in 37 and 35 patients on the PXRT and VCB/C arms, respectively.
With a median follow-up of 53 months, the 36 month RFS was 82% for both PXRT and VCB/C.
The 36-month OS was 91% versus 88% for PXRT and VCB/C, respectively.
No significant differences were noted between the two arms in terms of vaginal or distant failure. However, pelvic or para-aortic nodal recurrences were significantly more common in the VCB/C arm (25 vs 12), largely driven by the difference in pelvic nodal failure (20 vs 6 patients). There was no statistically significant treatment effect heterogeneity with respect to RFS among clinical-pathologic variables evaluated.
The GOG investigators concluded that PXRT remains an effective, well-tolerated, and acceptable adjuvant treatment in patients with high-risk, early-stage endometrial carcinoma.