Approximately 36,000 cases of endometrial cancer are diagnosed yearly in the USA. The majority of these are stage I. After standard hysterectomy, there is little randomized data on which to base decisions regarding the utility of post-operative radiation therapy (PORT). Creutzberg et al has previously examined the utility of PORT in a prospective randomized study in women undergoing incomplete surgical staging (no node sampling). For stage I patients, a small decrease in locoregional recurrence (LRR) was seen prompting them to recommend PORT only for their high risk sub-group: those with FIGO stage IB Grade II tumors or worse or any patient over 60 years of age. The goal of this study (GOG 99) was to investigate the utility of PORT in stage IB, IC and II (occult) patients having undergone complete surgical staging.
Materials and Methods
Eligible patients were the "intermediate risk" group from GOG 33, i.e. those presumed to have a 5 year LRR of 20-25%. These were node negative stage IB, IC and II (occult) patients. Histology was limited to adenocarcinoma. Clear cell and papillary serous histologies were excluded.
Total abdominal hysterectomy, bilateral salpingo-oophorectomy and bilateral selective pelvic and para-aortic node dissections were required. Laparascopic procedures were not allowed.
Randomization was whole pelvis RT vs. no further treatment (NFT). RT consisted of 5040cGy in 180cGy fractions to start within 8 weeks of surgery. No vaginal cuff boosts were given.
Primary outcome measure was Recurrence Free Interval (RFI). Overall Survival (OS) was a secondary endpoint.
After determining that the LRR risk in the study population was about half the expected risk, an unplanned sub-group analysis was performed. Patients were divided into "high intermediate risk," i.e. patients of any age with the following risk factors: grade II-III, lymphovascular invasion and outer 1/3 myometrial invasion, > 50 years old with two of the above risk factors or > 70 years old with one risk factor. All other patients were "low intermediate risk."
448 women were entered from 6/1987 through 7/1995. 392 were deemed eligible. The remainder were excluded mainly due to incomplete surgery. The arms were prognostically well balanced.
Median follow-up was 68 months.
PORT reduced the relative risk of LRR by 58% (95% CI 27% - 75%).
Overall, the majority of local recurrences were vaginal cuff only (15/21). 10/13 recurrences in the PORT arm were distant only. Of the 3 local recurrences in the PORT arm, 2 occurred in patients who refused all RT.
Because of insufficient patient numbers, a statistically significant survival difference was not seen. However, the data suggested that a survival benefit might exist.
The high intermediate risk group comprised only 33% of the patients but suffered 66% of the recurrences. The relative risk reduction for LRR with PORT was the same for both groups, but the absolute improvement in LRR was much greater in the high intermediate risk group.
Grade III and IV bowel toxicity rates were significantly greater in the PORT groups and included two deaths.
This is the only large, randomized study of PORT vs. NFT in a population undergoing complete surgical staging. Overall, PORT reduced LRR by 58%, primarily by greatly reducing the incidence of vaginal recurrences. The absolute risk reduction in the high intermediate risk group was 19% vs. 4% in the low intermediate risk group. There was a trend towards OS benefit in the PORT arm. However, because of the low overall recurrence rate and the impact of intercurrent disease in this elderly population this study was underpowered to detect a survival benefit. Based on unplanned sub-group analysis, the authors recommend PORT only for those stage I patients who fit their criteria for high intermediate risk.