Survival and failure analysis in stage II endometrial cancer using the revised 1988 FIGO staging system.

Fifty-one patients with Stage II endometrial carcinoma diagnosed between 1974 and 1987 were restaged according to the FIGO 1988 revisions for endometrial carcinoma. Patients were divided into Stage IIA, those patients with cervical glandular involvement without stromal invasion, and Stage IIB, those patients having stromal invasion of the cervix. Tumor grade was also assessed. Patients were treated with radiation therapy alone, pre-operative radiation therapy followed by a simple hysterectomy, or a simple hysterectomy followed by postoperative radiation therapy. The 5-year actuarial survival for Stage IIA was 86% and the 5-year actuarial survival for Stage IIB was 46% (p = 0.06). The 5-year local recurrence rate in each group was 9%. Stage IIA had a distant metastases rate of 14% whereas 44% of the patients in Stage IIB developed distant disease (p = 0.06) at 5 years. The grade of the tumor did not play a role in local recurrence. However, when tumor grade was analyzed with respect to distant disease, 14% of patients with grade 1 tumors developed distant metastases, 31% of patients with grade 2 tumors developed distant metastases, and most significantly, 63% of patients with grade 3 tumors developed distant metastases (p = 0.004). There was no statistically significant relationship between stromal invasion and tumor grade. This study concludes that grade is the greatest predictor of survival, with only 37% of grade 3 patients surviving at 5 years. As a predictor of survival, stromal invasion is of less significance than grade (p = 0.06 vs. p = 0.004). Death most often occurs because of distant metastases, and local failure is rare and is not dependent on the degree of cervical involvement or grade.