OBJECTIVE: To assess different methods of measuring the depth of myometrial invasion in endometrial carcinoma as a prognostic factor. METHODS: Eighty-eight cases of stage I or II endometrial carcinoma treated initially by hysterectomy between 1979-1989 were reviewed histologically. Three methods of measuring myometrial invasion were evaluated: 1) percentage of invaded tumor thickness to whole thickness of myometrium, 2) percentage of whole tumor thickness to total thickness of tumor and myometrium, and 3) distance from the tumor-myometrial junction to the uterine serosa. We evaluated the effect of several factors on prognosis by multivariate analysis using Cox regression models. RESULTS: Myometrial invasion determined by these three measurement methods was associated significantly with survival in a univariate analysis. When myometrial invasion assessed by each method and other prognostic factors were entered into a multivariate model, the distance from the tumor-myometrial junction to the uterine serosa, lymph-vascular space invasion, and cervical stromal involvement were identified as independently significant prognostic factors. CONCLUSION: This method of evaluating myometrial invasion by measuring the distance from the tumor-myometrial junction to the uterine serosa was most useful as a correlate with survival.
OBJECTIVE: To assess different methods of measuring the depth of myometrial invasion in endometrial carcinoma as a prognostic factor. METHODS: Eighty-eight cases of stage I or II endometrial carcinoma treated initially by hysterectomy between 1979-1989 were reviewed histologically. Three methods of measuring myometrial invasion were evaluated: 1) percentage of invaded tumor thickness to whole thickness of myometrium, 2) percentage of whole tumor thickness to total thickness of tumor and myometrium, and 3) distance from the tumor-myometrial junction to the uterine serosa. We evaluated the effect of several factors on prognosis by multivariate analysis using Cox regression models. RESULTS: Myometrial invasion determined by these three measurement methods was associated significantly with survival in a univariate analysis. When myometrial invasion assessed by each method and other prognostic factors were entered into a multivariate model, the distance from the tumor-myometrial junction to the uterine serosa, lymph-vascular space invasion, and cervical stromal involvement were identified as independently significant prognostic factors. CONCLUSION: This method of evaluating myometrial invasion by measuring the distance from the tumor-myometrial junction to the uterine serosa was most useful as a correlate with survival.