RT Journal Article
SR Electronic
T1 Pretreatment hemoglobin, platelet count, and prognosis in endometrial carcinoma
JF International Journal of Gynecologic Cancer
JO Int J Gynecol Cancer
FD BMJ Publishing Group Ltd
SP 236
OP 240
DO 10.1136/ijgc-00009577-200105000-00013
VO 11
IS 3
A1 K. F. Tamussino
A1 F. Gücer
A1 O. Reich
A1 F. Moser
A1 E. Petru
A1 H. S. Scholz
YR 2001
UL http://ijgc.bmj.com/content/11/3/236.abstract
AB We analyzed whether a low preteatment hemoglobin level is a prognostic factor in endometrial cancer and whether it is associated with thrombocytosis. Two hundred and twelve patients with endometrial cancer treated with surgery were reviewed. Data were analyzed with Pearson's chi-squared test, Fisher's exact test in contingency tables, the Mann–Whitney U-test, the Student's t-test, and Kaplan-Meier estimates. Multivariate analysis was performed with the log-rank test and the Cox proportional hazard model. Thirty-nine patients (18%) had a pretreatment hemoglobin value of <12.0 g/dL. These 39 patients had significantly higher rates of nonendometrioid histology, high-grade tumors, myometrial invasion of >50%, adnexal involvement, lymph-vascular space involvement, and advanced FIGO stage than patients with hemoglobin ≥12.0 g/dL. The rate of thrombocytosis was significantly higher in patients with a low hemoglobin level (36% vs. 8%, P < 0.01). The overall 5-year survival rate of patients with low pretreatment hemoglobin was 59% compared with 89% for those with hemoglobin ≥12 g/dL (P < 0.01). In the multivariate analysis age, thrombocytosis, nonendometrioid histology, high-grade histology, and advanced FIGO stage were significantly associated with a poor prognosis whereas adnexal involvement, lymph-vascular space involvement, low hemoglobin and myometrial invasion were not. These data indicate that low pretreatment hemoglobin is a prognostic factor in patients with endometrial cancer and that it is associated with thrombocytosis. Low hemoglobin was strongly associated with other unfavorable prognostic factors so that it was significant in the univariate but not the multivariate analysis.