HISTOPATHOLOGIC PROGNOSTIC FACTORS IN STAGE IIB CERVICAL-CARCINOMA TREATED WITH RADICAL HYSTERECTOMY AND PELVIC-NODE DISSECTION - AN ANALYSIS WITH MATHEMATICAL-STATISTICS

Of 107 patients with stage IIb cervical cancer who underwent laparotom y, 82 (77%) could be treated with radical hysterectomy (RAH) and pelvi c-node dissection (PND). The remaining 25 patients were unsuitable for radical surgery because of para-aortic lymph node metastases, direct cancer invasion into the bladder muscle, and/or fixed enlarged pelvic lymph nodes (PLN): Such patients were treated with radiation therapy a fter laparotomy. Fifty-nine of RAH patients were given postoperative p elvic radiation because they had PLN metastases, parametrial invasion, and/or full thickness cervical stromal invasion. The overall 5-year s urvival of the patients undergoing RAH was significantly better than t hat of those who could not be treated with RAH (P < 0.001). In the RAH patients, parametrial invasion, which clinically defines stage IIb, w as found only in 45%. Univariate analysis of histopathologic prognosti c factors revealed that PLN metastasis, parametrial invasion, adenocar cinoma, and lymph-vascular space invasion significantly affected survi val of the RAH patients (P < 0.05). Multivariate analysis using Cox's proportional hazards regression model, however, selected only PLN meta stasis as a strong prognostic factor (P < 0.001). Concerning PLN metas tasis patients with two or more positive nodal groups vs. 49%, P < 0.0 001). The logistic regression analysis revealed that tumor diameter, p arametrial invasion and lymph-vascular space invasion were independent ly correlated with PLN metastases in two or more nodal groups. The pre sent data suggest that (i) the patients with massive pelvic extension of cancer cannot be cured by radiation therapy alone, (ii) the strong determinant of the prognosis of the patients undergoing RAH and PND is PLN metastasis. Therefore, for these patients with poor prognosic fac tors, other treatment modalities should be considered. From the presen t study it seems that planning RAH and PND for patients with stage IIb disease might make it possible to select poor prognostic subgroups, w ho have extra cervical extension or PLN metastases in two or more grou ps, and be useful in individualizing treatment.