Significance of degree of destructive stromal invasion in early stage endocervical adenocarcinoma.

Abstract

e17023 Background: Recent evidence has suggested that stratifying endocervical adenocarcinomas based on degree of destructive stromal invasion and desmoplastic response may help further identify risk of lymph node metastasis (1,2). We sought to explore the prognostic influence of the above histopathologic stratification in women undergoing treatment for early stage cervical adenocarcinoma. Methods: Women with stage IA1 – IB2 endocervical adenocarcinoma who underwent radical hysterectomy were included. Demographics and clinico-pathologic data were collected. Three independent gynecologic pathologists performed pathology review for all specimens and tumors were categorized into patterns A, B, and C based on the criteria described by Silva et al (2). Briefly, these criteria include adenocarcinoma without destructive stromal invasion or desmoplastic response (pattern A), minimum destructive invasion (pattern B), and frank invasion (pattern C). Clinico-pathologic data were reviewed and all cases were stratified as either low, intermediate (GOG 92 criteria) or high-risk (positive nodes, parametria, margins). Chi square and Fisher exact test were used for statistical analysis. Results: 46 patients with cervical adenocarcinoma were identified with the following stage distribution: 6 stage IA, 34 stage IB1 and 6 stage IB2. All patients with stage IA disease had pattern A or B tumors. Ten patients with pattern A or B tumors had stage IB1 (N = 9) or IB2 (N = 1). 30 patients had pattern C tumor, all of which were stage IB1-IB2. LVSI was identified in 40% of pattern C versus 6% of pattern B tumors (p < 0.02). All patients who were high-risk for recurrence had pattern C tumors. 16/17 patients who were intermediate-risk had pattern C versus 1/17 with pattern B tumor (p < 0.01). Lymph node metastasis was identified in 11% of pattern C and none in patterns A and B tumors (p < 0.01). All recurrences were diagnosed in patients with pattern C tumors (p < 0.01). Conclusions: Recurrent cervical adenocarcinoma was seen only in patients with pattern C tumors, while no patient with pattern A or B tumors recurred. This histopathologic stratification could be utilized to propose individualized therapy for women with newly diagnosed endocervical adenocarcinoma.