Lymph node metastases in endocervical adenocarcinoma: Correlation with Silva pattern of invasion.

Abstract

e16518 Background: Endocervical adenocarcinoma (EA) is the 2nd most common uterine cervical malignancy (after squamous cell carcinoma). Early stage tumors are defined by depth of invasion (DOI); although used to guide nodal dissection, DOI not an entirely reliable criteria, and most patients have negative nodes. Thus, pathologic criteria that better predict risk of nodal metastases are needed. A prior study of 352 cases from 12 US and international institutions, with complete pathological and clinical follow-up averaging 52 months, found classifying EA by a newly proposed ‘Silva Pattern of Invasion’ (SPI) is the best predictor of nodal metastases. SPI are defined as A: no destructive stromal invasion or lymph-vascular invasion (LVI); B: small foci (<4 mm) of fragmented, angulated glands, solid aggregates or individual tumor cells associated with desmoplastia or LVI; and C: diffuse destructive invasion. Pattern A, 25% of cases, had no nodal metastases or recurrences. Approximately 25% were B, with 5% nodal metastases, 2% recurrence, and no tumor deaths. Pattern C cases: nodal metastases 23%, recurrences 21%, and death from tumor 9%. SPI appeared simple to diagnose, but a formal assessment of reproducibility is needed. Methods: Reproducibility for diagnosing SPI, and for the standard measurement of DOI is compared. Four Gynecological Pathologists, who have read the initial manuscript, but not involved in the original study nor from the participating institutions, and blinded to clinical information examine 1or 2 hematoxylin and eosin-stained glass slides from 90 cases from the original series. Results: 30 cases each of Pattern A, B, and C are reviewed. Kappa statistics will be estimated. With 90 samples, we will have >80% power to detect a true kappa of 0.7 in a test of the null hypothesis that kappa = 0.5 vs the alternative that kappa≠0.5 with equal frequencies in each SPI category. Conclusions: A good reproducibility outcome would indicate that the SPI could be applied in general practice. Accurate use of SPI requires examination of the entire lesion unless pattern B or C is found in the biopsy specimen. Use of the SPI system to guide nodal dissection will require validation in a large series.