Improved risk prediction in HPV-associated endocervical adenocarcinoma through assessment of binary silva pattern-based classification: international multicenter retrospective study of the international society of gynecological pathologists

Introduction Endocervical adenocarcinomas (EACs) are neoplasms associated with diverse pathogenesis, morphology, and clinical behavior. The Silva pattern-based classification categorizes HPV-associated EACs based on the morphology of the invasion and predicts lymph node metastasis and recurrence. Traditionally the Silva classification was a three-tier system (pattern A, B, and C). A two-tier/binary system has recently been proposed whereby tumors are classified into low risk (pattern A/pattern B without lymphovascular invasion (LVSI)) and high risk (pattern B with LVSI/pattern C). Our aim was to develop a prognostic model for surgically treated FIGO stage IA2-IB3 EACs that incorporates patient age, LVSI, FIGO stage and three- and two-tier Silva systems.

Methods The International Society of Gynecological Pathologists (ISGyP) established a multicenter consortium to pool de-identified individual patient data for patients with HPV-associated EACs. All participating pathologists completed mandatory online training.

Results Our cohort comprised 792 HPV-associated EACs (table 1). On multivariate analysis a binary Silva system was associated with recurrence-free and disease specific survival (p<0.05) while FIGO 2018 stage I substages were not. In the current three-tiered system, disease specific survival for patients with pattern B tumors did not significantly differ from those with pattern C tumors while those with pattern A tumors did (table 2).

Conclusion/Implications These findings highlight the need for future prospective studies to further investigate the prognostic significance of stage I HPV-associated EAC substaging and the inclusion of the binary Silva pattern of invasion classification, which includes LVSI status, as a component of treatment recommendations.