Reproducibility of Lymphovascular Space Invasion Assessment in Cervical Squamous Cell Carcinoma: Does Experience Matter?

Lymphovascular space invasion (LVSI) in cervical squamous cell carcinoma is associated with increased nodal metastasis and disease recurrence, and this parameter is incorporated into clinical risk assessment and management algorithms. Despite this, there is limited data regarding pathologist reproducibility for LVSI diagnosis in cervical carcinoma. Recently, we assessed interobserver agreement for LVSI diagnosis using digital whole slide images of early-stage cervical squamous cell carcinomas with a mixture of presumed true LVSI (n=20) and LVSI mimics (n=20). Ten experienced pathologists with an average of 24 years of independent reporting experience achieved substantial agreement (κ value=0.76), with an overall percent agreement (OPA) of 88%. Herein, we evaluate the impact of years of experience on diagnostic reproducibility of LVSI assessment using the same study set of 40 whole slide images. Fourteen participants, comprising a mix of pathology trainees (n=6) and early-career pathologists (n=8), the latter with an average of 1.25 years of independent reporting experience, achieved borderline substantial agreement (κ value = 0.64) and an OPA of 82.5% (trainee OPA 78% and κ= 0.55; early-career pathologist OPA 86% and κ=0.72). Although diagnostic concordance for LVSI assessment was lower among trainees and early-career pathologists compared with experienced pathologists, the finding lacked statistical significance. This study provides evidence that trainees and early-career pathologists can recognize LVSI with a relatively high degree of accuracy, even when evaluating challenging cases in a digital setting.