Risk stratification and outcomes in 210 gynecologic perivascular epithelioid cell tumors (PEComas) cases

To better understand the risk stratification and outcomes of gynecologic PEComas.

Clinicopathological features and outcomes of gynecologic PEComas cases reported in both English and Chinese literature before September, 2020 were evaluated. The efficacy of three proposed criteria were compared to verify their practicability in gynecologic PEComas. The Chi-square test and Cox proportional hazard model were performed for statistical analysis.

A total of 210 cases were retrieved: 95 from English literature and 115 from Chinese literature. The Flope criterion achieved an accuracy of 47% for detecting malignancy of gynecologic PEComas, 64.2% for the Schoolmeester criterion, and 63.8% for the WHO criterion. Both Chi-square test and uni-variate analysis showed that tumor size ≥ 5 cm, infiltrative growth pattern, mitotic rate ≥ 1/50 high per filed (HPF), high nuclear grade and cellularity, necrosis, and vascular invasion were significantly related to recurrence and/or metastasis (R/M) of gynecologic PEComas. Still only high mitotic rate (≥ 1/50 HPF), high nuclear grade and cellularity, and necrosis significantly influenced the long-term survival. Multi-variate analysis showed high nuclear grade and cellularity was an independent risk factor for R/M of gynecologic PEComas. No model was fitted for the death rate due to a small number of events. When defined malignant PEComas cases as meeting three or more out of six clinicopathologic features, the accuracy of such attempt was 62%, but the false-negative rate dropped by 37-55%.

Gynecologic PEComas with three or more high-risk factors may be considered as malignant. Further efforts should be invested to look for new potential prognostic factors.