Diagnostic accuracy of phosphatase and tensin homolog loss in differentiating between atypical endometrial hyperplasia/endometrioid intraepithelial neoplasia and non-atypical endometrial hyperplasia: A systematic review and meta-analysis.

BACKGROUND

Accurate distinction between atypical endometrial hyperplasia/endometrioid intraepithelial neoplasia (AH/EIN) and non-atypical endometrial hyperplasia (EH) determines whether patients require hysterectomy or can be managed conservatively. Phosphatase and tensin homolog (PTEN) loss has been proposed as an immunohistochemical (IHC) marker to support diagnostic evaluation.

OBJECTIVES

This study evaluates the diagnostic value of PTEN IHC for the diagnosis of AH/EIN and to explore how PTEN assessment methods affect diagnostic performance.

METHOD: PubMed, Cochrane Library, ClinicalTrials.gov, Scopus, Embase, MEDLINE, and Web of Science were searched from database inception up to June 1, 2025. Studies reporting PTEN IHC in women with histologically confirmed AH/EIN or non-atypical EH. Two reviewers independently screened records, extracted data, and assessed risk of bias with QUADAS-2. Pooled diagnostic estimates and summary receiver-operating characteristics (SROC) curves were calculated using a bivariate random-effects model with subgroup analysis and the IHC method.

RESULTS

Twenty-seven studies (2274 women) were included. Pooled sensitivity was 58.0% (95% confidence interval [CI], 40.8-73.4) and specificity 84.7% (95% CI, 67-93.8). The diagnostic odds ratio was 3.695 (95% CI, 2.501-5.460), with an area under the SROC of 0.679. The PTEN assessment method based on percentage of stained cells and staining intensity showed the highest the area under the curve of 0.744 among the methods evaluated in the subgroup analysis.

CONCLUSIONS

PTEN IHC assessment demonstrated good specificity and might serve as a complementary tool in diagnosing AH/EIN, helping to reduce the number of incorrect diagnoses of atypia and unnecessary hysterectomies in clinical practice. Its limited sensitivity precludes its use as a standalone marker. Standardized staining protocols and multi-marker strategies are needed.

PROSPERO REGISTRATION: CRD42024607623, registration date the 29 October 2024.