Introduction Uterine carcinosarcoma (UCS) is a histological subtype of endometrial cancer, with a biphasic morphology. This challenges diagnosing UCS accurately pre-operatively via aspiration biopsy, hysteroscopic biopsies, or dilatation and curettage. This is important as preoperative diagnosis of UCS could impact treatment choices. The aim of this study is to determine the accuracy of each method in diagnosing UCS.

Methods Pathology reports were acquired from the Dutch Nationwide Pathology Databank of patients diagnosed with UCS in pre- and/or postoperative histology between 2001 to 2021. Patients without available pre- or postoperative pathology reports were excluded. The postoperative histology was set as reference. A 2×2 table 1 was plotted to compute the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for each sampling method.

Results 1273 patients were included. Overall sensitivity and specificity of preoperative diagnostic sampling was 35.5% and 73.9% respectively. None of the methods were superior. The sensitivity for aspiration biopsy (n=714) was 30%, with a specificity of 82%, a PPV of 85%, and a NPV of 26%. Hysteroscopic biopsy (n=56) had a sensitivity of 30% and a specificity of 79%, while the sensitivity of dilatation and curettage (n=111) was 38% and specificity 67%. The PPV of hysteroscopic biopsy and dilatation and curettage were 73% and 78%, and the NPV of these sampling methods were 37% and 26% respectively.

Conclusion/Implications Preoperative sampling methods have a low accuracy in diagnosing UCS, irrespective of the sampling technique. This highlights the need for novel preoperative diagnostic or pathologic assessment, i.e. via p53 immunohistochemistry or hypermethylation profiles.