Cost-effective improvement in endometrial cancer diagnosis by the incorporation of molecular analysis: prospective pilot study

Results of a cost-effectiveness study of diagnostic algorithms incorporating GynEC®-DX have been presented in the 21st European Gynaecological Oncology Congress of the European Society of Gynaecological Oncology (ESGO).

The study, led by the Bellvitge University Hospital (Barcelona), concludes that the incorporation of themolecular test for diagnosing EC in postmenopausal women with abnormal uterine bleeding reduces the number of explorations. Consequently, the molecular algorithm might be more cost-effective than conventional algorithms.

ABSTRACT

Introduction/Background: The histopathology remains the gold standard to diagnose endometrial cancer (EC) from endometrial biopsy. Molecular tests have recently emerged as a useful tool to classify EC according to its prognosis. However, there is currently no published protocol that includes molecular diagnosis of postmenopausal women with abnormal uterine bleeding (AUB). We hypothesized that the incorporation of a molecular test in the management of postmenopausal women with AUB improves the cost-efectiveness of the diagnostic process.

Methodology: We present a prospective study performed in postmenopausal women who presented AUB between 2009–2014. Seven centers recruited the patients. Three of them follow the classical diagnosis algorithm (group 1) and four centers follow the one that incorporates a molecular test (GynEC®DX) performed on the remnants of aspirates (group 2). In group 2, when both the endometrial biopsy and the molecular test were negative, the consequent explorations were considered as ‘out of procotol’. Clinical data, number of biopsies, ultrasounds, hysteroscopies and visits were compared between groups. In addition, the sensitivity (S), specificity (E), positive and negative predictive values (PPV and NPV) of the molecular test were calculated.

Results: 94 patients were recruited. 51 vs 43 women were included in the classical and the molecular algorithm respectively. There were no differences in age, BMI, parity, use of tamoxifen or hormonal treatment. The detailed outcomes of explorations between classical vs molecular group are shown in table 1. 324 vs 229 explorations were performed respectively (table 2). In the molecular group, 92 explorations were considered ‘out of protocol’. The test validations showed a 100% S, 92.5% E, 50% PPV and 100% NPV.

Conclusion: According to our results, the incorporation of a molecular test for diagnosing EC in postmenopausal women who complained with AUB reduces the number of explorations. Consequently, the molecular algorithm might be more cost-effective than conventional algorithms.

See ePoster EP517