Interobserver diagnostic reproducibility in advanced-stage endometrial carcinoma

Abstract

Background: The accurate pathologic diagnosis and subtyping of high-grade endometrial carcinoma are often problematic, due to its atypical and overlapping histopathological features. Methods: Three pathologists reviewed 21 surgically resected cases of advancedstage endometrial carcinoma. The primary diagnosis was based only on hematoxylin and eosin stained slides. When a discrepancy arose, a secondary diagnosis was made by additional review of immunohistochemical (IHC) stains. Finally, three pathologists discussed all cases and rendered a consensus diagnosis. Results: The primary diagnoses were identical in 13/21 cases (62%). The secondary diagnosis based on the addition of IHC results was concordant in four of eight discrepant cases. Among four cases with discrepancies occurring in this step, two cases subsequently reached a consensus diagnosis after a thorough discussion between three reviewers. Next-generation sequencing (NGS) study was performed in two cases in which it was difficult to distinguish between serous carcinoma and endometrioid carcinoma. Based on the sequencing results, a final diagnosis of serous carcinoma was rendered. The overall kappa for concordance between the original and consensus diagnosis was 0.566 (moderate agreement). Conclusions: We investigated stepwise changes in interobserver diagnostic reproducibility in advanced-stage endometrial carcinoma. We demonstrated the utility of IHC and NGS study results in the histopathological diagnosis of advanced-stage endometrial carcinoma.

Risk Stratification of Endometrial Cancer Patients: FIGO Stage, Biomarkers and Molecular Classification

“…Later, these subtypes were shown to be associated with specific molecular aberrations, such as microsatellite instability (MSI), and mutations in PTEN, KRAS , and CTNNB1 (beta-catenin) mutations in type 1 and TP53 , HER2 , and loss of heterogeneity in type 2 [ 53 ]. However, even among expert gynaecological pathologists, poor to moderate reproducibility was shown in subtype classification in high-grade EC, especially if diagnosis was based on haematoxylin eosin slides only [ 54 , 55 , 56 ]. Immunohistochemical analysis, in specific p53 and the oestrogen and progesterone hormone receptors (ER, PR), has for a longer time been applied to support the subtyping in less straightforward cases [ 57 ].…”

Accuracy of endometrial sampling in the diagnosis of endometrial cancer: a multicenter retrospective analysis of the JAGO-NOGGO

“…However, the specified significance level narrowly missed reaching statistical significance in the two-tailed test ( p = 0.057). Interestingly, several studies have reported poor interobserver viability in endometrial cancer, particularly for high-grade tumors [ 22 – 24 ]. Notably, these studies focused on interobserver variability within single institutions.…”

Application of NGS molecular classification in the diagnosis of endometrial carcinoma: A supplement to traditional pathological diagnosis

“…In addition, with considerable interobserver variation, high‐grade EC cannot be reliably classified according to histomorphologic criteria, resulting in misdiagnosis due to mixed high‐grade components in histology. 12 , 13 , 14 Moreover, stage and lymphovascular space invasion (LVSI), parameters that are needed to define the risk group, are only available after definitive surgery, 15 , 16 such information comes too late for EC patients who may benefit from fertility‐sparing alternatives. This situation prompted research into developing biologically informative diagnostic tools to enhance the diagnosis and risk stratification of patients with EC.…”

“…Traditional classification based on pathologic assessment includes Bokhman classification and WHO 2014 classification 9–11 can be served as an important input for (adjuvant) treatment decisions, but has proved to be unreliable and has poor reproducibility. In addition, with considerable interobserver variation, high‐grade EC cannot be reliably classified according to histomorphologic criteria, resulting in misdiagnosis due to mixed high‐grade components in histology 12–14 . Moreover, stage and lymphovascular space invasion (LVSI), parameters that are needed to define the risk group, are only available after definitive surgery, 15,16 such information comes too late for EC patients who may benefit from fertility‐sparing alternatives.…”