Inter-observer variability in the histologic criteria of diagnosis of hydatidiform moles

In the event of encountering hydropic villi in products of conception specimens, pathologists will have to distinguish complete and partial hydatidiform mole (CHM & PHM) from hydropic abortion (HA). The histological diagnostic criteria are subjective and demonstrate considerable inter-observer variability.

This study evaluated the inter-observer variability in diagnosis of CHM, PHM and HA according to defined histologic criteria. Ninety abortus conception specimens were reviewed. Representative haematoxylin and eosin-stained slides were assigned independently to two pathologists who were asked to make a diagnosis of CHM, PHM or HA, and provide a report of the identified diagnostic histological criteria. Kappa value was calculated for the inter-observer agreement.

There was a total of 36.7% disagreement between two pathologists (K = 0.403, Strength of Agreement = moderate), of which 24.4% and 12.2%, were differentiating PHM from CHM and PHM from HA, respectively. Among defined diagnostic histological criteria, the highest rate of agreement was observed in the identification of cistern formation and hydropic changes (K = 0.746 and 0.686 respectively, Strength of Agreement = substantial).

There was moderate to substantial agreement rate between two pathologists in identification of two essential histologic criteria for diagnosis of molar pregnancies i.e. "hydropic change" and "trophoblastic proliferation".