The frequency and prognostic significance of the histologic type in early-stage ovarian cancer (OC) is not as well established as in advanced stages. In addition, histologic typing based only on morphologic features may be difficult, especially in high-grade tumors. In this study, we have analyzed a prospective cohort of 502 early-stage OCs to investigate their frequency, immunohistochemical characteristics, and survival of the 5 main histologic types. Histotype was assigned according to not only the morphologic features but also according to the expression pattern of WT1, p53, Napsin A, and progesterone receptors. In addition, an extended panel including p16, β-catenin, HER2, Arid1A, HINF1B, CK7, CDX2, and CK20 was used to refine the diagnosis in difficult cases. In this series, the frequency of the 5 major histologic types was as follows: endometrioid carcinoma, 32.7%; clear cell carcinoma, 25.1%; high-grade serous carcinoma (HGSC), 24.7%; mucinous carcinoma, 10.2%; low-grade serous carcinoma, 4.6%; and others, 2.8%. The combination of morphology and immunohistochemistry allowed the reclassification of 23% of OCs. The lowest concordance was found between samples initially diagnosed as endometrioid, but finally classified as high-grade serous tumors (22% error rate). Endometrioid carcinoma was the most favorable histologic type, whereas HGSC and low-grade serous carcinoma had the worst prognosis. Clear cell carcinoma with abnormal p53 immunostaining pattern also had poor prognosis. Although histologic grade was not a prognostic factor among early-stage endometrioid OCs, distinction between grade 3 endometrioid OC and HGSC is recommended, taking into account differences in prognosis and molecular alterations that can guide different treatments.
*Ramón y Cajal Health Research Institute
†CIBER-ONC, Carlos III Health Institute
§Department of Pathology, Hospital Ramón y Cajal
#Department of Pathology, Hospital Universitario La Paz, IdiPAZ
**Faculty of Medicine, Autonomous University of Madrid
¶¶Faculty of Medicine, Alcalá University, Madrid
‡Clinical Area of Gynecological Oncology, Valencian Institute of Oncology
¶Molecular Biology Laboratory, Valencian Institute of Oncology
§§Department of Pathology, Valencian Oncology Institute Foundation
††Department of Pathology, Hospital Universitario La Fe
‡‡Department of Pathology, University of Valencia
##Oncogynecologic Department, Initia Oncology, Hospital Quironsalud Valencia, Valencia
∥Department of Medical Oncology, Hospital de la Santa Creu i Sant Pau, Barcelona
∥∥Department of Pathology, Virgen del Rocío University Hospital, Sevilla, Spain
S.L. and I.R. contributed equally.
J.A.L-G., J.P. and A.P. contributed equally as corresponding authors.
Conflicts of Interest and Source of Funding: Supported by grants from the Instituto de Salud Carlos III (ISCIII) (PIE15/00050 and PI16/00887) and CIBERONC (CB16/12/00316), cofinanced by the European Development Regional Fund ‘A way to achieve Europe’ (FEDER), and by the Spanish Association Against Cancer Scientific Foundation (aecc) (Grupos Coordinados Traslacionales aecc 2018), and S.L. receives a grant from aecc (AIO-aecc 2016). This study has also been partially funded by the Spanish Group of Research in Ovarian Cancer (GEICO group). For the remaining authors none were declared.
Correspondence: José Palacios, MD, PhD, Hospital Ramón y Cajal, Carr. de Colmenar Viejo, km. 9,100, 28034 Madrid, Spain (e-mail: firstname.lastname@example.org).