Usefulness of significant morphologic characteristics in distinguishing between Merkel cell polyomavirus–positive and Merkel cell polyomavirus–negative Merkel cell carcinomas

• Published in: Human pathology 2013-09, Vol.44 (9), p.1912-1917
• Main Authors: Iwasaki, Takeshi, Matsushita, Michiko, MT, MA, Kuwamoto, Satoshi, MD, PhD, Kato, Masako, MD, PhD, Murakami, Ichiro, MD, PhD, Higaki-Mori, Hiromi, MD, Nakajima, Hideki, MD, PhD, Sano, Shigetoshi, MD, PhD, Hayashi, Kazuhiko, MD, PhD
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Antigens, Viral, Tumor - immunology; Biomarkers, Tumor - metabolism; Carcinoma, Merkel Cell - diagnosis; Carcinoma, Merkel Cell - metabolism; Carcinoma, Merkel Cell - virology; Carcinoma, Squamous Cell - diagnosis; Carcinoma, Squamous Cell - metabolism; Carcinoma, Squamous Cell - virology; Cell Nucleus - pathology; Diagnosis, Differential; Female; Histologic criteria; Humans; Infections; Male; Medical research; Merkel cell carcinoma; Merkel cell polyomavirus; Merkel cell polyomavirus - genetics; Merkel cell polyomavirus - immunology; Merkel cell polyomavirus - isolation & purification; Middle Aged; Morphologic prediction for MCPyV infection; Neoplasms, Multiple Primary; Pathology; Polyomavirus Infections - diagnosis; Polyomavirus Infections - metabolism; Polyomavirus Infections - virology; Reproducibility of Results; Sensitivity and Specificity; Skin Neoplasms - diagnosis; Skin Neoplasms - metabolism; Skin Neoplasms - virology; Software; Tumor Virus Infections - diagnosis; Tumor Virus Infections - metabolism; Tumor Virus Infections - virology
• ISSN: 0046-8177; 1532-8392
• DOI: 10.1016/j.humpath.2013.01.026

Summary Merkel cell polyomavirus (MCPyV) monoclonally integrates into genomes of approximately 80% of Merkel cell carcinomas (MCCs) and undergoes mutation. We previously demonstrated statistically significant differences in tumor cell morphology and biology between MCPyV-positive and MCPyV-negative MCCs. We reassessed the usefulness of our morphologic criteria in differentiating MCPyV-negative and MCPyV-positive MCCs for practical diagnosis. Two trainees and 4 pathologists challenged estimations (5-point confidence scale) of MCPyV infection in MCCs using hematoxylin and eosin–stained slides of 43 new MCC cases and 2 morphologic criteria: (1) nuclear polymorphism is higher and cytoplasm is more abundant in MCPyV-negative MCC cells, and (2) MCC combined with squamous cell carcinoma is defined as MCPyV negative, regardless of tumor cell morphology of MCC. Subsequently, immunohistochemistry for MCPyV large T antigen and polymerase chain reaction for MCPyV DNA yielded concordant results (MCPyV positivity was 30/43 and 32/43, respectively) for 41 (96%) of 43 cases. The mean accuracy, sensitivity, and specificity of the trainees and pathologists were 92.4% ± 1.5% and 81.5% ± 11.0%, 95.6% ± 6.2% and 90.2% ± 8.3%, and 83.3% ± 11.8% and 74.6% ± 14.1%, respectively. Values of the areas under the curve were 0.80 to 0.95, indicating good informative scores. Using our morphologic criteria, observers can predict the absence of MCPyV infection and diagnose MCPyV-negative MCCs with poor prognosis. Unexpectedly, the performance of trainees was superior to that of pathologists, implying that our morphologic criteria are useful even for practitioners having little experience. Our morphologic criteria will provide pathologists with convenient and reliable hallmarks for accurate MCC diagnosis.