An immunohistochemical study of cervical neuroendocrine carcinomas: Neoplasms that are commonly TTF1 positive and which may express CK20 and P63

Abstract

Cervical small cell neuroendocrine carcinoma (SCNEC) and large cell neuroendocrine carcinoma (LCNEC) are uncommon but highly aggressive neoplasms. From a diagnostic point of view, there may be problems both in distinguishing these from other neoplasms and in confirming a cervical origin. This is important as management is critically dependent on the correct histologic diagnosis. We undertook a detailed immunohistochemical analysis of a relatively large series of primary cervical SCNEC (n=13) and LCNEC (n=8). Cases were stained with AE1/3, chromogranin, CD56, synaptophysin, PGP9.5, TTF1, p16, p63, CK7, CK20, neurofilament, and CD99. CK20 and neurofilament staining was undertaken to investigate whether some of these neoplasms might exhibit a Merkel cell immunophenotype and CD99 staining to assess whether there is immunohistochemical overlap with neoplasms in the Ewing family of tumors (EFT). For all markers, staining was classified as negative, 1+ (<10% cells immunoreactive), 2+ (10 to 50% cells immunoreactive), or 3+ (>50% cells immunoreactive). Eleven and 6 SCNEC and LCNEC, respectively were positive with AE1/3. Chromogranin, CD56, synaptophysin, and PGP9.5 were positive in 11, 19, 19, and 9 cases, respectively. Altogether 15 cases (71%) (11 SCNEC, 4 LCNEC) exhibited nuclear positivity, often diffuse, with TTF1. All but 1 case was diffusely positive with p16. p63 was positive in 9 cases, including 5 with diffuse nuclear immunoreactivity. Ten and 4 neoplasms were positive with CK7 and CK20, respectively. Neurofilament was positive in 7 tumors. The 4 neoplasms that were CK20 positive were stained with the monoclonal antibody CM2B4, generated against an antigenic epitope on the Merkel cell polyomavirus T antigen; all were negative. CD99 was positive in 6 cases. In 2 cases, adjacent foci of adenocarcinoma in situ (AIS) contained scattered individual chromogranin positive cells, raising the possibility that some cervical neuroendocrine carcinomas arise from neuroendocrine cells in AIS. Four of 13 cases of pure AIS also contained scattered chromogranin positive cells. Our results illustrate that a proportion of cervical neuroendocrine carcinomas are negative with broad spectrum cytokeratins and some of the commonly used neuroendocrine markers. TTF1 positivity is extremely common and may be a useful marker of a neuroendocrine carcinoma. It is of no value in exclusion of a pulmonary primary. p16 is almost always positive in cervical neuroendocrine carcinomas, possibly owing to an association with oncogenic human papillomavirus, although other mechanisms of expression are also possible. Cervical neuroendocrine carcinomas may be p63 positive, illustrating that this marker is not specific for squamous differentiation. CK20 and neurofilament positivity in some cervical neuroendocrine carcinomas is in keeping with a Merkel cell immunophenotype, similar to that described in SCNECs in other organs. However, the absence of staining with CM2B4 argues against a true Merkel cell tumor. CD99 staining in a cervical neuroendocrine carcinoma should not result in misdiagnosis as a neoplasm in the Ewing family of tumors.