Neuroendocrine Tumors of the Lung With Proposed Criteria for Large-Cell Neuroendocrine Carcinoma

Based on our review of 35 cases and the literature, we found the spectrum of pulmonary neuroendocrine (NE) tumors to be too broad to fit into the traditional three-category classification scheme of typical carcinoid (TC), atypical carcinoid (AC), and small-cell lung carcinoma (SCLC). We found that a spectrum of high- and low-grade tumors exist between TC and SCLC and that in the past many of these tumors have been called AC. We chose to adhere to Arrigoni's definition of AC, as his original criteria characterized a low-grade tumor. For the higher grade non-small-cell tumors (NSCLC), we propose a fourth category of large-cell neuroendocrine carcinoma (LCNEC), which is characterized by: (a) light microscopic NE appearance; (b) cells of large size, polygonal shape, low nuclear-cytoplasmic ratio (N:C), coarse nuclear chromatin, and frequent nucleoli; (c) high mitotic rate [greater than 10/10 high-power fields (HPF)] and frequent necrosis; and (d) NE features by immunohistochemistry (IHC) or electron microscopy (EM). Thus, after deciding that a pulmonary NE tumor is high grade, the major diagnostic issue is separation of LCNEC from SCLC. This distinction is based not only on cell size, but on a variety of morphologic features. We studied 20 TC, six AC, five LCNEC, and four SCLC and characterized the clinical, light microscopic, EM, IHC, and flow cytometric features of each type of tumor. We did not find any advantage to IHC, EM, or flow cytometry over light microscopy in the subclassification or prediction of prognosis; however, these methods were useful in characterizing these four types of pulmonary NE tumors and in demonstrating their NE properties. LCNEC must be distinguished from a fifth category pulmonary NE tumor: NSCLC with NE features in which NE differentiation is not evident by light microscopy and must be demonstrated by EM or IHC. Although the prognosis of LCNEC appears to be intermediate between AC and SCLC, larger numbers of patients will be needed to demonstrate significant differences in survival.