P1-169: Multicentric phase II trial of Cisplatin plus Etoposide chemotherapy in advanced Large–Cell Neuro Endocrine Carcinoma of the Lung (LCNEC): preliminary results. Study 03–02 from the “Groupe Francais de Pneumo–cancerologie” (GFPC)

Abstract

A few prospective studies concerning diagnosis and treatment of large cell neuroendocrine lung cancer were published, as this tumor seems relatively rare. The objectives of GFPC 03-02 prospective study were to establish the efficiency of Cisplatinum /Etoposide chemotherapy and to valid value of large spectrum cytokeratins in the diagnosis of neuro-endocrine differnciation. Patients with stage IV and stage III B (with neoplasic pleural effusion) were included; treatment consisted in three cycles of Cisplatinum 80mg/m2 d1 / Etoposide 100mg/m2 d1d2d3 and two more cycles in case of objective response or stable disease. The primary objective is the efficiency of this chemotherapy, secondary objective the validation of the paranuclear granular expression of a large spectrum cytokeratin, as a characteristic of neuro-endocrine differentiation. Slides were controlled by a panel of pathologists. Diagnoses were based on Travis’s criterias. 27 patients were included since May 2004. The slides of 21 cases were reviewed. 14 cases were confirmed as LCNEC. 4 were excluded on the basis of the nuclear criteria or cell size, and classified as small-cell carcinomas (SCC). One was excluded, because immunohistochemistry was not available. Two other patients were excluded because of the misinterpretation of the pathologic report by the clinicians (one SCC, one Large cell Carcinoma). Among the 14 cases, two sub-types were observed: –8 LCNEC with neuroendocrine features and positive immunohistochemistry–6 undifferentiated carcinomas with immunohistochemical neuroendocrine differentiation. None case was diagnosed as large-cell carcinoma with neuroendocrine pattern and negative immunohistochemistry. In all cases, at least two of the three neuroendocrine markers (chromogranin A, synaptophysin, CD56) were positive. Interobserver variability is not observed on the criterias of neuroendocrine immunohistochemical differentiaton, but only on the criterias of size. Some small scratched samples led to difficulties to distinguish small-cell carcinoma from large-cell carcinoma, especially when neuroendo-crine morphology (rosettes, trabecular growth pattern,_) was not present. It seems necessary to definite how many of the Travis’s crite-rias are necessary to accept the diagnosis of LCNEC (cell size, chromatin, nucleoli, nuclear-to-cytoplasmic ratio, large and eosinophilic or granular cytoplasm) when neuroendocrine differentiation is proved. This interobserver variability is described in the literature (1). Value of the paranuclear granular expression with Cytokeratins must be evaluated on cases not yet tested. Panel of pathologists is useful to diagnose LCNEC. Disagrements were on cell size, some LCNEC were reclassified as SCC. Usefulness of cytokeratins needs some more patients tested.