Journal of Gastrointestinal Cancer

, Volume 43, Issue 1, pp 133–136

Gastric Adenocarcinoma? Looks Similar but Behaves Differently!


    • Department of Internal MedicineUniversity of Connecticut
  • Mridula Rai
    • Department of Internal MedicineUniversity of Connecticut
  • Todd J. Alekshun
    • Division of Hematology-Oncology, Department of Internal MedicineHartford Hospital
Letter to the Editor

DOI: 10.1007/s12029-010-9220-6

Cite this article as:
Rustagi, T., Rai, M. & Alekshun, T.J. J Gastrointest Canc (2012) 43: 133. doi:10.1007/s12029-010-9220-6

Case Presentation

A 63-year-old male presented with unintentional weight loss of 20 lb over a 4-month duration. He reported loss of appetite, intermittent post-prandial nausea, bloating, and early satiety. The patient also complained of dyspepsia and had been empirically treated for reflux disease during the previous 2 years. He denied vomiting, dysphagia, odynophagia, or abdominal pain. There was no history of overt gastrointestinal bleeding or alterations in bowel habits. Additionally, he denied fevers, night sweats, cough, or dyspnea. He had quit smoking 25 years ago, and denied alcohol use. His past medical history was significant for basal cell carcinoma treated with local curative therapy and he was without recurrence on surveillance. Pertinent family history included a paternal uncle with lung cancer at the age of 74. Physical examination was unremarkable except for occult heme-positive stools. Laboratory evaluation revealed elevated liver enzymes (ALT—112 U/L, reference range, 7–49 U/L; AST—81 U/L, reference range, 9–36 U/L; and alkaline phosphatase—364 U/L, reference range, 32–104 U/L).

Computed tomography (CT) of the chest, abdomen, and pelvis showed diffuse, heterogeneous enlargement of the liver with extensive nodularity, raising the concern for metastases (Fig. 1). Serum tumor markers including PSA, CEA, CA 19-9, and AFP were all within normal limits. Colonoscopy done for screening was normal but an esophagogastroduodenoscopy revealed a 3 × 2.5 cm, malignant-appearing, polypoidal lesion with superficial ulceration, involving the gastro-esophageal junction and gastric-cardia (Fig. 2). Histopathology showed an invasive gastric large cell carcinoma with positive immunostaining for neuroendocrine markers including synaptophysin (SYN) and chromogranin A (CGA; Fig. 3a, b). Immunohistochemical staining for TTF-1 was negative, excluding pulmonary origin of the tumor. Ultrasound-guided fine needle aspiration of the largest hepatic lesion revealed malignant cells with cytological features consistent with large cell type carcinoma and positive immunostaining for synaptophysin favoring neuroendocrine differentiation (Fig. 4). Subsequently, a positron emission tomography (PET)–CT demonstrated intense diffuse Fludeoxyglucose (FDG) uptake of the liver, suggestive of diffuse hepatic parenchymal infiltration by tumor (Fig. 5a). There was no evidence of lymphadenopathy, and the remaining intra-abdominal and intra-thoracic structures were unremarkable. However, there were multiple foci of intense osseous FDG uptake, including thoracic vertebrae T7, T8, and right femoral head, with corresponding osteolytic lesions seen on CT scan (Fig. 5b). The patient will receive palliative systemic therapy with a platinum-based regimen.
Fig. 1

CT scan of the abdomen showing diffuse, heterogeneous liver with extensive nodularity
Fig. 2

Endoscopy showing a polypoidal mass with superficial ulceration, involving the gastro-esophageal junction and gastric-cardia
Fig. 3

a Histopathology demonstrating an invasive gastric large cell carcinoma. b Immunohistochemistry showing positive staining for neuroendocrine marker-synaptophysin
Fig. 4

Ultrasound-guided fine needle aspiration of the hepatic lesion showing malignant cells with cytological features consistent with large cell type carcinoma
Fig. 5

a FDG-PET demonstrating intense diffuse FDG uptake of the liver, suggestive of diffuse hepatic parenchymal infiltration by tumor. b FDG-PET showing intense diffuse FDG uptake in the right femoral head


Neuroendocrine (NE) tumors arise from embryonal neural crest cells, which are abundant in the epithelia of the gastrointestinal tract and bronchopulmonary system [1]. In the current WHO [2] classification and several other worldwide histologic classifications [3, 4], well-differentiated endocrine neoplasm and poorly differentiated endocrine carcinoma are the two major categories defined for gastric neuroendocrine tumors. The well-differentiated endocrine neoplasm is essentially equivalent to a carcinoid, and the poorly differentiated endocrine carcinoma is a rare and highly aggressive neoplasm, usually with small cell histologic features [510]. However, the spectrum of gastric NE tumors is broader than has previously been recognized and a new distinct clinicopathologic entity termed “gastric large cell neuroendocrine carcinoma” (GLCNEC) has been described [11].

GLCNEC was first reported by Matsui et al. [9] in 1998 and described in detail by Jiang et al. [11] in 2006. In their retrospective histopathological review of 2,835 resected gastric cancers, GLCNECs accounted for less than 1.5% of all gastric cancers, of which only 10% were diagnosed as poorly differentiated NE carcinomas, whereas the remaining had been classified as conventional adenocarcinomas [11]. Morphologically, these tumors differ from both carcinoids and small cell carcinomas, and are confirmed immunohistochemically using NE markers, CGA and SYN. Jiang et al. used >50% tumor cell positivity of CGA/SYN as the cut-off value to diagnose GLCNEC [11]. However, they found that even >20% positivity of NE markers CGA and/or SYN was enough to characterize a true GLCNEC, as long as light microscopic NE morphology is present in the majority of the tumor [11].

As with its pulmonary counterpart, the greatest difficulty with the diagnosis of GLCNEC is how to recognize its neuroendocrine morphology. Most GLCNECs show solid growth, and the most important issue is to distinguish them from poorly differentiated solid adenocarcinomas [11]. In most tumors, organoid architecture, even if faint and barely recognizable, is a good initial indication of GLCNEC, and the subsequent discernment of typical rosettes by close scanning at higher magnification contributes to a correct diagnosis in most cases [11]. Immunohistochemistry, however, is necessary for confirmation of the diagnosis [1, 9, 11].

GLCNECs have been divided roughly into three subtypes: solid, tubular, and scirrhous, according to their main growth pattern [11]. Solid subtype is the most common (75%), in which the tumor cells are arranged mainly in sheets, nests, or broad trabeculae of various sizes. Much less prominent organoid architecture, and highly frequent mitoses, distinguished these tumors from carcinoids or well-differentiated endocrine carcinomas, and their low nuclear-cytoplasm ratio from small cell carcinomas [11]. GLCNECs of the tubular subtype have tubule-like structures as the dominant feature, reminiscent of the histology of moderately differentiated tubular adenocarcinomas. The scirrhous subtype has histology very similar to so-called scirrhous carcinoma, with separated or small clusters of tumor cells infiltrating desmoplastic stroma. Although organoid architectures or rosettes are completely absent, the deep eosinophilic and finely granular cytoplasm suggests diagnosis, with subsequent confirmation using immunostaining [11].

From the standpoint of a clinical oncologist, the addition of this new entity to an existing tumor classification is not just theoretical, but very practical with a therapeutic and prognostic significance. GLCNECs which had been previously diagnosed as adenocarcinomas have been clearly demonstrated to be highly malignant and portend a significantly worse prognosis than conventional adenocarcinomas. At the time of presentation, vast majority (90%) of GCLNECs are advanced cancers, with 70–75% having lymph node metastases, and 5–10% also having metastases to the liver [11]. The 5-year survival rates for the GLCNECs and conventional adenocarcinomas were 31.1% and 69.3%, respectively, and the overall survival was very significantly different between GLCNECs versus conventional adenocarcinomas (P < 0.0001) [11]. Furthermore, stage-matched survival analyses revealed that, within each stage, patients with GLCNEC had a significantly worse prognosis than those with AC [11]. Thus, GLCNEC must be carefully differentiated from conventional adenocarcinomas not only histopathologically, but should also be taken as a distinct clinical entity.

The clinical management of gastric LCNECs has not been clearly defined given that this is a rare malignancy, and experience in treating this tumor is limited. An overall regression rate of 67% was achieved in a small series of poorly differentiated gastroenteropancreatic NE carcinomas treated with a regimen of etoposide plus cisplatin [12]. Given its biological behavior which appears to be similar to small cell carcinomas, similar chemotherapeutic regimens that are used to treat small cell carcinomas have been employed to treat GLCNEC [13]. Kirii et al. reported a complete response for the gastric lesion and liver metastasis, and a partial response for lymph node metastases with CDDP+CPT-11 combination chemotherapy [13]. Standard criteria for diagnosis, inclusion as a separate well-defined category in the tumor classification, and larger studies evaluating chemotherapeutic regimens are required to elucidate the most effective therapy.

Financial Disclosure

Author(s) have nothing to disclose. All author(s) disclose that there are no potential conflicts (financial, professional, or personal) that are relevant to the manuscript.

Copyright information

© Springer Science+Business Media, LLC 2010