|
|||||
|
|
||||||
Journal of Clinical Oncology, Vol 22, No 11 (June 1), 2004: pp. 2240-2242 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.10.031
Paraneoplastic Syndromes in CancerCASE 1. Polycythemia As a Result of Ectopic Erythropoietin Production in Metastatic Pancreatic Carcinoid TumorDepartments of Medical Oncology and Hematology, Radiology, and Anatomopathyology, Oncologic Center, Laarbeeklaan, Belgium A 51-year-old woman first presented in February 1999 with asthenia, ascites, and altered liver function tests. The hemoglobin level was 17.2 g/dL (upper limit of normal, 14.5) with a hematocrit of 50.4% (upper limit of normal, 43.9). Platelet and WBC counts were normal. In 1994, all hematologic parameters were normal. A computed tomography scan of the abdomen revealed a heterogeneous, hypervascular, well-demarcated nodular mass with a diameter of 3 cm at the junction of the pancreatic corpus and tail. The liver showed widespread well-delineated homogenous hypervascular lesions in all segments. At the referring hospital, microscopic examination of a liver biopsy showed a nodular monomorphous tumor surrounded by normal parenchyma. Small, polygonal tumor cells displayed an epithelial growth pattern. Mitotic figures were rare. Immunohistochemistry was negative for cytokeratin 7 and cytokeratin 20. Although immunohistochemistry was also negative for several neuroendocrine markers (neuron specific enolase, synaptophysin, chromogranin), a diagnosis of nonsecretory carcinoid pancreatic tumor with diffuse liver metastases was made, based on the typical light microscopic findings. Octreotide scintigraphy showed heterogeneous tracer uptake in the complete right hepatic lobe. An intra-arterial digital substraction angiography showed clearly circumscribed intrahepatic dense blush images with varying diameters and accentuated venous drainage. The pancreatic tumor was visualized as a hypervascular lesion supplied through the arteria lienalis. These images were considered typical for a neuroendocrine tumor.1,2 The patient underwent a chemoembolization through several hepatic branches with contour particles of 150 to 250 and 250 to 355 microns, mixed with 14 mg mitomycin-C, followed by the application of microcoils in loco. Eight days after embolization, the patient clinically improved. The hemoglobin level dropped to 13.6 g/dL and the hematocrit dropped to 39.1% (Fig 1).
Between April 1999 and March 2000, another three chemoembolizations for progressive or new hepatic metastases were performed, each resulting in lower hemoglobin levels postprocedure. None of the chemoembolization procedures resulted in internal or external hemorrhages. After each embolization, liver function tests remained stable, with only mild bilirubin rises (1.5 to 2.0 mg/dL, both conjugated and unconjugated fractions). In May 2000, a laparoscopic distal pancreatectomy and splenectomy were performed, as was a liver biopsy. Pathologic examination of pancreatic and hepatic tissue confirmed the initial diagnosis of carcinoid tumor (Fig 2A). The hemoglobin level dropped from 15.4 g/dL to 12.5 g/dL postoperatively. In September 2001, a fifth chemoembolization was done, without postprocedure lowering of hemoglobin levels. In February 2002, a hepatic computed tomography scanning revealed diffuse progressive lesions. The hemoglobin level was 18.2 g/dL; hematocrit was 56.9%. Systemic chemotherapy was started (methotrexate 40 mg/m2 intravenous bolus and infusional fluroblastine 60 mg/kg body weight over 48 hours, every 2 weeks). At no time point was exogenous erythropoietin administered. The subsequent hemoglobin levels varied between 17.6 g/dL and 15.1 g/dL; hematocrit varied between 54.4% and 43.9%. A serum erythropoietin (EPO) level was 66.1 U/L (normal range, 11 to 30 U/L; EPO-Trac, 125I RIA-Kit, DiaSorin Inc, Stillwater, MN). In September 2002, serum EPO was 49.1 U/L.
The polycythemia, throughout the patient's clinical course, fluctuated with disease volume (Fig 1), decreasing after procedures aimed at reducing disease (chemoembolization and surgery) and progressively increasing in the interval periods. In the absence of a demonstrable primary or secondary cause for the polycythemia (echocardiography, pulmonary function tests, bone marrow examination, and imaging of the kidneys were normal), the possibility of ectopic EPO production was considered. Immunohistochemical staining of the biopsy specimens of both the primary pancreatic and metastatic hepatic tumors revealed specific positive EPO immunnoreactivity (Fig 2B), thus confirming the tumor cells as a direct source of EPO production. In the normal pancreatic tissue, rare Langerhans islet cells exhibiting EPO immunnoreactivity were observed (Fig 3). To our knowledge, this is the first report of erythropoietin production by a carcinoid tumor of the pancreas metastatic to the liver. No cause for a secondary or primary polycythemia could be found and immunohistochemistry supports the production of EPO by the tumor cells resulting in elevated serum EPO levels. In addition, the hemoglobin levels varied in concordance with other measures of disease remission and progression. While ectopic EPO production by a carcinoid tumor has not been described, myeloproliferative disorders and primary polycythemia have been reported as a paraneoplastic phenomenon.35 In none of these cases has the mechanism leading to the paraneoplastic phenomenon been elucidated. Carcinoid tumors typically have neuroendocrine characteristics, synthesizing numerous bioactive amines and peptides, including neuron specific enolase, 5-hydroxytryptamine, 5-hydroxytryptophan, synaptophysin, chromogranin A and C, insulin, glucagon, growth hormone, neurotensin, adrenocorticotropic hormone, beta-melanin stimulating hormone, gastrin, pancreatic polypeptide, vasoactive intestinal polypeptide, somatostatin, calcitonin, substance P, tachykinins, growth hormone releasing hormone, - and ß-human chorionic gonadotropin, bombesin, and various growth factors such as tumor growth factor beta, platelet-derived growth factor, and beta-fibroblast growth factor.6 Hormone synthesis often does not result in secretion and elevated serum levels or clinical symptoms. Surgery is a key treatment modality with regard to neuroendocrine tumors, not only for the primary site, but also for palliation of metastatic disease. Chemoembolization can result in symptom control and reduction in tumor burden.7 Medical treatment includes the long-acting somatostatin analog octreotide, which inhibits neuroendocrine cell growth and hormone release8 and fluorouracil-based chemotherapy regimens.9 In our patient, repeated treatments resulted in decreased hemoglobin levels, concurrent with radiologic disease regression (Fig 1). Despite the large hepatic disease volume at presentation and diminishing effectiveness of chemoembolization and chemotherapy with time, the patient is alive, 4 years later, representing a rather indolent long course, often seen in carcinoid tumors.
In our patient, normal Langerhans islet cells also displayed EPO immunoreactivity. (Fig 3) This has not as yet been described. We also stained pancreatic tissue from 10 organ donor patients for EPO. Only scant dispersed EPO positive exocrine and endocrine cells were found in the tissue examined. These findings also rule out a cross reaction of the erythropoietin antibody (Monoclonal Antihuman EPO Antibody, clone AE7A5; R&D Systems Inc, Marina Del Ray, CA) with other islet cell markers. It is possible that in our patient the islets, and consequently this particular carcinoid tumor, arose from a subset of pancreatic progenitor cells producing EPO physiologically. Since EPO has antiapoptotic properties, this could have resulted in neoplastic growth of the pancreatic endocrine tissue.10 Alternatively, the carcinoid cells could have acquired this feature as an alteration of their differentiation status caused by the process of malignant transformation. It thus appears that rarely carcinoid tumors, at least those arising from the pancreas, can ectopically secrete EPO in sufficient amounts to cause polycythemia. Authors' Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest.
REFERENCES 1. Boysen E: Pancreatic Anigography, in Abrams HL (ed): Vascular and Interventional Radiology. Philadelphia New York, Lippincott Williams and Wilkins, 1983, pp 142766 2. Baum S: Hepatic Arteriography, in Abrams HL (ed): Vascular and Interventional Radiology. Philadelphia New York, Lippincott Williams and Wilkins, 1983, pp 1479504 3. Nelson RL: The association of carcinoid tumors of the rectum with myelofibrosis. Dis Colon Rectum24:548549, 1981[CrossRef][Medline] 4. Lara JF, Rosen PP: Extramedullary erythropoiesis in a bronchial carcinoid tumor. Arch Pathol Lab Med114:12831285, 1990[Medline] 5. Sham RL, Von Doenhoff L, Dipoala J, et al: Polycythemia vera associated with metastatic carcinoid. Cancer Invest14:120123, 1996[CrossRef][Medline] 6. Jensen RT, Doherty GM: Carcinoid Tumors and the Carcinoid Syndrome, in De Vita VT (ed): Cancer: Principles and Practice of Oncology. Philadelphia New York, Lippincott Williams and Wilkins, 2001, pp 181333 7. Schell SR, Camp ER, Caridi JG, et al: Hepatic artery embolization for control of symptoms, octreotide requirements, and tumor progression in metastatic carcinoid tumors. J Gastrointest Surg6:664670, 2002[CrossRef][Medline] 8. Aparicio T, Ducreux M, Baudin E, et al: Antitumour activity of somatostatin analogues in progressive metastatic neuroendocrine tumours. Eur J Cancer37:10141019, 2001[CrossRef][Medline] 9. Oberg K: Neuroendocrine gastrointestinal tumorsa condensed overview of diagnosis and treatment. Ann Oncol10:S3S8, 1999 (suppl 2)[Abstract] 10. Fenjves ES, Ochoa MS, Cabrera O, et al: Human, nonhuman primate, and rat pancreatic islets express erythropoietin receptors. Transplantation75:13561360, 2003[CrossRef][Medline]
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|