Karel Pacak, MD, PhD, DSc, Head, Section on Medical Neuroendocrinology
Anna Kozupa, MD, Clinical Associate
Henri Timmers, MD, PhD, Visiting Fellow
Karen T. Adams, MSN, CRNP, Research Nurse
Edwin W. Lai, BS, Predoctoral Fellow
Shiromi M. Perera, BS, Technician
Daniel C. Solis, BA, NIH Academy
Stephanie Fliedner, MS, Predoctoral Visiting Fellow
Lucia Martiniova, MS, Predoctoral Visiting Fellow

We conduct patient-oriented, conceptually innovative research into the etiology, pathophysiology, genetics, diagnosis, prognosis, and treatment of pheochromocytoma. Projects include not only translational research that involves the application of basic science knowledge to clinical diagnosis, pathophysiology, and treatment but also "reverse translation research," whereby appreciation of clinical findings leads to new concepts that basic researchers can pursue in the laboratory. We emphasize bedside-to-bench projects and multidisciplinary and multi-institutional collaborations. Our work is organized around three objectives: (1) development and testing of novel methods and criteria to diagnose and localize pheochromocytoma cost-effectively; (2) development of better treatments for malignant pheochromocytoma; and (3) identification of molecular and genetic mechanisms of pheochromocytoma tumorigenesis and clinical manifestations of disease. We expect that studies directed at the last objective will have the most far-reaching consequences by leading to new strategies for diagnosis and therapy of pheochromocytoma, particularly in the case of metastatic disease.

Recent advances in, current approaches to, and clinical recommendations for pheochromocytoma

Pacak, Timmers, Adams, Kozupa, Raygada; in collaboration with Breza, Carrasquillo, Chen, Eisenhofer, Fojo, Ilias, Lenders, Libutti, Linehan, Mannelli, Reynolds, Udelsman, Wesley

Recently, we summarized new clinical, diagnostic, and localization procedures and approaches for benign and metastatic sporadic and familial pheochromocytomas. Plasma and urinary fractionated metanephrine assays are the most accurate screening procedures for biochemical diagnosis. Initial testing should include measurement of fractionated metanephrines in urine, plasma, or both when such assays are available, and reference intervals should favor high sensitivity over high specificity to avoid a missed diagnosis. Localization studies should be initiated only after establishing reasonably compelling clinical evidence of tumor, including signs and symptoms of catecholamine excess, strongly positive biochemical tests, hereditary predisposition, or history of tumor. When feasible, laparoscopic surgery is the treatment of choice for removal of the primary tumor, with some type of pre-operative blockade to protect against released catecholamines. Findings that about a quarter of tumors develop secondarily to germline mutations of five genes identified to date indicate a need to consider germ-line mutation testing of many, if not all, affected patients. However, the testing of every patient for every disease-causing gene is not currently cost-effective. An algorithm considering tumor location, multiple tumors, presence of metastases, and type of catecholamine secreted may be useful in deciding which genes to test. Inadequate methods for distinguishing malignant from benign tumors and lack of effective treatments for malignancy are important problems requiring further resolution. Given the lack of curative treatment for metastatic pheochromocytoma, we are testing the efficacy of radiotoxic treatment of malignant pheochromocytoma by using [131I]-metaiodobenzylguanidine ([131I]-MIBG); in particular, we are evaluating whether pretreatment with "enhancer" pharmaceuticals increases the efficacy of experimental [131I]-MIBG treatment in reducing the size and number of tumors.

Biochemical diagnosis of pheochromocytoma

Pacak, Timmers, Kozupa, Adams, Solis; in collaboration with Eisenhofer, Fojo, Goldstein, Lenders, Linehan, Mannelli

In collaboration with Graeme Eisenhofer and others, we developed a biochemical test involving measurements of plasma-free normetanephrine and metanephrine that offers advantages over other biochemical tests for diagnosis of pheochromocytoma. In contrast to catecholamines, metanephrines are produced continuously and independently of exocytotic catecholamine release. Recently, we reported that plasma concentrations of free metanephrines are relatively independent of renal function and therefore more suitable for diagnosis of pheochromocytoma among patients with renal failure than measurements of deconjugated metanephrines. We also found that measurements of plasma-free metanephrines not only provide information about the likely presence or absence of a pheochromocytoma but, when a tumor is present, can also help predict tumor size and location. Currently, we are also attempting to determine the diagnostic utility of measurements of plasma-free metanephrines in patients with metastatic pheochromocytoma before and after experimental treatment.

Patients with dopamine-secreting tumors are often normotensive and therefore pose a significant diagnostic challenge. Such patients first present with either symptoms reflecting local invasion of a tumor or metastatic disease. Recently, we introduced the O-methylated metabolite of dopamine, methoxytyramine, as an additional biochemical blood test to detect dopamine-secreting tumors. We found that methoxytyramine provides a better marker for these tumors than urinary dopamine, which is derived primarily from renal extraction and decarboxylation of circulating DOPA.

Imaging modalities in the evaluation of patients with pheochromocytoma

Pacak, Brouwers,1 Adams, Kozupa, Kaji2; in collaboration with Carrasquillo, Chen, Eisenhofer, Fojo, Goldstein, Ilias, Linehan, Ling, Reynolds

Pheochromocytoma constitutes a form of surgically curable hypertension, and failure to localize the tumor can result in lethal complications due to organ damage by catecholamine excess. Given the limitations of anatomic nonspecific imaging techniques for visualization of pheochromocytoma, we initiated a collaboration with the PET and Nuclear Medicine departments to evaluate 6-[18F]-fluorodopamine PET scanning for diagnostic localization of pheochromocytoma. Our results showed that, in patients with pheochromocytoma (including metastatic disease), 6-[18F]-fluorodopamine PET scanning could detect and localize pheochromocytomas with higher sensitivity than other functional imaging modalities. However, 6-[18F]-fluorodopamine does not accumulate in some metastatic pheochromocytomas. Therefore, we hypothesized that the use of 2-[18F]-fluoro-2-deoxy-D-glucose (FDG) could be considered an ancillary tool in patients with such pheochromocytomas, as previously described for [123/131I]-MIBG-negative patients. Retrospectively, in five patients with either extensive or more rapidly growing metastatic pheochromocytoma, FDG PET showed lesions that were not detected on 6-[18F]-fluorodopamine PET scanning. Therefore, we recommended FDG PET as an ancillary test for the diagnosis and localization of metastatic pheochromocytoma. Our findings led to important refinements in clinical diagnostic approaches in pheochromocytomas, which we have since adopted and continue to apply and improve. New imaging algorithms show how to use anatomical imaging studies together with functional imaging as a means of localizing benign and metastatic lesions in a highly cost-effective manner and with particularly high sensitivity.

Currently, we are performing a large prospective study to compare 6-[18F]-fluorodopamine PET scanning with [123I]-MIBG, Octreoscan, and [18F]-fluorodeoxyglucose PET scanning to determine whether 6-[18F]-fluorodopamine PET scanning may be considered as the "first-line" imaging method in the diagnosis of various pheochromocytomas, especially in patients in whom the tumor cannot be localized by other currently available imaging modalities.

The molecular genetic basis of tumorigenesis in malignant pheochromocytomas

Ohta, ³ Lai, Brouwers, ¹ Pacak; in collaboration with Breza, Chan, Eisenhofer, de Krijger, Ksinantova, Lubensky, Munson, Pang, Kvetnansky

About 50 percent of patients with metastatic pheochromocytoma die within five years of diagnosis because of the lack of appropriate therapy. Thus, there is a need for markers that can identify patients who may develop or have metastatic pheochromocytoma. In addition, these markers could be explored as potential new therapeutic targets. Metastasis suppressor genes affect the spread of several cancers and therefore may provide promise as prognostic markers or therapeutic targets for malignant pheochromocytoma. We hypothesized that the downregulation of metastasis suppressor genes in malignant pheochromocytoma may play a role in malignant behavior.

We applied quantitative real-time polymerase chain reaction to 11 metastasis suppressor genes known to be involved in the regulation of important cancer-related cell events, such as cell growth regulation and apoptosis (nm23-HI, TIMP-1, TIMP-2, TIMP-3, TIMP-4, TXNIP, and CRSP-3), cell-cell communication (BRSM-1), invasion (CRMP-1), and cell adhesion (E-Cad and KiSS1). Following cross-validation, the non-linear rule produced no errors in 10 malignant samples and three errors in 15 benign samples, with an overall error rate of 12 percent. The results suggest that downregulation of metastasis suppressor genes reflects malignant pheochromocytoma with a high degree of sensitivity.

In another study, we used oligonucleotide microarrays to examine gene expression profiles in 98 pheochromocytomas, of which 26 malignant tumors accounted for metastases and primary tumors from which metastases developed. Other subgroups of tumors included those defined by tissue norepinephrine as compared with epinephrine contents (i.e., noradrenergic versus adrenergic phenotypes), adrenal versus extra-adrenal locations, and the presence of underlying germline mutations of SDHB, RET, and VHL genes. Correcting for the confounding influence of noradrenergic versus adrenergic catecholamine phenotype by analysis of variance revealed a larger and more accurate number of genes that discriminated benign from malignant pheochromocytoma than when the confounding influence of catecholamine phenotype was not considered. Seventy percent of these differentially expressed genes were underexpressed in malignant versus benign tumors. Similarly, 89 percent of genes were underexpressed in malignant primary tumors versus benign tumors, suggesting that malignant potential is largely characterized by a dedifferentiated pattern of gene expression. The present database of differentially expressed genes provides a unique resource for mapping the pathways leading to malignancy and establishing new targets for treatment and diagnostic and prognostic markers of malignant disease. The database may also be useful for examining mechanisms of tumorigenesis and genotype-phenotype relationships. Further progress based on the database can benefit from application of bioinformatics approaches for data mining and pathways analyses, follow-up confirmatory quantitative PCR and proteomics studies, and testing in pheochromocytoma cell culture and animal model systems.

Recently, succinate dehydrogenase subunit B (SDHB) mutations have been associated with an increased rate of malignancy as compared with apparently sporadic non-syndromic pheochromocytomas. However, there is no report on the prevalence of the SDHB mutation in a population of patients with metastatic pheochromocytoma. We therefore investigated the prevalence of the SDHB mutation in such a population and discovered that more than 30 percent of the patients harbored an SDHB mutation. As a result, all pheochromocytoma patients with metastatic disease warrant testing for SDHB.

Molecular and other mechanisms that link different pheochromocytoma tumor cell phenotypes and clinical presentation of disease to specific underlying mutations and transporter expression

Pacak, Ohta,3 Lai, Quimby4; in collaboration with Chan, Cleary, Eisenhofer, Elkahloun, Green, Huynh, Lubensky, Morris, Munson, Phillips, Tischler, Vortmeyer, Zhuang

Recently, we found that many of the genes overexpressed in VHL as compared with MEN 2 tumors were clearly linked to the hypoxia-driven angiogenic pathways that are activated in VHL-associated tumorigenesis. Further studies in VHL-associated tumors suggested that VHL gene deficiency caused co-expression of erythropoietin (Epo) and its receptor (Epo-R), which facilitated tumor growth. The objective of our study was to understand the different mechanisms of tumorigenesis for VHL- and MEN 2-associated pheochromocytomas. We found co-expression of Epo and Epo-R in all VHL-associated pheochromocytomas, whereas we documented expression of Epo-R, but not of Epo, in all MEN 2-associated pheochromocytomas. Expression of Epo appears to result from VHL gene deficiency, possibly through activation of the hypoxia-inducible factor-1 (HIF-1) pathway, while Epo-R is an embryonal marker whose sustained expression in both VHL- and MEN 2-associated pheochromocytoma reflects an arrest or defect in development. These findings suggest alternative genetic mechanisms of tumorigenesis in VHL- and MEN 2-associated pheochromocytomas and suggest Epo as a clinical biomarker for differentiating the tumors.

Pheochromocytomas produce catecholamines and numerous secretory proteins and peptides, including neuropeptide Y (NPY), a vasoactive peptide that may contribute to the clinical presentation of the tumors. Our study examined influences of hereditary factors on expression of NPY in pheochromocytomas and relationships with production of epinephrine and norepinephrine. The lower levels of NPY in tumors from VHL patients versus tumors from all other groups of patients strongly suggest that the substantial differences in NPY expression among hereditary tumors reflect an influence of VHL gene mutations rather than any major influence of differences in epinephrine and norepinephrine production.

Animal model of metastatic pheochromocytoma

Martiniova, Ohta,3 Lai, Pacak, Brouwers1; in collaboration with Abu-Asab, Green, Morris, Tischler, Wesley

Successful treatment of any metastatic cancer depends on early detection and localization. However, despite excellent improvements in various imaging techniques, small metastatic lesions are often not detected because of suboptimal spatial resolution of current anatomical and functional imaging modalities. The same applies to animal models of metastatic cancer. Animal models are crucial for the development of new therapeutic approaches and treatments. Recently, we introduced a new model of metastatic pheochromocytoma resulting from tail vein-injected mouse pheochromocytoma cells that reproducibly generated several tumors. For the first time, we showed that microCT using hepatobillary-specific contrast and MRI may reveal liver metastasis as small as 0.35 mm as early as four weeks after initial injection of tumor cells. We also introduced the use of 6-[18F]-fluorodopamine PET in the same model for the detection of metastatic lesions. It will be useful for studies on the in vivo molecular biology and therapeutic strategies for treatment of malignant pheochromocytoma.

¹ Frederieke M. Brouwers, MD, former Postdoctoral Fellow
² Priya Kaji, BS, former Predoctoral Fellow
³ Shoichiro Ohta, MD, PhD, former Postdoctoral Fellow
⁴ Ernika Quimby, former HHMI Student

COLLABORATORS

Mones Abu-Asab, MD, PhD, Laboratory of Pathology, NCI, Bethesda, MD
Salvatore Alesci, MD, PhD, Clinical Neuroendocrinology Branch, NIMH, Bethesda, MD
Jan Breza, MD, PhD, DSc, Komensky University, Bratislava, Slovak Republic
Jorge Carrasquillo, MD, Nuclear Medicine Department, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Wai-Yee Chan, MD, PhD, Laboratory of Clinical Genomics, NICHD, Bethesda, MD
Clara Chen, MD, Nuclear Medicine Department, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Susannah Cleary, BS, Clinical Neuroscience Branch, NINDS, Bethesda, MD
Ronald de Krijger, MD, PhD, Josephine Nefkens Institute, Rotterdam, Netherlands
Graeme Eisenhofer, PhD, Clinical Neuroscience Branch, NINDS, Bethesda, MD
Abdel Elkahloun, PhD, Genome Technology Branch, NHGRI, Bethesda, MD
Tito Fojo, MD, PhD, Cancer Therapeutics Branch, NCI, Bethesda, MD
David Goldstein, MD, PhD, Clinical Neurosciences Program, NINDS, Bethesda, MD
Jeff Green, MD, PhD, Laboratory of Cell Regulation and Carcinogenesis, NCI, Bethesda, MD
Thanh-Truc Huynh, BS, Clinical Neurocardiology Section, NINDS, Bethesda, MD
Ioannis Ilias, MD, University of Patras, Patras, Greece
Elaine Jagoda, PhD, Positron Emission Tomography, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Lucia Ksinantova, MD, PhD, Cyclotron Center, Bratislava, Slovak Republic
Richard Kvetnansky, PhD, DSc, Instituite of Endocrinology, Slovak Academy of Sciences, Bratislava, Slovakia
Jacques Lenders, MD, PhD, Universiteit Nijmegen, Nijmegen, Netherlands
Stephen Libutti, MD, Center for Cancer Research, NCI, Bethesda, MD
W. Marston Linehan, MD, Urologic Oncology Branch, NCI, Bethesda, MD
Alexander Ling, MD, Radiology Department, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Irina A Lubensky, MD, Surgical Neurology Branch, NINDS, Bethesda, MD
Massimo Mannelli, MD, Università di Firenze, Florence, Italy
John Morris, MD, PhD, Metabolism Branch, NCI, Bethesda, MD
Peter J. Munson, PhD, Mathematical Computing Program, CIT, NIH, Bethesda, MD
Alan L.Y. Pang, MD, PhD, Laboratory of Clinical Genomics, NICHD, Bethesda, MD
Jacqueline Phillips, PhD, Murdoch University, Perth, Australia
Margarita Raygada, PhD, Laboratory of Clinical Genomics, NICHD, Bethesda, MD
James Reynolds, MD, PhD, Nuclear Medicine Department, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Arthur S. Tischler, MD, New England Medical Center, Boston, MA
Robert Udelsman, MD, Yale University, New Haven, CT
Alexander O. Vortmeyer, MD, PhD, Surgical Neurology Branch, NINDS, Bethesda, MD
Robert A. Wesley, PhD, Biostatistics and Epidemiology Service, Warren G. Magnuson Clinical Center, NIH, Bethesda, MD
Zhengping Zhuang, MD, PhD, Surgical Neurology Branch, NINDS, Bethesda, MD

For further information, contact karel@mail.nih.gov.