Pheochromocytoma

In September 2008, the International Symposium on Pheochromocytoma held its second conference in the beautiful surroundings of Queens’ College in Cambridge. Pheochromocytoma has long attracted and stimulated much larger interest than its frequency as a medical problem might appear to justify. The meeting in Cambridge was another example of how this rare condition brought together, and held in thrall, an eclectic mix of some two hundred doctors and scientists from around the world and across a wide range of disciplines.

This volume includes a selected number of invited articles, which describe work at the cutting edge of research into either the basic science or clinical management of pheochromocytoma. Some authors have also been asked to review the state of play or make tentative recommendations for the clinic. As is often the case in the endocrinology of rare syndromes, there is a dearth of rigorous data to underpin many of the clinical recommendations. On the other hand, it is becoming clear that rare conditions are likely to be the exceptional cases where we can acquire a sound molecular basis for understanding disease pathogenesis – and sometimes predict effective treatment more successfully than in more common diseases.

This millennium has seen substantial advances in our understanding of the pathogenesis of pheochromocytoma – especially the 20% or more that arise as part of genetic syndromes, and there has been some progress in extrapolation from genetic to sporadic pheochromocytoma. Qin's paper may prove a pivotal, if still evolving, unifying scheme, which suggests that mutations in a number of functionally unrelated genes lead to pheochromocytoma via a final common pathway of defective apoptosis in neural crest precursor cells [1]. One group of genes is now thought to lead to stabilisation of hypoxia inducible factors (HIF), and consequent angiogenesis and enhanced adherence to extracellular matrix components. These are the VHL and succinate dehydrogenase (SDH) genes, which in recent years have found themselves surprising bedfellows in pheochromocytoma tumorigenesis – explained by the central role of prolyl hydroxylase as the molecular O₂ sensor. The other group of genes which, as it were, Dahia's group now put in the same bed are RET, NF1, and the recently implicated pheochromocytoma susceptibility gene, KF1Bβ. These genes are considered to activate apoptosis through tyrosine kinase activation, for instance of the prolyl hydroxylase called Egln3 in neuronal precursor cells.

Qin's article also discusses the evidence that mutations in the genes required for oxygen sensing lead to a shift in tumour cells from oxidative metabolism to glycolysis, opening the door to novel therapeutic possibilities. A possible difference between the two groups of genes may be that whereas a classical two-hit mechanism (genomic+somatic mutation) may be required for tumour development in the VHL and SDH patients, haploinsufficiency in the tyrosine kinase activators is tumorigenic without a second, somatic hit.

Dahia's work lies at the crossroads of pathogenesis and genetic diagnosis, and will hopefully contribute both to new therapies and to improvements in diagnosis and prognosis. The main reason at present for genetic diagnosis is the obvious one of earlier screening and detection in affected family members. But the patients themselves increasingly benefit, partly through recognition of other tumours in the syndrome – for example, kidney or CNS tumours in VHL, and paragangliomas in SDH. Strikingly, the recognition of SDHB mutations is now robustly associated with a high, 60%, malignancy rate, whereas SDHD mutations are associated with multiple benign tumours. This is one of the main conclusions of the series described by Cascón et al., who also repeats the observation that up to 25% of apparently sporadic pheochromocytomas have a germ-line mutation when sought systematically [2]. Cascón's recommendations as to which patients should now be screened for mutations largely chime with the conclusions of the symposium during the meeting, chaired by Chatterjee and Mannelli. In Cascón's series only 2% of patients aged >45 were found to have a germ-line syndrome. The recommendation at present can be that, in the absence of any clear family history, age should be a strong determinant of threshold for screening, with <45 or <50 being a reasonable threshold. Resources can be stretched by selective screening, for instance omitting RET screening in patients aged >30, or with extra-adrenal tumours, or by concentrating on SDHB if family history is reliably negative. This emphasis on SDHB is justified partly because of the malignant potential, and partly because SDHB mutations are turning out to be less penetrant than other mutations – and therefore less likely to have previously manifested.

Both in the selected articles, and in the symposium on biochemical screening chaired by Peaston and Singh, there was strong re-inforcement of the 2005 recommendations to switch from catecholamine or vanillylmandelic acid (VMA) measurements to either or both of plasma and urine fractionated (free) metanephrines. The reason for the superiority of metanephrines is that pheochromocytoma tissue contains large amounts of catechol-O-methyltransferase (COMT), whereas sympathetic nerve endings contain mainly monoamine oxidase (MAO) but no (COMT) enzyme. Metanephrines are derived solely by methylation of norepinphrine and epinephrine, whereas VMA requires the action of both MAO and COMT. It has previously been shown that pheochromocytoma patients have relatively low ratios of deaminated norepinephrine (dihydroxyphenylglycol, DHPG, the product of MAO) to free norepinephrine, whereas the reverse is true of patients with increased noradrenergic nervous activity [3]. The latter patients – those who have symptoms of pheochromocytoma but have no tumour – may have high plasma norepinephrine levels, and elevated ratio of DHPG to norepinephrine. But importantly, the newer assays show them to have normal or even low metanephrine levels. By contrast, even a slight elevation of plasma or urine metanephrines is highly suggestive of pheochromocytoma, ignored at one's peril. Unger shows in his review that it is possible to achieve >90% sensitivity and specificity with normetanephrine measurements, especially in plasma [4]. Predictably, metanephrine levels – meaning the specific methylated metabolite of epinephrine – are often normal in pheochromocytoma, either because the tumour is extra-adrenal, or because large adrenal tumours switch from epinephrine to norepinephrine synthesis as the portocapillary delivery of cortisol declines, and the phenylethanolamine-N-methyltransferase (PNMT) enzyme is no longer induced. In addition, but still unexplained, the VHL and SDH syndromes seem to arise in norepinephrine-secreting cells of the adrenal medulla. This observation surely contains an important clue to pathogenesis.

Despite the imperative of switching routine screening assays to measurement of metanephrines, there is work to be done on standardising assays, and agreeing comparability between the HPLC, ELISA, and mass spectrometric methodologies. These difficulties are not, however, an excuse for hospitals to delay in either establishing, or finding, a laboratory to undertake metanephrine measurements. Recent implementation of inter-laboratory quality assurance programs provides a means for hospitals and clinicians to select a reliable laboratory.

Several of the articles discuss advances in management – both diagnosis and treatment – of malignant pheochromocytoma. This condition is mercifully rare even in the practice of most specialists attending ISP2008. Nevertheless, even one patient every few years can make a distressing impact on the team and take as much time as the other 90% or so of patients seen during the same time. While not quite unique among endocrine tumours in being usually benign but occasionally malignant, the particular challenge for pheochromocytoma is that benign and malignant tumours look alike macroscopically and microscopically. To add to the puzzle – but also, probably, the clues – the malignancy risk is not spread evenly among all types of pheochromocytoma, being greatest among SDHB carriers, whilst rarely or never occurring in patients with VHL or RET mutations despite all other manifestations of these being malignant.

Carlsen reports an analysis of 25 years worth of archival tissue, in which the retinoblastoma tumour was found to distinguish tumours assigned a high malignancy score on the PASS system for grading morphology [5]. Such studies add to the jigsaw of information, and for the moment are more likely to contribute to hypotheses regarding pathogenesis than to clinical diagnosis – where ultimately putative markers will need to be tested prospectively against the clinical yardsticks of malignancy, namely recurrence and spread to sites where further primary tumours do not occur. Papewalis' study of chromogranin is unlikely to add to understanding of pathogenesis, but proposes an empirical new approach to therapy, based on the well-recognised co-secretion of chromogranin (and other peptide hormones) with biogenic amines [6]. The authors have investigated vaccination using chromogranin peptides in various mouse models of pheochromocytoma. Powers’ paper on RET expression in pheochromocytoma is at the other end of the spectrum, in the sense of exploring pathogenesis without immediate therapeutic dividend [7]. The authors acknowledge that any over-expression in tumour tissue can be a secondary epiphenomenon. Yet co-expression of RET with PNMT offers an intriguing clue as to why increased epinephrine secretion is almost always a striking feature of MEN-associated pheochromocytomas – in contrast to the low epinephrine levels in most other germ-line syndromes.

The papers by Santarpia [8], Adjallé [9], Druce [10], and Bravo [11] concentrate on therapies for malignant pheochromocytoma: the tyrosine kinase inhibitor, sunitinib; the VEGF inhibitor, thalidomide; radiolabelled somatostatin analogues – for example, DOTATOC; the mammalian target of rapamycin (mTOR) inhibitor, everolimus; and the nucleoside incorporation inhibitor, temozolomide. These papers review the newer therapies alongside the role of older therapies, including 131I-MIBG, and ‘standard’ chemotherapy comprising cyclophosphamide, vincristine and dacarbazine (‘CVD’). The provisional conclusion is that most experience remains anecdotal or comes from very early-phase trials. None of the therapies look likely to be more than palliative. Effective treatment for pheochromocytoma is more likely to stem from the clues arising from its unique biology than from the stuttering attempts to mount prospective randomised controlled trials of known chemotherapeutic agents. The low penetrance of SDHB associated pheochromocytoma points to a likelihood that malignant pheochromocytoma – like most cancers – is a consequence of multiple molecular events, and the corollary that treatment too will need to be multi-targeted in order to be effective. The more optimistic view would be that pheochromocytoma provides one of the best experiments of nature for understanding how changes in oxygen sensing lead to a failure of apoptosis, and that such patients will one day be the beneficiaries of appropriate translational research.

But the subject of pheochromocytoma has been translational since long before the word was invented. Each patient is a walking textbook of how to take the bench to the bedside. No two patients are identical, but in most instances it is possible to explain the constellation of clinical manifestations through an understanding of catecholamine biochemistry and adrenoreceptor pharmacology. While ISP2008 emphasised the growing need for collaboration among centres in order to harness the power of modern technologies in studies of adequate power, we were still able to enjoy the individual clinical anecdotes. The doyen of these, and master of the protean manifestations of pheochromocytoma, has been William Manger. It is fitting that both the audience at ISP2008 and readers of this volume have been able to peer with him into his unique store of memories [12]. Fore-warned is fore-armed, and no amount of 21st century science will save the patient if the general physician fails to entertain the possibility that his once-in-a-lifetime diagnosis has arrived.

References

• 1 Qin Y, Buddavarapu K, Dahia PL. Pheochromocytomas: from genetic diversity to new paradigms.  Horm Metab Res. 2009;  41 664-671
• 2 Cascón A, López-Jiménez E, Landa I, Leskelä S, Leandro-García LJ, Maliszewska A, Letón R, Vega LD, García-Barcina MJ, Sanabria C, Alvarez-Escolá C, Rodríguez-Antona C, Robledo M. Rationalization of genetic testing in patients with apparently sporadic pheochromocytoma/paraganglioma.  Horm Metab Res. 2009;  41 672-675
• 3 Brown MJ. Simultaneous assay of noradrenaline and its deaminated metabolite, dihydroxyphenylglycol, in plasma: a simplified approach to the exclusion of pheochromocytoma in patients with borderline elevation of plasma noradrenaline concentration.  Eur J Clin Invest. 1984;  14 67-72
• 4 Unger N, Deutschbein T, Walz MK, Mann K, Petersenn S. The value of immunoassays for metanephrines in the biochemical diagnosis of pheochromocytomas.  Horm Metab Res. 2009;  41 676-679
• 5 Carlsen E, Abdullah Z, Kazmi SM, Kousparos G. Pheochromocyto-mas, PASS, and immunohistochemistry.  Horm Metab Res. 2009;  41 715-719
• 6 Papewalis C, Kouatchoua C, Wuttke M, Jacobs B, Scherbaum WA, Schott M. Chromogranin A as potential tumour antigen in pheochromocytoma.  Horm Metab Res. 2009;  41 707-709
• 7 Powers JF, Picard KL, Tischler AS. RET expression and neuron-like differentiation of pheochromocytoma and normal chromaffin cells.  Horm Metab Res. 2009;  41 710-714
• 8 Santarpia L, Habra MA, Jiménez C. Malignant pheochromocytomas and paragangliomas: molecular signaling pathways and emerging therapies.  Horm Metab Res. 2009;  41 680-686
• 9 Adjallé R, Plouin PF, Pacak K, Lehnert H. Treatment of malignant pheochromocytoma.  Horm Metab Res. 2009;  41 687-696
• 10 Druce MR, Kaltsas GA, Fraenkel M, Gross DJ, Grossman AB. Novel and evolving therapies in the treatment of malignant pheochromocytoma: experience with the mTOR Inhibitor everolimus (RAD001).  Horm Metab Res. 2009;  41 697-702
• 11 Bravo EL, Kalmadi SR, Gill I. Clinical utility of temozolomide in the treatment of malignant paraganglioma: a preliminary report.  Horm Metab Res. 2009;  41 703-706
• 12 Manger WM. The protean manifestations of pheochromocytoma.  Horm Metab Res. 2009;  41 658-663

Correspondence