Portal de Periódicos da CAPES

Screening children at risk of developing inherited endocrine neoplasia syndromes

2000; Wiley; Volume: 52; Issue: 2 Linguagem: Inglês

10.1046/j.1365-2265.2000.00956.x

1365-2265

J R Johnston, Chew, Peter Trainer, Reznek, Grossman, Besser, Monson, Savage,

Lung Cancer Research Studies

The familial endocrine neoplasia syndromes, multiple endocrine neoplasia (MEN) types 1 and 2 and von Hippel Lindau disease (VHL), are autosomal dominant disorders which most commonly present clinically in early adulthood and onwards. They can, however, cause significant disease in childhood. The discovery of the causative genes for these disorders now allows the identification, by genetic mutation analysis, of most individuals at risk, at least in principle. Subjects found to be carriers of a mutation are at risk of developing the disorder and can be clinically monitored and treated in childhood so that appropriate targeted management may prevent the devastating complications which may develop, e.g. haemorrhage from retinal angiomas, sudden death from phaeochromocytoma and the development of medullary thyroid carcinoma from pre-existing C cell hyperplasia. Furthermore, genetic screening allows the children from affected families who have not inherited the mutation to be reassured and avoid regular clinical monitoring. Genetic testing raises many issues over informed consent, counselling and confidentiality; these are reviewed elsewhere (Reilly et al., 1997). Epidemiological and clinical information on the penetrance of familial endocrine neoplasia in childhood is derived mainly from clinical case reports of children presenting with manifestations of these disorders and not from screening data. This paper will review the screening and clinical management strategies for MEN1, MEN2 and VHL. We discuss, using illustrative case reports from our own experience, the role and nature of DNA testing, and review the relevant investigations and their timing in children at risk of endocrine neoplasia syndromes. MEN1 is a rare autosomal dominant syndrome affecting the parathyroid glands (usually four gland hyperplasia), in association with adenomas of the pituitary gland and benign or malignant tumours of the pancreatic islets. Less commonly adreno-cortical adenomas, thymic and bronchial carcinoid tumours, lipomas and thyroid adenomas may occur (Trump et al., 1996). There is a high degree of penetrance; over 80% of cases manifest the disease by the fifth decade, 43% show evidence of the disorder by 20 years of age on biochemical screening, while the youngest cases have presented before the age of 10 years (Trump et al., 1996; Bassett et al., 1998). Previously, gastrointestinal haemorrhage secondary to gastrinoma was a leading cause of death in MEN1 patients, but with earlier detection of pancreatic tumours by screening these patients suffer fewer complications and the prognosis of MEN1 relates more to specific organ involvement, malignant transformation of tumours and to hormonal hypersecretion (Teh, 1998). The gene for MEN1 is located on the long arm of chromosome 11 and has recently been cloned and reported with pathogenic mutations (Chandrasekharappa et al., 1997). The gene product, menin, is a protein of unknown function which has no similarity to previously known proteins. Recent in vivo molecular studies have shown that menin functions principally as a nuclear protein and interacts with the AP1 transcription factor JunD to repress transcriptional activation (Agarwal et al., 1999; Huang et al., 1999). Many of the tumours associated with MEN1 have loss of heterozygosity which suggests that the MEN1 gene is a tumour suppressor gene (Chandrasekharappa et al., 1997). More than 70 mutations of the menin gene have now been identified across the 9 coding exons. There is no evidence at present for any genotype:phenotype correlation as the phenotype is highly variable, both between and within families with respect to age of onset, clinical presentation and natural history (Hoff & Gagel, 1997; Teh, 1998). Affected individuals can be characterized genetically, and children of an MEN1 patient should be screened to determine if they carry the same mutation as their affected parent. Individuals identified to be carriers of a mutation and therefore at risk of developing MEN1 should be monitored clinically on an annual basis, as should all children whose parents have MEN1 which is not related to an identified mutation of the menin gene. Screening should be started between the age of 10 and 15 years of age, as a considerable volume of blood is required to perform the large number of serum and plasma analyses. Screening need only be performed at a younger age in particular families with a history of early onset of disease (Table 1). Hyperparathyroidism occurs in 90–97% cases of MEN1 and is usually polyglandular, with four gland hyperplasia (Trump et al., 1996; Hoff & Gagel, 1997). The mean age of presentation is 19 years of age, but it can occur before the age of 10 (Hoff & Gagel, 1997). Annual serum calcium measurements should then be performed and serum PTH levels should be measured if the calcium is elevated. The timing and type of surgery is controversial and curative treatment is difficult because residual parathyroid tissue has a tendency to regrow. Some surgeons argue that total parathyroidectomy with autotransplantation can provide long-term relief of hypercalcaemia with a decreased risk of hypocalcaemia (Mallette, 1994). Other centres have a more reluctant approach to surgery, recommending subtotal parathyroidectomy when the patient becomes symptomatic, the serum calcium is greater than 2.8 mmol/l, or there is evidence of bone loss or nephrolithiasis (Hoff & Gagel, 1997). Postoperative hypocalcaemia can occur, but is readily managed with vitamin D analogues. Pancreatic islet cell tumours, which are usually multifocal, occur in 75–81% of cases and biochemical abnormalities are frequently found in adolescence (Skogseid et al., 1994, 1995). These tumours present at a younger age compared to sporadic pancreatic tumours (mean 28 years v 45 years), are multifocal, and have malignant potential with reports of 30–50% of patients developing hepatic metastases (Trump et al., 1996; Hoff & Gagel, 1997). Pancreatic polypeptide, insulin and proinsulin are typically secreted early in the natural history of these tumours with a predominance of gastrin secretion at later stages (Mutch et al., 1997; Hoff & Gagel, 1997). Children at risk of MEN1 should be screened with annual plasma measurement of pancreatic polypeptide and gastrin, and pancreatic imaging by transabdominal ultrasound, and every three years with fat-suppressed T1-weighted contrast-enhanced MRI (Chanson et al., 1997; Mergo et al., 1997). One potential cause of difficulty in interpreting results is coexistent medication which may artefactually elevate serum gastrin, also, a normal gastrin in the upper normal range does not exclude a gastrinoma because serum gastrin may hover either side of the normal range. Surgical treatment is controversial and must balance the risks of secondary diabetes mellitus and pancreatic exocrine failure with control of the tumour mass (Hoff & Gagel, 1997). Some centres have a reluctant approach to surgery as the tumours are multifocal, whereas others advocate a more aggressive approach recommending the removal of most of the pancreas although without causing diabetes mellitus or pancreatic exocrine insufficiency (Tagge et al., 1995). Pituitary tumours occur in 10–65% of MEN1 patients, with prolactinomas being the most common followed by somatotrophinomas which are much less common. These should be screened for in subjects at risk by annual measurements of serum prolactin and IGF-I in addition to pituitary MRI every five years. A five point serum growth hormone day curve and an oral glucose tolerance test, measuring suppression of serum GH, are indicated where serum IGF-I levels are elevated to confirm GH oversecretion (Trainer & Besser, 1995). Treatment of pituitary tumours in these patients is similar to that of sporadic tumours. Transsphenoidal surgery should be performed for somatotroph adenomas, nonsecretory tumours if they are enlarging and, if there is a poor response to medical therapy in prolactinomas. Any abnormal monitoring test result should be followed up closely in order to assess the extent of the tumour and its natural history. The surveillance guidelines recommended are somewhat arbitrary, but have been refined by clinical experience at this and other centres. The 16-year-old daughter of a male patient with gastrinoma and hyperparathyroidism was screened for MEN1 associated endocrinopathy. Clinical history and examination were normal and the only abnormality detected on biochemical screening was a slightly elevated corrected calcium (2.64 mmol/l). One year later she was screened again. At this time she reported headaches and examination revealed reduced visual acuity on the left (12/6 compared to 6/6 on the right) with left optic disc pallor. Visual fields were full to confrontation. Assessment of her pubertal status showed she had stage 3 breast development with minimal breast tissue, stage 5 pubic hair and had not experienced menarche. Serum prolactin was elevated to 75 000 mU/l and combined anterior pituitary function testing (ITT, LHRH, TRH) revealed isolated gonadotrophin deficiency. A CT scan demonstrated a pituitary macroadenoma with suprasellar and lateral extension. Treatment with bromocriptine was commenced and followed by radiotherapy when the tumour had shrunk into the pituitary fossa and the patient became intolerant of high doses of medical therapy. Twelve years later the patient has normal serum prolactin levels, remains gonadotrophin, GH and TSH deficient but has normal visual acuity. The patient remains asymptomatic from her hyperparathyroidism (maximum corrected serum calcium 2.72 mmol/l) and is therefore managed conservatively. Gut peptide measurements performed annually and abdominal imaging have all been normal to date. This case illustrates that pituitary tumours can occur at a young age in association with MEN1 and that screening allows prompt management, which in this case probably prevented permanent loss of visual function. There are three familial forms of MEN2, their common features being the occurrence of medullary thyroid carcinoma (MTC) (Chew & Eng, 1995). MEN2A is characterized by MTC in combination with the variable presence of phaeochromocytomas and parathyroid tumours. MEN2B consists of MTC, phaeochromocytoma, ganglioneuromatosis and a Marfanoid habitus. Familial MTC (FMTC) occurs without any additional endocrinopathy. These familial forms account for a quarter of the total incidence of MTC (MEN2A 17%, FMTC 5%, MEN2B 3%) (Raue et al., 1993; Raue, 1996). These autosomal dominant disorders result from gain of function mutations in the RET proto-oncogene which codes for a receptor tyrosine kinase thought to play a role in the development of the neural crest and its derivatives. In MEN2A 97% of kindreds have a mutation of one of the cysteine residues in the extracellular ligand binding domain, with 73–84% occurring at codon 634 (Mulligan et al., 1994; Tsai et al., 1994; Heshmati et al., 1997). These mutations allow dimerization of RET molecules in the absence of ligand which results in constitutive activation of the tyrosine kinase domain (Carlson et al., 1994; Asai et al., 1995; Santoro et al., 1995; Learoyd et al., 1997). Additional evidence that these mutations are pathogenic comes from the observations that they also exist in some sporadically occurring MTC and phaeochromocytoma tissue, and that in transgenic mice a mutation at codon 634 is associated with multifocal MTC (Learoyd et al., 1997). In FMTC, 77–86% of kindreds have extracellular domain cysteine mutations, predominantly at codon 618 (54%), although mutations can occur at the same codons as defects in MEN2A kindreds. The reason for the different phenotypes is not understood (Heshmati et al., 1997; Learoyd et al., 1997). Fifty percent of MEN2B cases are thought to be the result of spontaneous germline mutations. More than 90% of MEN2B patients have mutations at codon 918 of RET with substitution of a threonine for a methionine residue (Carlson et al., 1994; Eng et al., 1994; Hofstra et al., 1994; Rossel et al., 1995; Kahn et al., 1996). This gain of function mutation occurs in the intracellular tyrosine kinase domain of RET, and is believed to alter substrate specificity of this domain and therefore signalling and receptor regulation (Santoro et al., 1995). The penetrance of MEN varies between the three different syndromes with MEN2B being the most penetrant and FMTC the least. In MEN2A 50% of individuals with RET gene mutations will develop the disease by age 50 years and 70% by age 70 years (Learoyd et al., 1997). MTC associated with MEN2B can occur in the first year of life (Gagel et al., 1993). Genetic testing in children and first degree relatives of patients with MEN2B should therefore be performed within the first year of life and in children from MEN2A and FMTC kindreds by 5 years of age (Gagel et al., 1995). Regular clinical and biochemical surveillance should be performed in all children known to carry a RET mutation. In the future mutation specific management may be possible as certain genotypes, for example codon 634 mutations, are most often associated with the MEN2A phenotype. However, other genotypes, for example codon 768 and 804 mutations, have a more variable phenotype and in this situation mutation specific recommendations are difficult. All children from a MEN2 kindred in which the mutation has not yet been identified also require regular screening. In families where there is an index case of apparently sporadic MTC, genetic screening of the RET gene is indicated in the index case. If a mutation is detected it can be used to screen family members as outlined above. In families where no RET mutation is found in the index case there is a very low chance of familial disease and a decision to clinically screen the first degree relatives must be based on the clinical probability of familial disease and the likely risk-benefit ratio of screening. The clinical features of the index case which may give a higher probability of sporadic disease include unifocal MTC, absence of C-cell hyperplasia, no associated hyperparathyroidism or phaeochromocytoma and a later age of presentation (Raue et al., 1993). However, none of these features are absolute. Children from families with MEN2 have previously been screened for MTC using the pentagastrin stimulation test (0.5 μg/kg iv over 15 s, serum calcitonin measured at baseline, 2, 5 and 10 minutes), but the availability of accurate, cost-effective genetic screening has enabled earlier identification of affected children (Gagel et al., 1988). The pentagastrin test is less accurate and both false positive and false negative results have been reported (Tsai et al., 1994; Marsh et al., 1996). In a review of annual screening tests in children with MEN2, 50% of the abnormal pentagastrin test results were associated with C cell hyperplasia while 50% already had MTC when the test became abnormal (Gagel et al., 1993) (Table 2). In MEN2B metastatic medullary thyroid carcinoma has been detected shortly after birth and therefore prophylactic total thyroidectomy is recommended in all clinically or genetically identified cases at the earliest possible age (Gagel et al., 1995; Zedenius et al., 1995; Learoyd et al., 1997). In MEN2A and FMTC prophylactic thyroidectomy is recommended at 5 years of age, as C cell hyperplasia has been reported as young as age 3 years in this disorder (Pacini et al., 1995; Gagel et al., 1995; Gill et al., 1996; Skinner et al., 1996; Learoyd et al., 1997). It is important to screen for phaeochromocytoma before thyroid surgery is performed in order to prevent an intraoperative crisis. Comparison of children treated prophylactically with total thyroidectomy and those treated after an abnormal pentagastrin test confirm an increased cure rate (100% v 75% in MEN2A and 27% in MEN2B) and a decreased risk of metastatic disease and death after prophylactic surgery (Skinner et al., 1996). Genetic screening for RET mutations therefore provides a powerful method of detection, enabling thyroidectomy to be performed at an earlier stage of the disease, than with traditional biochemical screening. The pentagastrin test, however, remains useful both for screening families where the mutation is not yet known and postoperatively to assess whether all tumour tissue has been removed (Learoyd et al., 1997). Phaeochromocytoma occurs in around 50% MEN2A patients but the prevalence is very variable, affecting from 6 to 100% individuals in different kindreds (Gagel et al., 1993; Howe et al., 1993). It is commonly multifocal and is bilateral in at least 50% of those affected (Gagel et al., 1993). The average age at clinical presentation is reported to be 37 years and the earliest presentation in MEN2A is at 10 years of age (Gagel et al., 1988; Gagel et al., 1993). The aim of surveillance is to detect phaeochromocytomas as early as possible in order to prevent progression of the disease and the 10% associated risk of sudden death (Casanova et al., 1993). Twenty-four hour urine collections should be analysed for catecholamine concentrations annually from the age of 5 years. Imaging of the adrenals should also be performed as part of the surveillance protocol with ultrasound and MRI initially followed by three-yearly MRI. Ultrasound is of value in detecting larger adrenal masses and is therefore indicated as an initial test. If the ultrasound is negative, MRI should be performed as the adrenal lesions may be small. In MEN2B phaeochromocytoma occurs in more than 50% cases and can occur in early childhood, so urine testing should start as soon as practicable and imaging at 3 years of age. In these young children a plasma screening test would be useful. A recent study suggests that measurement of plasma free metanephrines is an extremely sensitive screening test for phaeochromocytoma (Eisenhofer et al., 1999). Plasma Chromogranin A has also been reported to be a sensitive screening test, but levels may be elevated by the presence of MTC, although this normally only occurs with advanced disease (Hsiao et al., 1990; Blind et al., 1992). Any abnormality in the results should be followed by 123I-MIBG scanning and MRI in order to confirm the presence of a phaeochromocytoma and to locate any extra-adrenal paragangliomata (Gagel et al., 1993). Somatostatin receptor scintigraphy offers an alternative to 123I-MIBG scanning and is preferred in some centres (Kennedy & Dluhy, 1997). Bilateral tumours are more common in syndromic phaeochromocytoma (> 50%) compared to isolated sporadic cases (around 10%) (Gagel et al., 1993). It is important to determine whether the tumours are bilateral or whether unilateral adrenalectomy, which would avoid the need for corticosteroid replacement, is indicated. Care must be taken not to miss small lesions and MRI alone may not be adequately sensitive (Chew, 1994). Further imaging with radiolabelled meta-iodo-benzylguanidine (123I-MIBG) or adrenal venous sampling, which is particularly good at detecting small contralateral lesions, may be necessary as they have increased sensitivity (see case report 4) (Chew, 1994). Venous catheterization with measurement of catecholamines and determination of noradrenaline to adrenaline in the adrenal vein allows accurate diagnosis and localization of both adrenal (unilateral and bilateral) and extra-adrenal phaeochromocytomas. Normally noradrenaline concentrations are equal to or less than adrenaline levels in venous samples, therefore a ratio of greater than one suggests the presence of a phaeochromocytoma (Newbould et al., 1991). Hyperparathyroidism occurs in 20–30% of patients with MEN2A but is extremely rare in MEN2B (Kraimps et al., 1996). In children, annual serum calcium measurements should be performed from age 10 years to screen for this complication; serum parathyroid hormone concentration needs to be measured if the calcium is elevated. It is important to note that hypercalcaemia can occur secondary to phaeochromocytoma, resolving after adrenalectomy in these cases, so urinary catecholamines should always be measured in MEN2 patients with hypercalcaemia (Gagel et al., 1993). The 8-year-old daughter of a male patient with MEN2A was invited for screening. Her father, who was known to have a mutation of RET codon 634, had received treatment for bilateral phaeochromocytomas and MTC. The medical history of the girl was otherwise unremarkable and examination was normal. Genetic screening confirmed she had inherited the RET mutation from her father so she was tested for MTC and phaeochromocytoma. Serum calcitonin levels during a pentagastrin stimulation test at 8 years of age rose from 0.05 μg/l at baseline to 0.18 μg/l at 2 minutes (normal < 0.08 μg/l). The following year her calcitonin levels were < 0.04 μg/l at baseline and 0.13 μg/l at 2 minutes and she was followed clinically for a further year. At the age of 10 years her calcitonin response had increased (0.04 μg/l at baseline to 0.32 μg/l at 2 minutes) so total thyroidectomy was recommended. Histology confirmed C cell hyperplasia in both lobes of the thyroid gland, but no evidence of MTC was found. Postoperatively, her serum calcitonin levels are undetectable (< 0.04 μg/l). All serum calcium and urinary catecholamine measurements, pre- and post-thyroidectomy, have been within normal limits. This case highlights the early age at which C cell hyperplasia of the thyroid occurs in MEN2A. This child did not develop MTC due to careful pentagastrin monitoring, but prophylactic thyroidectomy at the age of 5 years, had genetic testing been available then, may have prevented the development of the premalignant C cell hyperplasia. A 3-year-old girl was referred to the paediatric endocrine clinic for further assessment of her failure to thrive. She was born at 36 weeks gestation and antenatal ultrasound scans had been normal other than showing distended loops of large bowel. She failed to pass meconium and had an emergency laparotomy for bowel obstruction in the first week of life. A colostomy was fashioned and histology showed hyperganglionosis. At one year of age her colostomy was closed but she continued to suffer from severe constipation as a result of poor large bowel motility. An ileostomy was fashioned when she was 3-year-old after persistent problems with bowel obstruction. Further histology was examined and was consistent with MEN2B-associated hyperganglionosis. At her initial presentation to the paediatric endocrine clinic she was noted to have coarse facial features and short stature. She has subsequently developed mucosal neuromas and the full lips typical of MEN2B, but does not have the characteristic Marfanoid habitus. Genetic analysis confirmed she has a threonine to methionine substitution at codon 918 of the RET gene. Endocrine investigations were performed at age 3.5 years. She had normal serum calcium, normal plasma catecholamines (urine collections were not possible at that time due to incontinence following pelvic surgery) and normal adrenal glands on MRI. A pentagastrin test was abnormal with a baseline serum calcitonin of 0.1 μg/l increasing to 0.8 μg/l two minutes after the pentagastrin (normal < 0.08 μg/l). Thyroid ultrasound showed asymmetry but a technetium scan was normal. She proceeded to total thyroidectomy and histology confirmed the removal of a small medullary cell carcinoma of the thyroid. She has been monitored by urinary catecholamine measurements annually for phaeochromocytoma and to date has normal levels. This case demonstrates that MTC may develop at a very early age in association with MEN2B. Invasive neoplasia can thus be prevented by prophylactic surgery following mutation detection. The estimated birth incidence of VHL in the United Kingdom is around 1 in 40000 with an average age at presentation of 26 years and over 95% penetrance by 60 years of age (Maher et al., 1990, 1996; Maddock et al., 1996). The predominant features of this disorder are haemangioblastomas of the brain, spinal cord and retina, renal cysts which may develop into multiple clear cell carcinomas, and phaeochromocytomas. Minor features include pancreatic cysts, epididymal cystadenomata and hepatic and splenic cysts (Melmon & Rosen, 1964; Horton et al., 1976; Atuk et al., 1979; Hough et al., 1994). The VHL gene was cloned in 1993 and encodes two protein products, one from an internal initiation site (Iliopoulos et al., 1998), both of which reportedly function as tumour suppressors. These proteins appear to have several functions including down-regulation of the transcriptional elongation factor, elongin (Latif et al., 1993; Duan et al., 1995; Kibel et al., 1995), ordering of fibronectin matrices (Ohh et al., 1998) and the regulation of vascular endothelial growth factor (Iliopoulos et al., 1996). Around 20% of affected patients have germline deletions of the VHL gene while 47% have smaller mutations (Maher et al., 1996). Tumour development in patients with deletions or nonsense mutations of the VHL gene requires loss or inactivation of the remaining wildtype allele in the susceptible cells. The mutant protein produced by missense mutations appears to exert a dominant-negative effect and is associated with a higher incidence and earlier onset of phaeochromocytomas, cerebellar haemangioblastomas and renal cell carcinomas than seen in patients with deletions or nonsense mutations (Maher et al., 1996). Genetic mutation analysis of the index case is the key to identifying further members of the family who are at risk of VHL and is usually performed by a combination of Southern blotting, single stranded conformational polymorphism analysis (SSCP) and direct DNA sequencing (Maher et al., 1996). The timing of genetic testing of the children of the index case follows the reporting of disease onset in published series of VHL patients (see Table 3). We suggest this should be performed at 5 years of age to enable clinical surveillance of those children shown to carry a mutation. Retinal haemangioblastomas, which occur in around 60% of VHL cases, are usually peripheral, but can involve the optic disc and are best detected by direct and indirect fundoscopy (Horton et al., 1976; Maher et al., 1990). Patients should be referred to an ophthalmologist for annual screening with fundoscopy and fluorescein angiography to enable early identification of retinal lesions. Treatment of these lesions with laser photocoagulation can prevent the devastating loss of vision which may occur following haemorrhage (Choyke et al., 1995). Fifty-five percent of VHL cases have cerebellar haemangioblastomas and around 8–14% have spinal cord or medulla oblongata tumours (Horton et al., 1976; Maher et al., 1990). VHL-associated cerebellar haemangioblastomas generally occur at an earlier age than sporadic tumours and they may be cystic or solid and are often multiple (Neumann et al., 1989; Neumann et al., 1992; Choyke et al., 1995). Gadolinium enhanced, T1-weighted MRI is the best method of imaging the brain and spinal cord in VHL and should be performed by 10 years of age and repeated annually (Choyke et al., 1995). If a lesion is identified more frequent imaging should be performed to assess tumour progression. Treatment of these lesions, which should be guided by the clinical state of the patient and by the size of the tumour, is by surgery (often needed for predominantly cystic lesions) and/or radiotherapy (generally effective for more solid or vascular lesions). Phaeochromocytoma is less common, occurring in around 10% of VHL cases. It may present during the first decade of life and can be life threatening if not detected and treated early (Horton et al., 1976; Maher et al., 1990). The phaeochromocytomas mostly arise from the adrenal glands, but can be extra-adrenal and are frequently bilateral (Maher et al., 1990; Chew et al., 1995). The most sensitive method of detecting these lesions is by analysis of catecholamines in 24 h urine collections (Ross et al., 1993). This should be performed annually from 5 years of age in addition to initial ultrasonography and triennial MRI imaging of the adrenals with both T1 and T2-weighted sequences. A rise in catecholamine levels to above the 95% confidence limits for age should act as a warning to repeat sampling and imaging, and a rise to above the 99% confidence limits makes the diagnosis of phaeochromocytoma highly probable (Ross et al., 1993). As discussed earlier, it is important to investigate the possibility of bilateral phaeochromocytomas by venous sampling before surgery in a case with an apparent unilateral phaeochromocytoma on MRI and123I-MIBG imaging, Renal carcinoma is now the leading cause of death in VHL as patients are being successfully treated for CNS haemangioblastomas (Choyke et al., 1995). Renal cysts may occur and be detected by renal ultrasonography and MRI before 4 years of age. These cysts are commonly multifocal and their natural history is variable, with most simple lesions growing slowly (Choyke et al., 1992). Cysts, which may have all the ultrasound and radiographic characteristics of benignity when occurring outside this syndrome, must be viewed with suspicion in VHL and followed closely. They may gradually develop thicker walls and more solid elements. Cysts with a solid component or solid lesions herald the development of carcinoma; they may enlarge more rapidly with a doubling time in one study of 10 months (Choyke et al., 1992). Solid renal lesions are not usually seen in children under 16, but it is important to start screening early in order to follow the any change in size of cysts as the first reported clinical presentation of renal cell carcinoma was at 16 years of age (Choyke et al., 1992). The recommended frequency of VHL screening, based on the current literature, is outlined in Table 4. If any abnormality is detected then detailed investigation is indicated with imaging repeated at 6 monthly intervals to monitor the progression of the lesion. A 13-year-old boy was referred following the diagnosis of VHL in his father who had presented with phaeochromocytoma and renal carcinoma. The past medical history was unremarkable and clinical examination was normal; blood pressure was 110/50. Genetic testing confirmed that he had inherited the VHL gene mutation (C712T) from his father. The patient was referred to an ophthalmologist who found no retinal angiomas on fundoscopy. An MRI scan of the brain and renal CT and MRI were both normal. Two 24 h urine collections and blood were taken for catecholamine measurement. The urinary noradrenaline was raised on both occasions (2570 and 3000 nmol/24 h) above the 99% confidence limits (814 nmol/24 h). The plasma noradrenaline was 9.9 nmol/l, which is also above the 99% confidence limit (< 5.67 nmol/l). An abdominal MRI scan showed a 4.5 × 2.2 cm mass in the right adrenal gland with a normal left adrenal. 123I-MIBG imaging confirmed the focal area on the right side at both 18 and 42 h. At venous catheterization the noradrenaline to adrenaline ratios were greater than 1 on both sides (right 541, left 3), confirming the presence of bilateral phaeochromocytomas. The patient underwent bilateral adrenalectomy and histology confirmed the findings of the venous catheter with a 9-mm phaeochromocytoma in the left adrenal gland and a 50-mm tumour on the right. The patient is now on replacement therapy and undergoes annual screening as indicated in Table 4. This case demonstrates that phaeochromocytoma can occur at a young age and that it is important to detect bilateral tumours thereby saving the patient a second operation. Genetic testing has revolutionized the management of these inherited disorders. At-risk individuals can be now be identified genetically in early childhood in most cases. Clinical surveillance protocols, undertaken in childhood, offer the opportunity to detect and treat the manifestations of these disorders as early as possible and to prevent the devastating complications, some of which can be fatal. Equally, individuals who do not have the causative mutation may be identified at an early age and do not need to be subjected to invasive and expensive clinical monitoring. Although the majority of mutations can be identified, in some families the genetic defect remains elusive and all children of these patients should undergo regular clinical screening in childhood. The availability of nonionizing forms of imaging (MRI and ultrasound) make surveillance safer. Screening and further management of these inherited endocrine neoplasia syndromes should be undertaken by an integrated team consisting of clinical geneticists, paediatric and adult endocrinologists and endocrine surgeons. The results of screening children who are at risk of inherited endocrine neoplasia will provide much information about the risks and benefits of screening as well as the natural history of these different syndromes in childhood, and it is envisaged that with this information screening protocols can be further refined in the future.