Distinct Clinical Features of Paraganglioma Syndromes Associated With SDHB and SDHD Gene Mutations FREE

Pheochromocytoma and paraganglioma are tumors of the autonomic nervous system. Terminology in science and clinical practice is divergent. Herein, we use the term pheochromocytoma for location in the adrenal glands, extra-adrenal abdominal, and thoracic locations (eg, where nearly all tumors are endocrinologically active). In contrast, the term paraganglioma is only used for tumors in the head and neck area where most tumors are nonfunctioning. All these tumors have been described as sporadic and as hereditary entities.^1- 3 Estimated yearly incidence of pheochromocytomas and paragangliomas is about 1 in 300 000.⁴

Classic syndromes associated with pheochromocytomas are multiple endocrine neoplasia type 2 due to mutations of the RET gene, von Hippel-Lindau disease (VHL), and neurofibromatosis type 1.^2,5- 6 Recently, the paraganglioma syndromes (PGL) have attracted attention especially after identification of the succinate dehydrogenase subunit D (SDHD) gene as the susceptibility gene for PGL type 1 (PGL-1), and succinate dehydrogenase subunit B (SDHB) gene as the susceptibility gene for PGL type 4 (PGL-4).^7- 8 A considerable number of cases with germline mutations in 1 of these 2 genes has been reported in both population-based and referral-based series of cases presenting with pheochromocytomas, and from hospital referral–based and selected series of cases presenting with head and neck paragangliomas.^7- 14 In contrast, only 4 PGL type 3 (PGL-3) families have been identified with a germline mutation of the succinate dehydrogenase subunit C (SDHC) gene, whereas the susceptibility gene for PGL type 2 (PGL-2) remains unidentified.^15- 18

Our research on SDHB and SDHD gene mutations started with analyses of SDHD in 17 blood-tumor pairs, resulting in the first description of SDHD germline mutations in patients with pheochromocytoma.³ Subsequently, we extended our work to blood DNA of all available patients with pheochromocytoma and characterized germline mutations of the RET, VHL, SDHB, and SDHD genes in nonsyndromic pheochromocytomas.¹⁰ Although it is accepted that carriers of SDHB and SDHD mutations are at risk for tumors of the entire paraganglial system, systematic clinical investigations of mutation carriers have not been reported to date.

Our current endeavor was therefore to clinically characterize the diseases based on mutations of the SDHB and SDHD genes using a complex approach based on our updated Freiburg-Warsaw Pheochromocytoma Registry, which includes Germany and central Poland, and a newly founded German Head and Neck Paraganglioma Registry, which includes all of Germany. We examined a population-based series of registrants with pheochromocytomas and/or paragangliomas and their relatives for mutations in SDHB and SDHD and systematically clinically characterized all carriers found among index cases and their first-degree and second-degree relatives.

Our study used patients who were registered to 2 population-based registries (Figure 1). The updated Freiburg-Warsaw Pheochromocytoma Registry, as of May 15, 2004, comprised 487 patients with adrenal and abdominal or thoracic extra-adrenal pheochromocytomas. We systematically included patients who presented with symptomatic disease and histologically confirmed pheochromocytoma from Freiburg, Germany, and Warsaw, Poland, since 1985, from Essen, Germany, and Würzburg, Germany, since 1995, from Padova, Italy, since 1998, and German pediatric patients from 1979-1999, who came for diagnosis, treatment, or reevaluation, and who consented to participate in scientific research studies. All individuals presenting with symptomatic pheochromocytoma or paraganglioma in their respective geographic regions were registered. In addition, we included 27 patients from whom we received blood DNA throughout Germany and from clinicians abroad.

For this study, syndromic features and known family history were exclusion criteria. We excluded 153 patients with pheochromocytoma who had clinical, familial, and/or molecular genetic evidence for multiple endocrine neoplasia type 2, VHL disease, or neurofibromatosis type 1 based on extended personal history, pedigree evaluation, and genetic analyses of the RET and VHL genes. We also excluded patients with mutations of the SDHC gene because of the rarity of the condition. After exclusions, the study population consisted of 334 cases compared with the Freiburg-Warsaw pheochromocytoma registry¹⁰ of 228 cases as of December 1, 2001, which included 191 unrelated German, 113 Polish, and 30 index cases from other countries.

In January 2000, we founded a new German registry that ascertained the population base by head and neck paraganglioma presentations. This registry comprised 77 unrelated patients, as of May 15, 2004, and an additional 6 paraganglioma index cases from other countries. For this registry, 100 otorhinolaryngology departments throughout Germany participated. Patients who developed head and neck paragangliomas as well as adrenal and extra-adrenal pheochromocytomas entered the registry after the first symptomatic tumor presentation.

In total, 417 unrelated patients were studied from April 1, 2000, until May 15, 2004, for germline mutations of the SDHB and SDHD genes. Mutation analysis for the 8 exons and flanking intronic regions of SDHB and the 4 exons and flanking intronic regions of SDHD was performed as previously described using a combination of single-strand confirmation polymorphism and direct sequencing.^3,8 Missense mutations were regarded as pathogenetically relevant if not found in 300 control patients, whereas truncating mutations are predicted to be deleterious because of structure-function consequences. The 300 blood DNA controls came from white healthy blood donors from blood banks in Freiburg, Germany, and Warsaw, Poland, and from Switzerland (ie, region-matched, race-matched normal controls), as well as a small minority from Columbus, Ohio.

The research protocols were approved by the ethical committees of the University of Freiburg, the Institute of Cardiology, Warsaw, Poland, and the Human Subjects' Protection Committee, The Ohio State University, Columbus. All participants gave oral or written informed consent.

Identification of mutation carriers among eligible registrants was followed by genetic screening of the family members. Once a germline mutation was identified, we extended our work in 2 directions. First, the mutation carriers were reevaluated in depth. The clinical screening program included magnetic resonance imaging (MRI) of the neck and skull base, MRI or computed tomography (CT) of the thorax, and MRI or CT of the abdomen and 24-hour urine assays for norepinephrine, epinephrine, and vanillylmandelic acid. Second, we offered all first-degree relatives of mutation-positive index cases molecular genetic testing for the mutation identified in the index patient. When a germline mutation was detected in a relative of the index case, the same clinical screening procedure was used.

We used demographic data, including age, sex, as well as number, location, and benign or malignant status of the tumor specimens. Criterion for malignancy was only presence of distant metastases. Clinical screening results enabled us to calculate penetrance for the development of tumors (the percentage of mutation carriers who had developed a tumor). Only index cases and relatives who underwent clinical screening, or previous MRI and CT results that were positive, have been included for penetrance calculations. For calculation of the registry-based prevalence of SDHB and SDHD gene mutations in patients with pheochromocytoma, paraganglioma, or both, we only included all cases from Germany and Poland but excluded those from other countries to avoid any bias. Differences in clinical parameters among SDHB-associated and SDHD-associated pheochromocytomas were compared by 2-tailed Fisher exact test. Penetrances of SDHB-related and SDHD-related tumors were estimated by cumulative incidence functions, by the method of Kaplan-Meier but substituting patients' age for survival time. For comparison of age distributions and age-related penetrances, Wilcoxon signed rank test and Cox-Mantel test were used, respectively. P<.05 was regarded as significant. To avoid spurious positive results due to multicomparison, Bonferroni adjustment was applied to P values for differences in tumor location and malignancy. The software Mathematica version 5 (Wolfram Research Inc, Champaign, Ill) was used for all statistical analyses.

In the combined pheochromocytoma (n = 334) and paraganglioma (n = 83) registries, 49 patients (12%) showed a mutation in the SDHB or SDHD gene. A total of 25 index cases had 18 different germline mutations of the SDHB gene (Table 1) and 24 index cases had 15 different germline mutations in the SDHD gene (Table 2). No patients had more than 1 germline mutation. The prevalence of mutations for both genes was approximately 10% overall (5% SDHB and 5% SDHD) (Table 3). SDHB and SDHD gene mutation frequencies were similar among all registries: SDHD mutation frequencies in the Freiburg and Warsaw Pheochromocytoma Registries and the German Head and Neck Paraganglioma Registry were 4% for each registry; SDHB mutation frequencies were 6%, 4%, and 5%, respectively. The prevalence data did not include 36 study patients from countries outside Germany and Poland (30 with pheochromocytomas and 6 with head and neck paragangliomas) of whom 10 were shown to be carriers of SDHB or SDHD mutations (Table 1 and Table 2). In the SDHB gene, mutations occurred in exons 1 to 7 but not in exon 8, whereas in the SDHD gene, the mutations were distributed throughout all 4 exons but tended to cluster in exon 1. In both genes, missense, nonsense, frameshift, and splice site mutations were found. In addition, in the SDHB gene, a single codon insertion was noted. The spectra of mutations did not differ statistically between the 2 genes: 50% (9 of 18) of SDHB mutations compared with 27% (4 of 15) of SDHD mutations were missense (P = .16). Eight SDHB mutations and 5 SDHD mutations have not been described previously. Five SDHB and 11 SDHD mutation carriers had available consenting parents who could be tested for the presence of mutations. Among these, there were 2 confirmed cases with de novo mutations (defined as first cases in a family, both parents without mutations) in SDHD (c. 33 C/A), which did not occur on the same haplotype. Mean (SD) ages at diagnosis of disease of the index cases were 29.8 (15.2) years for SDHB mutation carriers and 30.6 (14.3) years for SDHD mutation carriers and thus not significantly different (P = .77).

All 161 eligible first-degree relatives of mutation-positive index cases were offered genetic testing. Of these, 103 proceeded with testing, 43 for SDHB and 60 for SDHD. Among these 103 relatives, 26 were newly identified as SDHB mutation carriers and 21 as SDHD mutation carriers. In addition, we included 4 further first-degree relatives with known neck paragangliomas for whom we did not have DNA or paraffin-embedded surgical specimens. Thus, the study comprised a total of 53 SDHB and 47 SDHD mutation carriers. In the parental generation, SDHD mutations were only found in fathers, consistent with known maternal imprinting. Maternal imprinting is caused by methylation and hence silencing of the maternal allele. Therefore, individuals who inherit a mutation from the mother would not manifest tumors.¹⁹

After comprehensive clinical investigation in mutation carriers, we detected nonfunctioning tumors of the neck in 7 carriers of an SDHD mutation, of the thorax in 1 SDHB and 1 SDHD carrier each, and of the abdominal paraganglia or adrenal gland in 3 SDHD carriers. No tumors at all have been found in 10 carriers of an SDHB mutation and 9 carriers of an SDHD mutation; all the latter were classified as maternally imprinted cases. The mean (SD) age of carriers at tumor diagnosis was not significantly different from those without tumors (SDHB: 31.3 [15.4] years vs 34.7 [15.8] years, P = .49; and SDHD: 32.4 [16.1] years vs 47.6 [25.3] years, P = .08).

Blood pressure and catecholamine levels at the time of initial pheochromocytoma diagnosis were available for 41 patients (n = 21 for SDHB and n = 20 for SDHD mutation carriers). All had at least paroxysmal hypertension. Levels of catecholamine excretion were available from 20 patients: norepinephrine level was increased in 19 patients and normal in 1 case; whereas, epinephrine level was increased in 7 and normal in 13 patients, respectively. None of the patients with head and neck paragangliomas in the absence of pheochromocytoma had elevated blood pressure or catecholamine excretion.

For purposes of comparing various tumor characteristics among SDHD and SDHB mutation carriers, we excluded 9 carriers of SDHD mutations from further calculations. This included 7 patients without evidence of tumors who likely (n = 6 fathers; mean [SD] age, 61.8 [7.3] years; range, 51-71) or definitely (n = 1 mother) have inherited the mutation from their mothers, and 2 children of mothers who have been recognized as mutation carriers by this study but will not develop tumors because of maternal imprinting of the SDHD gene.

A total of 55 mutation carriers, including 20 relatives, had complete clinical screening; in an additional 19 carriers, positive findings were already available for the neck or thorax, and/or abdominal area, or all 3. In total, information was available for head and neck tumors in 67 (including 20 index cases operated for symptomatic neck tumors), for thoracic tumors in 56 (including 5 index cases operated for symptomatic thoracic tumors), and for abdominal tumors in 68 (including 35 index cases operated for symptomatic abdominal tumors) mutation carriers.

Table 4 summarizes the percentages of tumor locations and characteristics in relation to the total number of patients with tumors. Relating these numbers to all mutation carriers who underwent adequate clinical workup for the single body regions, a very similar picture can be drawn. Multiple tumors occurred in 34 patients, 9 of 32 carriers with SDHB and 25 of 34 with SDHD mutations (P<.001). Fifty-three percent (18 of 34) of SDHD mutation carriers were found to have adrenal pheochromocytomas compared with 28% (9 of 32) of SDHB mutation carriers (P = .049). Interestingly, 79% (27 of 34) of all adequately investigated SDHD mutation-positive individuals had head and neck paragangliomas compared with only 31% (10 of 32) in SDHB mutation carriers (P<.001). For non-SDHB/SDHD mutation carriers with pheochromocytoma or paraganglioma, the studied features are all statistically significantly different (multiple tumors, adrenal location, head and neck paragangliomas, for all P<.001).

Eleven SDHB mutation–positive individuals (34% of tumor patients with an SDHB mutation) but no SDHD mutation carriers were found to have malignant pheochromocytomas or paragangliomas. All 11 individuals were found to have distant metastases, defined as tumors in the lungs or bones (Table 5). There was a difference in the prevalence of malignant disease between SDHB and SDHD mutation carriers (11 of 32 vs 0 of 34, P<.001). Malignant pheochromocytoma was present in 14 of 368 non-SDHB/SDHD mutation carriers (P<.001 for SDHB vs non-SDHB/SDHD mutation carriers, P = .28 for SDHD vs non-SDHB/SDHD mutation carriers). In addition, there were 3 SDHD mutation–positive patients who were initially misdiagnosed with malignant disease because of multiple abdominal tumors in 2 cases and abdominal and thoracic tumors in another case. Because of our clinical investigations, these patients were actually found to have multifocal benign tumors.

Tumors of the extraparaganglial system were observed in 5 carriers of SDHB mutations. Two carriers of the SDHB c. 847-50 del TCTC mutation, belonging to 1 family, were found to have clear cell renal carcinoma at ages 21 and 26 years. Tumor tissue (paraffin blocks) showed loss of the wild-type allele.²⁰ Two patients, 1 carrier of the SDHB c. 328 T/C mutation and 1 carrier of the SDHD c. 14 G/A mutation, showed a papillary thyroid carcinoma at ages 14 and 26 years, respectively.

Age-related penetrance based on symptomatic and asymptomatic tumors is shown for SDHB mutation carriers and for SDHD mutation carriers in Figure 2. Ten (24%) of 42 carriers of an SDHB mutation with adequate clinical investigation had no tumors, in contrast with SDHD mutation carriers, all of whom (34 of 34) were found to have tumors (P = .002). Fifty percent of SDHB mutation carriers were estimated to develop at least 1 tumor by 35 years (SDHB mutations have 50% penetrance by 35 years). Penetrance increases to 77% by 50 years. In comparison, SDHD mutations confer 50% penetrance by 31 years and 86% by 50 years. The overall age-related penetrance for SDHB and SDHD mutations was not statistically different (P = .67). Interestingly, there were significant differences in the age-related penetrance of tumor manifestations by site. Adrenal pheochromocytomas appeared more frequently and at an earlier age in SDHD mutation carriers (P = .03), and there was also a significant earlier onset for head and neck paraganglioma in SDHD mutation carriers (P = .007).

The paraganglioma syndromes have been relatively newly delineated as unique entities. Although paraganglioma has been clinically recognized for more than 40 years, only in the last 4 years have they been classified based on molecular genetics: SDHD mutations predispose to PGL-1, mutations in an unidentified gene on chromosome 11 to PGL-2, SDHC mutations to PGL-3, and SDHB mutations to PGL-4.²¹ Our population-based study of apparently sporadic symptomatic pheochromocytoma presentations revealed that approximately 25% of such individuals carry unsuspected germline mutations in 1 of 4 genes, including SDHB and SDHD.¹⁰ However, to date, carriers of these germline mutations were known to have a risk for paragangliomas, pheochromocytomas, or both but detailed clinical information, such as gene-specific clinical features, demographics, and penetrance, for purposes of genetic counseling, treatment, and follow-up were not known. We could not examine the gene for PGL-2, as it has yet to be identified, but we performed molecular genetic exclusion of SDHC mutation carriership (PGL-3); these mutations seem to be extremely rare.^15- 17,22

Our observations demonstrate that individuals carrying germline SDHB and SDHD mutations have some features in common. For example, mean age at diagnosis was 29 years for both genes. Prevalence of mutation carriers for each gene in the population-based registries of 2 countries was similar, between 4% and 6%. There were significant clinical differences between carriers of SDHB mutations compared with those with SDHD mutations.

The apparent age-related penetrance of tumors was not significantly different for SDHB and SDHD mutation carriers. Nevertheless, it is remarkable that 10 carriers of SDHB mutations did not develop a tumor, whereas all SDHD carriers, who were not likely to be subject to maternal imprinting, did have tumors. With a longer follow-up, it is quite probable that this apparent difference might also become statistically significant. Age-related penetrance for adrenal locations was higher among SDHD mutation carriers compared with penetrance for SDHB mutation carriers.

Head and neck paragangliomas were statistically more prevalent among SDHD carriers compared with those with SDHB mutations, although intra-abdominal extra-adrenal tumors were more prevalent among SDHB mutation carriers. Although most SDHD mutation carriers presented with multiple tumors (74%) compared with SDHB mutation carriers (28%), malignant tumors are more frequent in SDHB mutation–positive individuals (11 of 32 vs 0 of 34 in SDHD mutation carriers, P<.001). Similarly, in SDHB mutation carriers, a high rate of distant metastases (4 of 8 cases) has been reported recently by Giminez-Roqueplo et al²³ and no malignant pheochromocytoma or paraganglioma has yet been reported in an SDHD mutation carrier in the literature to date. Consistent with the apparently aggressive nature of SDHB dysfunction, 5 mutation carriers in our study were also found to have extraparaganglial malignancies (eg, renal cell carcinoma and thyroid papillary carcinoma). Kidney carcinomas are considered oncocytic tumors (replete with mitochondria) and thus, the involvement of a mitochondrial complex II gene in kidney carcinogenesis may be explained. The apparently more aggressive nature of the tumors in SDHB mutation carriers may be postulated to be a consequence of the prevention of assembly of the catalytic complex that normally comprises SDHA and SDHB, thus leaving only complexes of the structural SDHC and SDHD moieties.²⁴

Based on these observations, preliminary guidance for genetic counseling and surveillance is possible. Although our population-based study established an approximately 25% germline mutation frequency in apparently sporadic symptomatic presentations of pheochromocytoma, this may involve any 1 of 4 genes.¹⁰ Which gene(s) to begin testing is often a practical question for clinicians. Our present data suggest that individuals presenting with head and neck paragangliomas, multifocal tumors, or both should be targeted for SDHD testing in the first instance. Extra-adrenal abdominal presentations, malignant disease, renal cell carcinoma, and thyroid carcinoma may suggest SDHB testing initially. An older age of onset may suggest SDHB/SDHD-related PGL syndromes compared with VHL disease, the latter of which is usually characterized by early-onset pheochromocytoma.¹⁰

Conversely, if an individual was newly found to carry an SDHD mutation, there is a likelihood of developing head and neck paragangliomas and penetrance is relatively high in a lifetime. Thus, asymptomatic carriers should be offered 3 body region clinical screening, including MRI of the neck, thorax, and abdomen/pelvis; 18 fluorodopa or 18 fluorodopamine positron emission tomography might be an acceptable alternative.^25- 27 In addition, measurement of catecholamines, preferentially of plasma metanephrines, should be performed.^28- 29 Annual intervals may be considered, although rate of tumor growth is currently unknown. The only exception is perhaps the children of female SDHD mutation carriers who likely do not require clinical surveillance because the axiom of maternal imprinting of SDHD has never been reported to be violated in any case to date. Although SDHB mutation carriers should be subjected to annual clinical surveillance, patients should be counseled that while their disease might be less penetrant, multifocal disease, malignant disease, and early-onset renal cell carcinoma are possible.²⁰

Whether thyroid malignancies are also components of SDHB- or SDHD-related disease awaits further confirmation. This may be germane to surgical decision making, if our data can be independently replicated. The current standard of care in our consortium institutions for hereditary forms of intra-adrenal pheochromocytoma is to offer adrenal sparing tumor resection, typically endoscopic resections of the tumor(s) because of the possibility of multifocal metachronous disease, and morbid consequences of total adrenalectomies.^30- 33 The 2 major genes contributing to the heritable pheochromocytoma-paraganglioma syndromes, PGL-1 through PGL-4, are SDHB (PGL-4) and SDHD (PGL-1). SDHC (PGL-3) has only been found in 4 unrelated families^15- 18 and the fourth locus (PGL-2), without an isolated gene, is germane only to 1 family in the Netherlands. Our continuing studies have suggested that clinical features are based on gene type, and therefore, knowing which gene, SDHB or PGL-4 vs SDHD or PGL-1, would be important for not only diagnosis but further clinical management and genetic counseling.