This retrospective study involved a total of 110 individuals diagnosed with PPGLs who were assessed at UHN’s cancer genetics program between 2011 and 2023. All individuals consented and underwent standard-of-care gene panel testing. Complete information on the genes tested for each individual is outlined in Supplementary Table 1.
Figure 1 summarizes germline and tumor IHC findings in the study population. WGS was performed for seven individuals with tumor IHC SDH deficiency and negative/inconclusive results from cancer panel germline genetic testing.
Fig. 1
Germline and tumor immunohistochemistry findings in the study population. Legend - CHCS: comprehensive hereditary cancer syndromes; VUS: variant of unknown significance; IHC: immunohistochemistry; WGS: whole genome sequencing
Solved CasesAmong these 110 individuals, 31 (28.18%) had a P/LP variant in a known PPGL predisposition gene (including SDHA, SDHB, SDHC, SDHD, RET, VHL, and NF1) or other genes in their initial reports. Six VUS findings (including five known PPGL predisposition genes and one incidental BARD1 finding) were also reclassified as P/LP variants (Fig. 2). Of the remaining individuals, 17 (15.45%) had a VUS in a PPGL-associated gene or other genes, and 62 (56.36%) had negative results, suggesting their disease may be caused by P/LP variants in non-traditional PPGL genes, variants undetectable by current technologies or panel-based approaches, or may reflect a non-hereditary etiology [21, 22].
Fig. 2
Frequency of likely pathogenic / pathogenic variants in PPGL genes in the study population
The status of expression for SDHB protein in the tumor was assessed by IHC in our population of 110 individuals. SDHB protein expression was lost in 28 individuals’ tumors (25.5%), intact in 52 (47.3%), and unknown in 30 (27.3%) (Supplementary Fig. 1). Interestingly, eight individuals with loss of SDHB expression had negative genetic testing results, and two had a VUS (one in SDHC and the other in RET) (Supplementary Table 1).
Self-reported ethnicity information was accessible for 101/110 participants (Supplementary Fig. 2). White self-reported ethnicity was reported in 49.5% of cases, East Indian in 14.9%, Asian in 9.9%, Black or African in 9.9%, Ashkenazi Jewish in 7.9%, Middle Eastern in 5.0%, Native Hawaiian or other Pacific Islander in 2.0%, and Hispanic or Latino in 1.0%.
Of the 17 variants initially reported as VUS, seven (16.28%, 7/43) occurred in the non-European group, including three individuals of Asian self-reported ethnicity, three of East Indian self-reported ethnicity, and one of Black or African self-reported ethnicity. The germline variants of two non-European individuals (one Asian, one East Indian) initially reported as VUS were later reclassified as pathogenic/likely pathogenic on ClinVar [26], representing a VUS-to-P/LP reclassification rate of 28.57% (2/7) in the non-European group. This contrasts with the European group (n = 58) in whom 10 VUS (17.24%) were initially identified and four (40.0%, 4/10) were subsequently reclassified as P/LP on ClinVar [26](Supplementary Table 1). Of eight SDH deficient tumors in germline-negative individuals, four were identified in the non-European group and (50%) and three in the European group. Self-reported ethnicity was not available for one individual in this group. Conversely, of five individuals whose tumors had normal (intact) SDHB expression and harbored germline P/LP variants in PPGL predisposition genes, one (20%) was of non-European self-reported ethnicity (East Indian). Figure 3 illustrates the distribution of self-reported ethnicities across germline finding categories.
Fig. 3
Distribution of different self-reported ethnicities within each category of germline findings
In this study, several cases diverged from typical genotype–phenotype associations (Table 1). Although SDHD pathogenic variants are more commonly associated with multiple head and neck paragangliomas [27, 28], an individual of Native Hawaiian or other Pacific Islander self-reported ethnicity with an abdominal pheochromocytoma harbored a pathogenic germline variant in SDHD c.3G > C p.Met1? (M1?). RET pathogenic variants are usually associated with adrenal pheochromocytomas, bilateral in over half of cases [29], yet a germline pathogenic RET c.2372 A > T (p.Tyr791Phe) variant was identified in an individual of European self-reported ethnicity with a cardiac paraganglioma, alongside a pathogenic variant in SDHB c.287G > A (p.Gly96Asp) (usually associated with abdominal paragangliomas). VHL pathogenic variants are typically associated with bilateral adrenal pheochromocytomas [30], yet a germline pathogenic VHL c.461 C > G (p.Pro154Arg) variant was detected in an individual of East Indian self-reported ethnicity with a para-pancreatic paraganglioma. No statistically significant associations were found between self-reported ethnicity and tumor SDHB status or metastatic disease.
Incidental FindingsExpanded hereditary cancer panels and complementary genomic analysis additionally identified pathogenic or likely pathogenic variants in TSC1, PALB2, BARD1, and APC. These were regarded as incidental findings, ascertained outside targeted PPGL panel testing, and the contribution of these genes to PPGL pathogenesis remains unknown.
An individual of Asian self-reported ethnicity with SDH-deficient left adrenal pheochromocytoma and bilateral glomus jugular paragangliomas had a BARD1 exon 1–6 duplication, initially classified as a VUS and subsequently reclassified as LP on ClinVar [26]. This was detected through an 84-gene comprehensive cancer panel, which was offered due to patient‘s complex history and uninformative PPGL panel NGS testing results. Similarly, an individual of Ashkenazi Jewish self-reported ethnicity with an SDHB-intact adrenal pheochromocytoma and negative 12-gene PPGL panel was found to have a LP APC variant(c.3902T > A; p.Ile1307Lys) through a 76-gene comprehensive cancer panel, offered in light of his complex personal and family history three years later. Another incidental LP germline variant was found in an individual of European ethnicity diagnosed with an SDH-deficient carotid body paraganglioma whose initial NGS panel testing was uninformative. This was a novel heterozygous frameshift variant in TSC1 (NM_000368.5: c.286_290del, p.Val96Lysfs*9) discovered through WGS (Delly v0.8.1 or Manta v1.6.0). This variant was absent from control population databases and not previously reported in the literature. TSC1 is associated with autosomal dominant Tuberous sclerosis complex (TSC), a multisystem disorder characterized by hamartomas in multiple organ systems [33]. This patient also harbored a heterozygous exon 12 deletion in PALB2 [31]. While no direct association between PALB2 and PPGL pathogenesis has been established, this deletion represents a clinically actionable secondary finding with implications for cancer surveillance and cascade genetic testing in biological relatives [32]. Notably, the PALB2 deletion was not identified by either automated SV callers employed in the WGS pipeline (Delly v0.8.1 or Manta v1.6.0), suggesting a limit of detection for small exon level deletions within standard WGS-based workflows [33].
Candidate FindingsAmong those with SDH deficient tumor tissues and negative/inconclusive results from standard-of-care panel genetic testing, seven individuals consented to germline WGS (Table 2). Beyond clinically relevant genes, we found six P/LP variants in six individuals in genes that are usually not part of standard germline PPGL panels, including DBH, HBB, C1QA, GJB2, and XDH.
Table 2 Clinical and genetic testing details of individuals who consented for whole genome sequencing (WGS)Analysis of SDHx genes by WGS revealed a complex SV involving 33,740 bp in the SDHA gene in one individual, which was confirmed using OGM (Fig. 4), but no additional variants were identified.
Fig. 4
Structural delineation of the SDHA inverted duplication in individual INS-022. A) Bionano VIA (v7.1) visualization of chromosome 5p15.33 region. Blue bars indicate the inversion, insertion, and duplication breakpoints. Genes in the region are shown above and indicate that the structural variant spans the 3’ end of PLEKHG4B and the 5’ end of SDHA. B) The inverted duplication is shown in the bionano genome brower view. The alignment of the individual sample’s consensus map (blue bar) with the hg38 human reference sequence of chromosome 5 (green bar) illustrates the inverted tandem duplication. The dashed arrows mark the duplicated and inverted segment. The dark green bar denotes the inversion region and the genes involved
Individual INS-027 demonstrated a rare heterozygous LP variant in the DBH gene (NM_000787.4: c.339 + 2T > C, rs74853476); this splice-site variant is present in population databases at a frequency of 0.1% (gnomAD) and has been previously submitted to ClinVar as pathogenic (Variation ID: 1750), although no submission was associated with PPGL. This variant is associated with autosomal recessive Orthostatic Hypotension 1, and characterized by profound autonomic failure, ptosis, nasal stuffiness, impaired ejaculation, and a neonatal history of delayed eye opening. As expected, this phenotype was not reported in the study individual, given the autosomal recessive inheritance pattern and only one heterozygous variant identified. A heterozygous nonsense variant in C1QA (NM_015991.4: c.622 C > T, p.Gln208*) was observed in related individuals, INS-021 and INS-021-F. This variant is associated with autosomal recessive C1q deficiency but no link with PPGL has been established, suggesting familial segregation of an immunological risk allele without relevance to PPGL pathogenesis. The remaining rare P/LP variants identified in HBB, GJB2 and XDH genes had no previously reported association with PPGLs or associated tumors at the time of this study.
Comments (0)