SDH-deficient gastrointestinal stromal tumors (GIST) account for 20–40% of all KIT/PDGFRA-negative GIST (1). Several evidences have suggested that SDH-deficient GIST exclusively arise from the stomach with mainly multifocal primary localization, frequently present lymph node involvement, generally affect younger population, and above all, have an indolent behavior even with metastatic disease (2). SDH deficiency in GIST is defined by the loss of expression of SDHB protein at immunohistochemistry and is mainly due to mutations in the four SDH mitochondrial complex subunits: SDHA, SDHB, SDHC, and SDHD (3). SDHA mutations are the most frequent among SDH alterations in GIST, accounting approximately for half of the cases. Other rare epigenetic events involve the recurrent aberrant DNA methylation of SDHC seen in GIST associated to the Carney triad (CT), which is a rare condition with synchronous or metachronous occurrence of GIST, paragangliomas, and pulmonary chondromas (4–6).

Most of SDH mutations in GIST are germline, in particular germline mutations in SDHB, SDHC, and SDHD occur in about 20–30% of SDH-deficient disease and may be referred to as a Carney-Stratakis syndrome (CSS) (4). This syndrome was firstly described in 2002 as a hereditary condition characterized by the occurrence of GIST and paraganglioma. Germline SDHA pathogenic variants have been rarely described in apparently sporadic cases (7, 8). Currently, germline testing is recommended for all SDH-deficient GIST, but no clear guidelines for genetic counseling and follow-up of SDH mutation carriers and relatives have yet been released (9). Moreover, regarding SDHA mutations, many issues are still unclear such as the germline status and the clinical implications that are not yet linked to well-defined hereditary syndrome. The aim of this work is to evaluate the presence and the prevalence of SDHA variants in the germline lineage in an SDH-deficient GIST population harboring SDHA somatic mutations.

In a population of SDH-deficient GIST harboring SDHA somatic mutations, we found germline SDHA variants in all cases for which normal DNA was available, and these findings underline that germline mutations in SDHA are highly frequent in SDHA-deficient GIST. Besides the non-sense and missense mutations, we also found two cases harboring damaging splice-site mutations (c.457-2_457 and c.1663+3 G>C) leading to a predicted premature protein sequence truncation. Overall, in our series we found seven truncating germline variants (five non-sense and two splice-site) present in all but one case in young patients. Conversely, four out of seven germline missense variants, whose effect on protein function and stability is harder to be assessed, were identified in older patients (more than 50 years old at diagnosis). Moreover, we did not find a high frequency of the Arg31* variant, whose recurrent identification in previous studies even led to the assumption of a putative founder effect (3, 7, 14, 16–18). Apart from the recurrent Arg31*, the high diversity of the variant type and position seen in our and in previous series supports a view in which SDHA can be inactivated by mutations scattered throughout its whole genomic sequence, suggesting that SDHA genetic analysis must be performed entirely and by qualified high-volume molecular diagnostic centers, as already suggested in other settings (19).

Lastly, all the described patients carried two mutational events at the SDHA locus, either the loss of the wild-type allele or a second somatic event in compound heterozygosity, in full agreement with the two-hit hypothesis of tumor suppressor genes inactivation. Therefore, this suggests the possibility of a germline first mutation also in the two cases of our series for which the matched normal DNA counterpart was not available. Interestingly, an in-depth biochemical study of the functional effects of SDHA variants of unknown significance supported the pathogenicity of two somatic mutations (p.Arg451Cys e p.Gly419Arg) classified as uncertain by ACMG recommendations, while it did not support the loss of function of the Arg171His variant (20).

Furthermore, our study definitely confirms the highly relevant association between germline SDHA pathogenic variants and GIST onset, which is supported by many previous studies reporting the development of GIST as the only cancer disease in SDHA germline-variant carrier population (3, 7, 13, 14, 18). In our series, only one case reported having been affected by paraganglioma more than 10 years before, but the diagnosis was not now revised. Recently, a large case-control study has estimated the penetrance of SDHx variants for PPGL, indicating a 1.7% lifetime disease penetrance for SDHA pathogenic variants (21). Up to now, the onset of other tumor types in the context of SDHA-mutant GIST appears to be very rare and limited to the concurrent description of paraganglioma in one case, of pulmonary chondroma in two cases, and of the full Carney triad in one patient (8, 14). Indeed, it was shown that mutations in SDHA account for around 30% of all KIT/PDGFRA-WT GIST (10), while being responsible for <1% of all pheochromocytoma and paraganglioma cases (PPGL), as opposed for example to SDHB that accounts for about 10% of all PPGL (22, 23). The only study that up to now suggesting the possible role of SDHA as a predisposing factor for other tumor types reported the development of neuroblastoma in one SDHA-mutation carrier showing the inactivation of the second allele (17). Therefore, our result, coupled with previous data and with the identification of a higher-than-expected frequency of SDHA germline pathogenic variants in a cohort of cancer patients with respect to healthy individuals from the Exome Aggregation Consortium (ExAC) database (17), further endorses the view of SDHA as a cancer gene mainly predisposing to GIST development. This aggregate result underscores the different tumor spectrum of SDHA-germline variant carriers and Carney Stratakis Syndrome patients, which are known to harbor SDHB, SDHC, or SDHD mutations, and to develop invariably the association of GIST and paraganglioma during their life course (24).

Actually, in literature and in our experience, a clear syndrome has not been clearly defined since all these cases appear to be sporadic without other neoplasms except from GIST, and therefore it still remains unclear if SDHA mutations should be accounted in the CSS. CSS by definition is due to SDHB, SDHC, and SDHD mutations and is characterized by the combination of GIST and paraganglioma inherited in an apparently autosomal dominant manner and with incomplete penetrance. Paragangliomas were described to be multicentric and GIST multifocal. Both these clinical presentations support the inherited nature of this tumor predisposition (4). Currently, germline testing is recommended for all SDH-deficient GIST including SDHA in some clinical practice guidelines (9, 25), but no clear protocol for genetic counseling and follow-up of SDHx variant carriers and relatives has been released, especially for those carrying a germline SDHA pathogenic variant not yet linked to well-defined hereditary syndrome.

Regarding the clinical findings in GIST patients with SDHA germline mutation, our series confirms the stomach as unique site of GIST onset but underlines a higher heterogeneity for the multifocality development and also for the metastatic presentation at diagnosis. Larger series and international efforts should be required to better define a phenotype/genotype correlation in this subset of disease between GIST features, other tumor development, and SDH genotyping in tumor/germline. No conclusive consideration can be made regarding the survival rates because of the heterogeneity of clinical presentations and the multiple treatments received in cases with metastatic disease. After all, our series confirms the indolent clinical course and long survival expectations since half of population presents a follow-up higher than 10 years, and four of these patients present a metastatic disease.

This study therefore shows that, differently from previously stated, the SDHA-mutant GIST patient is almost exclusively a germline SDHA-variant carrier that is prone to develop the tumor throughout his entire life and not just in his early adulthood (7, 10–12, 14). In fact, 38% of the patients from our series were older than 50 years, with peaks up to 70 years old. Actually, this result is supported by previous studies that occasionally reported age ranges up to 70 years old and subsets of SDHA-mutant GIST patients of more than 50 years old (3, 7, 13, 14, 16). This fact has very important implications supporting the need for genetic counseling, since even though complete pedigree analysis of SDHA mutation carrier families is still lacking and the clinical insights suggest a low penetrance of SDHA mutation in affected families (12, 26), the notion of a lifelong risk for GIST development urges the need for genetic testing and permanent clinical surveillance of SDHA-variant carriers.

As the process of referral to genetic counseling of the patients included in the study is still ongoing, conclusive information on documented history of cancer in relatives, segregation of the variant in the family, and clinical assessment of carriers, who will be included in management programs according to updated recommendations, are not yet available. Now, this may represent a formal limitation of the study, however not modifying its substantial conclusions.

In conclusion, germline pathogenic variants in SDHA are highly frequent in SDHA-deficient GIST in young and adult patients, and the disease may occur also in older adulthood. Genetic counseling of SDHA-variant carriers and relatives should be planned, and their clinical follow-up should be accurately defined.

The datasets presented in this study was submitted to ClinVar (SUB10460221). Variants identified are available under accession numbers from SCV002026125 to SCV002026144.

The studies involving human participants were reviewed and approved by Ethical Committe of Pol. Sant’Orsola Malpighi, University of Bologna. The patients/participants provided their written informed consent to participate in this study.

Conceptualization: MP, MU, AAs, and VI. Data curation: VI, AAs, MN, and MSa. Formal Analysis; MU, VI, AS, AAs, and GR. Funding acquisition: MP and MN. Investigation: MU, VI, AS, AAs, BV, AB, GG, MC, EF, EC, MSa, AL, and GT. Methodology: MU, VI, AS, AAs, EG, AAl, and GR. Project administration: MP. Software: VI, GR, and MU. Supervision: MP, AAs, AAr, and PS. Validation: MU, VI, AS, AA, GR, AL, EG, AAl, and DB. Visualization: DT, MSe, and GT. Writing—original draft: MP, AAs, and VI. Writing—review and editing: MP, AAs, MU, and VI. All authors contributed to the article and approved the submitted version.

This work was supported by Fondazione Carisbo, Bologna (Bando Ricerca medica traslazionale e clinica 2019), Italy, and by a research donation in memory of Alberto Arenghi, Caravaggio, BG, Italy. The funders have no role in the design of the study, the collection, analysis, and interpretation of data, the writing of the manuscript, or submission of the manuscript for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.