SDHB variant type impacts phenotype and malignancy in pheochromocytoma-paraganglioma

Preferential MGMT hypermethylation in SDH-deficient wild-type GIST

AIMS

Wild-type gastrointestinal stromal tumours (wtGIST) are frequently caused by inherited pathogenic variants, or somatic alterations in the succinate dehydrogenase subunit genes (SDHx). Succinate dehydrogenase is a key enzyme in the citric acid cycle. SDH deficiency caused by SDHx inactivation leads to an accumulation of succinate, which inhibits DNA and histone demethylase enzymes, resulting in global hypermethylation. Epigenetic silencing of the DNA repair gene MGMT has proven utility as a positive predictor of the therapeutic efficacy of the alklyating drug temozolomide (TMZ) in tumours such as glioblastoma multiforme. The aim of this study was to examine MGMT promoter methylation status in a large cohort of GIST.

METHODS

MGMT methylation analysis was performed on 65 tumour samples including 47 wtGIST (33 SDH-deficient wtGIST and 11 SDH preserved wtGIST) and 21 tyrosine kinase (TK) mutant GIST.

RESULTS

MGMT promoter methylation was detected in 8 cases of SDH-deficient (dSDH) GIST but in none of the 14 SDH preserved wild-type GIST or 21 TK mutant GIST samples analysed. Mean MGMT methylation was significantly higher (p 0.0449) and MGMT expression significantly lower (p<0.0001) in dSDH wtGIST compared with TK mutant or SDH preserved GIST. No correlation was identified between SDHx subunit gene mutations or SDHC epimutation status and mean MGMT methylation levels.

CONCLUSION

MGMT promoter hypermethylation occurs exclusively in a subset of dSDH wtGIST. Data from this study support testing of tumour MGMT promoter methylation in patients with dSDH wtGIST to identify those patients who may benefit from most from TMZ therapy.

Hypothesis: Why Different Types of SDH Gene Variants Cause Divergent Tumor Phenotypes

Despite two decades of paraganglioma-pheochromocytoma research, the fundamental question of how the different succinate dehydrogenase (SDH)-related tumor phenotypes are initiated has remained unanswered. Here, we discuss two possible scenarios by which missense (hypomorphic alleles) or truncating (null alleles) SDH gene variants determine clinical phenotype. Dysfunctional SDH is a major source of reactive oxygen species (ROS) but ROS are inhibited by rising succinate levels. In scenario 1, we propose that SDH missense variants disrupt electron flow, causing elevated ROS levels that are toxic in sympathetic PPGL precursor cells but well controlled in oxygen-sensing parasympathetic paraganglion cells. We also suggest that SDHAF2 variants, solely associated with HNPGL, may cause the reversal of succinate dehydrogenase to fumarate reductase, producing very high ROS levels. In scenario 2, we propose a modified succinate threshold model of tumor initiation. Truncating SDH variants cause high succinate accumulation and likely initiate tumorigenesis via disruption of 2-oxoglutarate-dependent enzymes in both PPGL and HNPGL precursor tissues. We propose that missense variants (including SDHAF2) cause lower succinate accumulation and thus initiate tumorigenesis only in very metabolically active tissues such as parasympathetic paraganglia, which naturally show very high levels of succinate.