Metabologenomics of Phaeochromocytoma and Paraganglioma: An Integrated Approach for Personalised Biochemical and Genetic Testing

Abstract

The tremendous advances over the past two decades in both clinical genetics and biochemical testing of chromaffin cell tumours have led to new considerations about how these aspects of laboratory medicine can be integrated to improve diagnosis and management of affected patients. With germline mutations in 15 genes now identified to be responsible for over a third of all cases of phaeochromocytomas and paragangliomas, these tumours are recognised to have one of the richest hereditary backgrounds among all neoplasms. Depending on the mutation, tumours show distinct differences in metabolic pathways that relate to or even directly impact clinical presentation. At the same time, there has been improved understanding about how catecholamines are synthesised, stored, secreted and metabolised by chromaffin cell tumours. Although the tumours may not always secrete catecholamines it has become clear that almost all continuously produce and metabolise catecholamines. This has not only fuelled changes in laboratory medicine, but has also assisted in recognition of genotype-biochemical phenotype relationships important for diagnostics and clinical care. In particular, differences in catecholamine and energy pathway metabolomes can guide genetic testing, assist with test interpretation and provide predictions about the nature, behaviour and imaging characteristics of the tumours. Conversely, results of genetic testing are important for guiding how routine biochemical testing should be employed and interpreted in surveillance programmes for at-risk patients. In these ways there are emerging needs for modern laboratory medicine to seamlessly integrate biochemical and genetic testing into the diagnosis and management of patients with chromaffin cell tumours.

Figure 1.

Timeline outlining a half century of advances in genetics (A) and biochemical diagnosis (B) of phaeochromocytomas and paragangliomas. Abbreviations: NF1, neurofibromatosis type 1; VHL, von Hippel-Lindau; MEN 2, multiple endocrine neoplasia type 2; PGL, paraganglioma; RET, rearranged during transfection proto-oncogene; SDHA, SDHB, SDHC, SDHD, succinate dehydrogenase subunits A, B, C, & D; TMEM127, transmembrane domain protein 127; SDHAF2, succinate dehydrogenase complex assembly factor 2; MAX, MYC-associated factor X; HIF2α, hypoxia inducible factor 2; FH, fumarate hydratase; MDH2, malate dehydrogenase 2; PHD1, PHD2, prolyl hydroxylase 1 & 2; (PHD1 and PHD2), KIF1Bβ, kinesin family member 1B (genes shown in grey have been described exclusively or near exclusively with somatic mutations); VMA, vanillylmandelic acid.

Figure 2.

Pathways of catecholamine metabolism in the absence of a chromaffin cell tumour (A) and the presence of a secreting noradrenergic tumour (B) or a non-secreting adrenergic tumour (C). Numbers at the head of pathway arrows indicate rates of entry (nmol/min) of catecholamines and their metabolites into the bloodstream compartment from sympathoneuronal, extraneuronal, adrenal chromaffin and tumour cell compartments. Values at the base of each panel indicate combined rates of entry into the plasma compartment for noradrenaline (NA), adrenaline (A), normetanephrine (NMN) and metanephrine (MN). Values in panel B illustrate the mathematical explanation for the larger diagnostic signal strength for NMN than NA (i.e. 4.5 vs 2.0-fold increase) derived from a noradrenergic tumour that secretes NA and NMN in the same proportions as from adrenal chromaffin cells. Values in panel C illustrate the advantages of measuring metanephrines over catecholamines in a tumour that does not secrete catecholamines or secretes catecholamine intermittently. Abbreviations: COMT, catechol-O-methyltransferase; DHPG, 3,4-dihydroxyphenylglycol; MAO, monoamine oxidase; MHPG, 3-methoxy-4-hydroxyphenylglycol; PNMT, phenylethanolamine-N-methyltransferase.

Figure 3.

Catecholamine biosynthetic pathway within an adrenal chromaffin cell. Abbreviations: TYR, tyrosine; DOPA, dihydroxyphenylalanine; DA, dopamine; NA, noradrenaline; A, adrenaline; MTY, methoxytyramine; NMN, normetanephrine; MN, metanephrine; TH, tyrosine hydroxylase; AADC, aromatic amino acid decarboxylase; DBH, dopamine beta-hydroxylase; PNMT, phenylethanolamine N-methyltransferase; COMT, catechol-O-methyltransferase; VS, vesicular sequestration; VL, vesicular leakage.

Figure 4.

Distributions of chromaffin cell tumours according to tumour tissue contents of adrenaline as a percent of total catecholamine contents (A), relationship of tumour tissue adrenaline contents with tumour-derived increases of plasma metanephrine (MN) as a percent of combined increases of normetanephrine (NMN) and metanephrine (B) and scattergram of plasma concentrations of metanephrine with tumour-derived increases of plasma metanephrine as a percent of combined increases of normetanephrine and metanephrine (C). In panel C designation of an adrenergic phenotype requires both an elevated plasma concentration of metanephrine above 62 pg/mL (0.31 nmol/L) and a tumour-derived increase of metanephrine of more than 6% of the combined increases of normetanephrine and metanephrine.

Figure 5.

Plasma concentrations of normetanephrine (A), metanephrine (B) and methoxytyramine (C) in patients with PPGLs due to VHL, SDHB, SDHD RETand NF1 mutations and separation of groups illustrated by principal components (D) and discriminant analyses (E). Dashed horizontal lines in panels A, B and C indicate upper cut-offs of reference limits, which for normetanephrine vary between 100 and 200 pg/mL depending on advancing age.

Figure 6.

Phenotypic features that distinguish cluster 1 versus cluster 2 groups of PPGLs as related to pathways of chromaffin cell tumourigenesis. Pathway components impacted by mutations of tumour susceptibility genes are indicated by stars (✸) and include genes for succinate dehydrogenase subunits (SDHx), including SDHA, SDHB, SDHC, SDHD and the accessory protein, SDHAF2. Cluster 1 tumours are proposed to develop from immature chromaffin progenitors that express HIF2α and are subsequently susceptible to mutations of genes that promote stabilisation of HIF2α at the protein level. Stabilised expression of HIF2α contributes to arrested differentiation characterised by immature phenotypic features and an earlier age of disease presentation compared to cluster 2 tumours, which are derived from mature chromaffin tumour progenitors that do not express HIF2α.

Figure 7.

Tumour tissue contents of catecholamines (A), rates of secretion of catecholamines from tumours (B) and catecholamine secretory rate constants (C) for PPGLs from patients with hereditary (VHL, SDHx) and sporadic (SPOR) noradrenergic tumours versus hereditary (NF1, RET) and sporadic (SPOR) adrenergic tumours. Secretory rate constants (day⁻¹) illustrate that for noradrenergic tumours over a third of all catecholamines in stores are secreted within one day, whereas for adrenergic tumours less than a twentieth of stores are secreted within one day.

Figure 8.

Algorithms for PPGL genetic testing by Sanger sequencing (A) or next-generation sequencing (B). The conventional feature driven algorithm (A) dictates sequential targeted analysis (Sanger sequencing) based on tumour location, biochemical phenotype, multiple or single tumours, age at diagnosis and presence of metastatic disease. The proposed algorithm for next generation sequencing (NGS) using multi-gene panels (B) similarly includes targeted testing for patients with a syndromic presentation. If no pathogenic mutation is found, other genes included in the panel are explored. Specimens from patients with a non-syndromic presentation are directly analysed using the NGS panel. After identifying a variant of unknown significance (VUS), those features that compose the genome-metabolome-phenome provide tools to assist classification of variants as pathogenic or benign. In the absence of any mutation, the phenome-metabolome-genome should be carefully reviewed. If the profile is characteristic of an established PPGL susceptibility gene, other mutational mechanisms that may be missed by routine screening techniques (e.g. promoter hypermethylation) could be explored. If the profile is not characteristic of an established PPGL susceptibility gene, the patients are candidates for the search of mutations in new genes. Any patient carrying a VUS, or in whom no variants are found may be re-evaluated at a later date according to new knowledge.

Figure 9.

PPGL tumour tissue ratios of succinate to fumarate (A), pyruvate to cis-aconitate (B) and use of ratios of those and other energy pathway metabolites to distinguish groups of patients with PPGLs with and without (none) different mutations of tumour susceptibility genes (i.e. VHL, HIF2α, SDHB, SDHD, SDHC RET, NF1and TMEM127) according to discriminant analyses that illustrates almost complete separation of patients with mutations of genes for subunits of SDH (SDHx) from other groups. Dashed horizontal lines in panels A and B serve to indicate upper cut-offs of reference limits for ratios.

Figure 10.

Changes in plasma concentrations of normetanephrine (A) and noradrenaline (B) over a time period of up to 4 years of blood sampling in 7 patients with adrenal phaeochromocytomas secondary to VHL mutations. The dashed horizontal lines indicate upper cut-offs (UC) of reference intervals, which for normetanephrine are age-specific dependent on ages of each of the 7 patients. Note that for most patients (exception patient 7) normetanephrine steadily climbs before tumours are resected, whereas for noradrenaline there is no clear trend and concentrations mainly remain below upper cut-offs.