Exploring the link between tumour metabolism and succinate dehydrogenase deficiency: A ¹⁸F-FDOPA PET/CT study in head and neck paragangliomas

Abstract

Objectives:

Nuclear imaging findings by virtue of phenotyping disease heavily depend on genetic background, metabolites, cell membrane specific targets and signalling pathways. PPGL related to succinate dehydrogenase subunits mutations (SDHx mutations) are less differentiated than other subgroups and therefore may lack to concentrate ¹⁸F-FDOPA, a precursor of catecholamines biosynthesis. However, this ¹⁸F-FDOPA negative phenotype has been reported mostly in SDHx-PPGL of sympathetic origin, suggesting that both genotype status and location (from sympathetic vs parasympathetic paraganglia; adrenal vs extra-adrenal) could influence ¹⁸F-FDOPA uptake. The aim of this study was to test if SDHx drives ¹⁸F-FDOPA uptake in presence of normal epinephrine/norepinephrine concentrations.

Design:

Retrospective study

Patients:

A cohort of 86 head and neck PPGL patients (including three metastatic) with normal metanephrines underwent ¹⁸F-FDOPA PET/CT. The relationships between ¹⁸F-FDOPA uptake and tumour genotype were evaluated.

Results:

In nonmetastatic HNPGL (50 non-SDHx/33 SDHx), no significant difference was observed between these two groups for SUVmax (P = .256), SUVmean (P = .188), MTV 42% (P = .596) and total lesion uptake (P = .144). Metastatic HNPGL also had high elevated uptake values.

Conclusions:

Our results suggest that SDH deficiency or metastatic behaviour have no influence on ¹⁸F-FDOPA uptake in HNPGL probably due to their very-well differentiation status, even at metastatic stage. The potential prognosticator value of ¹⁸F-FDOPA uptake would need to be further explored in the setting of metastatic PPGL of sympathetic origin.

1 |. INTRODUCTION

Phaeochromocytoma and paraganglioma (PGGL) are the most commonly inheritable neuroendocrine tumours. In recent years, advances in genetics have profoundly helped to gain comprehensive picture about PPGL pathogenesis with description of three main subgroups associated with activation of specific signalling pathways. One of the most frequent and fascinating causes is linked to mutations in the SDH genes (SDHA-D and SDHAF2, collectively named SDHx) that encode for subunits of the succinate dehydrogenase (SDH), a key respiratory enzyme complex that converts succinate to fumarate in tricarboxylic acid (TCA) cycle and also functions in the mitochondrial electron transport chain. SDHx mutations result in SDH deficiency, disruption of the TCA cycle with subsequent accumulation of succinate and metabolic reprogramming that promotes tumorigenesis.

In the recent years, nuclear medicine has also emerged at the forefront of tools for PPGL staging and phenotyping.¹ This is of prime importance as alteration of mitochondrial metabolism through inactivation of the TCA cycle and electron transport chain are also found in several cancers, signifying that these metabolic changes constitute an emerging metabolic hallmark of cancer.² In the setting of PPGL, ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) avidity was found to be associated to SDHx mutations,^3–6 possibly due to the effects of succinate on stabilization of hypoxia-inducible factors⁷ and/or on microenvironment metabolism.^{8,9 18}F-fluorodopa (¹⁸F-FDOPA) phenotyping is more complex since it is influenced by both tumour genotype and catecholamines secretion/content, and both likely being interconnected. ¹⁸F-FDOPA PET/CT was found to be a sensitive imaging tool for sporadic and SDHx-related HNPGL (mostly linked to SDHD mutations).¹⁰ A strong correlation between ¹⁸F-FDOPA- derived metabolic indices and urinary/plasma metanephrines have been found in pheochromocytoma.^11,12

However, beyond secretion, patients with certain genetic mutations have been found as an important component of ¹⁸F-FDOPA phenotype since false-negative findings are mainly observed in patients with metastatic SDHx-PPGL.⁵ Interestingly, it was shown that this negative ¹⁸F-FDOPA PET phenotype may also occur in PPGLs associated with the sympathetic nervous system which are often catecholamines secreting tumours.¹³

In order to ascertain if SDHx drives ¹⁸F-FDOPA uptake in absence of catecholamines excess, we evaluated the relationship between ¹⁸F-FDOPA-derived metabolic indices and genotype in head and neck paraganglioma (HNPGL) which are characterized by absence of production of catecholamines.

2 |. PATIENTS AND METHODS

2.1 |. Patient population

We performed a comprehensive search in our database to identify all patients evaluated by PET/CT over 10 years for a presumed head and neck paraganglioma (HNPGL).¹⁴ Among these patients, only those who fulfilled the following criteria were included: (a) final diagnosis of HNPGL (b) evaluation by ¹⁸F-FDOPA PET/CT, (c) genetic testing for at least germline mutations in the SDHB/SDHD/SDHC genes (including large deletions) (d) normal plasma and/or urinary metanephrines.

The study was approved by the local ethical committee of Aix Marseille University. All patients gave informed consent for the use of anonymous personal data extracted from their medical records for research purposes.

2.2 |. ¹⁸F-FDOPA PET/CT imaging protocol and assessment of ¹⁸F-FDOPA uptake indices

A combined PET/CT scanner was used using Discovery ST or Discovery 710 GE Medical Systems. All patients fasted for at least 3 hours. Image acquisition started at approximately 60 minutes after injection of 3.5 Megabecquerels (MBq)/kilogram (kg) of IASOdopa (1.8–5 MBq/kg, median 3.4 MBq/kg).

Volumetric regions of interest were placed over the areas of significant ¹⁸F-FDOPA uptake (>background) in the HNPGL. SUVmax, SUVmean and MTV 42% (for tumour with highest SUVmax values) were obtained. Total lesion (TL) uptake was calculated as the product of tumour SUVmean and MTV 42% (defined as the region enclosed by a 42% isocontour around the maximum PET voxel). Tumour size was not reported since our PET/CT was performed without contrast enhancement.

2.3 |. Gold standard

Surgical findings and pathological examination were considered as the gold standard for PGL. In cases, where surgical resection was not performed, the diagnosis of PGL and its location was made by the consensus between experienced radiologists and nuclear physicians taking into account all imaging studies (including MRI with angiographic sequences, petrous bone CT, cervico-thoraco-abdominal CT, PET with other radiopharmaceuticals) and follow-up data.

2.4 |. Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics version 20 (IBM SPSS Inc). Comparisons of means values between two groups were performed using Student’s t test or Mann-Whitney U test. Comparison of percentages was performed using chisquare test or Fisher’s exact test. The statistical significance was defined as P < .05.

3 |. RESULTS

3.1 |. Patient and tumour characteristics

One hundred and four patients were evaluated by ¹⁸F-FDOPA for HNPGL during the study period. All of these patients had positive ¹⁸F-FDOPA PET/CT. Eighteen patients were excluded for further analysis due to lack of original DICOM (Digital Imaging and Communications in Medicine) files for quantitative image analysis (n = 15:4 SDHx and 11 non-SDHx cases) or presence of variants of unknown significance pathogenicity (n = 3). Eighty-six patients with HNPGLs were therefore included in this retrospective study. Thirty three patients (40%) carried mutations in one of the SDH genes (5 SDHB, 8 SDHC, 20 SDHD). The mean age (±SD), gender distribution and the percentages of multifocality for nonmetastatic SDHx and non-SDHx, respectively were: 48 ± 13 vs 64 ± 13 years, 16F/17M vs 41F/9M, 54.5% vs 8%. Only one SDHx patient had a concomitant phaeochromocytoma. Three additional patients with metastatic HNPGL (1 SDHB, SDHD and non-SDHx) were evaluated. There was no difference in the start time of the PET acquisitions between both groups (55.6 ± 13.1 in the SDHx group vs 54.7 ± 14.1 in the non-SDHx group, P = .7).

3.2 |. Relationships between SDH status and ¹⁸F-FDOPA uptake

In nonmetastatic HNPGL, the median uptake values of SUVmax, SUVmean, MTV42% and total lesion (TL) uptake for SDHx-HNPGL and nonmutated cases were: 14 (quartile 1: Q1 = 6.4; quartile 3: Q3 = 20.2) vs 9.65 (Q1 = 5.3; Q3 = 13.9), 9 (Q1 = 4.1; Q3 = 12.3) vs 5.8 (Q1 = 3.1; Q3 = 8.7), 10.1 (Q1 = 3.7; Q3 = 15.7) vs 8.2 (Q1 = 3.8; Q3 = 11.1) and 82.5 (Q1 = 19.9; Q3 = 130.3) vs 46.7 (Q1 = 11.6; Q3 = 67.8), respectively. On univariate analysis, age, sex and mutlifocality were found to be associated with SDHx mutation status. No statistical difference between groups was found for ¹⁸F-FDOPA-derived metabolic indices: for SUVmax: P = .256, SUVmean: P = .188, MTV 42%: P = .596 and TL: P = .144 (Figure 1, Table. 1). SDHC-related HNPGL was found to have higher ¹⁸F-FDOPA-derived uptake parameters compared to their non-SDHx counterparts (P = .033 for SUVmax, P = .031 for SUVmean and P = .004 for TL) and higher TL compared to SDHD (P = .033; Figure 2).

FIGURE 1.

Comparison of ¹⁸F-FDOPA-derived indices across genotypes

TABLE 1.

Comparison of patients and tumours’ characteristics in SDHx vs nonSDHx HNPGL (univariate analysis)

Gene Mutated	SDHx mutated n = 33	SDHx negative n = 50	P
Age	48 ± 13.4	64.1 ± 12.5	<.001
Sex ratio (Male/Female)	1.06	0.22	.001
Multifocality, n (%)	54.6	8	<.001
SUVmax	14 ± 13.3	9.7 ± 5.6	.256
SUVmean	9 ± 8.9	5.8 ± 3.3	.188
MTV42%	10.1 ± 10.2	8.2 ± 7.6	.596
TL	82.5 ± 95.9	46.7 ± 44.1	.144

FIGURE 2.

Examples of ¹⁸F-FDOPA PET/CT findings in HNPGL of various genotypes. A, SDHC, B, SDHB, C, SDHD, D, Sporadic case

In the three metastatic patients included, the tumours exhibited highly elevated uptake values with SUVmax, SUVmean, MTV42% and TL of 33.5 ± 34.3; 20 ± 20.4, 2.5 ± 2.5 and 33.7 ± 39.6, respectively.

4 |. DISCUSSION

The present study confirms that HNPGLs are highly avid for ¹⁸F-FDOPA despite absence of catecholamines production but more importantly shows that ¹⁸F-FDOPA uptake is not decreased in SDHx-HNPGL compared to their sporadic counterparts. This finding somewhat contrasts with data obtained in sympathetic and/or metastatic PPGL which may lack to concentrate this radiopharmaceutical.^5,13,15 Although imaging phenotypes are now well-defined across genotypes, until now, there was a paucity of knowledge regarding the association between quantitative imaging and genetic backgrounds. The novelty of this study lies in the fact that since a majority of HNPGLs (67%) do not produce catecholamines,¹⁶ their evaluation allow to only weigh the genetic background of these tumours. One limitation is that we did not assess 3-methoxytyramine (a dopamine metabolite) and that can be elevated in a subset of HNPGLs.¹⁶

Another possible limitation is that an apparently sporadic patient may have a mutation in an SDHx gene (ie SDHA, SDHAF2) that was not included in our gene panel testing (SDHB-D).

To the best of our knowledge, this is the first study that compares ¹⁸F-FDOPA-derived quantitative uptake parameters between SDHx and non-SDHx patients in the setting of HNPGL. The principal conclusion that can be drawn is that no significant difference was demontrated between these two groups in terms of tumour SUVmax/SUVmean/MTV 42%/TL. On subgroup analysis, SDHC-related PGLs were found to have higher ¹⁸F-FDOPA uptake parameters, more particulary TL uptake compared to sporadic and SDHD ones. In our opinion, this finding is probably linked to the larger volume of SDHC-related PGL that might have decreased the partial effect volume but this requires further exploration.

The inter- and intra-patient differences observed on ¹⁸F-FDOPA PET/CT in SDHx mutation carriers is still unexplainable. The presented data suggest that presence of SDH mutations is not enough to drive the ¹⁸F-FDOPA imaging phenotype in all PPGL but can influence tumour metabolism in a certain cell lineage of tumours equipped with catecholamine biosynthesis/storage/turnover. Over the past few years, novel data regarding SDHx-related patients has been described covering their clinical behaviour, biochemical phenotype and imaging characteristics.⁹ It was shown that SDHx mutated tumours are classically associated with lack of uptake of specific radiopharmaceuticals such as ¹⁸F-FDOPA or ¹²³I-MIBG.¹⁷ It is also remarkable to note from published data that in SDHx patients with multifocal disease, some tumours may exhibit highly elevated ¹⁸F-FDOPA uptake values whereas others completely lacked uptake.¹⁸ From the molecular genetics standpoint, SDH deficiency was found to be associated with silencing of chromaffin-specific genes.⁷ Thus, dedifferentiation could play a critical role in decreased ¹⁸F-FDOPA uptake. A case series showed that lack of catecholamines secretion can be observed in SDHB-related PPGL and this finding was not due to defects in the mechanisms of storage or secretion of catecholamines but, instead, reflects a defect in catecholamine synthesis resulting in a near complete absence of releasable catecholamine stores. The defect was identified as an absence of tyrosine hydroxylase, the enzyme that catalyses the initial and rate-limiting step in catecholamine biosynthesis.¹⁹ HNPGLs are classically positive for tyrosine hydroxylase on immunohistochemistry. Since HNPGLs still concentrate ¹⁸F-FDOPA and do not secrete catecholamines, lack of ¹⁸F-FDOPA uptake may be related to a disruption in early steps of catecholamines biosynthesis such as lack of expression/phosphorylation of tyrosine hydroxylase. In this scenario, SDH deficiency may facilitate (induce) complete loss of the machinery to produce catecholamines and ¹⁸F-FDOPA uptake as parts of tumour dedifferentiation and this could mainly occur in sympathetic-derived PPGL. This is also in agreement with our previous work showing that down-regulation of LAT1/CD98hc is unlikely to be the cause of decreased ¹⁸F-FDOPA uptake.²⁰ Another hypothesis for ¹⁸F-FDOPA negative PPGL phenotype being undifferentiated could be that they develop from undifferentiated chromaffin precursors cells lacking expression of key enzymes involved in catecholamines biosynthesis. Overall, ¹⁸F-FDOPA uptake is possibly associated with maintenance of differentiation status. In order to further explore various scenarios, analysis of the molecular machinery of catecholamine biosynthesis and secretion and screening for potential somatic mutations in ¹⁸F-FDOPA negative compared to ¹⁸F-FDOPA positive tumours would be of particular interest.

It has also been shown that SDHB-related PPGL which are mostly of sympathetic origin are characterized by a greater propensity towards metastasis compared to other SDHx subgroups,^21,22 suggesting that beyond SDH deficiency common to SDHA/B/C/D mutants, additional events may promote an aggressive behaviour. The analysis of our three cases of metastatic HNPGL suggests that loss of ¹⁸F-FDOPA uptake (immature phenotype) is not a prerequisite for metastatic spread and may occur at distant sites in host tissues.

In conclusion, our results suggest that SDH deficiency or metastatic behaviour did not decrease ¹⁸F-FDOPA uptake in HNPGL probably due to their well-differentiated status, even at metastatic stage. The potential prognosticator value of ¹⁸F-FDOPA uptake would need to be further explored in the setting of metastatic PPGL of sympathetic origin.