Genotype–phenotype correlations in pheochromocytoma and paraganglioma: a systematic review and individual patient meta-analysis

in Endocrine-Related Cancer
View More View Less
  • 1 Department of Medical Sciences, Uppsala University, Uppsala, Sweden
  • | 2 Section on Medical Neuroendocrinology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland, USA
  • | 3 Department of Medical Oncology, The Christie NHS Foundation Trust (ENETS Centre of Excellence), Manchester, UK

Correspondence should be addressed to J Crona: joakim.crona@medsci.uu.se
Free access

Pheochromocytoma and paraganglioma (PPGL) can be divided into at least four molecular subgroups. Whether such categorizations are independent factors for prognosis or metastatic disease is unknown. We performed a systematic review and individual patient meta-analysis aiming to estimate if driver mutation status can predict metastatic disease and survival. Driver mutations were used to categorize patients according to three different molecular systems: two subgroups (SDHB mutated or wild type), three subgroups (pseudohypoxia, kinase signaling or Wnt/unknown) and four subgroups (tricarboxylic acid cycle, VHL/EPAS1, kinase signaling or Wnt/unknown). Twenty-one studies and 703 patients were analyzed. Multivariate models for association with metastasis showed correlation with SDHB mutation (OR 5.68 (95% CI 1.79–18.06)) as well as norepinephrine (OR 3.01 (95% CI 1.02–8.79)) and dopamine (OR 6.39 (95% CI 1.62–25.24)) but not to PPGL location. Other molecular systems were not associated with metastasis. In multivariate models for association with survival, age (HR 1.04 (95% CI 1.02–1.06)) and metastases (HR 6.13 (95% CI 2.86–13.13)) but neither paraganglioma nor SDHB mutation remained significant. Other molecular subgroups did not correlate with survival. We conclude that molecular categorization accordingly to SDHB provided independent information on the risk of metastasis. Driver mutations status did not correlate independently with survival. These data may ultimately be used to guide current and future risk stratification of PPGL.

Abstract

Pheochromocytoma and paraganglioma (PPGL) can be divided into at least four molecular subgroups. Whether such categorizations are independent factors for prognosis or metastatic disease is unknown. We performed a systematic review and individual patient meta-analysis aiming to estimate if driver mutation status can predict metastatic disease and survival. Driver mutations were used to categorize patients according to three different molecular systems: two subgroups (SDHB mutated or wild type), three subgroups (pseudohypoxia, kinase signaling or Wnt/unknown) and four subgroups (tricarboxylic acid cycle, VHL/EPAS1, kinase signaling or Wnt/unknown). Twenty-one studies and 703 patients were analyzed. Multivariate models for association with metastasis showed correlation with SDHB mutation (OR 5.68 (95% CI 1.79–18.06)) as well as norepinephrine (OR 3.01 (95% CI 1.02–8.79)) and dopamine (OR 6.39 (95% CI 1.62–25.24)) but not to PPGL location. Other molecular systems were not associated with metastasis. In multivariate models for association with survival, age (HR 1.04 (95% CI 1.02–1.06)) and metastases (HR 6.13 (95% CI 2.86–13.13)) but neither paraganglioma nor SDHB mutation remained significant. Other molecular subgroups did not correlate with survival. We conclude that molecular categorization accordingly to SDHB provided independent information on the risk of metastasis. Driver mutations status did not correlate independently with survival. These data may ultimately be used to guide current and future risk stratification of PPGL.

Introduction

The Cancer Genome Atlas (TCGA) proposed that neuroendocrine tumors of adrenal paraganglia, pheochromocytomas (PCCs) and extra-adrenal paraganglia paragangliomas (PGLs, together denoted PPGL) can be divided into three main molecular subgroups that have been linked to distinct driver genes (Fishbein et al. 2017): pseudohypoxia (SDHA, SDHB, SDHC, SDHD, SDHAF2, FH, VHL, EPAS1 and EGLN1), Wnt-altered (CSDE1 or MAML3) and kinase signaling (RET, NF1, TMEM127, MAX, HRAS, FGFR1 and MET) (Letouze et al. 2013, Castro-Vega et al. 2015, Toledo et al. 2016, Bausch et al. 2017, Fishbein et al. 2017, Welander et al. 2018). Previous data also support that the pseudohypoxic group can be further divided into two subclusters: tricarboxylic acid (TCA) cycle-related (SDHA-SDHD, SDHAF2 or FH) and those with VHL/EPAS1-related (VHL/EPAS1/EGLN1) PPGLs (Burnichon et al. 2011, Letouze et al. 2013, Fliedner et al. 2016). Each subgroup is named after their molecular hallmarks and are thought to be associated with distinct biochemical and clinical phenotypes (reviewed in Crona et al. 2017, Neumann et al. 2018). All pseudohypoxic PPGLs secrete norepinephrine and those related to the TCA cycle are more predominantly PGLs with relatively high proportion having dopamine secretion. The TCA cycle subgroup and particularly SDHB carriers are associated with the highest proportion of metastatic disease (Eisenhofer et al. 2011a,b). On the other end of the spectrum is the kinase signaling subgroup that has a more well-differentiated phenotype with epinephrine secretion, predominantly adrenal location and rarely develop metastatic disease. PPGLs related to the Wnt-altered subgroup are thought to display intermediate characteristics in terms of catecholamine secretion (mixed noradrenergic and adrenergic) and frequency of metastatic or recurrent disease (Fishbein et al. 2017). It has also been proposed that PPGL with somatic abberations in genes related to telomere maintenance (inactivation of ATRX or transcriptional activation of TERT) as well as chromatin maintenance (SETD2) could have more aggressive features and may thus be disease modifiers (Fishbein et al. 2015, 2017, Job et al. 2018).

The predominant cause of death in patients with PPGL is metastasis that occurs in about 10–20% of cases (Timmers et al. 2008, Hamidi et al. 2017a). Tumor location (PGL versus PCC), germline SDHB mutations (SDHB mutated versus SDHB wild type), ATRX mutation, TERT overexpression, catecholamine secretion (noradrenergic or dopaminergic versus adrenergic) and large size of the primary tumor have all been independently associated with metastasis (Ayala-Ramirez et al. 2011, Welander et al. 2011, Eisenhofer et al. 2012, Turkova et al. 2015, Assadipour et al. 2017, Cho et al. 2018, Job et al. 2018). The disease course of those with metastatic disease is heterogeneous in terms of tumor aggressiveness and overall survival (Hamidi et al. 2017b). Size of the primary tumor, gender, SDHB mutation, catecholamine profile, ATRX mutation and TERT overexpression are suggested to be prognostic factors for survival (Amar et al. 2007, Ayala-Ramirez et al. 2011, Zelinka et al. 2011, Hamidi et al. 2017b, Job et al. 2018). Recent data published after submission of this article interestingly showed that among patients with metastatic disease, SDHB was not a negative prognostic factor for survival (Hescot et al. 2019).

Although at least 16 common driver genes have been identified in PPGL, the only disease driver that showed a robust correlation to metastatic disease and outcome has been SDHB. We and others have proposed that the improved characterization of PPGL driver mutations provide additional information beyond the dichotomous categorization based on SDHB. However, due to disease rarity and extensive genetic heterogeneity, interpretation of findings are currently restricted due to low statistical power. We hypothesized that a systematic review and individual patient meta-analysis could overcome these challenges and provide information on correlation between driver mutation status and clinical parameters. We particularly focused on predictive factors for metastatic disease and prognostic factors for survival.

Methods

This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) workflow (Liberati et al. 2009). The study reviewed and analyzed published data, these activities fall under an approval by the Regional Ethics Committee in Uppsala, Sweden (Dnr 2015/544).

Search strategy

One investigator (J C) performed a systematic search of PubMed to identify relevant reports published between 2007-01-01 and 2017-12-01. The following search terms were used: ‘pheochromocytoma’ and ‘paraganglioma’. Reports were initially screened by title for relevance and potentially relevant reports had its abstract reviewed. Case reports, review articles and editorials as well as those publications in other languages than English were not considered. Potentially relevant studies were assessed for eligibility through review of the full-text article.

Study selection and data extraction

Studies fulfilling the following eligibility criteria were included. Criteria (1), genetic sequencing and reporting of PPGL disease drivers: germline mutations, SDHA, SDHB, SDHD, TMEM127; germline and somatic mutations, VHL, RET, NF1, MAX; and somatic mutations, HRAS. Criteria (2), shared data on genetic mutations and clinical characteristics on the individual patient level for both mutation-positive and mutation-negative cases. Criteria (3), patient identification numbers for cross-validation between different studies from the same study site. Publications were grouped into cohorts based on the study site to allow for reconstruction of each study cohort. Two investigators (J C and S G) reviewed the papers independently and transferred the data into one study database. Values that did not overlap between the two investigators were re-assessed to reach a common conclusion.

Data collection and cleaning

Patients without available PPGL tissue for analysis were excluded. For patients with multiple primary tumors, the one that occurred at the earliest age was selected or if the same time point, the row that occurred first in the original data was chosen. In patients with a conflict of data between multiple publications, the most recent value was used. Collected data points and definitions are provided in the Supplementary Methods and Results (see section on supplementary data given at the end of this article).

Definition of PPGL driver gene subgroups

With the TCGA publication as a starting point and taking into account the available literature, we selected three different systems for driver gene categorization. Firstly, a two-molecular subgroup system accordingly to SDHB mutation status: SDHB mutated or SDHB wild type. Secondly, a system with three-molecular subgroups categorized according to the presence of germline/somatic driver mutations/gene fusions: pseudohypoxia (SDHA, SDHB, SDHC, SDHD, SDHAF2, FH, VHL, EPAS1 or EGLN1), kinase signaling (NF1, RET, TMEM127, MAX, HRAS, MET or FGFR1) and Wnt/unknown (CSDE1 or MAML3). PPGL with driver mutations associated with different molecular subgroups as well as those without a driver mutation was classified as Wnt/unknown. The cortical admixture subgroup, originally reported by the TCGA project is thought to be defined by non-tumoral cells and was not included (Crona et al. 2017, Fishbein et al. 2017). Thirdly, a system with four-molecular subgroups was used to take into account the distinct features of TCA-cycle-related (SDHA, SDHB, SDHC, SDHD, SDHAF2 and FH) and VHL/EPAS1-related (VHL, EPAS1 and EGLN1) PPGLs (Burnichon et al. 2011, Flynn et al. 2015, Fliedner et al. 2016).

Risk of bias assessment

Risk of bias was determined by two investigators (J C, S G), and in cases of discrepant assessment, the original papers were re-evaluated to reach a common conclusion. Bias assessment was designed based on a modified Newcastle–Ottawa tool for bias assessment adopted by Hamidi et al. (2017a) that was further modified to this study. Criteria for bias assessment is available in the Supplementary Methods and Results.

Statistical analyses

Nominal data are presented as number of patients and percentages and were analyzed with chi-square test. Scaled data were presented as median and range or 95% confidence interval (CI) and were analyzed with Mann–Whitney U or Kruskal–Wallis tests. Logistic regression (univariate/multivariable) was used as appropriate. Survival analysis was performed using log-rank, Kaplan–Meier and Cox regression analyses. P values <0.05 were defined as statistically significant. Variables identified as significant in univariate analysis were included in multivariable analysis (applicable for logistic and for Cox regression). Statistical analyses were performed using SPSS version 22 (IBM) and Stata version 12. Figures were drawn with Prism 6.0h (GraphPad Software Inc) and Stata version 12.

Results

A PubMed search generated 7689 results, and 118 manuscripts were selected for review of eligibility (Supplementary Fig. 1). A total of 97 publications did not meet the criteria on method for genetic sequencing (n = 82) or individual patient data availability (n = 13). Two studies were excluded as individual patients could not be matched to previous studies. Twenty-one publications matched study criteria and allowed reconstruction of seven cohorts (Table 1). These seven cohorts represented 948 individual patients, 32 had data on multiple tumor lesions. Two hundred forty-five patients were excluded as there was no tumor tissue available. Seven hundred three patients remained, 274 were analyzed with exome sequencing and 429 with a targeted re-sequencing approach.

Table 1

Study cohorts.

ReferencesPatients (n)Study siteWES (n)Targeted re-sequencing (n)
Welander et al. (2012, 2018), Welander et al. (2014a,b), Juhlin et al. (2015), Stenman et al. (2016a,b)137Karolinska University Hospital, Stockholm, Sweden; Linköping University Hospital, Linköping, Sweden; Haukeland University Hospital, Bergen, Norway19118
Wilzen et al. (2016)9Sahlgrenska University Hospital, Gothenburg, Sweden90
Flynn et al. (2015, 2016, 2017), Dwight et al. (2018)48The Peter MacCallum Cancer Centre and University of Melbourne, Australia; Kolling Institute and University of Sydney, Australia; Royal Brisbane Hospital, Brisbane, Australia; Royal North Shore Hospital, Sydney, Australia; The Children’s Hospital at Westmead, Sydney, Australia444
Burnichon et al. (2011, 2012), Favier et al. (2012), Letouze et al. (2013), Castro-Vega et al. (2014, 2015)190INSERM, Hôpital Europe ́en Georges Pompidou and Université Paris Descartes, Sorbonne Paris Cité French, Paris, France. Cortico et Médullosurrénale: les Tumeurs Endocrines (COMETE) Network, France29161
Fishbein et al. (2017)173National Institutes of Health, United states and multiple collaborating institutions of the Cancer Genome Atlas, Pheochromocytoma and Paraganglioma project1730
Curras-Freixes et al. (2015)118Centro Nacional de Investigaciones Oncológicas, Madrid, Spain and multiple collaborating institutions throughout Spain0118
Toledo et al. (2016)28University of Texas Health Science Center at San Antonio, San Antonio, United States280

WES, whole-exome sequencing.

Risk of bias assessment

Risk of bias assessments was performed for each study cohort, and it is presented in Supplementary Fig. 2. All studies performed retrospective characterization of case series. Assessment of genetic results (7/7 studies low risk) and method coverage (5/5 studies low risk) showed relatively low risk of bias. Clinical data and particularly hormone assessment (6/7 studies high or unclear risk) and follow-up time (7/7 studies high or unclear risk) had a high risk of bias.

Baseline characteristics

Clinical characteristics of the reviewed patients are presented in Table 2. PPGL-related driver mutations were detected in 437 patients (62.6%, 95% CI 58.5–65.7, Supplementary Fig. 3 and Supplementary Table 1) that were confirmed as germline in 178 (25.3%, 95% CI 22.3–28.7) and somatic in 237 (33.7%, 95% CI 30.3–37.3). The frequency of mutations in the different driver genes are shown in the Supplementary Methods and Results, and Supplementary Table 1.

Table 2

Clinical characteristics of the reviewed patients.

Patients (n = 703) Frequency %
GenderMale30743.7
Female 39255.7
Data not available 40.6
Age at diagnosisMedian (range) 46 (7–84)
Tumor size (mm)Median (range) 45 (10–160)
Stage Non-metastatic61887.9
Metastatic 8512.1
Catecholamine profileEpinephrine16122.9
Norepinephrine13919.8
Dopamine 243.4
Data not available32453.9
WHO 2004 Pheochromocytoma57281.4
Paraganglioma12718.1
Data not available 40.5
WHO 2017 PCC57281.4
Sympathetic PGL 9613.7
Head and neck PGL 273.8
Data not available 81.1
Time on follow-up (months) Median (range) 33 (0–316)
Status at the end of follow-up Alive49470.3
Dead 405.7
Data not available16924.0
ATRX mutation statusATRX mutated45063.9
ATRX wild type 172.4
Data not available23733.7
2-molecular subgroupsSDHB wild type64591.7
SDHB mutated 588.3
3-molecular subgroupsPseudohypoxia17725.2
Kinase signaling24534.9
Wnt/unknown28139.9
4-molecular subgroupsPseudohypoxia TCA-cycle 7911.3
Pseudohypoxia VHL/EPAS1 9813.9
Kinase signaling24534.9
Wnt/unknown28139.9

Data on age was missing in 3 patients, on tumor size on 291 patients and on follow-up length in 167 patients.

DA, dopamine; E, epinephrine; F, female; HNPGL, head and neck PGL; M, male; NA, not available; NE, norepinephrine; PCC, pheochromocytoma; PGL, paraganglioma; sPGLs, sympathetic PGL; TCA, tricarboxylic acid.

Patients were categorized into three different molecular systems. Two-molecular subgroups: SDHB mutated 58 patients (8.3%, 95% CI 6.4–10.5, Table 2) and SDHB wild type 645 patients (91.8%, 95% CI 89.5–93.6). Three molecular subgroups: pseudohypoxia, 177 patients (24.9%, 95% CI 21.8–28.2); kinase signaling, 245 patients (34.9%, 95% CI 31.4–38.5); and Wnt/unknown, 281 patients (39.9%, 95% CI 36.4–43.6). In the four-molecular subgroup system, the pseudohypoxia subgroup was further divided into TCA cycle, 79 patients (11.2%, 95% CI 9.1–13.8) and VHL/EPAS1 related, 98 patients (13.9%, 95% CI 11.6–16.7).

Clinical correlations to molecular subgroups

An overview of clinical correlations to the different molecular systems are presented in Fig. 1, Supplementary Fig. 3 and Supplementary Table 2. Gender, catecholamine profile, WHO classification (PCC versus PGL), metastatic stage, age at diagnosis as well as tumor size were all differently distributed (P values <0.05) among subgroups of all three molecular systems. Detailed descriptive data are available in the Supplementary Methods and Results.

Figure 1
Figure 1

Clinical correlations to modified molecular subgroups from the Cancer Genome Atlas. Epi, epinephrine; kinase, kinase signaling subgroup; Norepi, norepinephrine; PCC, pheochromocytoma; PGL, paraganglioma; TCA, tricarboxylic acid; UK, unknown; VHL/EPAS1, pseudohypoxia VHL/EPAS1 related. ***Chi-square test including all four molecular subgroups had a significance level <0.001.

Citation: Endocrine-Related Cancer 26, 5; 10.1530/ERC-19-0024

Predictive factors of metastatic disease

Frequency of metastatic disease in the cohort was 12.1% (85/703 patients). Categorization accordingly to the two-, three- and four-molecular subgroup systems as well as catecholamine profile, WHO classification and ATRX mutation status correlated with metastatic disease in univariate Cox regression analyses (Fig. 1 and Table 3). Those with SDHB-mutated PPGLs had metastatic disease in 46.6% (27/58 patients, OR 8.81 (95% CI 4.92–15.78); P < 0.001) that was higher compared to SDHB wild type 8.9% (58/645 patients) PPGLs. In the three-molecular subgroup system, metastasis was more common in pseudohypoxia 24.3% (43/177 patients, OR 2.49 (95% CI 1.51–4.13) P < 0.001) and less frequent in kinase signaling 4.1% (10/245 patients, OR 0.33 (95% CI 0.16–0.69) P = 0.003) compared to Wnt/unknown 11.4% (32/281 patients). In the four-molecular subgroups classification, metastatic PPGLs occurred more often in TCA cycle 40.5% (32/79 patients, OR 5.29 (95% CI 2.96–9.47)) but was not different in VHL/EPAS1-related PPGLs 11.2% (11/98 patients, OR 0.98 (95% CI 0.78–2.04)) compared to the Wnt/unknown group.

Table 3

Factors related with increased risk of metastatic disease.

Logistic regression (univariate analysis)Logistic regression (multivariable analysis)Logistic regression (multivariable analysis)Logistic regression (multivariable analysis)
Model 1Model 2Model 3
metastatic disease patients (%)OR (95% CI); P valueOR (95% CI); P valueOR (95% CI); P valueOR (95% CI); P value
GenderFemale11.2%1 (Ref)
Male13.4%1.22 (0.77–1.92); 0.393
Age at diagnosisContinuous variableNA0.99 (0.98–1.01); 0.307
Tumor size (mm)≤5010.9%1 (Ref)
>5015.7%1.52 (0.76–3.05); 0.236
HormoneEpinephrine3.1%1 (Ref)1 (Ref)1 (Ref)1 (Ref)
Norepinephrine12.9%4.64 (1.68–12.86); 0.0033.01 (1.02–8.79); 0.0452.88 (0.94–8.82); 0.0653.12 (1.02–9.56); 0.046
Dopamine29.2%12.85 (3.67–44.93); <0.0016.39 (1.62–25.24); 0.0087.86 (2.03–30.4); 0.0036.32 (1.58–25.30); 0.009
WHO 2004PCC7.9%1 (Ref)1 (Ref)1 (Ref)1 (Ref)
PGL29.1%4.81 (2.95–7.85); <0.0011.52 (0.52–4.43); 0.4363.09 (1.20–7.97); 0.0191.76 (0.57–5.42); 0.324
WHO 2017PCC7.9%1 (Ref)***
Sympathetic PGL29.2%4.82 (2.82–8.23); <0.001***
Head and Neck PGL25.9%4.09 (1.65–10.21); 0.002***
ATRX mutation statusATRX mutated58.5%13.18 (4.78–36.36); <0.001***
ATRX wild type9.8%1 (Ref)***
2-Molecular subgroupsSDHB wild type8.9%1 (Ref)1 (Ref)£&
SDHB mutated46.6%8.81 (4.92–15.78); <0.0015.68 (1.79–18.06); 0.003£&
3-Molecular subgroupsWnt/unknown11.4%1 (Ref)$1 (Ref)&
Pseudohypoxia24.3%2.49 (1.51–4.13); <0.001$0.92 (0.35–2.43); 0.861&
Kinase signaling4.1%0.33 (0.16–0.69); 0.003$0.49 (0.13–1.91); 0.305&
4-Molecular subgroupsWnt/unknown11.4%1 (Ref)$£1 (Ref)
Pseudohypoxia TCA-cycle40.5%5.29 (2.96–9.47); <0.001$£2.65 (0.83–8.48); 0.101
Pseudohypoxia VHL/EPAS1-11.2%0.98 (0.78–2.04); 0.965$£!
Kinase signaling4.1%0.33 (0.19–0.69); 0.003$£0.45 (0.12–1.74); 0.246

ATRX mutation status correlated with increased frequency of metastasis in univariate analysis. But, it was not included in the multivariate model due to a lack of clinical annotations in cases that had analysis of ATRX mutation status. Identification of threshold for tumor size is described in the Supplementary Materials and Results

*Not included, WHO 2004 classification included in the multivariable analysis; $three- and four-molecular subgroup systems not included in model 1 of the multivariable analysis (two-molecular subgroups included instead); £two- and four-molecular subgroups not included in model 2 of the multivariable analysis (three-molecular subgroups classification included instead); &two- and three-molecular subgroups not included in model 3 of the multivariable analysis (four-molecular subgroups classification included instead); !could not be calculated due to lack of observations.

95% CI, 95% confidence interval; DA, dopamine; E, epinephrine; HNPGL, head and neck PGL; NA, not available; NE, norepinephrine; OR, odds ratio; PCC, pheochromocytoma; PGL, paraganglioma; Ref, reference; sPGLs, sympathetic PGL; TCA, tricarboxylic acid.

The three different molecular systems were analyzed separately for association with metastatic disease in multivariate models. Each model included other significant variables in univariate analyses: catecholamine profile (norepinephrine or dopamine compared to epinephrine) and WHO classification (PGL compared to PCC). While ATRX-mutated PPGL showed a positive correlation with metastases in the univariate analysis, information on ATRX mutation status was only available in a subset of patients (467/703) that also lacked complete clinical annotations. As such, ATRX mutation status was not included in the multivariate models. In model 1 (exploring the role of two-molecular subtype; Table 3, Column B), SDHB mutation (OR 5.68 (95% CI 1.79–18.06); P = 0.003) as well as norepinephrine (OR 3.01 (95% CI 1.02–8.79); P = 0.045) and dopamine (OR 6.39 (95% CI 1.62–25.24); P = 0.008) secretion but not WHO classification were associated with metastatic disease. In model 2 (exploring the role of three-molecular subtype; Table 3, Column C), dopamine secretion (OR 7.86 (95% CI 2.03–30.4), P = 0.003), PGL (OR 3.09 (95% CI 1.20–7.97); P = 0.019) but not the three-molecular subgroup system were associated with metastatic disease. In model 3 (exploring the role of four-molecular subtype; Table 3, Column D), norepinephrine (OR 3.12 (95% CI 1.02–9.56) P = 0.046), dopamine (OR 6.32 (95% CI 1.58–25.3) P = 0.009) but not the four-molecular classification system nor WHO classification showed association with metastasis. Thus, in the context of clinical characteristics, the only relevant molecular biomarker for predicting metastasis was categorization accordingly to SDHB mutation status.

Prognostic information

Median survival time for the entire cohort was 240 months (95% CI 202–not reached). Age at diagnosis (hazard ratio (HR) 1.04 (95% CI 1.01–1.05); P = 0.019), metastatic stage (HR 6.63 (95% CI 3.46–12.7); P < 0.001), PGL (HR 2.6 (95% CI 1.32–5.15); P = 0.006), SDHB mutation (HR (95% CI 1.32–5.94); P = 0.007), pseudohypoxia TCA cycle (HR 2.28 (95% CI 1.03–5.08); P = 0.043) and ATRX mutation (HR 9.44 (95% CI 3.29–27.15); P < 0.001) correlated with worse survival in univariate Cox regression analyses (Table 4 and Supplementary Fig. 4). In multivariate model 1 (exploring the role of two-molecular subtype; Table 4, Column B), age (HR 1.04 (95% CI 1.02–1.06); P = 0.001) and metastases (HR 6.13 (95% CI 2.86–13.13); P < 0.001) but not PGL nor SDHB mutation remained significant for survival. In multivariate model 2 (exploring the role of four-molecular subtype; Table 4, Column C), age (HR 1.04 (95% CI 1.02–1.06); P < 0.001) and metastases (HR 5.85 (95% CI 2.69–12.71); P < 0.001) but not PGL or categorization according to the four-molecular subgroup system remained significant for survival.

Table 4

Survival analysis.

 Median survival (months)Cox regression (univariate analysis)Cox regression (multivariable analysis)Cox regression (multivariable analysis)
(95% CI)HR (95% CI); P valueHR (95% CI); P valueHR (95% CI); P value
Model 1Model 2
GenderFemale240 (202–nr)1 (Ref)
MaleNr (!)1.14 (0.59–2.19); 0.686
Age at diagnosisContinuous variableNA1.02 (1.01–1.05); 0.0191.04 (1.02–1.06); 0.0011.04 (1.02–1.06); <0.001
Tumor size (mm)≤50240 (nr–nr)1 (Ref)
>50Nr (!)0.82 (0.34–1.99); 0.654
Stage Non-metastatic240 (202–nr)1 (Ref)1 (Ref)1 (Ref)
Metastatic 156 (84–nr)6.63 (3.46–12.70); <0.0016.13 (2.86–13.13); <0.0015.85 (2.69–12.71); <0.001
CatecholamineEpinephrine240 (nr–nr)1 (Ref)
NorepinephrineNr (192–nr)1.15 (0.39–3.35); 0.800
Dopamine168 (156–nr)3.06 (0.85–11.04); 0.088
WHO 2004PCC240 (202–nr)1 (Ref)1 (Ref)1 (Ref)
PGL192 (156–nr)2.60 (1.32–5.15); 0.0061.56 (0.64–3.81); 0.3321.47 (0.56–3.87); 0.440
WHO 2017PCC240 (202–nr)1 (Ref)**
Sympathetic PGL192 (117–nr)2.76 (1.37–5.58); 0.005**
Head and neck PGLNr (!)1.54 (0.21–11.53); 0.672**
ATRX mutation statusATRX mutated100 (3–nr)9.44 (3.29–27.15); <0.001**
ATRX wild typeNr (!)1 (ref)**
2-Molecular subgroupsSDHB wild type240 (202–nr)1 (Ref)1 (Ref)£
SDHB mutated168 (117–nr)2.80 (1.32–5.94); 0.0071.45 (0.47–4.44); 0.514£
3-Molecular subgroupsWnt/unknownNr (!)1 (Ref)
Pseudohypoxia202 (156–nr)1.66 (0.79–3.48); 0.180
Kinase signaling240 (nr–nr)0.66 (0.28–1.55); 0.335
4-Molecular subgroupsWnt/unknownNr (!)1 (Ref)$1 (Ref)
Pseudohypoxia TCA cycle168 (117–nr)2.28 (1.03–5.08); 0.043$1.43 (0.45–4.55); 0.543
Pseudohypoxia VHL/EPAS1202 (192–nr)0.86 (0.25–2.98); 0.814$0.88 (0.24–3.24); 0.851
Kinase signaling240 (nr–nr)0.66 (0.28–1.56); 0.341$0.79 (0.33–1.95); 0.616

ATRX mutation status correlated to survival in univariate analysis. But, it was not included in the multivariate model due to a lack of clinical annotations in cases that had analysis of ATRX mutation status.

!Could not be calculated due to lack of observations; *not included, WHO 2004 classification included in the multivariable analysis; $four-molecular subgroups system not included in model 1 of the multivariable analysis (two-molecular subgroups included instead); £two-molecular subgroups system not included in model 2 of the multivariable analysis (four-molecular subgroups classification included instead).

95% CI, 95% confidence interval; DA, dopamine; E, epinephrine; HNPGL, head and neck PGL; HR, hazard ratio; NA, not available; NE, norepinephrine; Nr, not reached; PCC, pheochromocytoma; PGL, paraganglioma; Ref, reference; sPGLs, sympathetic PGL; TCA, tricarboxylic acid.

A subgroup analysis of patients with metastatic disease (n = 57) did not show any clinical or molecular factors associated with survival in univariate Cox regression analysis (Supplementary Table 3). Even though there was a trend toward worse overall survival on patients with PPGLs classified as pseudohypoxia TCA cycle related and Wnt/unknown identified in Kaplan–Meier curves (Supplementary Fig. 5), such differences did not reach statistical significance due to limited power and number of events (log-rank test, P = 0.1620).

Discussion

We performed a meta-analysis on data from a systematic review of 703 PPGL patients published by 21 genome sequencing studies, and this is to our knowledge, the largest review in the literature. We focused on identifying predictive factors of metastatic disease, the major determinant of outcome from PPGL disease. While tumor location, biochemical phenotype and the driver gene classifications all showed different frequencies of metastatic disease in the univariate analyses, the only categorization accordingly to a driver gene that remained significant in the multivariate models was SDHB mutation status. In the univariate analyses, age, tumor location, metastatic disease, SDHB and TCA cycle-related PPGL showed difference in survival. But, no molecular information remained significant for survival in the multivariate model.

The aggregated frequency of driver mutations presented in our review was 62.2% – 24.6% in germline and 32.9% on the somatic level. This number is slightly lower than the frequencies observed in the included TCGA study (27% germline and 39% somatic driver mutations) that used the most comprehensive genetic analysis of all included studies (Fishbein et al. 2017). Major driver genes in the reviewed studies were NF1, VHL, RET, SDHB and HRAS that were mutated in 45.2% in of PPGL. A second group of driver genes, EPAS1, SDHD, SDHA, MAML3, MAX and TMEM127 occurred less frequently and had a cumulative frequency of 8.8%. A third group of genes were only found to be mutated in a minority of patients, cumulative frequency 2.8%: CSDE1, FGFR1, MET, SDHC, SDHAF2, FH and EGLN1. It should be noted that MAML3, CSDE1, FGFR1 and MET were recently discovered in this disease and were therefore only partially included in the sequencing analyses of the reviewed studies.

In order to correlate these findings to patient phenotype, we categorized PPGLs into subgroups accordingly to the biological hallmarks of the tumor as per driver mutation status. A novel category, Wnt/unknown, was created to allow for groups with adequate patient numbers for the statistical analyses. We recognize that Wnt/unknown represents a diverse group of PPGLs that is likely to be dissected as investigators employ more comprehensive methods for genome sequencing in a near future. Such improved categorization could include additional data on newly discovered PPGL driver genes, such as EGLN2 (Yang et al. 2015), SLC25A11 (Buffet et al. 2018), MDH2 (Cascon et al. 2015), DNMT3A (Remacha et al. 2018), H3F3A (Toledo et al. 2016) as well as information on disease-modifying genes related to telomere maintenance as well as chromatin modification (Fishbein et al. 2015, 2017, Job et al. 2018).

Tumor location, biochemical phenotype and molecular subgroup are three interconnected factors that are all known to be associated with PPGL metastasis (reviewed in Crona et al. 2017). Welander et al. reviewed the frequency of metastatic disease in patients with hereditary PPGL: RET, 2.9%; VHL, 3.4%; SDHD, 3.5%; and SDHB, 30.7% (Welander et al. 2011). A systematic review later showed that metastasis occurred in 17% of SDHB and 8% of SDHD carriers (van Hulsteijn et al. 2012). The findings in our review and meta-analysis corroborate these studies that define SDHB (46.6%), pseudohypoxia (24.3%) and pseudohypoxia TCA-cycle-related (40.5%) PPGL as having a relatively high risk of metastatic disease. Different from previous studies, PPGLs related to either VHL/EPAS1 (11.2%) or Wnt/unknown (11.4%) subgroups had an intermediate frequency of metastasis, whereas the kinase signaling subgroup was validated as having a relatively low frequency of metastatic disease (4.1%). However, only molecular categorization according to SDHB mutation status, but not other molecular systems or mutations, was associated with metastasis in the multivariate models.

SDHB has been proposed as a negative prognostic factor for survival in metastatic PPGL (Amar et al. 2007, Turkova et al. 2015, Assadipour et al. 2017). However, the recent study showed that SDHB mutation was not an independent prognostic factor for survival in a large cohort of metastatic PPGL (Hescot et al. 2019). Although our survival analysis did not show significant results for the molecular subgroups, Kaplan–Meier curves clearly indicate trend toward worse outcome on both TCA cycle and Wnt/unknown-related PPGL. Remarkably, no deaths occurred in patients with pseudohypoxia VHL/EPAS1 as well as kinase signaling PPGLs. This information must be considered with caution since the number of events was very low but indicate that patient stratification beyond SDHB mutation status should be investigated in future studies.

Our review and analysis has a number of limitations: clinical annotations in general, and hormone evaluations in particular, showed a high risk of bias. Lack of data on ATRX inactivation or TERT expression is also a relevant limitation as it has been associated with higher frequency of metastasis as well as poor survival (Fishbein et al. 2015, 2017, Job et al. 2018). Selection bias is also likely as a majority of reviewed manuscripts comes from well-recognized groups at tertiary centers. Another bias may have been incorporated from our exclusion of patients without available tumor tissue, which could include a selection bias that excludes a relevant subgroup of patients (Roman-Gonzalez et al. 2018). The analysis of survival in the whole study cohort is likely skewed by the higher age in patients with sporadic PPGLs, which are less likely to have metastasis, compared to the pseudohypoxia group that develop disease earlier mainly due to genetic predisposition. Disease-related survival would be a preferred measurement, even though it could not be explored due to lack of data. Finally, there was a significant loss of patients for the multivariate analysis due to incomplete clinical annotations, which the subsequent limited statistical power that this implies.

Our findings demonstrated SDHB as independently associated with PPGL metastasis and do not favor the use of information on other driver genes as it was not independently correlated to metastatic disease. Due to relatively low patient number and various risks of bias, we predict that the observed trends for both metastasis and survival still indicate that there is a potential of molecular information to yield relevant information on PPGL outcome in future. To test this hypothesis large, preferably prospective, series with very complete clinical and genetic annotation will be required (Kimura et al. 2014, Turkova et al. 2015, Koh et al. 2017).

Conclusion

Our review and individual patient meta-analysis validated previous phenotype correlations including different frequencies of metastasis in-between PPGL driver genes. However, only SDHB mutation status remained significant in the multivariate model. Instead, the biochemical profile including dopamine secretion emerged as a more useful predictor of metastatic disease. Categorization accordingly to a driver gene mutation was not an independent factor associated with survival in this study.

Supplementary data

This is linked to the online version of the paper at https://doi.org/10.1530/ERC-19-0024.

Declaration of interest

J C received lecture honoraria from Novartis. The other authors have nothing to disclose.

Funding

This work was supported by grants from Akademiska Sjukhuset, Uppsala, the Paradifference foundation (http://www.paradifference.org) and Lions Cancerforskningsfond, Uppsala and by the National Cancer Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Angela Lamarca was partially funded by the ASCO Conquer Cancer Foundation Young Investigator Award.

References

  • Amar L, Baudin E, Burnichon N, Peyrard S, Silvera S, Bertherat J, Bertagna X, Schlumberger M, Jeunemaitre X, Gimenez-Roqueplo AP, et al.2007 2007 Succinate dehydrogenase B gene mutations predict survival in patients with malignant pheochromocytomas or paragangliomas. Journal of Clinical Endocrinology and Metabolism 92 38223828. (https://doi.org/10.1210/jc.2007-0709)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Assadipour Y, Sadowski SM, Alimchandani M, Quezado M, Steinberg SM, Nilubol N, Patel D, Prodanov T, Pacak K & Kebebew E 2017 SDHB mutation status and tumor size but not tumor grade are important predictors of clinical outcome in pheochromocytoma and abdominal paraganglioma. Surgery 161 230239. (https://doi.org/10.1016/j.surg.2016.05.050)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Ayala-Ramirez M, Feng L, Johnson MM, Ejaz S, Habra MA, Rich T, Busaidy N, Cote GJ, Perrier N, Phan A, et al.2011 Clinical risk factors for malignancy and overall survival in patients with pheochromocytomas and sympathetic paragangliomas: primary tumor size and primary tumor location as prognostic indicators. Journal of Clinical Endocrinology and Metabolism 96 717725. (https://doi.org/10.1210/jc.2010-1946)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bausch B, Schiavi F, Ni Y, Welander J, Patocs A, Ngeow J, Wellner U, Malinoc A, Taschin E, Barbon G, et al.2017 Clinical characterization of the pheochromocytoma and paraganglioma susceptibility genes SDHA, TMEM127, MAX, and SDHAF2 for gene-informed prevention. JAMA Oncology 3 12041212. (https://doi.org/10.1001/jamaoncol.2017.0223)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Buffet A, Morin A, Castro-Vega LJ, Habarou F, Lussey-Lepoutre C, Letouze E, Lefebvre H, Guilhem I, Haissaguerre M, Raingeard I, et al.2018 Germline mutations in the mitochondrial 2-oxoglutarate/malate carrier SLC25A11 gene confer a predisposition to metastatic paragangliomas. Cancer Research 78 19141922. (https://doi.org/10.1158/0008-5472.CAN-17-2463)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Burnichon N, Vescovo L, Amar L, Libe R, de Reynies A, Venisse A, Jouanno E, Laurendeau I, Parfait B, Bertherat J, et al.2011 Integrative genomic analysis reveals somatic mutations in pheochromocytoma and paraganglioma. Human Molecular Genetics 20 39743985. (https://doi.org/10.1093/hmg/ddr324)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Burnichon N, Buffet A, Parfait B, Letouze E, Laurendeau I, Loriot C, Pasmant E, Abermil N, Valeyrie-Allanore L, Bertherat J, et al.2012 Somatic NF1 inactivation is a frequent event in sporadic pheochromocytoma. Human Molecular Genetics 6 6.

    • Search Google Scholar
    • Export Citation
  • Cascon A, Comino-Mendez I, Curras-Freixes M, de Cubas AA, Contreras L, Richter S, Peitzsch M, Mancikova V, Inglada-Perez L, Perez-Barrios A, et al.2015 Whole-exome sequencing identifies MDH2 as a new familial paraganglioma gene. Journal of the National Cancer Institute 107 djv053. (https://doi.org/10.1093/jnci/djv053)

    • PubMed
    • Search Google Scholar
    • Export Citation
  • Castro-Vega LJ, Buffet A, De Cubas AA, Cascon A, Menara M, Khalifa E, Amar L, Azriel S, Bourdeau I, Chabre O, et al.2014 Germline mutations in FH confer predisposition to malignant pheochromocytomas and paragangliomas. Human Molecular Genetics 23 24402446. (https://doi.org/10.1093/hmg/ddt639)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Castro-Vega LJ, Letouze E, Burnichon N, Buffet A, Disderot PH, Khalifa E, Loriot C, Elarouci N, Morin A, Menara M, et al.2015 Multi-omics analysis defines core genomic alterations in pheochromocytomas and paragangliomas. Nature Communications 6 6044. (https://doi.org/10.1038/ncomms7044)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Cho YY, Kwak MK, Lee SE, Ahn SH, Kim H, Suh S, Kim BJ, Song KH, Koh JM, Kim JH, et al.2018 A clinical prediction model to estimate the metastatic potential of pheochromocytoma/paraganglioma: ASES score. Surgery 164 511517. (https://doi.org/10.1016/j.surg.2018.05.001)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Crona J, Taieb D & Pacak K 2017 New perspectives on pheochromocytoma and paraganglioma: toward a molecular classification. Endocrine Reviews 38 489515. (https://doi.org/10.1210/er.2017-00062)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Curras-Freixes M, Inglada-Perez L, Mancikova V, Montero-Conde C, Leton R, Comino-Mendez I, Apellaniz-Ruiz M, Sanchez-Barroso L, Aguirre Sanchez-Covisa M, Alcazar V, et al.2015 Recommendations for somatic and germline genetic testing of single pheochromocytoma and paraganglioma based on findings from a series of 329 patients. Journal of Medical Genetics 52 647656. (https://doi.org/10.1136/jmedgenet-2015-103218)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Dwight T, Flynn A, Amarasinghe K, Benn DE, Lupat R, Li J, Cameron DL, Hogg A, Balachander S, Candiloro ILM, et al.2018 Tert structural rearrangements in metastatic pheochromocytomas. Endocrine-Related Cancer 25 19. (https://doi.org/10.1530/ERC-17-0306)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Eisenhofer G, Lenders JW, Timmers H, Mannelli M, Grebe SK, Hofbauer LC, Bornstein SR, Tiebel O, Adams K, Bratslavsky G, et al.2011a Measurements of plasma methoxytyramine, normetanephrine, and metanephrine as discriminators of different hereditary forms of pheochromocytoma. Clinical Chemistry 57 411420. (https://doi.org/10.1373/clinchem.2010.153320)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eisenhofer G, Pacak K, Huynh TT, Qin N, Bratslavsky G, Linehan WM, Mannelli M, Friberg P, Grebe SK, Timmers HJ, et al.2011b Catecholamine metabolomic and secretory phenotypes in phaeochromocytoma. Endocrine-Related Cancer 18 97111. (https://doi.org/10.1677/ERC-10-0211)

    • Search Google Scholar
    • Export Citation
  • Eisenhofer G, Lenders JW, Siegert G, Bornstein SR, Friberg P, Milosevic D, Mannelli M, Linehan WM, Adams K, Timmers HJ, et al.2012 Plasma methoxytyramine: a novel biomarker of metastatic pheochromocytoma and paraganglioma in relation to established risk factors of tumour size, location and SDHB mutation status. European Journal of Cancer 48 17391749. (https://doi.org/10.1016/j.ejca.2011.07.016)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Favier J, Buffet A & Gimenez-Roqueplo AP 2012 HIF2A mutations in paraganglioma with polycythemia. New England Journal of Medicine 367 21612162. (https://doi.org/10.1056/NEJMc1211953)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fishbein L, Khare S, Wubbenhorst B, DeSloover D, D'Andrea K, Merrill S, Cho NW, Greenberg RA, Else T, Montone K, et al.2015 Whole-exome sequencing identifies somatic ATRX mutations in pheochromocytomas and paragangliomas. Nature Communications 6 6140. (https://doi.org/10.1038/ncomms7140)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Fishbein L, Leshchiner I, Walter V, Danilova L, Robertson AG, Johnson AR, Lichtenberg TM, Murray BA, Ghayee HK, Else T, et al.2017 Comprehensive molecular characterization of pheochromocytoma and paraganglioma. Cancer Cell 31 181193. (https://doi.org/10.1016/j.ccell.2017.01.001)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Fliedner SM, Shankavaram U, Marzouca G, Elkahloun A, Jochmanova I, Daerr R, Linehan WM, Timmers H, Tischler AS, Papaspyrou K, et al.2016 Hypoxia-inducible factor 2alpha mutation-related paragangliomas classify as discrete pseudohypoxic subcluster. Neoplasia 18 567576. (https://doi.org/10.1016/j.neo.2016.07.008)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Flynn A, Benn D, Clifton-Bligh R, Robinson B, Trainer AH, James P, Hogg A, Waldeck K, George J, Li J, et al.2015 The genomic landscape of phaeochromocytoma. Journal of Pathology 236 7889. (https://doi.org/10.1002/path.4503)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Flynn A, Dwight T, Harris J, Benn D, Zhou L, Hogg A, Catchpoole D, James P, Duncan EL, Trainer A, et al.2016 Pheo-type: a diagnostic gene-expression assay for the classification of pheochromocytoma and paraganglioma. Journal of Clinical Endocrinology and Metabolism 101 10341043. (https://doi.org/10.1210/jc.2015-3889)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Flynn A, Dwight T, Benn D, Deb S, Colebatch AJ, Fox S, Harris J, Duncan EL, Robinson B, Hogg A, et al.2017 Cousins not twins: intratumoural and intertumoural heterogeneity in syndromic neuroendocrine tumours. Journal of Pathology 242 273283. (https://doi.org/10.1002/path.4900)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hamidi O, Young WF Jr, Gruber L, Smestad J, Yan Q, Ponce OJ, Prokop L, Murad MH & Bancos I 2017a Outcomes of patients with metastatic phaeochromocytoma and paraganglioma: a systematic review and meta-analysis. Clinical Endocrinology 87 440450. (https://doi.org/10.1111/cen.13434)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hamidi O, Young WF Jr, Iniguez-Ariza NM, Kittah NE, Gruber L, Bancos C, Tamhane S, Bancos I & Bancos I 2017b Malignant pheochromocytoma and paraganglioma 2017 malignant pheochromocytoma and paraganglioma: 272 patients over 55 years. Journal of Clinical Endocrinology and Metabolism 102 32963305. (https://doi.org/10.1210/jc.2017-00992)

    • Search Google Scholar
    • Export Citation
  • Hescot S, Curras-Freixes M, Deutschbein T, van Berkel A, Vezzosi D, Amar L, de la Fouchardiere C, Valdes N, Riccardi N, Do Cao C, et al.2019 Prognosis of malignant pheochromocytoma and paraganglioma (MAPP-prono study): an ENS@T retrospective study. Journal of Clinical Endocrinology and Metabolism [epub]. (https://doi.org/10.1210/jc.2018-01968)

    • Search Google Scholar
    • Export Citation
  • Job S, Draskovic I, Burnichon N, Buffet A, Cros J, Lepine C, Venisse A, Robidel E, Verkarre V, Meatchi T, et al.2018 Telomerase activation and ATRX mutations are independent risk factors for metastatic pheochromocytoma and paraganglioma. Clinical Cancer Research 25 760770. (https://doi.org/10.1158/1078-0432.CCR-18-0139)

    • Search Google Scholar
    • Export Citation
  • Juhlin CC, Stenman A, Haglund F, Clark VE, Brown TC, Baranoski J, Bilguvar K, Goh G, Welander J, Svahn F, et al.2015 Whole-exome sequencing defines the mutational landscape of pheochromocytoma and identifies KMT2D as a recurrently mutated gene. Genes, Chromosomes and Cancer 54 542554. (https://doi.org/10.1002/gcc.22267)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kimura N, Takayanagi R, Takizawa N, Itagaki E, Katabami T, Kakoi N, Rakugi H, Ikeda Y, Tanabe A, Nigawara T, et al.2014 Pathological grading for predicting metastasis in phaeochromocytoma and paraganglioma. Endocrine-Related Cancer 21 405414. (https://doi.org/10.1530/ERC-13-0494)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Koh JM, Ahn SH, Kim H, Kim BJ, Sung TY, Kim YH, Hong SJ, Song DE & Lee SH 2017 Validation of pathological grading systems for predicting metastatic potential in pheochromocytoma and paraganglioma. PLoS ONE 12 e0187398. (https://doi.org/10.1371/journal.pone.0187398)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Letouze E, Martinelli C, Loriot C, Burnichon N, Abermil N, Ottolenghi C, Janin M, Menara M, Nguyen AT, Benit P, et al.2013 SDH mutations establish a hypermethylator phenotype in paraganglioma. Cancer Cell 23 739752. (https://doi.org/10.1016/j.ccr.2013.04.018)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J & Moher D 2009 The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Medicine 6 e1000100. (https://doi.org/10.1371/journal.pmed.1000100)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Neumann HP, Young WF Jr, Krauss T, Bayley JP, Schiavi F, Opocher G, Boedeker CC, Tirosh A, Castinetti F, Ruf J, et al.2018 65 YEARS OF THE DOUBLE HELIX: Genetics informs precision practice in the diagnosis and management of pheochromocytoma. Endocrine-Related Cancer 25 T201T219. (https://doi.org/10.1530/ERC-18-0085)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Remacha L, Curras-Freixes M, Torres-Ruiz R, Schiavi F, Torres-Perez R, Calsina B, Leton R, Comino-Mendez I, Roldan-Romero JM, Montero-Conde C, et al.2018 Gain-of-function mutations in DNMT3A in patients with paraganglioma. Genetics in Medicine 20 16441651. (https://doi.org/10.1038/s41436-018-0003-y)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roman-Gonzalez A, Zhou S, Ayala-Ramirez M, Shen C, Waguespack SG, Habra MA, Karam JA, Perrier N, Wood CG & Jimenez C 2018 Impact of surgical resection of the primary tumor on overall survival in patients with metastatic pheochromocytoma or sympathetic paraganglioma. Annals of Surgery 268 172178. (https://doi.org/10.1097/SLA.0000000000002195)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Stenman A, Juhlin CC, Haglund F, Brown TC, Clark VE, Svahn F, Bilguvar K, Goh G, Korah R, Lifton RP, et al.2016a Absence of KMT2D/MLL2 mutations in abdominal paraganglioma. Clinical Endocrinology 84 632634. (https://doi.org/10.1111/cen.12884)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stenman A, Welander J, Gustavsson I, Brunaud L, Backdahl M, Soderkvist P, Gimm O, Juhlin CC & Larsson C 2016b HRAS mutation prevalence and associated expression patterns in pheochromocytoma. Genes, Chromosomes and Cancer 55 452459. (https://doi.org/10.1002/gcc.22347)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Timmers HJ, Brouwers FM, Hermus AR, Sweep FC, Verhofstad AA, Verbeek AL, Pacak K & Lenders JW 2008 Metastases but not cardiovascular mortality reduces life expectancy following surgical resection of apparently benign pheochromocytoma. Endocrine-Related Cancer 15 11271133. (https://doi.org/10.1677/ERC-08-0049)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Toledo RA, Qin Y, Cheng ZM, Gao Q, Iwata S, Silva GM, Prasad ML, Ocal IT, Rao S, Aronin N, et al.2016 Recurrent mutations of chromatin-remodeling genes and kinase receptors in pheochromocytomas and paragangliomas. Clinical Cancer Research 22 23012310. (https://doi.org/10.1158/1078-0432.CCR-15-1841)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turkova H, Prodanov T, Maly M, Martucci V, Adams K, Widimsky J Jr, Chen CC, Ling A, Kebebew E, Stratakis C, et al.2015 Characteristics and outcomes of metastatic SDHB and sporadic pheochromocytoma/paraganglioma: an NIH study. Endocrine Practice 22 302314.

    • Search Google Scholar
    • Export Citation
  • van Hulsteijn LT, Dekkers OM, Hes FJ, Smit JW & Corssmit EP 2012 Risk of malignant paraganglioma in SDHB-mutation and SDHD-mutation carriers: a systematic review and meta-analysis. Journal of Medical Genetics 25 25. (https://doi.org/10.1136/jmedgenet-2012-101192)

    • Search Google Scholar
    • Export Citation
  • Welander J, Soderkvist P & Gimm O 2011 Genetics and clinical characteristics of hereditary pheochromocytomas and paragangliomas. Endocrine-Related Cancer 18 R253R276. (https://doi.org/10.1530/ERC-11-0170)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Welander J, Larsson C, Backdahl M, Hareni N, Sivler T, Brauckhoff M, Soderkvist P & Gimm O 2012 Integrative genomics reveals frequent somatic NF1 mutations in sporadic pheochromocytomas. Human Molecular Genetics 21 54065416. (https://doi.org/10.1093/hmg/dds402)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Welander J, Andreasson A, Brauckhoff M, Backdahl M, Larsson C, Gimm O & Soderkvist P 2014a Frequent EPAS1/HIF2alpha exons 9 and 12 mutations in non-familial pheochromocytoma. Endocrine-Related Cancer 21 495504. (https://doi.org/10.1530/ERC-13-0384)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Welander J, Andreasson A, Juhlin CC, Wiseman RW, Backdahl M, Hoog A, Larsson C, Gimm O & Soderkvist P 2014b Rare germline mutations identified by targeted next-generation sequencing of susceptibility genes in pheochromocytoma and paraganglioma. Journal of Clinical Endocrinology and Metabolism 99 E1352E1360. (https://doi.org/10.1210/jc.2013-4375)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Welander J, Lysiak M, Brauckhoff M, Brunaud L, Soderkvist P & Gimm O 2018 Activating FGFR1 mutations in sporadic pheochromocytomas. World Journal of Surgery 42 482489. (https://doi.org/10.1007/s00268-017-4320-0)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Wilzen A, Rehammar A, Muth A, Nilsson O, Tesan Tomic T, Wangberg B, Kristiansson E & Abel F 2016 Malignant pheochromocytomas/paragangliomas harbor mutations in transport and cell adhesion genes. International Journal of Cancer 138 22012211. (https://doi.org/10.1002/ijc.29957)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • Yang C, Zhuang Z, Fliedner SM, Shankavaram U, Sun MG, Bullova P, Zhu R, Elkahloun AG, Kourlas PJ, Merino M, et al.2015 Germ-line PHD1 and PHD2 mutations detected in patients with pheochromocytoma/paraganglioma-polycythemia. Journal of Molecular Medicine 93 93104. (https://doi.org/10.1007/s00109-014-1205-7)

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zelinka T, Musil Z, Duskova J, Burton D, Merino MJ, Milosevic D, Widimsky J Jr & Pacak K 2011 Metastatic pheochromocytoma: does the size and age matter? European Journal of Clinical Investigation 41 11211128. (https://doi.org/10.1111/j.1365-2362.2011.02518.x)

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation

 

Society for Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 1886 0 0
Full Text Views 1791 1017 48
PDF Downloads 931 600 47
  • View in gallery

    Clinical correlations to modified molecular subgroups from the Cancer Genome Atlas. Epi, epinephrine; kinase, kinase signaling subgroup; Norepi, norepinephrine; PCC, pheochromocytoma; PGL, paraganglioma; TCA, tricarboxylic acid; UK, unknown; VHL/EPAS1, pseudohypoxia VHL/EPAS1 related. ***Chi-square test including all four molecular subgroups had a significance level <0.001.