CHARACTERISTICS AND OUTCOMES OF METASTATIC SDHB AND SPORADIC PHEOCHROMOCYTOMA/PARAGANGLIOMA: AN NATIONAL INSTITUTES OF HEALTH STUDY

Abstract

Objective

Overall about 10 to 20% of pheochromocytomas/paragangliomas (PHEOs/PGLs) are metastatic, with higher metastatic potential observed in succinate dehydrogenase subunit B/ fumarate hydratase (SDHB/FH)-related tumors. Due to the improved availability of biochemical and genetic testing and the frequent use of anatomical/functional imaging, there is currently a higher detection rate of metastatic PHEO/PGL.

Methods

A retrospective analysis of 132 patients (27 children, 105 adults) with metastatic PHEO/PGL diagnosed and treated from 2000 to 2014 was conducted.

Results

Seventy-seven (58%) males and 55 (42%) females were included; 39 (30%) have died, with no sex preference. Seventy-three (55%) patients had SDHB mutations; 59 (45%) patients had apparently sporadic tumors (AST). SDHB patients had an average age at primary tumor diagnosis of 31 ± 16 years compared to 40 ± 15 years in AST patients (P<.001). The average metastatic interval (MI) decreased with increasing age in both SDHB and AST patients (P = .013 for both). Only 16% of all primary tumors were smaller than 4.5 cm. Eleven percent of patients had biochemically silent disease, more with SDHB. Of SDHB patients, 23% had metastatic tumors at first diagnosis, compared to 15% of AST patients. Five- and 10-year survival rates were significantly better for metastatic AST than SDHB patients (P = .01). Overall survival was significantly different between children and adults (P = .037); this was mostly attributed to the SDHB patients, in whom children had statistically significantly longer survival than adults (P = .006). The deceased patients all died due to the PHEO/PGL and mainly had noradrenergic phenotypes.

Conclusion

In children, metastatic PHEOs/PGLs are mainly due to SDHB mutations; in adults they are equally distributed between in SDHB mutations and AST, with better 5- and 10-year survival rates for ASTs. In SDHB patients, children survive longer than adults. Primary metastatic tumors, most presenting as noradrenergic PGLs, are larger than 4.5 cm in >80% of patients. The frequency of metastatic tumors from primary AST increases with age, including a decreased MI compared to SDHB tumors. These results support several recommendations that are summarized in the Discussion.

INTRODUCTION

The 2004 World Health Organization (WHO) classification of endocrine tumors defines pheochromocytoma (PHEO) as a tumor arising from catecholamine-producing chromaffin cells in the adrenal medulla – an intra-adrenal paraganglioma (PGL) (1). Closely related tumors of the extra-adrenal sympathetic and parasympathetic paraganglia are classified as extra-adrenal PGLs.

Currently, about 35 to 40% of PHEOs/PGLs occur in the context of several major inherited tumor syndromes, mainly including multiple endocrine neoplasia type 2 (MEN2, with rearranged during transfection (RET) mutations), von Hippel-Lindau (VHL) disease (caused by mutations in VHL), neurofibromatosis type 1 (NF1, with NF1 mutations), and mutations in succinate dehydrogenase subunits (SDHx) (2–4). SDHx syndromes are associated with four different genes that encode subunits A, B, C, and D (5,6). Germline mutations in SDHAF2 (encoding SDH complex assembly factor 2) have been linked to an inherited tumor syndrome (7). Newly discovered mutations in transmembrane protein 127 (TMEM127) and myc-associated factor X (MAX) genes are also associated with PHEO/PGL, currently without a specific syndromic presentation (8–10). Most recently, somatic gain-of-function mutations in the gene encoding hypoxia-inducible factor 2α (HIF2A) have been discovered, mainly in patients presenting with a new syndrome of multiple PGLs and duodenal somatostatinomas associated with polycythemia in females (11–15). Five pathogenic germline mutations in the FH gene encoding fumarate hydratase were identified. These reveal a new role for FH in susceptibility to malignant and/or multiple PHEO/PGL (16,17).

The WHO classification defines the presence of metastases, but not local invasion, as the only accepted criterion for the diagnosis of metastatic PHEO/PGL (1). Metastatic lesions are found at sites where chromaffin tissue is normally absent (e.g., the lymph nodes, bone, lung, and liver). The prevalence of malignancy in PHEOs/PGLs is reported to range from about 2 to 26% (18,19). SDHB mutations are associated with more aggressive tumor behavior and a higher rate of malignancy than other PHEO/PGL types, in some studies with rates up to 50 to 90% (20–25). Mutations in the MAX and FH genes were also found to be associated with a higher rate of metastatic PHEO/PGL (26,26).

In addition to the presence of SDHB mutations, several other independent factors exist that seem to be closely associated with metastatic behavior of primary tumors, including their location (extra-adrenal), size (mainly over 5–6 cm), the presence of high levels of methoxytyramine, and likely some histopathological criteria (e.g., necrosis, a high mitotic rate) (19, 27–30). Currently, surgery is the only curative treatment for some patients with metastatic disease (31–34).

At present, due to the improved availability of state-of-the-art biochemical (plasma or urine metanephrines [MNs] and plasma methoxytyramine) and genetic testing, as well as the frequent use of anatomical (computed tomography [CT] or magnetic resonance imaging [MRI]), functional (positron emission tomography [PET]), or other nuclear medicine imaging modalities, patients with metastatic PHEO/PGL are detected more frequently than in the past (35). More information is available to these patients as well as to physicians, especially oncologists. Thus, to provide medical professionals with updated and clinically useful information and recommendations about these tumors, a large population of patients, preferably from centers with extensive experience in this disease, is desirable to make such recommendations.

The present study summarizes 15 years of experience with 132 patients with metastatic PHEO/PGL diagnosed or referred to the National Institutes of Health (NIH). The present study was designed to provide a comprehensive and updated clinical, genetic, and biochemical characterization of metastatic PHEOs/PGLs.

METHODS

Patients

One hundred thirty-two subjects with metastatic PHEO/PGL were seen consecutively at the NIH in Bethesda, Maryland between 2000 and 2014, and their data were retrospectively reviewed. All patients provided written informed consent for clinical, genetic, biochemical, imaging, and treatment studies under Eunice Kennedy Shriver National Institute of Child Health and Human Development/NIH protocol 00-CH-0093.

For more detailed evaluation of trends, changes, and outcomes, we divided our patients into 2 main subgroups: (1) children (0–19 years) and (2) adults (20 years and older). The adult group was again divided into 3 subcategories: (a) young adults (20–30 years), (b) adults (31–50 years), and (c) older adults (≥51 years).

Genetics of PHEO/PGL

Genetic testing for mutations in major PHEO/PGL susceptibility genes (SDHB, SDHC, SDHD, VHL, RET, MAX, and TMEM127) was performed at either the NIH, Mayo Medical Laboratories (Rochester, MN), the Division of Molecular Diagnostics at the University of Pittsburgh Medical Center, or the Department of Genetics of the Children’s Hospital of Philadelphia (36). NF1 was diagnosed based on the presence of typical clinical features. Genetic testing was performed by polymerase chain reaction-based bidirectional sequencing of the 8 coding exons and adjacent intronic regions as well as portions of the 5’-untranslated and 3’-untranslated regions of the SDHB gene by the Division of Molecular Diagnostics at the University of Pittsburgh Medical Center (Pittsburgh, PA) (37). Genotyping was performed also in collaboration with the Mayo Clinic, Rochester, MN, and the Department of Genetics, the European Georges Pompidou Hospital, Paris, France as previously published (38).

Biochemical Testing and Diagnosis

The biochemical phenotype of the tumors was characterized based on increased catecholamines (epinephrine [EPI], norepinephrine [NE], or dopamine [DA]) or MNs (NMN or MN) in plasma and/or urine, as well as on plasma methoxytyramine, when available. At the NIH, plasma catecholamines and MNs were measured by liquid chromatography with electrochemical detection, as previously described (39–41). The reference intervals for plasma concentrations were established at the NIH, as previously described (28,41).

The confirmation of the diagnosis of metastatic PHEO/PGL was based on the clinical presentation and positive biochemistry or histopathologic examination of tumors and a previous history of PHEO/PGL. Positivity on PHEO/PGL-specific functional imaging studies further supported the diagnosis of metastatic PHEO/PGL (42).

Localization of the Primary Tumor and Metastases

CT and MRI from the neck to the pelvis were carried out in all patients. Most of the patients underwent PET or PET/CT scanning using 2 different radiopharmaceuticals and/or [^123/131I]-MIBG scintigraphy. CT scans of the neck, chest, abdomen, and pelvis were performed using initially single-channel or multichannel spiral CT machines (GE Healthcare Technologies, Chicago, IL), and since early 2003 exclusively multichannel helical CT equipment (General Electric Healthcare Technologies; Philips Medical Systems, Amsterdam, the Netherlands; Siemens Medical Solutions, Munich, Germany). MRI scans of the neck, chest, abdomen, and pelvis were obtained with 1.5- or 3- Tesla scanners (General Electric Healthcare Technologies and Philips Medical Systems). For PET and PET/CT scanning, the patients fasted for at least 4 hours before intravenous (i.v.) injection of [¹⁸F]-fluorodopamine ([¹⁸F]-FDA) (1 mCi) or [¹⁸F]-fluorodeoxyglucose ([¹⁸F]-FDG) (15 mCi). [¹⁸F]-FDA PET scans performed before March 2005 used an Advance Scanner (General Electric Medical Systems) with a 15-cm axial field of view. Subsequent [¹⁸F]-FDA and all [¹⁸F]-FDG scans were done using a PET/CT scanner (Siemens) with a 15 cm axial field of view. For [^123/131I]-MIBG scanning, patients were imaged at 24 h (and 48 h in some cases) following i.v. administration of 10 mCi (370 MBq) of [¹²³I]-MIBG or 0.5 mCi (18.5 MBq) of [¹³¹I]-MIBG.

Data Analysis

Continuous data are presented as arithmetic means and SDs. Group comparisions were based on Student’s 2-sample t tests, Mann-Whitney U tests, and analysis of variance followed by Sidak’s test for multiple comparisons. A linear regression model was used to investigate the relationship between tumor size and a set of potential predictors. Categorical data are given as absolute and relative frequencies (percentages). The differences in proportions between groups were analyzed using Fisher exact test and its generalization.

The probability of survival after PHEO/PGL surgery was estimated using the Kaplan–Meier method, and the equality of survivor functions was tested by a log-rank test. Cox’s regression model was used to assess the influence of treatment on survival. All statistical tests were treated as 2-sided and evaluated at a significance level of 0.05. Statistical analysis was performed by Stata statistical software, release 9.2 (Stata Corp LP, College Station, TX, USA).

Follow-up

All patients with metastatic PHEO/PGL were followed up every 3 to 12 months with clinical, biochemical, and imaging evaluations followed by treatment recommendations and/or follow-up plans. Moreover, all patients were also contacted by phone between November 2012 and March 2015 to ensure that the most accurate data is presented in this report. Deaths were recorded until March 2015 and included in the final statistics in this report. All reported deaths were due to metastatic PHEO/PGL.

RESULTS

Patient Demographics

The present study included 132 patients with metastatic PHEO/PGL, 77 (58%) males and 55 (42%) females. Thirty-nine (29.5%) patients died between 2000 and March 2015, 21 (54%) males and 18 (46%) females. At initial presentation, there were 98 (74%) patients with solitary tumors; 11 (8%) with bilateral adrenal, neck, or multiple primary tumors; and 23 (18%) with metastatic tumors. The patients include both those with a primary diagnosis and those with recurrence, some diagnosed before 2000 but subsequently referred to the NIH.

Hereditary and Apparently Sporadic Metastatic PHEOs/PGLs

Overall, 73 (55%) patients had mutations in SDHB, and 59 (45%) patients presented with apparently sporadic tumors (AST), no mutations in SDHC, SDHD, RET, or VHL were included. Mutations in other genes were not investigated. In children, 22 (81%) had mutations in SDHB (90% were PGLs); in adults, 51 (49%) had mutations in SDHB (80% were PGLs). The percentage of SDHB tumors in children was significantly higher than in adults (P<.001, Table 1).

Table 1.

Characteristics of Primary and Metastatic Tumors by Age at Primary Diagnosis, Location, and Sex

	S	Children						Adults
Total patient numbers		All		SDHB		AST								20–30 years				31–50 years				≥51 years
								All		SDHB		AST		SDHB		AST		SDHB		AST		SDHB		AST
	All	27		22		5		105		51		54		16		7		28		33		7		14
	M	20		18		2		57		30		27		11		2		15		18		4		7
	F	7		4		3		48		21		27		5		5		13		15		3		7
Age ± SD
APTD	All	12 ± 4		12 ± 4		13 ± 2		41 ± 13		38 ± 12		43 ± 12		25 ± 3		24 ± 3		40 ± 6		40 ± 5		60 ± 8		59 ± 9
AMTD	All	21 ± 10		19 ± 8		29 ± 13		46 ± 13		42 ± 12		49 ± 12		30 ± 8		32 ± 6		44 ± 7		47 ± 8		61 ± 8		62 ± 9
MI	All	9 ± 9		7 ± 7		15 ± 14		5 ± 6		4 ± 5		6 ± 6		5 ± 7		8 ± 6		4 ± 4		7 ± 7		1 ± 1		3 ± 5
Site		A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA	A	EA
	All	1	26	1	21	0	5	37	68	9	42	28	26	3	13	2	5	6	22	19	14	0	7	7	7
	M	1	19	1	17	0	2	19	38	5	25	14	13	3	8	0	2	2	13	10	8	0	4	4	3
	F	0	7	0	4	0	3	18	30	4	17	14	13	0	5	2	3	4	9	9	6	0	3	3	4

Abbreviations: A = adrenal; AMTD = age of metastatic tumor diagnosis; APTD = age of primary tumor diagnosis; AST = apparently sporadic tumors; EA = extra-adrenal; F = female, M = male; MI = metastatic interval; S = sex; SDHB = succinate dehydrogenase subunit B.

Age-Related Characteristics of Primary and Metastatic PHEOs/PGLs

The overall age of patients at the time of the primary tumor diagnosis (APTD) was 35 ± 16 years, and the age at the initial metastatic tumor diagnosis (AMTD) was 41 ± 16 years. The overall length of time between the primary tumor and metastatic disease (metastatic interval: MI) was 6 ± 7 years. There were no statistical differences between males and females (Table 1). Table 1 shows the APTD, AMTD, and MI in children and adults. The MI in children (9 ± 9 years) was statistically longer than in adults (5 ± 6 years; P = .010).

When we compared the SDHB and AST groups, the APTD (31 ± 16 vs. 40 ± 15 years, P<.001) and AMTD (35 ± 15 vs. 48 ± 13 years, P<.001) differed significantly, while the MI did not (5 ± 6 vs. 7 ± 7 years, P = .111). No differences were seen between sexes or between children and adults (Table 1).

Subgroup analysis showed that the average MI significantly decreased with increasing age in both the SDHB (P = .013) and AST groups (P = .013), calculated by the Jonckheere-Terpstra test for Ordered Alternatives. There were no statistical differences between sexes (Table 1).

Overall the proportion of primary EA versus primary A tumors was as follows: children: EA: 26 (96%), A: 1 (4%) and adults: EA: 68 (65%), A: 37 (35%); P = .003. For AST: children: EA: 5 (100%), A: 0% and adults: EA: 26 (48%), A: 28 (52%); P = .050. For SDHB tumors: children: EA: 21 (95%), A: 1 (5%) and adults: EA: 42 (82%), A: 9 (18%); P = .040 (Table 1).

Primary Tumor Size

The size of the primary tumor has recently been introduced as an important independent risk factor for metastatic disease, with tumors 6 cm or larger more frequently associated with the development of metastatic disease (43). Recently, Schovanek et al established optimal primary tumor size cut-off for development of metastases for SDHB-related tumors as 4.5 cm (44). In the present study, based on the histopathologic examination, 16% and 28% of all primary tumors were smaller than 4.5 or 6 cm, respectively; 70% and 58% were equal to or larger than 4.5 and 6 cm, respectively; and 14% were of an unknown size (either surgery was not performed or the tumor diameter was not recorded after the tumor was surgically removed).

Of 59 primary AST, 19% and 22% were smaller than 4.5 or 6 cm, respectively; 64% and 61% were equal to or larger than 4.5 or 6 cm, respectively; and 17% were of an unknown size. Forty-nine primary AST with known sizes had an average size of 8.4 cm (children: 7.7 cm, adults: 8.4 cm) and a median of 8.0 cm (children: 7.0 cm, adults: 8.1 cm). In patients with SDHB mutations, 15% and 35% of primary tumors were smaller than 4.5 or 6 cm, respectively; 74% and 54% were equal to or larger than 4.5 or 6 cm, respectively; and 11% were of an unknown size. Sixty-five primary SDHB tumors with known sizes had an average size of 7.4 cm (children: 6.3 cm, adults: 7.8 cm) and a median of 6.8 cm (children: 5.5 cm, adults: 7.6 cm) (Table 2).

Table 2.

Primary Tumor in Patients by Age and Genetic Mutation

Age (n)	Mut.	#	Size <4.5 cm	Size ≥4.5 cm	Size not known
Children (27)	SDHB	22	3 (13%)	14 (64%)	5 (23%)
	AST	5	1 (20%)	2 (40%)	2 (40%)
Young adults (23)	SDHB	16	3 (19%)	13 (81%)	0
	AST	7	2 (29%)	4 (57%)	1 (14%)
Adults (61)	SDHB	28	4 (14%)	21(75%)	3 (11%)
	AST	33	6 (18%)	21 (64%)	6 (18%)
Older adults (21)	SDHB	7	1 (14%)	6 (86%)	0
	AST	14	2 (14%)	11 (79%)	1 (7%)

Abbreviations: AST = apparently sporadic tumors; Mut. = mutation; SDHB = succinate dehydrogenase subunit B.

When these groups were compared, there was no statistically significant difference in the average size of the primary tumors (P = .203). In both the SDHB and AST groups, the size of the primary tumor statistically correlated with the APTD (P = .004), with a higher percentage of primary tumors ≥4.5 cm and also ≥6 cm found in older patients (P = .031 for size ≥ 6 cm) (Table 2).

Location of the Metastases

In 132 patients, 290 metastatic lesions were detected (165 in SDHB and 125 in AST patients, Table 3). In patients with SDHB mutations, metastases were most often found in the bones (78%), lungs and mediastinum (45%), lymph nodes (36%), and liver (35%); only 1 patient had metastases detected only in lymph nodes. AST patients had metastases most frequently in the bones (64%), followed by the lungs and mediastinum (47%), lymph nodes (36%), and liver (32%); again only 1 patient had detected metastases only in lymph nodes. In all age groups, the bones were the most common sites of metastatic disease. Liver lesions were more common with increasing age, regardless of sex, while other metastatic locations showed no preference for age or sex. In SDHB patients, there were approximately twice as many metastases in each location in males than in females regardless of age, except for metastases in the liver. There were no statistical differences in the location of metastases between males and females in AST patients.

Table 3.

Classification of PHEOs/PGLs by Genetic Mutation, Tumor Type, Location, and Patient Age^a,b

	Localization		Total		Children		20–30 years		31–50 years		≥51 years
All pts			132	%	27	%	23	%	61	%	21	%
SDHB			73	55%	22		16		28		7
Solitary	EA		45	62	21	95	6	38	14	50	4	57
		Abd/ret.	36		18	90	5	83	10	72	2	50
		Pelvis	4		1	5	1	17	2	14
		Other	5		1	5			2	14	2	50
	A		8	11	1	5	2	12	4	14	1	14
Metastatic			14	19			5	31	7	25	2	29
Multiple			6	8			3	19	3	11
AST			59	45%	5		7		33		14
Solitary	EA		18	30	5	100	4	57	6	18	3	21
		Abd/ret.	15		3	60	3	75	6	100	3	100
		Head/neck	2		1	20	1	25
		Unknown	1		1	20
	A		27	46			2	29	18	55	7	50
Bilateral			1	2					1	3
Multiple			4	7			1	14	3	9
Metastatic			9	15					5	15	4	29

Abbreviations: A = adrenal; abd/ret. = abdomen/retroperitoneum; all pts = all patients; AST = apparently sporadic tumors; EA = extra-adrenal.

^aBold: total number of patients

^bUnderlined bold: total number of patients in different mutation groups

Plasma Biochemical Results in Primary Tumors

Of 132 patients (27 children and 105 adults), 111 (84%) had positive plasma biochemistry. The remaining 21 (16%) patients had plasma biochemistry within normal ranges (4 children and 17 adults without preference to sex and age range in adults). In patients with SDHB mutations, 46 (73%) of patients with extraadrenal PGLs had noradrenergic) phenotype; 16 (35%) of these patients also had elevated DA. Three (5%) had elevations of only DA; 14 (22%) were biochemically silent. In 10 adrenal PHEOs in SDHB patients, 6 (60%) had noradrenergic phenotype, 2 (20%) had adrenergic and noradrenergic phenotype; 3 (38%) of adrenergic and noradrenergic phenotype had also elevated DA. Two (20%) were biochemically silent (Table 4). In the 59 patients with AST, 31 had extraadrenal PGLs: 27 (87%) had noradrenergic phenotype; 5 (19%) of these AST patients also had elevated DA. Four (13%) were biochemically silent. In 28 adrenal PHEOs in AST patients, 13 (46%) had noradrenergic phenotype, 14 (50%) had elevated EPI and NE; 6 (22%) of those with elevated EPI and NE had also elevated DA. One (4%) were biochemically silent (Table 4).

Table 4.

Plasma and Urine Biochemistry in all Patients Due to Mutation Type and Tumor Localization

Plasma biochemistry
		SDHB				AST
Local	Bioch. type	#	Σ of bioch. type	%	with DA	%	#	Σ of bioch. type	%	with DA	%
PGL	Total	63		86			31		53
	NE	5	46	73	16	35 of NA	1	27	87	5	19 of NA
	NMN	6					7
	NE + NMN	35					19
	Silent	14		22			4		13
	Only DA	3		5			0
PHEO	Total	10		14			28		47
	NMN		6	60	3	38 of A + NA	2	13	46	6	22 of A + NA
	NE + NMN	6					11
	MN/NMN	1	2	20			4	14	50
	EPI/NE/MN/NMN	1					10
	Silent	2		20			1		4
Urine biochemistry
		SDHB				AST
		#	Σ of bioch. type	%	with DA	%	#	S of bioch. Type	%	with DA	%
PGL	Total	63		86			31		53
	NE	5	41	65	12	29 of NA	2	23	74	5	23 of NA
	NMN	10					5
	NE + NMN	26					16
	Silent	22		35			8		26
PHEO	Total	10		14			28		47
	NE		6	60	1	2 of NA	1	11	39	1	4 of A + NA
	NMN						2
	NE + NMN	6					8
	EPI, MN						2	2	7
	MN/NMN	1	2	20			3	10	36
	EPI/NE/MN/NMN	1					7
	Silent	2		20			5	5	18

Abbreviations: A = adrenergic phenotype; AST = apparently sporadic tumors; Bioch. type = type of produced chemical substance (catecholamine or metanephrine); DA = dopamine; EPI = epinephrine; Local = tumor localization; MN = metanephrine; NA = noradrenergic phenotype; NE = norepinephrine; NMN = normetanephrine; PGL = paraganglioma; PHEO = pheochromocytoma; SDHB = succinate dehydrogenase subunit B; silent = biochemically silent in plasma or urine; Σ of bioch. type = sum of patients with specified biochemical phenotype.

Urine Biochemical Results in Primary Tumors

Ninety-five (72%) patients had elevated urine biochemistry; 37 patients had urine biochemistry within normal ranges (7 children and 30 adults without preference for sex or age range in adults).

In 73 patients with SDHB mutations, 41 (65%) of the 63 PGLs were noradrenergic; 12 (29%) of these SDHB patients also had elevated DA. Twenty-two (35%) were biochemically silent. In the SDHB patients with PHEOs, 6 (60%) had noradrenergic phenotype, 2 (20%) had elevated EPI and NE; 1 (13%) of them also had elevated DA. Two (20%) were biochemically silent (Table 4). There was no preference to age or sex.

In 59 patients with AST, 23 (74%) of the 31 PGLs were noradrenergic; 5 (22%) also had elevated DA. Eight (26%) were biochemically silent. In the 28 PHEOs in AST patients, 11 (39%) had noradrenergic phenotype, 2 (7%) had adrenergic phenotype, and 10 (36%) had elevated EPI and NE; 1 (4%) of them had also elevated DA. Five (18%) were biochemically silent (Table 4). There was no preference of biochemistry type based on age or sex.

Biochemically Silent Primary Tumors in Plasma and Urine

Of 132 patients included in this study, 15 (11%) patients had tumors that were biochemically silent, with no elevations of any catecholamines or MNs detected in the plasma and urine together. Based on genetic testing, 11 (15%) SDHB patients (2 children and 9 adults without preference to sex) and 4 (7%) AST patients (1 child and 3 adults) were biochemically silent in plasma and urine. Based on age, 3 (11%) children and 12 (11%) adults were biochemically silent in plasma and urine. Based on localization, 2 (18%) SDHB tumors were PHEOs, and 9 (82%) were PGLs. In AST, 1 (25%) tumor was PHEO and 3 (75%) were PGLs.

In total, 11% of tumors in the adults and children combined were biochemically silent in plasma and urine. Twenty percent of children with AST and 10% of children with SDHB tumors were biochemically silent; 6% of adult AST and 18% of adult SDHB tumors were biochemically silent in both plasma and urine. Based only on plasma biochemistry, 20% of children with AST and 14% with SDHB tumors were biochemically silent, compared to 7% of AST and 25% of SDHB tumors ina dults. Based on urine biochemistry, 40% of AST and 23% of SDHB tumors in children were biochemically silent; in adults 20% of patients with AST and 37% with SDHB tumors were biochemically silent in urine. Based on age, 15% of tumors in children were biochemically silent in blood, and 26% were silent in urine biochemistry. In adults, 16% of tumors were biochemically silent in blood and 29% in urine.

Biochemical Phenotype in Plasma and Urine Based on Survival

PGLs in SDHB and AST patients who were alive at the time of this study had a plasma (or urine) noradrenergic phenotype in 67% (62%) and 85% (73%), respectively. SDHB PGLs had a dopaminergic phenotype in 7% (0%). SDHB tumors were biochemically silent in 26% (38%), and AST were biochemically silent in 15% (27%).

SDHB or AST patients with PGLs who died had noradrenergic phenotype in plasma (urine) in 83% (71%) or 100% (80%). They were biochemically silent in plasma (urine) in 17% (29%) in SDHB, while no patient with AST had a biochemically silent PGL. Eight (40%) of noradrenergic phenotype SDHB patients with PGL also had elevated DA in plasma, but 6 (35%) of noradrenergic phenotype in urine had also elevated DA. Three (60%) patients with AST PGLs with noradrenergic phenotype also had elevated DA, but none in urine (Table 5).

Table 5.

Comparison of Plasma and Urine Biochemistry in Patients Based on Survival^a

Plasma biochemistry
		ALIVE				DIED
		SDHB		AST		SDHB		AST
		Number	%	Number	%	Number	%	Number	%
PGL	total	39	62	26	84	24	38	5	16
	NE	26	57	22	81	20	43	5	19
	silent	10	71	4	100	4	29	0
	DA	3	100	0		0		0
PHEO	total	5	50	23	82	5	50	5	18
	NE	1	17	8	62	5	83	5	38
	EPI + NE	2	100	14	100	0		0
	silent	2	100	1	100	0		0
Urine biochemistry
		ALIVE				DIED
		SDHB		AST		SDHB		AST
		Number	%	Number	%	Number	%	Number	%
PGL	total	39	62	26	84	24	38	5	16
	NE	24	59	19	83	17	41	4	17
	silent	15	68	7	88	7	32	1	12
	DA	0	0	0	0	0	0	0	0
PHEO	total	5	50	23	82	5	50	5	18
	NE	1	17	8	73	5	83	3	27
	EPI	0		2	100	0		0
	A + NA	2	100	9	90	0		1	10
	silent	2	100	4	80	0		1	20

Abbreviations: A = adrenergic phenotype; ALIVE = surviving patients; AST = apparently sporadic tumors; DA = dopaminergic phenotype; DIED = patients who died; EPI = epinephrine; NA = noradrenergic phenotype; NE = norepinephrine; PGL = paraganglioma; PHEO = pheochromocytoma; SDHB = succinate dehydrogenase subunit B; silent = biochemically silent in plasma or urine.

^aThe percentage amount of total number of PGL (PHEO) with SDHB mutation or AST who were alive at the time of study are percentages of all of the SDHB or AST PHEOs and PGLs. Surviving patients are percentages of the NE, silent, or DA subgroup, not of the whole number of SDHB or AST patients. Deceased patients in all subgroups are 100% minus alive patients.

PHEOs (adrenal tumors) in SDHB patients who were alive at the time of the study had a plasma (urine) noradrenergic phenotype in 20% (20%), elevated EP and NE in 40% (40%), and biochemically silent phenotypes in 40% (40%). In surviving patients with AST PHEO, a noradrenergic phenotype in plasma (urine) was found in 35% (35%), elevated EPI and NE in 61% (39%), and only elevated EPI in urine in 9%. Biochemically silent AST PHEOs were found in 4% (17%).

SDHB or AST patients with PHEOs who died had a plasma (urine) noradrenergic phenotype in 100% (100%) or 100% (60%), respectively. In urine, patients with AST tumors also elevated EPI and NE in 20%, and were biochemically silent in 20%. One (20%) of SDHB PHEO with noradrenergic phenotype also had elevated DA only in plasma (Table 5).

In conclusion, 10 patients who died had primary A (5 SDHB, 5 AST), and 29 had primary EA tumors (24 SDHB, 5 AST). All of the patients who died had a noradrenergic phenotype in plasma and urine, except in patients with SDHB PGLs, in whom 17% were biochemically silent in plasma and 29% in urine. Only 1 AST PHEO had elevated EPI and NE in urine, and 1 AST PHEO was biochemically silent only in urine (with noradrenergic phenotype in plasma). There were no differences between males and females (AST: P = .349, SDHB: P = .272), and no differencies in biochemical phenotype were seen based on age. SDHB patients with noradrenergic phenotype also had elevated DA in plasma (urine) in 28% (27%); patients with AST with noradrenergic phenotype also had elevated DA only in plasma in 30% (Table 5).

Data about blood and urine biochemistry within metastases was not available, so it was not possible to compare biochemical profiles in primary tumor with biochemical profiles in metastases.

Patient Survival

Of 132 patients, 39 (29.5%) patients died between 2000 and 2015. They died 0 to 29 years after the diagnosis of a primary tumor. Twenty-nine of these patients had SDHB mutations; 10 were AST patients. There was a statistically significant difference in survival between patients with SDHB mutations and those with AST (P = .006, Fig. 1A). Kaplan–Meier estimates of 5- and 10-year survival rates were 91.8% (95% confidence interval [CI]: 82.696.2%) and 75.5% (95% CI: 63.5–84.1%) in the SDHB mutation group and 94.8% (95% CI: 84.8–98.3%) and 86.3% (95% CI: 73.1–93.3%) in the group of patients with AST, respectively. Overall, all deaths were because of metastatic PHEO/PGL. However, therapeutic interventions did not affect this difference (P = .535). Patients were treated by [¹³¹I]-MIBG; cyclophosphamide, vincristine and dacarbazine (CVD); external radiation; Octreotide; experimental therapy; or a combination of the above. Some patients did not receive any therapy. Furthermore, treatment was not associated with any differences in survival between the SDHB and AST patient groups (P = .483).

Fig. 1.

(A) Kaplan-Meier survival curves of 73 patients with SDHB mutations and 59 patients with AST; there was a statistically significant difference in survival between patients with SDHB mutations and those with AST (P = .006).

(B) Kaplan-Meier survival curves for SDHB versus AST patients, divided by sex. No statistically significant difference in survival probability was seen between males and females in either the AST or SDHB groups (AST: P = .349, SDHB: P = .272)

(C) Kaplan-Meier survival curves for children (0–19 years) versus all adults (age 20 or older). Overall survival was significantly different between children and adults (P = .037).

AST = apparently sporadic tumors; SDHB = succinate dehydrogenase subunit B.

D. Kaplan-Meier survival curves for children (0-19 years) versus all adults (age 20 or older).

There was no statistically significant differences in survival probability between males and females in either the AST or SDHB groups (AST: P = .349, SDHB: P = .272) (Fig. 1B). Nevertheless, overall survival was significantly different between children and adults (P = .037) (Fig. 1C); this was mostly attributed to the SDHB patients, since they, but not AST patients, had statistically significantly longer survival in children than adults (P = .006).

DISCUSSION

In the present study of 132 patients with metastatic PHEO/PGL, we assessed several important clinical presentations of these tumors. First, we did not find any major differences in the frequency of SDHB versus AST adult patients with metastatic disease, suggesting that other genetic and epigenetic factors besides SDHB mutations play an important role in the pathogenesis of metastatic PHEO/PGL. In contrast, in children, there were five times more patients with SDHB mutations than with AST. Second, we found that adult patients with SDHB mutations had statistically worse survival than those without them. Third, in the SDHB group, children had a statistically longer survival than adults. Fourth, we found that the size, location, and genetic background of the primary tumor, as well as the age at first diagnosis, were important variables in clinical behavior and outcomes of these tumors. Most importantly, primary metastatic tumors, most of them presenting as noradrenergic PGLs, were larger than 4.5 cm in more than 80% of patients. The frequency of metastatic tumors from primary AST increased with age, including a decreased MI compared to SDHB tumors.

From previous and current studies, several factors are known to contribute to the development of metastatic PHEO/PGL (45,46). These include: SDHB gene mutation (21), older age at initial diagnosis (47), location of primary lesions (48,49), tumor size/weight (18), and a noradrenergic and/or dopaminergic biochemical phenotype (29,38,43).

Currently, SDHB mutations, although associated with a low penetrance, are considered one of the most important risk factors for developing metastatic PHEO/PGL with a poor prognosis, including in children (3,21,22,46,50,51). In a paper by Amar et al, it was found that patients with PHEOs/PGLs with SDHB mutations have a high relative risk (71.4%) of developing metastatic tumors, which was later confirmed by several other investigators (20,23,44,52). Furthermore, Brouwers et al found that SDHB mutations are responsible for about 50% of all metastases originating from primary extra-adrenal abdominal PGLs (23). A meta-analysis of 12 studies published from 2000 to August 2011 found a higher incidence and prevalence of metastatic PGL in SDHB than in SDHD mutation carriers (51).

In the present study, in adult patients, we did not find a more frequent association between the presence of SDHB-related metastatic PHEO/PGL versus AST (Table 3). These findings are suported by the previous work of Brouwers et al as described above (23). Although we function as a tertiary center for the diagnosis and treatment of these patients, there is no bias in the acceptance of patients with or without hereditary or metastatic PHEO/PGL. Thus, we believe that the present results on this large population of patients with metastatic PHEO/PGL suggest that there are additional important factors besides SDHB mutations, including other genetic and epigenetic mechanisms, that play a crucial role in the development of metastases, at least in adults. By using the most updated genomic and proteomic approaches, such genetic and epigenetic changes may be discovered in the near future (26,53). Other than the presence of SDHB mutations, the size and location of the primary tumor were also found to play an important and most likely independent role in the devolopment of metastases in the present study, as described below.

First, for tumor size, O’Riordain et al found that a tumor size >5 cm was a strong predictor of persistent or recurrent disease and of mortality (54). More recently, in a large study including 142 patients, a primary tumor size of 6 cm or larger was found to be an important independent risk factor for metastatic disease (43). Schovanek et al confirmed in SDHB-related PHEO/PGL that the size of the primary tumor predicts development of metastatic disease and also affects patient survival. Patients with tumors smaller than 4.5 cm developed metastases significantly later than patients with larger tumors, and patients with smaller tumors than 4.5 cm had significantly better survival than patients with larger tumors (44). Furthermore, the primary tumor size of both PHEOs and PGLs was also found to be larger in patients with SDHB PHEO/PGL than in patients with AST (29,45).

In the present study, we confirmed the previous data that most primary AST and SDHB-related tumors that developed metastatic disease were over 4.5 cm. Interestingly, larger tumors were found in older patients with both AST and SDHB-related tumors. The explanation for these findings is unclear at present, but it is possible that in older patients these tumors may grow faster than in younger patients due to more profound immune or other system disturbances. Another explanation is that these tumors may be more frequently missed in older patients, since at least 30 to 40% of the older population suffers from hypertension, and antihypertensive drugs can mitigate the symptoms and signs of catecholamine excess. Some of these facts may also reflect the present findings that MI significantly decreased with increasing age in both the SDHB and AST groups.

In about 15 to 20% of patients with PHEO/PGL, basal plasma or urine catecholamines are within normal limits; these are called nonsecretory tumors (55). However, most of these “silent” tumors synthesize and metabolize catecholamines to metanephrines and may show elevations in plasma catecholamines only during paroxysmal attacks. Also, patients with SDHB-related tumors may present with biochemically silent phenotypes due to defective catecholamine synthesis resulting from the absence of tyrosine hydroxylase (56). These tumors are considered to be less differentiated. Methoxytyramine is a novel biomarker for metastatic PHEO/PGL introduced several years ago and has been a part of our workup since that time (43). However, since methoxytyramine levels was not available for all patients in this series, we felt this would represent incomplete data and therefore chose not to include them. Furthermore, metastatic PHEOs/PGLs are known to produce predominantly NE and/or NMN and usually have an exclusively noradrenergic phenotype (28,45). In the present study, 71% of the patients with SDHB had a noradrenergic phenotype in plasma and 64% in urine. SDHB tumors were more frequently biochemically silent or more frequently had a combined noradrenergic and dopaminergic phenotype in comparison to AST. AST more frequently secreted EPI and/or MN than SDHB tumors, either alone or in combination with NE, NMN, or DA. This data may suggest that metastatic nonhereditary PHEOs/PGLs are more differentiated than those with SDHB mutations, as discussed above.

We also found that 11% of tumors were biochemically silent in plasma and urine in this group of patients. In children, 15% of tumors were biochemically silent in blood and 26% silent in urine biochemistry. In adults, 16% of tumors were biochemically silent in blood and 29% in urine. These findings show there are no differences in catecholamine or MN production based on age, and metastatic tumors are the similarly agressive in children and adults.

Ten patients who died had primary adrenal (5 SDHB, 5 AST), and 29 had primary extra-adrenal tumors (24 SDHB, 5 AST). All of the patients who died had a noradrenergic phenotype in plasma and urine, except in some patients with SDHB PGLs, in whom 17% were biochemically silent in plasma and 29% in urine. There were no differences between males and females and no differences in biochemical phenotype due to age. SDHB patients with noradrenergic phenotype also had elevated DA in plasma (urine) in 28% (27%); patients with AST with noradrenergic phenotype also had an elevated DA only in plasma in 30%. The DA phenotype was mostly in PGLs. This supports the hypothesis that SDHB tumors are less differentiated and also more aggressive, especially PGLs.

We also confirmed previous observations that the main sites of metastatic spread of PHEO/PGL are lymph nodes, bones, liver, and lungs (29,31). Interestingly, liver lesions were more common with increasing age. Furthermore, in SDHB patients, there were approximately twice as many metastases in each location in males than in females, except for in the liver. The explanations for these observations are unclear at present, although differences in the hormonal status of males and females may be one of several explanations. The most common metastastic site was bones in both SDHB and AST; only 4 patients with SDHB and 2 AST patients did not have metastases in bones.

CONCLUSION

In summary, this large study including only SDHB and AST patients, has found that metastatic PHEO/PGL are equally distributed between SDHB and AST, with better 5- and 10-year survival rates for ASTs (P = .01), but currently with appropriate follow-up, about 75% of SDHB patients and 86% of AST patients survive 10 years. Overall survival was significantly different between children and adults (P = .037); this was mostly attributed to the SDHB patients, who had statistically significantly longer survival in children than adults (P = .006). In children, 81% (22) had mutations in SDHB (90% PGLs); in adults, only 49% (51) had mutations in SDHB (80% PGLs). The average MI decreased with increasing age in both SDHB and AST patients. Only 17% of all primary tumors were smaller than 4.5 cm.

This supports the establishment of certain recommendations for practicing oncologists and other physicians: (1) carefully examine patients who present with PHEO/PGL in childhood, including testing all of them for SDHB mutations and following them for many years due to the late development of metastases; (2) increase the follow-up frequency for patients with tumor sizes over 4.5 cm and consider performing whole-body imaging, especially in SDHB patients since one-fifth of patients present with metastatic disease at the initial diagnosis, mainly in bones; (3) increase follow-up frequency in older patients; (4) assure patients that with appropriate follow-up and treatment the 10-year survival rates are currently 75% or 86% for SDHB and AST types, respectively; and (5) follow-up visits should preferably be done in a collaborative oncology-endocrinology combined practice.

Abbreviations

A

adrenal

AMTD

age at the initial metastatic tumor diagnosis

APTD

age of patients at the time of the primary tumor diagnosis

AST

apparently sporadic tumor

CI

confidence interval

CT

computed tomography

DA

dopamine

EA

extra-adrenal

EPI

epinephrine

[18F]-FDA

[18F]-fluorodopamine

[18F]-FDG

[18F]-fluorodeoxyglucose

FH

fumarate hydratase

HIF2A

hypoxia-inducible factor 2α

MAX

myc-associated factor X

MI

metastatic interval

MIBG

metaiodobenzylguanidine

MN

metanephrine

MRI

magnetic resonance imaging

NE

norepinephrine

NF1

neurofibromatosis type 1

NIH

National Institutes of Health

NMN

normetanephrine

PET

positron emission tomography

PGL

paraganglioma

PHEO

pheochromocytoma

RET

rearranged during transfection

SDHA

succinate dehydrogenase subunit A

SDHAF2

encoding SDH complex assembly factor 2

SDHB

succinate dehydrogenase subunit B

SDHC

succinate dehydrogenase subunit C

SDHD

succinate dehydrogenase subunit D

TMEM127

transmembrane protein 127

VHL

von Hippel-Lindau