A male patient with PCC passed at age 71, 11 years after his initial diagnosis. He initially presented in 1999 with diaphoresis, palpitations, left-sided varicocele, and uncontrolled hypertension (HTN) despite pharmacological intervention. An abdominal ultrasound and subsequent computed tomography (CT) scan found a 14 cm left adrenal tumor. Subsequent biochemical testing revealed elevated plasma norepinephrine (2850 pg/mL [<175 pg/mL]). He underwent a left adrenalectomy and nephrectomy. Histopathology reported a well-encapsulated PCC with no invasion into the kidney or surrounding structures. Three years later, he experienced recurrent night sweats, irritable bowel symptoms, and HTN (up to 170/110 mmHg), with elevated plasma normetanephrine (681 pg/mL [<175 pg/mL]) and chromogranin A (CgA) (244 ng/mL [<76 ng/mL]). CT, ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET)/CT, ¹²³I-metaiodobenzylguanidine (¹²³I-MIBG) scintigraphy, and technetium-99m methylene diphosphonate (MDP) bone scintigraphy detected multiple metastatic lesions, including lungs, lymph nodes, and bones. Due to ¹²³I-MIBG-avid metastatic lesions, he began ¹³¹I-MIBG radiotherapy with interval improvement in periaortic and multiple abdominal lymph nodes. The following year, widespread metastatic progression was noted, involving the lungs, lymph nodes, and bones. He was referred to our team at the NIH and underwent ¹⁸F-fluorodopa (¹⁸F-FDOPA) PET/CT, ¹⁸F-fluorodopamine (¹⁸F-FDA) PET, and anatomic imaging, which confirmed extent of metastatic involvement. Upon our recommendations, he underwent a right middle lobectomy and mediastinal node dissection. Plasma normetanephrine (313 pg/mL [18–122 pg/mL]) and CgA (635 ng/mL [<225 ng/mL]) remained elevated. Due to progressive and extensive metastases, he began treatment with cyclophosphamide, vincristine, and dacarbazine (CVD). He completed 3 cycles, complicated by neutropenia and peripheral neuropathy. Further progression of the disease was observed in the lungs, lymph nodes, and bone, leading to the discontinuation of CVD therapy. He planned to receive radiation therapy to a growing right hilar mass in his home state; however, he was lost to follow-up and the NIH was notified of the patient’s death 2 years later from extensive metastatic disease.

A 28-year-old male initially presented to the emergency room with a pounding headache, abdominal pain, and nausea in 1987 and was admitted to the hospital with hypertensive crisis. Magnetic resonance imaging (MRI) identified a right adrenal mass, for which he underwent a right adrenalectomy. Histopathology reported a PCC measuring 13 cm. He was lost to follow-up, and twenty-six years later, MRI, CT, ¹⁸F-FDG, ¹⁸F-FDOPA, and ⁶⁸Ga-DOTA(0)-Tyr(3)-octreotate (⁶⁸Ga-DOTATATE) PET/CT, and ¹²³I-MIBG scintigraphy revealed multiple tumors in the abdomen, lungs, bones, and an aortocaval mass. Biochemical evaluation showed elevated plasma normetanephrine (1462 pg/mL [18–122 pg/mL]), dopamine (25 pg/mL [0–24 pg/mL]), and CgA (1349 ng/mL [<93 ng/mL]). He subsequently underwent an exploratory laparotomy, open cholecystectomy, and resection of the right aortocaval mass. The following year, he developed recurrence in the right adrenalectomy bed, and progression of metastatic lesions in the lungs, lymph nodes, and bones. One year later, he underwent a resection of a C2 paraspinal mass ( Figure 1 ) and intensity-modulated radiation therapy (IMRT, 5400 cGy) to the right sacral area with shrinkage of the lesion. Due to continued progression, he began Sandostatin^® long-acting release (LAR) for 4 months, followed by sunitinib for 5 months. Sunitinib was discontinued due to uncontrolled HTN, flushing, ageusia, and weight loss. Due to progressive disease and recurrence of his C2 paraspinal mass, he was evaluated at the NIH and enrolled in our NCT03206060 protocol: Lu-177-DOTATATE (Lutathera^®) in Therapy of Inoperable Pheochromocytoma/Paraganglioma Clinical Trial. He completed 4 cycles of treatment without progression and shrinkage in several lung, liver, and lymph node metastatic lesions. After 2 years, his CgA levels began to rise, and after 3 years, pulmonary progression was seen on CT imaging. He restarted ¹⁷⁷Lu-Dotatate and completed 4 additional cycles. Since then, disease remains stable on imaging, and he remains in close follow-up with the NIH.

A male patient with PCC passed at age 68, two years after his initial diagnosis. He initially presented with longstanding HTN, right back pain, pounding headaches, vertigo, fatigue, and multiple subcutaneous tumors. During routine abdominal ultrasound for risk of abdominal aortic aneurysm due to cigarette smoking, an incidental 1.7 x 1.6 x 1.4 cm hypoechoic lesion in the left liver lobe was identified, along with an enlarged paraaortic lymph node. He was evaluated at the NIH with anatomic and functional imaging scans confirming these lesions, along with a large 8.7 x 7 cm right adrenal mass, multiple masses in the liver, lungs, bone, and enlarged mesenteric and retroperitoneal lymph nodes. Histopathology from a liver biopsy specimen showed a well-differentiated neuroendocrine neoplasm, consistent with metastatic PCC. Biochemical evaluation showed elevated plasma normetanephrine (8592 pg/mL [<148 pg/mL]) and metanephrine (1261 pg/mL [<57 pg/mL]). Subsequently, he underwent a laparoscopic right adrenalectomy, with histopathologically confirmed PCC and Ki-67 < 3%. The following year, increasing size and number of metastatic lesions were noted in the lungs, liver, lymph nodes, and bones. Based on the avidity of his metastatic lesions on ¹²³I-MIBG scintigraphy, he planned to start treatment with high-dose iobenguane-¹³¹I-MIBG (Azedra^®). Approximately one year later, he was admitted to the emergency department experiencing shortness of breath and ultimately passed away due to complications from pneumonia and sepsis. It is believed that his advanced pulmonary metastatic disease played a role in his death.

A male patient with PCC passed at age 64, twenty years after diagnosis. The patient initially presented with progressive HTN, palpitations, flushing, and uncontrolled type 2 diabetes mellitus. He was found to have an elevated vanillylmandelic acid; metanephrines were unavailable. He underwent an abdominal ultrasound and CT scan showing a large left-sided adrenal mass. Due to suspicion of PCC, he underwent a left radical adrenalectomy and nephrectomy without evidence of recurrence until 9 years later. At that time, he developed progressive HTN and uncontrolled hyperglycemia. He was evaluated at the NIH with MRI, CT, ¹⁸F-FDG, and ¹²³I-MIBG scintigraphy, which revealed pulmonary, liver, lymph node, peritoneal, and left mesocolon metastasis and recurrence in the left surgical adrenal bed. Biochemical analysis revealed elevated plasma normetanephrine (3307 pg/mL [18–122 pg/mL]), norepinephrine (4949 pg/mL [80–498 pg/mL]), dopamine (59 pg/mL [3–46 pg/mL]), and CgA (1940 ng/mL [≤ 225 ng/mL]). He subsequently underwent resection of left adrenal surgical bed recurrence and metastatic lesions located in the subdiaphragmatic lymph node and mesocolon. Three years later, he presented with worsening HTN and glucose management. Imaging detected pulmonary, liver, lymph node, and bone metastatic lesions. He subsequently underwent exploratory laparotomy with lysis of adhesions and resections of the left periaortic and retropancreatic lymph node masses. Approximately 3 years later, due to progressive disease, he completed 4 cycles of ¹⁷⁷Lu-DOTATATE, followed by 14 months of lanreotide injections, and subsequently 3 cycles of CVD, all without stabilization of disease. The patient ultimately succumbed to metastatic disease.

The study protocol was approved by the Intramural Program of the National Institutes of Health, the Eunice Kennedy Shriver National Institute of Child Health and Human Development Institutional Review Board (NCT00004847). Written informed consent was obtained for all clinical, genetic, biochemical, and imaging studies. Our institution complies with all applicable laws, regulations, and policies concerning privacy and confidentiality.

Among NIH study participants without known genetic etiology for a personal history of PCC/PGL, 192 patients underwent exome sequencing (germline, somatic, or both) through the Center for Cancer Research Sequencing Facility. Next-generation sequencing was run on a NovaSeq 6000 S2 using Agilent SureSelect Human All Exon V7 in paired-end sequencing mode. DRAGEN was used for mapping to the reference genome hg38, and variant calling. Germline and somatic variants in published PCC/PGL-associated genes (including SDHA, SDHB, SDHC, SDHD, SDHAF2, VHL, RET, NF1, IDH1, KIF1B, HRAS, EPAS1, EGLN1, EGLN2, MAX, TMEM127, FH, BAP1, MDH2, ATRX, DLST, ACO1/IRP1, MAML3, GOT2, and DNMT3A) identified by next-generation sequencing (NGS) were confirmed by targeted Sanger sequencing. After removal of false positive results (those not confirmed by Sanger sequencing) and variants classified as benign or likely benign, four males had somatic putative loss-of-function ATRX variants, including three variants that were classified as likely oncogenic and one as a variant of uncertain significance (VUS), as determined by application of the Standards for the Classification of Pathogenicity of Somatic Variants in Cancer (Oncogenicity) (25) ( Table 1 ). Each of the four patients described in detail here underwent paired exome sequencing (i.e., germline and somatic sequencing with only one tumor used for each patient), and none were found to have germline pathogenic or likely pathogenic variants, nor any suspicious variants of uncertain significance in the genes analyzed. There were no other relevant somatic variants in the genes analyzed that passed confirmation by Sanger sequencing except for those listed explicitly in the Results section. Rare Exome Variant Ensemble Learner (REVEL) scores are derived from the aggregate of multiple computational tools to predict pathogenicity of missense variants as a probability, ranging from 0 to 1 (26).

Biochemical evaluation was based on data sent to and performed at the NIH, which included plasma normetanephrine, norepinephrine, metanephrine, epinephrine, 3-methoxytyramine, dopamine, and chromogranin A. Response to systemic therapies was determined by comparing the last biochemical value pre- and post-treatment.

Biochemical values were compiled throughout patients’ metastatic disease course. Given inconsistencies in laboratory reported reference ranges, fold changes above the upper reference limit (URL) were reported. If the patient’s plasma metanephrine amounted to less than 5% of their total metanephrine, normetanephrine, and 3-methoxytyramine sum, corrected by their respective upper reference limits, they were described as having a noradrenergic biochemical phenotype (27).

Per-lesion and per-patient analyses were performed using the following scans available at the NIH: ¹⁸F-FDOPA PET or PET/CT, ¹⁸F-FDA PET or PET/CT, ⁶⁸Ga-DOTATATE PET/CT, ¹⁸F-FDG PET/CT, and ¹²³I-MIBG scintigraphy. The duration between functional imaging scans in all patients was less than 3 months except in patient 2, where ¹⁸F-FDA was performed about a year prior to other scans. Histologic proof of all metastatic lesions was not feasible. Therefore, a composite of all the imaging studies served as a reference standard for the calculation of detection rates. A lesion was classified as true positive if found to be positive on at least two functional imaging modalities (¹⁸F-FDOPA, ¹⁸F-FDG, ⁶⁸Ga-DOTATATE, ¹⁸F-FDA, and ¹²³I-MIBG). A positive lesion found on one functional imaging modality and negative on all others was considered a false-positive ( Table 2 ). A patient was considered positive regardless of the number of positive lesions present; counting of metastatic lesions was limited to a maximum of 15 lesions per region (lymph nodes, lungs, mediastinum, liver, and abdomen). Ratios were defined by the number of lesions detected by functional imaging compared to the composite reference standard. McNemar’s test was used to determine significance (p-value <0.05) between ¹⁸F-FDOPA and other imaging modalities as it detected all the lesions on imaging comparator.

Immunohistochemistry (IHC) for ATRX was performed in tumor specimens from the four patients to assess functional correlation of the protein. A Sigma-Aldrich anti-ATRX antibody (polyclonal, HPA001906, St. Louis, MO) was used at a dilution of 1:200 on unstained slides from formalin fixed paraffin embedded (FFPE) tissue using the Ventana Benchmark XT (Ventana, Tucson, AZ, USA). A Santa Cruz Biotechnology anti-FH antibody (monoclonal, sc-100743, Dallas TX) was used at a dilution of 1:100 on unstained slides from formalin fixed paraffin embedded (FFPE) tissue using the Ventana Benchmark XT (Ventana, Tucson, AZ, USA). Validation of the ATRX staining was performed on the daily clinical laboratory control, the same day these patient tumor samples were stained, by the surgical pathologist on clinical service at the Laboratory of Pathology, National Cancer Institute.

Patient characteristics are summarized in Supplementary Table 1 . Four male PCC patients had somatic ATRX (NM_000489.6) variants detected in primary (patient 3) and metastatic lesions (patients 1, 2, and 4). Patients 2 and 4 had additional somatic VHL (NM_000551.4) and FH (NM_000143.4) pathogenic variants, respectively, revealed years after their initial diagnosis when comprehensive genetic analysis was re-initiated at the NIH. Age at initial diagnosis ranged from 28 to 66 years (median=52.5 years). PCC sizes ranged from 8.7 to 14 cm (median=12 cm). Two patients (50%, patients 1 and 4) had left PCC, and the remaining (50%, patients 2 and 3) had right PCC. All patients developed metastatic disease.

Clinical characteristics of primary and metastatic disease are summarized in Supplementary Table 1 . Initially, all patients underwent surgical resection of their PCCs. One patient (25%, patient 3) presented with synchronous metastases, in which PCC and distant metastases were found within 6 months, and the remaining three had metachronous metastases (75%, patients 1, 2, and 4). Time between primary resection and development of metastasis in these three patients ranged from 4–26 years (median=9 years). Two patients (50%, patients 2 and 4) had recurrence in their primary adrenalectomy bed ranging between 9–28 years (median=18.5 years). Patient 2 had recurrence of a C2 paraspinal mass 2 years after resection. This patient, the youngest in our cohort, exhibited the most prolonged interval from primary surgical resection to metastatic disease on imaging. However, it is noteworthy that the patient did not adhere to regular follow-up throughout those 26 years.

All patients presented with multiple metastatic lesions to bones, lungs, and lymph nodes. Three patients (75%, patients 2, 3, and 4) had metastases to the liver. Patient 4 (25%) had metastasis to the mesocolon. Those with pulmonary, hepatic, and osseous disease had multiple metastatic lesions in these locations. Two patients (50%, patients 1 and 2) presented with synchronous organ and osseous metastases, while these were metachronous for the remaining two (50%, patients 3 and 4).

Three of four somatic ATRX variants were classified as likely oncogenic. Patients 1 and 4 both had missense variants in exon 20: c.5248C>G (p.Pro1750Ala) and c.5229G>T (p.Arg1743Ser), respectively, both in the ATP-binding SNF2 protein domain (20). REVEL scores for these variants in patients 1 and 4 were 0.94 and 0.709, respectively, providing “strong” and “supporting” levels of computational evidence, respectively, of pathogenicity for these variants (28). Given that the germline classification of the ATRX variant for patient 1 would be considered likely pathogenic, this variant was considered likely oncogenic in the somatic context (25, 29). Of all these somatic ATRX variants, only this one (p.Pro1750Ala) had been reported previously in the Catalogue of Somatic Mutations in Cancer (COSMIC: https://cancer.sanger.ac.uk/cosmic), found in a uterine leiomyosarcoma. Patients 2 and 3 had frameshift truncating variants in exon 9: c.3230del (p.Ser1077del) and c.2018dup (p.Thr674fs), respectively, the latter of which is located in the RBR domain (20). Both of these variants met criteria to be classified as likely oncogenic ( Supplementary Table 2 ).

In addition to each hemizygous somatic ATRX oncogenic variant, two patients (2 and 4) were found to have additional somatic variants: VHL c.505C>G (p.Leu169Val) and FH c.305C>A (p.Ala102Glu), respectively, both of which were classified as likely oncogenic missense variants ( Table 1 ; Supplementary Table 2 ). None of these variants (in ATRX, VHL, or FH) were present in ClinVar.

ATRX staining for patient 4 tumor sample showed loss of nuclear ATRX in the tumor samples with faintly positive stromal cells serving as an internal control ( Figure 2 ). ATRX staining in tumor samples from patients 2 and 3 failed to display an adequate internal control staining in the stromal cell nuclei, and the tumor sample from patient 1 showed subsets of tumor cell nuclei with ATRX loss and ATRX retention. Samples from patients 1, 2, and 3 were determined to be inadequate for interpretation due to these possible technical concerns in the staining pattern. Loss of FH protein in the tumor cells was confirmed by IHC ( Supplementary Figure 1 ) in patient 4, with validation performed on the same day clinical laboratory control sample.

All patients underwent biochemical analysis, as described above ( Supplementary Table 3 ). However, assessment was not available at regular intervals since patients were largely managed and diagnosed outside the NIH.

Three patients (patients 1, 3, and 4) have succumbed to metastatic disease. Overall survival from the time of diagnosis ranged from 2 to 26 years with a median of 11 years. One patient (25%, patient 2) is alive 36 years after initial diagnosis with stable disease 2 years after completion of 8 cycles of Lutathera^®.

Patients 1, 2, and 4 received various therapies, including cytotoxic chemotherapy, targeted radiotherapies, external beam radiation, and somatostatin analogs. Patient 1 received ¹³¹I-MIBG therapy, with mixed response, specifically, shrinkage in one lymph node and progression in pulmonary disease, ultimately requiring surgical pulmonary debulking. Upon continued progression and widespread metastatic disease, he received 3 cycles of CVD with progression on therapy leading to death from metastatic disease.

Patient 2 received IMRT (5400 cGy) to the right sacral area with shrinkage and was also treated with Sandostatin^® LAR and sunitinib for widespread organ metastases, during which progression was noted. He subsequently received four cycles of Lutathera^®, leading to shrinkage of liver, lung, and lymph node lesions, along with stabilization of other areas for 3 years. Once progression was noted, he restarted Lutathera^® for an additional 4 cycles and continues to have stable disease 2 years after treatment.

Lastly, patient 4 received 4 cycles of Lutathera^®, followed by lanreotide and CVD, with progression, leading to death from metastatic disease.

Overall, treatment with cytotoxic chemotherapy and a tyrosine kinase inhibitor (TKI) were not effective in stabilizing or shrinking tumors. Systemic radiation with ¹³¹I-MIBG and IMRT targeted to sacral metastasis showed response in a periaortic and multiple abdominal lymph nodes (patient 1, 25%) and a sacral osseous metastasis (patient 2, 25%), respectively. Lanreotide treatment in two patients (patients 2 and 4) showed the most durable response with stabilization and shrinkage of tumors in patient 2.