Pheochromocytoma Characteristics and Behavior Differ Depending on Method of Discovery

Abstract

Context

Modern pheochromocytomas (PHEOs) are often discovered by incidental finding on cross-sectional imaging or mutation-based genetic case detection testing. Little is known about how these PHEOs behave.

Objective

To describe the characteristics and behavior of PHEOs discovered incidentally on imaging or through mutation-based genetic case detection testing.

Design

Retrospective study.

Setting

Referral center.

Patients

Consecutive patients with pathology-confirmed PHEOs, treated from 2005 to 2016.

Main Outcome Measure(s)

Tumor size, plasma/urine fractionated metanephrines and catecholamines, and preoperative management.

Results

Two hundred seventy-one patients (52% women, median age 52.0 years) presented with 296 PHEOs. Discovery method was most often incidental finding on cross-section imaging (61%) rather than PHEO-related symptoms (27%) or mutation-based case detection testing (12%). Patients with incidentally discovered PHEOs were older than symptomatic and mutation-based case detection testing patients (median age 56.6 vs 43 vs 35 years, P < 0.0001). Mutation-based case detection PHEOs were smaller than those discovered due to symptoms (median size 29.0 vs 50.5 mm, P = 0.0027). Patients with PHEOs discovered due to symptoms had the highest median concentration of 24-hour urinary metanephrines and total plasma metanephrines (P < 0.0001). These patients required a higher cumulative phenoxybenzamine dose than patients with incidental or case detection PHEO (median 450 vs 375 vs 270 mg, P = 0.029).

Conclusions

PHEOs are primarily discovered due to incidental finding on cross-sectional imaging rather than PHEO-related symptoms. PHEOs discovered through mutation-based genetic case detection testing were smaller and required less α-adrenergic blockade preoperatively compared with PHEOs found due to symptoms, which supports routine case detection testing for patients genetically predisposed for PHEOs.

Pheochromocytoma (PHEO) is a rare neuroendocrine tumor that originates in the adrenal medulla. Historically, PHEOs were diagnosed when patients presented with typical signs and symptoms caused by hypersecretion of catecholamines. These symptoms classically were the triad of headache, palpitations, and diaphoresis. However, the breadth of symptoms associated with PHEO ranges from paroxysmal hypertension, episodic anxiety, and tremors to devastating acute heart failure and flash pulmonary edema. One of the difficulties in recognizing that a patient has a PHEO is the wide variety of symptoms and clinical manifestations, which can be caused by this tumor and its slow growth rate over time. Thus, many patients may have a PHEO for years before being diagnosed based on symptoms.

For decades, a symptomatic presentation that may be consistent with PHEO was a common reason for patient referral to endocrinologists. However, PHEO recognition has changed in recent years with the advent and increasingly routine use of computed cross-sectional imaging. Among patients with incidentally discovered masses (adrenal incidentalomas), PHEO accounted for 17.6% in a recent study (1).

Another more modern means of PHEO discovery is mutation-based case detection testing. It is well established that certain genetic conditions predispose patients to develop PHEOs. Most commonly, RET protooncogene mutations (MEN2), neurofibromatosis type 1 (NF1), von Hippel-Lindau (VHL), and succinate dehydrogenase enzyme mutations (primarily subunits SDHB, SDHC, and SDHD) are associated with PHEO. It is estimated that 40% to 50% of PHEOs occur in patients with a genetic predilection (2, 3). Patients with these genetic conditions tend to present at a younger age and more commonly have bilateral disease. Given this association, common practice dictates periodic biochemical testing for PHEO in patients with known disease-causing germline mutations. The method and frequency of this testing can vary widely among institutions but is becoming an accepted practice.

Despite the obvious shift in PHEO discovery from symptom-based to incidental and mutation-based case detection testing, little data exists about how the means of discovery correlates to tumor characteristics and behavior. Several studies have implied that patients with incidentally discovered PHEOs have less severe symptoms (4, 5). Our objective was to describe the characteristics and behavior of the “modern PHEO.” We hypothesized that incidentally discovered PHEOs are smaller in size, have less biochemical activity and more subtle symptoms, and require less preoperative α-adrenergic blockade. We also suspected that tumors found through mutation-based case detection testing might behave similarly to incidentally discovered PHEOs.

Methods

We performed a retrospective review of electronic medical records for patients with confirmed PHEO on pathology at the Mayo Clinic in Rochester, Minnesota, from 2005 through 2016. We used a search function through our electronic medical records to identify patients who had pathology for PHEO. Each medical chart was reviewed to ensure inclusion criteria were met. Patients who presented with recurrent or metastatic disease (diagnosed prior to 2005) were excluded. We performed a secondary review of PHEOs diagnosed from 1995 to 2004 to compare the method of discovery to the latter 12 years (2005 to 2016).

All biochemical testing was performed at our facility, with the exception of several instances of plasma metanephrines/normetanephrine measurements. PHEO was considered functioning if either plasma or urine fractionated metanephrines or catecholamines were above the upper limit of their respective reference ranges. For tumors with evidence of both adrenergic and noradrenergic function, a sum of the metanephrine and normetanephrine in urine or plasma was used for analysis. The cross sectional imaging was reviewed by a staff radiologist who specializes in abdominal imaging (R.P.H.). Patients with metachronous bilateral PHEOs were analyzed independently for each unique presentation, while patients with synchronous bilateral PHEO were analyzed as a single event.

All continuous data were summarized as median and ranges, while categorical data were summarized as a number (%). Associations between variables were assessed using the Student t test and ANOVA for continuous variables and Wilcoxon test for categorical variables. Statistical significance was defined as P value <0.05. Data were analyzed using JMP software, version 10 (SAS, Cary, NC). This study was approved by the Mayo Clinic Institutional Review Board.

Results

Patients

We identified 271 patients [52% women, median age of diagnosis 52.0 years (range, 5 to 84)] who presented with 296 PHEOs. PHEOs were most commonly discovered as an incidental finding on cross-sectional imaging, accounting for 169 patient presentations (62%), followed by discovery due to PHEO-related symptoms (n = 73, 26%) and mutation-based case detection testing (n = 32, 12%) (Table 1). Regardless of the mode of discovery, 185 (68%) patients had symptoms at presentation, including some patients whose PHEO was discovered due to incidental finding or mutation-based case detection testing. The most common presenting symptoms were palpitations (n = 103, 55.7%), headaches (n = 80, 43.2%), diaphoresis (n = 77, 41.6%), paroxysmal hypertension (n = 73, 39.5%), and spells (n = 69, 37.3%). Other symptoms (anxiety, nausea, etc.) were present in less than 20% of our cohort. Two of three patients with metachronous bilateral PHEO had symptoms with both presentations.

Genetic testing was performed in 113 (41.7%) of our cohort, and this testing was offered at the discretion of the providers specialized in PHEO. Younger age, positive family history for PHEO, and bilateral disease were important considerations for testing. Sixty-six patients (24.4%) had a disease-causing germline mutation. Nineteen of these patients were identified as being genetically predisposed to PHEO only after the tumor was identified. An additional 28 patients were encouraged to have genetic testing, but did not have this performed often due to cost. RET protooncogene (MEN2) was the most common genetic mutation (n = 29, 43.9%), followed by NF1 (n = 20, 30.3%), VHL (n = 13, 20.0%), SDHD (n = 2, 3.0%), and SDHB (n = 2, 3.0%). Bilateral PHEOs were identified in 25 (9.2%) patients, with only three having metachronous disease. Patients with bilateral disease often had an identified germline mutation (n = 20, 80%), with four of the remaining patients having negative testing for known mutations. One patient with bilateral disease did not have genetic testing due to expense. As expected, patients with bilateral disease were diagnosed at a younger age compared with unilateral disease (median age 34.5 vs 54.0 years, P < 0.0001). The genetic mutations noted in bilateral disease were RET protooncogene (n = 10), VHL (n = 6), and NF1 (n = 4).

Tumor characteristics

Patients presented with 137 left and 159 right PHEOs, of which 22 patients had synchronous bilateral disease and 3 patients had metachronous bilateral disease. The median tumor size was 40.0 mm (range, 8 to 260). Unenhanced CT imaging was available for review in 101 patients. The median precontrast CT attenuation was 35.3 Hounsfield units (HU) (range, 17.9 to 57.6), and the CT attenuation following contrast delivery was 82.6 HU (range 22.4 to 187.0).

Surgical outcomes

Preoperative biochemical data were available for 255 patients accounting for 258 patient events (3 patients with metachronous bilateral disease), and almost all demonstrated biochemical evidence of catecholamine excess (n = 244, 94.6%). In patients with functioning PHEOs, the biochemical profile was noradrenergic in 76 (31%) and adrenergic in 168 (69%) patients. All patients but one were treated with adrenalectomy. One of the tumors was not removed due to extensive metastatic disease found during open surgery. Detailed information on medical preparation for surgery was available for 262 surgical procedures. Most of the patients in our cohort were treated with α-adrenergic blockade prior to surgery either with phenoxybenzamine (n = 243, 92.7%) or doxazosin/prazosin (n = 13, 5%), two patients being on both phenoxybenzamine and doxazosin. The median duration of α-adrenergic blockade treatment was 13 days (range, 1 to 210). The median dose of phenoxybenzamine was 40 mg daily (range, 10 to 180), and for doxazosin it was 6 mg (range, 2 to 18). β-Adrenergic blockade was provided to most of our cohort as well (224, 85%) with type of β-adrenergic blocker being highly variable (atenolol, propranolol, metoprolol, carvedilol, or labetalol). A minority of patients required additional treatment with metyrosine (n = 37, 14%).

Of 273 adrenalectomies, 22 (8%) were bilateral. The laparoscopic surgical approach was used for 208 tumors (76.2%), and open surgery was used to remove 65 tumors (23.8%). The median hospital stay for our cohort was 3 days (range, 1 to 32), with patients undergoing laparoscopic surgery having a shorter duration of hospital stay with a median of 2 days (range, 1 to 9) vs a median of 5 days (range, 1 to 32) for open surgery (P ≤ 0.0001). No patients died of complications of adrenalectomy.

Excluding patients with <12 months follow-up, 179 patients were followed for a median of 3 years (range, 1 to 11) and a median disease-free survival of 3 years (range, 1 to 11). Metastatic disease was identified in 11 patients (4.0%), and 16 (5.9%) patients died (4 due to complications of metastatic disease) during follow-up. Metastatic disease was found at the time of diagnosis for eight patients (synchronous disease), with three additional patients developing metastases at 11, 33, and 34 months after initial primary tumor diagnosis.

Effect of the mode of PHEO discovery on outcomes

To investigate how the mode of discovery influenced patient outcomes, we divided our cohort into three groups: patients discovered with PHEO incidentally; patients who were discovered with PHEO based on symptoms related to PHEO; and patients discovered with PHEO-based case detection testing because of a known predisposing mutation (Table 1).

We found significant differences in patients’ presentation based on the mode of discovery. Patients discovered incidentally were older at the time of diagnosis with a median age 56.5 years vs median ages of 43.0 and 35.0 years in patients discovered based on symptoms or mutation-based case detection testing, respectively (P < 0.0001, Table 1). A higher proportion of bilateral PHEOs was discovered in patients undergoing mutation-based case detection testing: 25% vs 4.2% and 9.6%, respectively, for incidental and symptom-based discovery (P = 0.0009, Table 1). Patients discovered proactively based on mutation-based case detection testing presented with smaller PHEOs (median size of 29.0 mm), and patients discovered based on symptoms had larger PHEOS (median size of 50.5 mm, P = 0.0027, Table 1)

Patients with PHEOs detected based on symptoms presented primarily with headaches, palpitations, and spells. Notably, even patients discovered with PHEO incidentally or due to mutation-based case detection were often symptomatic (83.5% and 63.6%, respectively). The symptoms were primarily palpitations and sweating in the incidental group and palpitations and tremor in the mutation-based case detection group. Patients with PHEOs discovered due to symptoms reported a greater number of PHEO symptoms when compared with patients discovered incidentally or through mutation-based case detection (median number of symptoms 4.04 vs 3.02 vs 2.86, P < 0.0001). Interestingly, three patients had PHEO discovered because of symptoms, but biochemical testing was normal. All three had resolution of their symptoms following adrenalectomy.

Patients discovered with PHEOs based on symptoms had the highest median concentrations of total 24-hour urinary metanephrines and total plasma metanephrines, followed by patients with incidentally discovered PHEOs and patients discovered by mutation-based case detection testing (P < 0.0001, Table 1). Moreover, patients discovered due to symptoms vs incidentally vs by mutation-based case detection required a higher median cumulative phenoxybenzamine dose (450 vs 375 vs 270 mg, P = 0.029). Mode of PHEO discovery did not influence the type of adrenalectomy.

Of the 11 patients with metastatic disease, 5 patients were diagnosed because of symptoms and 6 patients were diagnosed after incidental discovery of PHEO. None of the patients diagnosed through mutation-based case detection testing developed metastases throughout the time of follow-up. Nevertheless, two patients who were diagnosed with synchronous metastatic disease after incidental discovery of an adrenal mass had a known genetic predisposition for PHEO (VHL, SDHD mutation). Overall, patients with genetic predisposition for PHEO (n = 66, with only 48.5% discovered based on case detection testing) had a slightly lower rate of metastatic disease compared with 205 patients who did not have identified mutations, although this is not statistically significant (3.0% vs 5.4%, P = 0.47).

Effect of PHEO size on outcomes

To investigate the effect of tumor size on patient’s outcomes, we have divided our cohort into two groups: patients with PHEO diameter ≤4 cm and patients with PHEO diameter >4 cm (based on median tumor size in our cohort of patients). In those patients with bilateral synchronous disease, a sum of diameters was used. We found that patients with PHEO >4 cm demonstrated higher levels of 24-hour urine total metanephrines and total plasma metanephrines (P < 0.0001, Table 2). Correspondingly, patients with tumor size >4 cm needed a higher cumulative phenoxybenzamine dose (420 vs 335 mg, P = 0.02) and were more likely to have a longer hospital stay (median 2 vs 3 days, P = 0.0002) and more likely to have an open surgery (33% vs 14%, P = 0.0002) (Table 2).

Biochemically silent PHEOs

Eleven patients had no biochemical activity suggestive of PHEO, four with an incidentally found tumor. Four patients presented with symptoms suggestive of PHEO, but had normal biochemical testing and imaging consistent with PHEO (Table 3). Patient number 1 had a history of right PHEO 4 years prior to presentation. Her previous PHEO was also biochemically silent, but had similar symptoms to her presentation 4 years later. Patient number 2 had a history of symptomatic paraganglioma and presented several years later with similar symptoms and normal biochemistries. Patient number 4 presented with metastatic disease at the time of PHEO diagnosis. He had extensive metastases in the mediastinum, liver, and bone. Patient number 6 was found to have a PHEO through case detection testing. He was followed for 2 years with biochemistries and imaging. His biochemistries remained normal, but he developed anxiety, palpitations, and hypertension, which prompted right adrenalectomy.

Shift in method of discovery over last decade

Between 1995 and 2004, we identified 120 patients with 140 PHEOs. The majority of PHEOs were discovered due to symptoms (n = 72, 51.4%), followed by imaging (n = 44, 31.4%) and mutation-based case detection testing (n = 6, 4.3%) (Fig. 1). Metastatic disease was identified in 35 of the patients, and 28 patients are deceased, 9 from metastatic disease. No patients with metastatic disease were diagnosed through mutation-based case detection testing.

Discussion

In this study of a large cohort of patients evaluated for PHEO after 2005, we found PHEOS are most commonly discovered because of incidental finding on imaging performed for another reason. This represents a “modern era” shift when compared with earlier years, when most patients were discovered based on symptoms, especially with paroxysmal or sustained hypertension (6, 7). Prior to 1990, none of the patients were discovered with PHEO incidentally, although starting in the mid 1990s, our records show increasing incidental discovery of PHEO. Similar trends were reported in studies at other institutions (5, 8). This is not surprising, as cross-sectional imaging became a more main-stay test in the last few decades, leading to more incidental findings in general. We also found that mutation-based case detection screening was more frequent in the last decade. The first reported case of PHEO found secondary to mutation-based case detection screening was reported between 1970 and 1980 at our institution, but routine testing was not a common practice at that time (7). The knowledge, availability, and affordability of genetic testing has improved over the last 2 decades—likely accounting for our increased number of PHEOs related to mutation-based case detection screening. Notably, the number of PHEOs diagnosed in previous decades was significantly lower compared with the last decade at our institution, which is probably reflective of patients not being recognized with a PHEO rather than an increase in PHEO incidence (Fig. 1) (6, 7). We also demonstrate that survival for PHEO is improving in recent years, perhaps related to earlier diagnosis through case detection testing and incidental finding on imaging.

We hypothesized that incidentally discovered PHEO behaved differently than those discovered due to symptoms. Indeed, we found that tumors discovered due to symptoms were larger and more biochemically active, suggesting they were discovered later in their natural history. These patients also needed a more intense preoperative medical preparation with higher cumulative doses of α-adrenergic blockade required, likely related to their higher biochemical activity. It is interesting to note that nearly one-half of the patients with mutation-based case detection and incidental PHEO were actually symptomatic. The severity of the symptoms was difficult to assess, but there was a slightly different spectrum of symptoms depending on the method of discovery. Patients discovered with PHEO based on symptoms tended to have more suggestive symptoms (headache and spells), although patients discovered incidentally or through mutation-based case detection testing had more vague symptoms (palpitations, sweating, and tremor). Again, this shows that PHEO can be insidious in onset and have nonspecific symptoms, making it difficult to recognize until disease is advanced.

In concordance to other studies, we found no sex predilection, and age of diagnosis varied based on the mode of discovery (being lowest in the group discovered due to mutation-based case detection testing) (9). The previously accepted “rule of 10’s” for PHEO (10% familial, 10% malignant, 10% bilateral) has fallen out of favor given advances in genetic testing. We show genetic association in about one-fourth of our cohort; however, it is important to note that many patients declined genetic testing, thus a much larger proportion of our cohort could harbor a predisposing genetic mutation. Other studies reported familial PHEOs in about 40% of patients with PHEOs (2, 3). In our cohort, RET protooncogene (MEN2) mutation was most common, followed by VHL and NF1. Other studies show that VHL is the most common mutation in PHEO cohorts (10, 11); both of these studies had multiple patients who were diagnosed before age 20 years. Our cohort had only 6 patients who presented at age 20 or younger, which could account for VHL mutations being less common in our study. We showed that bilateral tumors occur in less than 10% of patients with PHEO, but are quite common in hereditary PHEO, similar to other studies (10, 11).

We found the tumor size at presentation is an important factor affecting patients’ outcomes. PHEOs larger than 4 cm were more biochemically active and required a higher cumulative dose of phenoxybenzamine before adrenalectomy. We also found that one-third of PHEOs >4 cm were removed through open surgery, which was associated with longer duration of hospitalization stay. Other studies have shown similar results with larger PHEOs being more biochemically active as well as having increased risks during surgery, including greater blood loss, intraoperative hemodynamic instability, and adverse perioperative events (12–14).

Our study is limited by its retrospective design and relies heavily on documentation in the medical record. In addition to a large sample size of our cohort, a strength of this study is that most patients were evaluated by staff endocrinologists with a similar approach to α-adrenergic blockade.

This study has a potential to influence clinic practice in several ways. We showed that PHEOs found incidentally and by mutation-based case detection behave differently than PHEOs identified through symptoms. For example, our data suggest that tumors found proactively through mutation-based case detection testing are smaller and less biochemically active and require less α-adrenergic blockade prior to surgery. Consequently, these patients experience a quicker recovery postoperatively given the shorter length of stay. Moreover, we found no patients discovered with PHEO based on mutation-based case detection testing who developed metastases during 1995 to 2017. This information supports the importance of case detection testing, though conclusive recommendations are difficult to make in the absence of large prospective studies. Even though PHEOs are slow-growing tumors, the lower symptom burden, less premedication, shorter hospitalization, and potentially lower risk of metastatic disease with earlier diagnosis can have a substantial impact on patients who have known genetic predisposition for PHEO. We also showed that symptomatic tumors are larger than tumors discovered by other means. This has implications for clinical management, as our study shows larger PHEOs (>4 cm) require careful surgical planning, particularly allowing adequate time for α-adrenergic blockade titration. Clinicians should also expect the doses of α-adrenergic blockade to be greater in patients with tumors >4 cm. Specialized endocrine surgeons should be involved in the preoperative planning as well as the surgery because these larger tumors are more likely to require open surgery and longer hospital stays.

With advances in technology and laboratory testing, the way PHEOs are detected has changed. The “modern” PHEO is most commonly diagnosed following incidental discovery on cross-sectional computed imaging rather than due to symptoms. Also, case detection testing for patients with genetic predisposition for PHEO has also become a common means of diagnosis. As expected, PHEOs discovered proactively by mutation-based case detection testing tend to be smaller in size and less biochemically active and require less preoperative α-adrenergic blockade and have shorter length of stay compared with patients with PHEOs found due to symptoms.. This has implications for diagnosing PHEO earlier in the disease course for patients who have known genetic predisposition. Additionally, PHEOs >4 cm in diameter require higher dosages of α-adrenergic blockade and have longer length of hospital stay, pointing to the importance of allowing adequate time for titration of medications and surgical planning.

Abbreviations:

Abbreviations:
 
• HU

  Hounsfield unit

 
• PHEO

  pheochromocytoma

Acknowledgments

Disclosure Summary: The authors have nothing to disclose.

References