- Email: [email protected]
Biopsy of pheochromocytomas and paragangliomas: Potential for disaster
Biopsy of pheochromocytomas and paragangliomas: Potential for disaster Kimberly A. Vanderveen, MD, MAS,a Scott M. Thompson, BA,b Matthew R. Callstrom, MD,c William F. Young Jr, MD, MSC,d Clive S. Grant, MD,a David R. Farley, MD,a Melanie L. Richards, MD,a and Geoffrey B. Thompson, MD,a Rochester, MN
Background. Pheochromocytomas and paragangliomas are highly vascular neuroendocrine neoplasms that often secrete catecholamines. Percutaneous biopsy has been associated with life-threatening hemorrhage, hypertensive crisis, capsular disruption with tumor implantation, and death. We sought to determine the outcomes of biopsy in 20 consecutive patients. Methods. We reviewed retrospectively the medical records of patients with biopsied pheochromocytomas and paragangliomas referred to our Endocrine Division for subsequent management between 1995 and 2005. Biopsy complications, operative findings, and outcomes were reviewed. Results. Twenty patients (14 pheochromocytomas and 6 paragangliomas) were biopsied percutaneously prior to referral. Mean tumor diameter was 6.4 cm (range, 1--15). Pre-biopsy biochemical testing was not performed in 90% of patients, and was negative in the remainder. Fourteen patients (70%) developed a biopsy-related complication, including: increased difficulty of the definitive operation in 7 of 17 (41%) operative cases with 1 patient requiring conversion to an open procedure; severe hypertension (15%); hematoma (30%); incorrect or inadequate biopsy (25%); severe pain (25%); and delay in surgical treatment (15%). Mean follow-up was 58 months, with 4 tumor-related deaths and 2 recurrences. Conclusion. Biochemical testing prior to biopsy of adrenal or suspicious retroperitoneal masses is critical to exclude a catecholamine-secreting neoplasm because serious complications and increased operative difficulty can result. (Surgery 2009;146:1158-66.) From the Departments of Surgery,a Radiology,c and Endocrinology, Diabetes, Metabolism, and Nutrition,d Mayo Clinic College of Medicine, Mayo Medical School,b Rochester, MN
PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS are rare, catecholamine-secreting neuroendocrine neoplasms originating from the chromaffin cells of the adrenal medulla and sympathetic ganglia, respectively. The annual incidence is approximately 2--8 per million people.1 Pheochromocytomas, which account for 90% of catecholamine-secreting tumors, have a low malignant potential (less than 10%); whereas the rate of malignancy is greater in paragangliomas (upwards of 40%). Various symptoms, often characterized as ‘‘spells,’’ may occur from uncontrolled secretion of catecholamines, including increased heart rate and blood pressure, palpitations, diaphoresis, pallor, anxiety, and headaches. Spells may be either spontaneous or precipitated by postural change, anxiety, Accepted for publication September 17, 2009. Reprint requests: Geoffrey B. Thompson, MD, Mayo Clinic College of Medicine, 200 First Street SW, Mayo W-18, Rochester, MN 55905. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2009 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2009.09.013
1158 SURGERY
medications (eg, metoclopramide, anesthetic agents), exercise, or maneuvers that increase intra-abdominal pressure (eg, change in position, lifting, defecation, exercise, colonoscopy, pregnancy, trauma), or biopsy. Such classic ‘‘spells,’’ however, are present in the minority of patients. Pheochromocytomas may present in the context of an incidentally discovered adrenal mass. A recent study reported a prevalence for adrenal incidentaloma of 4.4% on abdominal computed tomography (CT).2 Approximately 1 in 20 (5%) incidentalomas prove to be pheochromocytomas.3 They are highly vascular tumors with characteristic, but not pathognomonic, radiographic features that can be shared with primary and secondary adrenal neoplasms. Imaging features that raise strong suspicion of pheochromocytoma include: increased Hounsfield units (>20) on non-contrast CT, intravenous contrast enhancement with delayed contrast washout, central cystic necrosis, and bright intensity on T2-weighted magnetic resonance imaging (MRI). Despite well-established guidelines and principles for managing incidentally discovered adrenal
Surgery Volume 146, Number 6 masses,4-8 some patients with newly diagnosed adrenal masses on imaging continue to undergo percutaneous fine needle aspiration (FNA) or core biopsy prior to biochemical evaluation for a functional tumor. Adrenal biopsy is often done to exclude metastatic disease to the adrenal gland, but can be life-threatening in the setting of pheochromocytoma. The literature contains numerous case reports from the 1970s to the 1990s of devastating complications related to adrenal biopsy, such as: hemorrhage, capsular disruption with tumor implantation, hypertensive crisis, myocardial infarction, arrhythmia, stroke, or death.8-18 Therefore, biopsy of an incidentally discovered adrenal mass without prior biochemical testing for pheochromocytoma is specifically contraindicated in practice guidelines.19 In the current era of high-resolution CT, however, an increasing number of incidentalomas are detected. The overall safety and diagnostic accuracy of image-guided biopsy has also greatly improved over the last 10 years, and radiology reports are often now recommending biopsy of indeterminate lesions.8,9,16 This study was undertaken to review retrospectively the recent outcomes in a large series of patients who underwent image-guided percutaneous needle biopsies of unsuspected chromaffin neoplasms. METHODS After obtaining approval from our institutional review board, we identified patients referred to our tertiary endocrine practice who had undergone biopsy of a pheochromocytoma or paraganglioma from January 1995 to December 2005. A retrospective medical record review and independent rereview of available imaging was performed. Data recorded included general demographic data, medical history, pre-biopsy symptoms and presentation, pre-biopsy imaging and laboratory results, histologic interpretation of biopsy findings, biopsyrelated clinical outcomes, including complications, post-biopsy evaluation and findings, type of treatment/surgery, surgical findings/complications, final pathology, and available follow-up. Neoplasms were considered malignant based on local invasion or metastatic disease (not histologic criteria). If the operative note documented technical difficulty during the procedure due to hematoma, inflammation, or scarification at the biopsy site, this finding was considered a biopsy-related effect. If the operative note documented difficulty during the procedure due to tumor size, anatomy, or local invasion, this was not considered biopsy-related.
Vanderveen et al 1159
RESULTS Twenty-two patients were identified who underwent biopsy of a pheochromocytoma or paraganglioma prior to referral. Two patients were excluded who were seen in follow-up consultation only (after biopsy and definitive operation had been completed elsewhere), whose records were not available for review. We, therefore, identified 20 patients (14 pheochromocytomas and 6 paragangliomas) who were included in our study. Seven of the 20 patients had their biopsy performed at our institution. Mean follow-up was 58 months (range, 8--141). Seven patients (35%) had expired---4 from progressive tumor-related disease, 2 from other malignancies (lung cancer, leukemia), and 1 of unknown causes. Demographics of the study group are presented in Table I. Mean tumor diameter was 6.4 cm (range, 1--15). While nearly half (45%) of these masses were identified during evaluation of abdominal symptoms, 20% were discovered during cancer surveillance, 30% were discovered incidentally, and 1 patient (5%) had a palpable abdominal mass (Fig 1). Only 6 patients (30%) had a history of hypertension, and just one presented with classic pheochromocytoma ‘‘spells.’’ However, after diagnosis, 3 additional patients retrospectively offered histories consistent with catecholamine excess (diaphoresis, palpitations, headache). Of the 6 patients with a history of hypertension, only 2 patients had fluctuating (labile) hypertension, 1 had poorly controlled hypertension on multiple medications, and 1 patient was diagnosed with ‘‘pre-eclampsia’’ during pregnancy, which resolved post-partum. Six patients had a history of malignancy (4 with multiple cancers), including: leukemia and duodenal cancer; high-grade gastrointestinal stromal tumor (GIST); melanoma; squamous cell cancer and non-small cell lung cancer (NSCLC); NSCLC and prostate cancer; duodenal and prostate cancers. Findings on pathology and complications from percutaneous biopsy are shown in Table II. Overall, 15 patients (75%) had a biopsy complication, most commonly a hematoma (30%) or severe pain (25%). Two patients underwent a second needle biopsy for inconclusive results (lack of adequate tissue) on first biopsy, both of which were consistent with a chromaffin neoplasm on second biopsy. Seventeen patients underwent subsequent resection of their neoplasms. There were 9 of 17 operations that were planned for laparoscopic resection; 1 required hand-assistance due to tumor size and limited exposure, and 1 required
1160 Vanderveen et al
Surgery December 2009
Table I. Patient characteristics and biopsy complications Age, mean (range) Sex, male:female Pheochromocytomas: paragangliomas Maximal tumor diameter, mean (range) Benign:malignant at presentation Pheochromocytomas Paragangliomas Pre-biopsy evaluation Catecholamine/ metanephrine biochemical testing CT-scan Ultrasound MRI
56 y (33–83) 12:8 14:6 6.4 cm (1–15)
10:4 3:3 2/20 (2 false negative) 20/20 3/20 3/20
CT, Computed tomography; MRI, magnetic resonance imaging.
Fig 1. Presenting symptoms in 20 patients with biopsied pheochromocytoma or paraganglioma. Number of patients with each symptom is noted to the right of bars.
conversion to an open procedure due to biopsyrelated inflammatory changes. On review of the operative reports, the surgeon noted that the difficulty of the definitive operation was increased as a result of the pre-operative biopsy in 7 of the 17 resected cases (41%). Seven patients underwent fine needle aspirates (FNA), 8 underwent core needle biopsy (with or without FNA), and the remaining 5 patients did not have documentation of biopsy type. No correlation between biopsy type with documented hematoma or operative difficulty was observed. All biopsy reports reviewed documented multiple needle passes. Three patients had a delay in the time to their operation resulting from the biopsy. One patient’s biopsy diagnosis was incorrect (epithelial neoplasm) as he was presumed to have metastatic lung cancer. Because of the error in diagnosis, he was followed for 9 months, at which point he developed increases in serum adrenal biochemistries and tumor growth (further described below). Two additional patients who developed moderatesized retroperitoneal hematomas had their operation delayed 3 and 8 weeks, respectively. Pre-biopsy biochemical testing was not performed in 18 patients (90%), and was negative in the remaining 2 patients. Post-biopsy biochemical testing was obtained in all but 1 patient (who was thought to have a primary hepatic malignancy), and was diagnostic in 12 of 13 patients with pheochromocytoma, and 3 of 6 patients with paraganglioma. Of those tested postoperatively, 93% (14 of 15) had normalization of their biochemical markers after operation. The 1 patient
with persistently increased postoperative catecholamines had known metastatic disease prior to operation (described below). Two patients (both with pheochromocytomas) underwent pre-biopsy biochemical testing for pheochromocytoma with false negative results. The first had a history of hypertension, prostate cancer, and prior pancreatoduodenectomy for a stage II duodenal cancer 3 years prior to presentation, and was found to have an adrenal mass on cancer surveillance imaging. His CT report noted a new 1-cm indeterminate adrenal mass that was ‘‘concerning for metastasis.’’ He received a full biochemical workup prior to biopsy, with normal plasma fractionated metanephrines and 24-hour urinary fractionated metanephrines and catecholamines. After biopsy, repeat biochemical testing was similarly negative. His biopsy was interpreted as ‘‘metastatic neuroendocrine neoplasm with features of a carcinoid neoplasm.’’ He was offered adrenalectomy and refused, and was therefore observed for 9 months with a slow increase in neoplasm size and repeat biochemical testing suggestive of a functional neoplasm (increased urine cortisol, but persistently normal fractionated metanephrines and catecholamines) at which point he consented to adrenalectomy. An open posterior approach was used due to his prior operation. Intraoperative manipulation of the neoplasm resulted in severe hypertension (systolic blood pressure rose to 260 mm Hg) which abated on removal of the neoplasm, raising the suspicion of a pheochromocytoma. His blood pressure was controlled intra-operatively without sequelae. Final
Surgery Volume 146, Number 6
Table II. Pathologic interpretations and complications of percutaneous biopsy Pathologic diagnosis Pheochromocytoma 8 Paraganglioma 4 Inadequate specimen* 2 Low grade 1 neuroendocrine neoplasm--most consistent with paraganglioma Atypical cells 1 Malignant undifferentiated tumor 1 Lymphoma 1 Epithelial neoplasm 1 Metastatic neuroendocrine tumor with 1 features of carcinoid Biopsy complications Hematoma 6/20 Pain 5/20 Hypertension 3/20 Delay in surgery 3/20 Error in diagnosis 1/20 Difficult surgical resection 7/17 Conversion to open adrenalectomy 1/9 Total complications 14/20 (70%) *Two patients underwent a second biopsy after an inadequate specimen on initial biopsy; repeat biopsy in both patients was ‘‘suspicious for pheochromocytoma.’’
pathology confirmed pheochromocytoma. He ultimately developed lung and bone lesions with biopsies showing ‘‘metastatic undifferentiated neuroendocrine neoplasm’’ and died of disease 6.5 years after his adrenalectomy. The second patient with false-negative biochemical testing for pheochromocytoma had no history of hypertension and presented at age 39 with abdominal pain, panic attacks, and palpitations with Wolf-Parkinson-White Syndrome. She had no prior history of malignancy and a 2.8-cm left adrenal mass was identified on CT scan. Urinary case detection testing was negative. An uneventful biopsy was performed. At the time of referral, interpretation of her images was documented as ‘‘consistent with pheochromocytoma,’’ showing a hypervascular adrenal nodule with a necrotic cystic center. Post-biopsy plasma fractionated metanephrines were mildly elevated. She underwent uneventful laparoscopic adrenalectomy and remains alive and well 84 months post-resection. CT was performed on all patients prior to biopsy, with 9 patients having undergone multiple imaging studies (including ultrasound, n = 3; MRI, n = 3; and positron emission tomography [PET] with 18F-fluorodeoxyglucose [FDG], n = 4). Prebiopsy imaging reports suggested ‘‘metastatic disease’’ or ‘‘neoplasm’’ in 12 of 15 reports that
Vanderveen et al 1161
were available for review. Only 3 original reports specifically cited pheochromocytoma in the differential diagnosis. Thirteen of the patients had prebiopsy images available for re-review at the time of our study (6 of 14 pheochromocytoma and 5 of 6 paraganglioma patients). On re-review, all 6 of the pheochromocytomas had a classic appearance (examples: Fig 2). Of the remaining 8 patients, all are reported to display characteristics suspicious for pheochromocytoma (eg, contrast enhancement, central necrosis) except in the following 3 cases: (1) a 15-cm tumor mistakenly thought to arise within the liver in a woman with abdominal pain and palpable mass; (2) a 2-cm indeterminate adrenal mass found on surveillance imaging in a patient with a history of prostate cancer and a newly-diagnosed lung cancer (Fig 2); and (3) a 1-cm indeterminate adrenal mass found on surveillance imaging in a patient with a history of prostate and duodenal cancers (described above). The paraganglioma images were more variable in their appearance---of the 5 patients available for review, 3 had neoplasms that should have raised suspicion of a chromaffin tumor. One patient’s tumor was in the right retroperitoneum and had the classic appearance of an adrenal-based pheochromocytoma (Fig 3). Another also showed a classic appearance but was located in the pelvis and should have raised suspicion of paraganglioma arising from the Organ of Zuckerkandl. The third patient had a history of a high-grade GIST and presented on surveillance imaging with an aortocaval nodule, which was FDG-avid on PET. This was presumed to be a lymph node metastasis. Biopsy was performed (Fig 4), which was complicated by a hematoma. Biopsy indicated paraganglioma, and he underwent operative excision. He was later diagnosed with Carney Triad and has been diseasefree for 66 months. The remaining 3 patients (2 with available images) had massive peri-aortic disease that mimicked lymphoma. One patient undergoing preoperative abdominal CT for a complex hernia, had incidentally discovered extensive ‘‘peri-aortic adenopathy,’’ deemed suspicious for lymphoma (Fig 5). Core biopsy was uneventful, and initial pathology suggested lymphoma. After additional immunohistochemistry was performed, the diagnosis of paraganglioma was made. On retrospective review, he had hypertension, diabetes, and night sweats which all resolved after resection. He also had markedly increased urinary and plasma fractionated metanephrines on post-biopsy testing.
1162 Vanderveen et al
Surgery December 2009
Fig 3. Paraganglioma in the right upper quadrant (arrow). A 39-year-old man with abdominal pain found to have a right upper quadrant mass on CT. The radiology report suggested: ‘‘differential diagnosis for this mass would include primary leiomyosarcoma of the inferior vena cava, retroperitoneal sarcoma and metastatic disease’’ but did not list pheochromocytoma, despite the classic appearance of an adrenal-based lesion, and a poorly visualized right adrenal gland. Biopsy resulted in inflammatory response and a technically difficult operation. He remains alive and without evidence of recurrence at 78 months post-resection.
1 remains alive and well 14 years after initial diagnosis, having undergone multiple metastectomies for symptom control. Fig 2. Pheochromocytoma found on cancer staging. An 83-year-old man followed for prostate cancer was found to have a new lung mass (A, arrow). Imaging, with positron emission tomography with 18F-fluorodeoxyglucose (FDG) showed a FDG-avid mass in the right adrenal described as ‘‘worrisome for metastatic disease,’’ and CT was also obtained (B, arrow on adrenal mass). Biopsy was undertaken, resulting in a large retroperitoneal hemorrhage which delayed treatment. He underwent a technically difficult laparoscopic adrenalectomy followed by thoracotomy and resection of his lung cancer 3 weeks later. He remains alive and well 41 months later.
Three of the 6 paraganglioma patients and 3 of 14 pheochromocytoma patients had clinical evidence of malignant disease at the time of presentation, based on locally invasive neoplasms (n = 3), distant metastases on imaging (n = 2), or both (n = 2). Of these, 2 of the 3 malignant paraganglioma patients have died (at 9 and 16 mo), and the third has slowly progressive disease (at 45 mo) despite treatment on multiple clinical trials. Of the malignant pheochromocytoma patients, 2 have died (at 8 and 81 mo), and
DISCUSSION Clinically important biopsy-related complications were encountered in our cohort, and our series likely underestimates true complication rates, because only 7 of the 20 patients were biopsied at our center. Furthermore, 6 of 7 (86%) patients biopsied at our center had documented complications, although no mortalities. Those patients that were referred after biopsy elsewhere, by definition, survived their biopsies and did not have complications that prevented their referral. In our series, biopsies resulted in clinically important inflammation and increased difficulty of resection in almost half of patients eligible for laparoscopic approach, with 1 patient requiring conversion to an open procedure. This finding brings to question the low reported rate of hematomas, because uncomplicated biopsy would not be expected to result in relevant inflammation or adhesions; however, even small areas of contained capsular disruption after multiple needle passes---even without hematoma---can result in retroperitoneal fixation, thus complicating resection.
Surgery Volume 146, Number 6
Vanderveen et al 1163
Fig 5. Paraganglioma imaging is often confused with lymphoma. A 48-year-old asymptomatic man underwent a CT scan prior to repair of a complex abdominal wall hernia (arrow = hernia). He was found to have what appeared to be periaortic ‘‘mass or adenopathy’’ (T = tumor). He relayed symptoms of weight loss, sweats, and pallor---raising a suspicion of lymphoma. His biopsy was uncomplicated, and initially interpreted as ‘‘consistent with lymphoma’’; when further staining studies were performed, the diagnosis was revised to ‘‘pheochromocytoma.’’ He remains alive and asymptomatic 47 months postoperatively.
Fig 4. Paraganglioma confused with lymphatic metastases. A 37-year-old man with a history of a prior high grade GIST was found to have a new mass in the retroperitoneum (A, arrow), that on PET with 18 F-fluorodeoxyglucose had high uptake. The radiology reports documented the mass as ‘‘compatible with regional lymph node tumor.’’ Needle biopsy was performed (B, arrow on biopsy needle entering from posterior approach, T = tumor) resulting in severe flank pain and a moderate sized retroperitoneal hemorrhage (C, arrow). Biopsy was consistent with paraganglioma. After surgical excision he is alive and without evidence of recurrence 5 years postoperatively. He was later diagnosed with Carney Syndrome.
We did not assess length of hospital stay, although other authors have seen prolonged hospitalization as a result of biopsying a pheochromocytoma.20 Hypertension was seen in several patients, but is also likely underreported as the majority of the patients underwent their biopsies elsewhere. Others have reported that biopsy results are often misleading, inaccurate, and rarely affect treatment strategy.16 Our results confirm this bias. Only 4 patients in our series had adrenal neoplasms less than 4 cm; of these 3 of the 4 were clearly functional on biochemical testing, and the last increased in size on follow-up imaging, with subsequent development of positive biochemical testing. Even if these had represented isolated solitary metastases, surgical resection would have been offered. For the patients with paragangliomas, half (3 of 6) had suggestive imaging and should have undergone case detection testing for catecholaminesecreting neoplasms, which would have obviated the need for biopsy. The other 3 patients had neoplasms with atypical appearances and clinical presentations; all were very suggestive of lymphoma. One patient’s symptoms (night sweats, weight loss) were misinterpreted as ‘‘B-symptoms’’
1164 Vanderveen et al
of lymphoma. Needle biopsy was the key diagnostic maneuver that led to the diagnosis in 2 of 3 unusual-appearing neoplasms---markedly affecting their treatment strategy (resection vs chemotherapy for lymphoma). We do not suggest that biopsy of a suspected paraganglioma is acceptable---certainly any suspicion of a catecholamine-secreting neoplasm warrants pre-biopsy biochemical testing with either plasma fractionated metanephrines or 24-hour urine fractionated metanephrines and catecholamines. A patient who is suspected strongly of having a pheochromocytoma or paraganglioma (eg, history of a familial syndrome) with negative biochemical testing may be considered for 123 I-meta-iodobenzylguanidine scintigraphy prior to biopsy. This approach may have helped confirm the diagnosis without biopsy in a few of these patients, but likely would not have affected surgical management because we now use preoperative pharmacologic blockade for all adrenal masses with an imaging phenotype suggestive of pheochromocytoma, regardless of biochemical data. Unusual presentations, negative biochemical testing, and atypical imaging were seen in several of the patients in this series, demonstrating the importance of maintaining a high degree of clinical suspicion. The majority of the patients did not relate typical symptoms of catecholamine excess, but rather presented with imaging incidentalomas or vague abdominal discomfort. Several patients had known prior or concurrent malignancies. These presentations highlight the importance of imaging phenotype (highly vascular, enhancing, and often with cystic necrosis) and reinforce the importance of case detection testing prior to biopsy of any suspicious adrenal or retroperitoneal mass, regardless of symptomatology. Those patients with negative case detection testing and classic-appearing lesions on imaging, again demonstrate the value of imaging phenotype in guiding clinical management. Although it is rare, these cases demonstrate that symptomatology, imaging characteristics, and case detection testing are all fallible. Based on our limited follow-up to date, we have not had any unanticipated recurrences in the tumor bed suggestive of needle tract seeding from the biopsy. The 1 patient with recurrent disease in the surgical bed had a 17-cm malignant neoplasm, initially thought to be within the liver. The other patient with biochemical recurrence was a patient with a paraganglioma (Fig 5), who did not have any biopsy-related complications. Longer follow-up will be necessary to address recurrence risk in these patients.
Surgery December 2009
At the Mayo Clinic, we perform an average of 25 to 30 CT-guided adrenal biopsies per year, at a cost to the payer of approximately $6,000. The cost of 24-hour urinary fractionated metanephrines and catecholamines is less than $300, and plasma fractionated metanephrines is less than $200. Measurement of plasma fractionated metanephrines is very sensitive and convenient for the patient, because it eliminates the 24-hour urine collection period. In one multicenter study, sensitivity for plasma fractionated metanephrines was 99% and the specificity was 89%. Although the concern of false positives remains, it is far outweighed by the risk of biopsy of an unanticipated chromaffin neoplasm.21,22 Our current strategy is to use plasma fractionated metanephrines as the case detection test in patients with high clinical suspicion (suggestive symptoms, family history, or imaging) and urinary fractionated catecholamines and metanephrines in patients with low clinical suspicion. Ultimately, the choice of the case detection test---plasma or urine---is matter of debate, but the need for pre-biopsy biochemical testing should not be controversial. It is difficult to calculate the overall cost of biochemical testing for pheochromocytoma in this setting because it is not known what the denominator is for patients being screened versus the number referred for CT-guided biopsy. There is also the issue of additional testing for borderline or false positive results in case detection tests. Our belief, however, is that the cost of biochemical case detection testing is minimal relative to the potential morbidity and mortality resulting from the unnecessary and risky biopsy of an undiagnosed chromaffin neoplasm. Biopsy-related hemorrhage may be life threatening, and even when occult, may increase the difficulty of the definitive surgery or result in conversion from a laparoscopic to open procedure. Furthermore, our series by its design, under-reports the incidence of the most serious biopsy-related complications. This study is about safety in clinical practice. Adrenal neoplasms are common, and pheochromocytomas comprise a larger percentage of asymptomatic adrenal masses than was thought previously. Given the risk of a catastrophic complication, and the relative lack of clinically relevant information that biopsy provides in the diagnosis of a chromaffin neoplasm, we discourage strongly image-guided biopsy of any adrenal or suspicious retroperitoneal mass without pre-biopsy biochemical case detection testing. As others have reported, language in radiology reports that ‘‘biopsy is warranted’’ or ‘‘biopsy may be performed’’ should
Surgery Volume 146, Number 6 be avoided.16 Whether the responsibility to verify normal biochemical testing lies with the referring provider, or with the radiologist performing the biopsy, is a matter that must be addressed by individual centers, and will require effective communication between providers utilizing a systems-based approach or protocol.
REFERENCES 1. Stenstro¨m G, Sva¨rdsudd K. Pheochromocytoma in Sweden 1958-1981. An analysis of the National Cancer Registry Data. Acta Med Scand 1986;220:225-32. 2. Bovio S, Cataldi A, Reimondo G, Sperone P, Novello S, Berruti A, et al. Prevalence of adrenal incidentaloma in a contemporary computerized tomography series. J Endocrinol Investig 2006;29:298-302. 3. Young WF Jr. Management approaches to adrenal incidentalomas. A view from Rochester, Minnesota. Endocrinol Metab Clin North Am 2000;29:159-85. 4. Bu¨low B, Ahre´n B. Swedish Research Council Study Group of Endocrine Abdominal Tumours. Adrenal incidentaloma--experience of a standardized diagnostic programme in the Swedish prospective study. J Intern Med 2002;252: 239-46. 5. Grumbach MM, Biller BM, Braunstein GD, Campbell KK, Carney JA, Godley PA, et al. Management of the clinically inapparent adrenal mass (‘‘incidentaloma’’). Ann Intern Med 2003;138:424-9. 6. Young WF Jr. Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 2007;356:601-10. 7. Montali G, Solbiati L, Bossi MC, De Pra L, Di Donna A, Ravetto C. Sonographically guided fine-needle aspiration biopsy of adrenal masses. AJR Am J Roentgenol 1984; 143:1081-4. 8. Paulsen SD, Nghiem HV, Korobkin M, Caoili EM, Higgins EJ. Changing role of imaging-guided percutaneous biopsy of adrenal masses: evaluation of 50 adrenal biopsies. AJR Am J Roentgenol 2004;182:1033-7. 9. Arellano RS, Harisinghani MG, Gervais DA, Hahn PF, Mueller PR. Image-guided percutaneous biopsy of the adrenal gland: review of indications, technique, and complications. Curr Probl Diagn Radiol 2003;32:3-10. 10. Baguet JP, Hammer L, Tremel F, Mangin L, Mallion JM. Metastatic phaeochromocytoma: risks of diagnostic needle puncture and treatment by arterial embolisation. J Hum Hypertens 2001;15:209-11. 11. Casola G, Nicolet V, vanSonnenberg E, Withers C, Bretagnolle M, Saba RM, Bret PM. Unsuspected pheochromocytoma: risk of blood-pressure alterations during percutaneous adrenal biopsy. Radiology 1986;159:733-5. 12. Dalal T, Maher MM, Kalra MK, Mueller PR. Extraadrenal pheochromocytoma: a rare cause of tachycardia and hypertension during percutaneous biopsy. AJR Am J Roentgenol 2005;185:554-5. 13. Ford J, Rosenberg F, Chan N. Pheochromocytoma manifesting with shock presents a clinical paradox: a case report. CMAJ 1997;157:923-5. 14. Lembke T, Greenberg H. Cystic pheochromocytoma with inadvertent needle biopsy and glucagon administration. Can Assoc Radiol J 1987;38:232-3. 15. McCorkell SJ, Niles NL. Fine-needle aspiration of catecholamine-producing adrenal masses: a possibly fatal mistake. AJR Am J Roentgenol 1985;145:113-4.
Vanderveen et al 1165
16. Quayle FJ, Spitler JA, Pierce RA, Lairmore TC, Moley JF, Brunt LM. Needle biopsy of incidentally discovered adrenal masses is rarely informative and potentially hazardous. Surgery 2007;142:497-502. 17. Sood SK, Balasubramanian SP, Harrison BJ. Percutaneous biopsy of adrenal and extra-adrenal retroperitoneal lesions: beware of catecholamine secreting tumours!. Surgeon 2007; 5:279-81. 18. Yu R, Ginsburg E. Natural history of a sporadic pheochromocytoma. Endocr Pract 2008;14:391. 19. State-of-the-Science Conference Statement on management of the clinically inapparent adrenal mass (‘‘incidentaloma’’). NIH Consens State Sci Statements. 2002 Feb 4-6; 19:1-25. ID:14768652. Available from: http://consensus. nih.gov/2002/2002AdrenalIncidentalomasos021html.htm. [Accessed 28 Feb 2009]. 20. Yu R, Nissen NN, Chopra P, Dhall D, Phillips E, Wei M. Diagnosis and treatment of pheochromocytoma in an academic hospital from 1997 to 2007. Am J Med 2009; 122:85-95. 21. Lenders JW, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, et al. Biochemical diagnosis of pheochromocytoma: which test is best? JAMA 2002;287:1427-34. 22. Eisenhofer G, Goldstein DS, Walther MM, Friberg P, Lenders JW, Keiser HR, et al. Biochemical diagnosis of pheochromocytoma: how to distinguish true- from falsepositive test results. J Clin Endocrinol Metab 2003;88: 2656-66.
DISCUSSION Dr Scott Wilhelm (Cleveland, OH): Obviously you have shown us, which I think is really important, that these things should not be routinely biopsied. Jeff Moley’s group did a similar paper last year which was nicely done as well and certainly agrees with Nancy Perrier’s studies about the pathologic adrenal METS really turn our 15% of the field, so it’s very important that this message gets out and I commend you on continuing to get the message out. My question to you is, since you are doing 30 biopsies a year there, what are you biopsying for and what patients are you screening for? Are these people who you think have metastases that maybe have other disease elsewhere, and so you are biopsying to try to rule out doing surgery on them? Is that why you are biopsying? What’s your indication? Dr Kimberly Vanderveen (Rochester, MN): Most of these patients are actually coming through medical oncology, although there are some that come through internal medicine and primary care. Any that come through the endocrine division are automatically biochemically tested as part of our protocol. But there are patients that miss us because they have other suspected disease processes and are biopsied as part of a staging workup. That would be the most common indication, and we just don’t get to see all of these patients in our endocrine division prior to biopsy. Our plan is to take this data back to our medical oncologists and again reinforce that we have got to be biochemically testing these patients prior to biopsy, even if they have a malignancy elsewhere.
1166 Vanderveen et al
Dr Kresimira M. Milas (Cleveland, OH): My question is, this is very much an audience that supports those conclusions, so what internal conversation takes place among colleagues from different specialties for the seven cases at your own institution? And then as you query literature to include in your references, have you seen the other specialties like the radiologists performing these biopsies have addressed this topic in their own societies and publications? Dr Kimberly Vanderveen (Rochester, MN): To the first question, we have been talking with the radiologists, who are finally picking up on this, but obviously the problem is if they get an order for a biopsy they perform it. There’s not a check point there to make sure that biochemical testing has been done. So we are trying to internally create some systems and discussion on how to do this properly. Again, talking with the medical oncologists in their conferences we hope will help as well. In regards to your second question, in looking at the literature, there has been a fair amount of literature in the radiology series. But their primary interest is: ‘‘What’s the safety of adrenal biopsy?’’ not ‘‘Why are we biopsying pheochromocytomas and what are the outcomes?’’ So those studies really do not address why those patients were biopsied, whether or not they were biochemically tested or what their longterm outcomes were, and specifically not what the surgical outcomes were. So I think this is the first series to really address that. Dr Christopher R. McHenry (Cleveland, OH): We have had patients who actually developed hypertensive crisis and sustained a stroke as a result of a biopsy. So there are even more serious issues other than difficulty with the operation. So my question to you is, when you present this in a quality assurance setting and your morbidity and mortality conference, this would be classified, I would think, as a system problem and it needs a system correction. And the correction here, the stop point really is the radiologist. So I wonder how are we going to deal with it---how are we going to correct this problem at the level of the radiologist? And the radiologists tend to do other things, too, when they recommend procedures. Instead of just interpreting the film, they often recommend things that are incorrect. So we have to have a way of correcting this system problem. And I wondered if you might be able to give us some insight into how we do that.
Surgery December 2009
Dr Kimberly Vanderveen (Rochester, MN): I m certainly open to suggestions, because I think it is a systems problem. I think the systems are different in different places and how those lesions get to biopsy varies from institution to institution. So again, internally we are working on it. The vast majority of adrenal lesions do at some point make it into our endocrine clinic because of the way we are structured. But again we are still missing some. And again, I don’t know how to speak to how other institutions will do it, but I think this does, as you say, point out that it is a systems issue. We have to find a way to stop it at that radiology check point. And how we all do that I think is a great discussion to have. Dr Jean-Christophe Lifante (New York, NY): Yes, we have the same experience with the neuromonitoring of the EBSLN. Sometimes, the external branch of the superior laryngeal nerve looks like a vessel. With neuromonitoring, we were able to avoid EBSLN injury. Dr Gregory Randolph (Boston, MA): I appreciate you work, and thank you for demonstrating both rates of increased visualization and improved voice quality. You can measure EMG amplitude in the muscle. You watch the twitch of the muscle, which is a very good, accurate measure. Did you have any electrode within the muscle measuring its amplitude? It can be difficult to do because the amplitude is smaller and the latency is very short. Did you measure any EMG? Second, is the upper head or laryngeal head of the sternal thyroid muscle a very reliable landmark for the external branch? Did you find that helpful in either group? Dr Jean-Christophe Lifante (New York, NY): For all patients with neuromonitoring we have the EMG of the cricothyroid muscle thanks to the paired electrodes. In our opinion, the analysis of the cricothyroid muscle EMG after the entire dissection of the upper pole of the thyroid allows us to determine the preservation of the EBSLN. The study does not allow us to answer the following question: Does the EBSLN neuromonitoring decrease the frequency of the nerve injury? Indeed, we did not perform a stroboscopy. Our study shows only that there exists a better recovery of the voice with the use of neuromonitoring. But the cricothyroid muscle EMG that we obtained at the end of the operation argues for the preservation of the nerve.
Background. Pheochromocytomas and paragangliomas are highly vascular neuroendocrine neoplasms that often secrete catecholamines. Percutaneous biopsy has been associated with life-threatening hemorrhage, hypertensive crisis, capsular disruption with tumor implantation, and death. We sought to determine the outcomes of biopsy in 20 consecutive patients. Methods. We reviewed retrospectively the medical records of patients with biopsied pheochromocytomas and paragangliomas referred to our Endocrine Division for subsequent management between 1995 and 2005. Biopsy complications, operative findings, and outcomes were reviewed. Results. Twenty patients (14 pheochromocytomas and 6 paragangliomas) were biopsied percutaneously prior to referral. Mean tumor diameter was 6.4 cm (range, 1--15). Pre-biopsy biochemical testing was not performed in 90% of patients, and was negative in the remainder. Fourteen patients (70%) developed a biopsy-related complication, including: increased difficulty of the definitive operation in 7 of 17 (41%) operative cases with 1 patient requiring conversion to an open procedure; severe hypertension (15%); hematoma (30%); incorrect or inadequate biopsy (25%); severe pain (25%); and delay in surgical treatment (15%). Mean follow-up was 58 months, with 4 tumor-related deaths and 2 recurrences. Conclusion. Biochemical testing prior to biopsy of adrenal or suspicious retroperitoneal masses is critical to exclude a catecholamine-secreting neoplasm because serious complications and increased operative difficulty can result. (Surgery 2009;146:1158-66.) From the Departments of Surgery,a Radiology,c and Endocrinology, Diabetes, Metabolism, and Nutrition,d Mayo Clinic College of Medicine, Mayo Medical School,b Rochester, MN
PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS are rare, catecholamine-secreting neuroendocrine neoplasms originating from the chromaffin cells of the adrenal medulla and sympathetic ganglia, respectively. The annual incidence is approximately 2--8 per million people.1 Pheochromocytomas, which account for 90% of catecholamine-secreting tumors, have a low malignant potential (less than 10%); whereas the rate of malignancy is greater in paragangliomas (upwards of 40%). Various symptoms, often characterized as ‘‘spells,’’ may occur from uncontrolled secretion of catecholamines, including increased heart rate and blood pressure, palpitations, diaphoresis, pallor, anxiety, and headaches. Spells may be either spontaneous or precipitated by postural change, anxiety, Accepted for publication September 17, 2009. Reprint requests: Geoffrey B. Thompson, MD, Mayo Clinic College of Medicine, 200 First Street SW, Mayo W-18, Rochester, MN 55905. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2009 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2009.09.013
1158 SURGERY
medications (eg, metoclopramide, anesthetic agents), exercise, or maneuvers that increase intra-abdominal pressure (eg, change in position, lifting, defecation, exercise, colonoscopy, pregnancy, trauma), or biopsy. Such classic ‘‘spells,’’ however, are present in the minority of patients. Pheochromocytomas may present in the context of an incidentally discovered adrenal mass. A recent study reported a prevalence for adrenal incidentaloma of 4.4% on abdominal computed tomography (CT).2 Approximately 1 in 20 (5%) incidentalomas prove to be pheochromocytomas.3 They are highly vascular tumors with characteristic, but not pathognomonic, radiographic features that can be shared with primary and secondary adrenal neoplasms. Imaging features that raise strong suspicion of pheochromocytoma include: increased Hounsfield units (>20) on non-contrast CT, intravenous contrast enhancement with delayed contrast washout, central cystic necrosis, and bright intensity on T2-weighted magnetic resonance imaging (MRI). Despite well-established guidelines and principles for managing incidentally discovered adrenal
Surgery Volume 146, Number 6 masses,4-8 some patients with newly diagnosed adrenal masses on imaging continue to undergo percutaneous fine needle aspiration (FNA) or core biopsy prior to biochemical evaluation for a functional tumor. Adrenal biopsy is often done to exclude metastatic disease to the adrenal gland, but can be life-threatening in the setting of pheochromocytoma. The literature contains numerous case reports from the 1970s to the 1990s of devastating complications related to adrenal biopsy, such as: hemorrhage, capsular disruption with tumor implantation, hypertensive crisis, myocardial infarction, arrhythmia, stroke, or death.8-18 Therefore, biopsy of an incidentally discovered adrenal mass without prior biochemical testing for pheochromocytoma is specifically contraindicated in practice guidelines.19 In the current era of high-resolution CT, however, an increasing number of incidentalomas are detected. The overall safety and diagnostic accuracy of image-guided biopsy has also greatly improved over the last 10 years, and radiology reports are often now recommending biopsy of indeterminate lesions.8,9,16 This study was undertaken to review retrospectively the recent outcomes in a large series of patients who underwent image-guided percutaneous needle biopsies of unsuspected chromaffin neoplasms. METHODS After obtaining approval from our institutional review board, we identified patients referred to our tertiary endocrine practice who had undergone biopsy of a pheochromocytoma or paraganglioma from January 1995 to December 2005. A retrospective medical record review and independent rereview of available imaging was performed. Data recorded included general demographic data, medical history, pre-biopsy symptoms and presentation, pre-biopsy imaging and laboratory results, histologic interpretation of biopsy findings, biopsyrelated clinical outcomes, including complications, post-biopsy evaluation and findings, type of treatment/surgery, surgical findings/complications, final pathology, and available follow-up. Neoplasms were considered malignant based on local invasion or metastatic disease (not histologic criteria). If the operative note documented technical difficulty during the procedure due to hematoma, inflammation, or scarification at the biopsy site, this finding was considered a biopsy-related effect. If the operative note documented difficulty during the procedure due to tumor size, anatomy, or local invasion, this was not considered biopsy-related.
Vanderveen et al 1159
RESULTS Twenty-two patients were identified who underwent biopsy of a pheochromocytoma or paraganglioma prior to referral. Two patients were excluded who were seen in follow-up consultation only (after biopsy and definitive operation had been completed elsewhere), whose records were not available for review. We, therefore, identified 20 patients (14 pheochromocytomas and 6 paragangliomas) who were included in our study. Seven of the 20 patients had their biopsy performed at our institution. Mean follow-up was 58 months (range, 8--141). Seven patients (35%) had expired---4 from progressive tumor-related disease, 2 from other malignancies (lung cancer, leukemia), and 1 of unknown causes. Demographics of the study group are presented in Table I. Mean tumor diameter was 6.4 cm (range, 1--15). While nearly half (45%) of these masses were identified during evaluation of abdominal symptoms, 20% were discovered during cancer surveillance, 30% were discovered incidentally, and 1 patient (5%) had a palpable abdominal mass (Fig 1). Only 6 patients (30%) had a history of hypertension, and just one presented with classic pheochromocytoma ‘‘spells.’’ However, after diagnosis, 3 additional patients retrospectively offered histories consistent with catecholamine excess (diaphoresis, palpitations, headache). Of the 6 patients with a history of hypertension, only 2 patients had fluctuating (labile) hypertension, 1 had poorly controlled hypertension on multiple medications, and 1 patient was diagnosed with ‘‘pre-eclampsia’’ during pregnancy, which resolved post-partum. Six patients had a history of malignancy (4 with multiple cancers), including: leukemia and duodenal cancer; high-grade gastrointestinal stromal tumor (GIST); melanoma; squamous cell cancer and non-small cell lung cancer (NSCLC); NSCLC and prostate cancer; duodenal and prostate cancers. Findings on pathology and complications from percutaneous biopsy are shown in Table II. Overall, 15 patients (75%) had a biopsy complication, most commonly a hematoma (30%) or severe pain (25%). Two patients underwent a second needle biopsy for inconclusive results (lack of adequate tissue) on first biopsy, both of which were consistent with a chromaffin neoplasm on second biopsy. Seventeen patients underwent subsequent resection of their neoplasms. There were 9 of 17 operations that were planned for laparoscopic resection; 1 required hand-assistance due to tumor size and limited exposure, and 1 required
1160 Vanderveen et al
Surgery December 2009
Table I. Patient characteristics and biopsy complications Age, mean (range) Sex, male:female Pheochromocytomas: paragangliomas Maximal tumor diameter, mean (range) Benign:malignant at presentation Pheochromocytomas Paragangliomas Pre-biopsy evaluation Catecholamine/ metanephrine biochemical testing CT-scan Ultrasound MRI
56 y (33–83) 12:8 14:6 6.4 cm (1–15)
10:4 3:3 2/20 (2 false negative) 20/20 3/20 3/20
CT, Computed tomography; MRI, magnetic resonance imaging.
Fig 1. Presenting symptoms in 20 patients with biopsied pheochromocytoma or paraganglioma. Number of patients with each symptom is noted to the right of bars.
conversion to an open procedure due to biopsyrelated inflammatory changes. On review of the operative reports, the surgeon noted that the difficulty of the definitive operation was increased as a result of the pre-operative biopsy in 7 of the 17 resected cases (41%). Seven patients underwent fine needle aspirates (FNA), 8 underwent core needle biopsy (with or without FNA), and the remaining 5 patients did not have documentation of biopsy type. No correlation between biopsy type with documented hematoma or operative difficulty was observed. All biopsy reports reviewed documented multiple needle passes. Three patients had a delay in the time to their operation resulting from the biopsy. One patient’s biopsy diagnosis was incorrect (epithelial neoplasm) as he was presumed to have metastatic lung cancer. Because of the error in diagnosis, he was followed for 9 months, at which point he developed increases in serum adrenal biochemistries and tumor growth (further described below). Two additional patients who developed moderatesized retroperitoneal hematomas had their operation delayed 3 and 8 weeks, respectively. Pre-biopsy biochemical testing was not performed in 18 patients (90%), and was negative in the remaining 2 patients. Post-biopsy biochemical testing was obtained in all but 1 patient (who was thought to have a primary hepatic malignancy), and was diagnostic in 12 of 13 patients with pheochromocytoma, and 3 of 6 patients with paraganglioma. Of those tested postoperatively, 93% (14 of 15) had normalization of their biochemical markers after operation. The 1 patient
with persistently increased postoperative catecholamines had known metastatic disease prior to operation (described below). Two patients (both with pheochromocytomas) underwent pre-biopsy biochemical testing for pheochromocytoma with false negative results. The first had a history of hypertension, prostate cancer, and prior pancreatoduodenectomy for a stage II duodenal cancer 3 years prior to presentation, and was found to have an adrenal mass on cancer surveillance imaging. His CT report noted a new 1-cm indeterminate adrenal mass that was ‘‘concerning for metastasis.’’ He received a full biochemical workup prior to biopsy, with normal plasma fractionated metanephrines and 24-hour urinary fractionated metanephrines and catecholamines. After biopsy, repeat biochemical testing was similarly negative. His biopsy was interpreted as ‘‘metastatic neuroendocrine neoplasm with features of a carcinoid neoplasm.’’ He was offered adrenalectomy and refused, and was therefore observed for 9 months with a slow increase in neoplasm size and repeat biochemical testing suggestive of a functional neoplasm (increased urine cortisol, but persistently normal fractionated metanephrines and catecholamines) at which point he consented to adrenalectomy. An open posterior approach was used due to his prior operation. Intraoperative manipulation of the neoplasm resulted in severe hypertension (systolic blood pressure rose to 260 mm Hg) which abated on removal of the neoplasm, raising the suspicion of a pheochromocytoma. His blood pressure was controlled intra-operatively without sequelae. Final
Surgery Volume 146, Number 6
Table II. Pathologic interpretations and complications of percutaneous biopsy Pathologic diagnosis Pheochromocytoma 8 Paraganglioma 4 Inadequate specimen* 2 Low grade 1 neuroendocrine neoplasm--most consistent with paraganglioma Atypical cells 1 Malignant undifferentiated tumor 1 Lymphoma 1 Epithelial neoplasm 1 Metastatic neuroendocrine tumor with 1 features of carcinoid Biopsy complications Hematoma 6/20 Pain 5/20 Hypertension 3/20 Delay in surgery 3/20 Error in diagnosis 1/20 Difficult surgical resection 7/17 Conversion to open adrenalectomy 1/9 Total complications 14/20 (70%) *Two patients underwent a second biopsy after an inadequate specimen on initial biopsy; repeat biopsy in both patients was ‘‘suspicious for pheochromocytoma.’’
pathology confirmed pheochromocytoma. He ultimately developed lung and bone lesions with biopsies showing ‘‘metastatic undifferentiated neuroendocrine neoplasm’’ and died of disease 6.5 years after his adrenalectomy. The second patient with false-negative biochemical testing for pheochromocytoma had no history of hypertension and presented at age 39 with abdominal pain, panic attacks, and palpitations with Wolf-Parkinson-White Syndrome. She had no prior history of malignancy and a 2.8-cm left adrenal mass was identified on CT scan. Urinary case detection testing was negative. An uneventful biopsy was performed. At the time of referral, interpretation of her images was documented as ‘‘consistent with pheochromocytoma,’’ showing a hypervascular adrenal nodule with a necrotic cystic center. Post-biopsy plasma fractionated metanephrines were mildly elevated. She underwent uneventful laparoscopic adrenalectomy and remains alive and well 84 months post-resection. CT was performed on all patients prior to biopsy, with 9 patients having undergone multiple imaging studies (including ultrasound, n = 3; MRI, n = 3; and positron emission tomography [PET] with 18F-fluorodeoxyglucose [FDG], n = 4). Prebiopsy imaging reports suggested ‘‘metastatic disease’’ or ‘‘neoplasm’’ in 12 of 15 reports that
Vanderveen et al 1161
were available for review. Only 3 original reports specifically cited pheochromocytoma in the differential diagnosis. Thirteen of the patients had prebiopsy images available for re-review at the time of our study (6 of 14 pheochromocytoma and 5 of 6 paraganglioma patients). On re-review, all 6 of the pheochromocytomas had a classic appearance (examples: Fig 2). Of the remaining 8 patients, all are reported to display characteristics suspicious for pheochromocytoma (eg, contrast enhancement, central necrosis) except in the following 3 cases: (1) a 15-cm tumor mistakenly thought to arise within the liver in a woman with abdominal pain and palpable mass; (2) a 2-cm indeterminate adrenal mass found on surveillance imaging in a patient with a history of prostate cancer and a newly-diagnosed lung cancer (Fig 2); and (3) a 1-cm indeterminate adrenal mass found on surveillance imaging in a patient with a history of prostate and duodenal cancers (described above). The paraganglioma images were more variable in their appearance---of the 5 patients available for review, 3 had neoplasms that should have raised suspicion of a chromaffin tumor. One patient’s tumor was in the right retroperitoneum and had the classic appearance of an adrenal-based pheochromocytoma (Fig 3). Another also showed a classic appearance but was located in the pelvis and should have raised suspicion of paraganglioma arising from the Organ of Zuckerkandl. The third patient had a history of a high-grade GIST and presented on surveillance imaging with an aortocaval nodule, which was FDG-avid on PET. This was presumed to be a lymph node metastasis. Biopsy was performed (Fig 4), which was complicated by a hematoma. Biopsy indicated paraganglioma, and he underwent operative excision. He was later diagnosed with Carney Triad and has been diseasefree for 66 months. The remaining 3 patients (2 with available images) had massive peri-aortic disease that mimicked lymphoma. One patient undergoing preoperative abdominal CT for a complex hernia, had incidentally discovered extensive ‘‘peri-aortic adenopathy,’’ deemed suspicious for lymphoma (Fig 5). Core biopsy was uneventful, and initial pathology suggested lymphoma. After additional immunohistochemistry was performed, the diagnosis of paraganglioma was made. On retrospective review, he had hypertension, diabetes, and night sweats which all resolved after resection. He also had markedly increased urinary and plasma fractionated metanephrines on post-biopsy testing.
1162 Vanderveen et al
Surgery December 2009
Fig 3. Paraganglioma in the right upper quadrant (arrow). A 39-year-old man with abdominal pain found to have a right upper quadrant mass on CT. The radiology report suggested: ‘‘differential diagnosis for this mass would include primary leiomyosarcoma of the inferior vena cava, retroperitoneal sarcoma and metastatic disease’’ but did not list pheochromocytoma, despite the classic appearance of an adrenal-based lesion, and a poorly visualized right adrenal gland. Biopsy resulted in inflammatory response and a technically difficult operation. He remains alive and without evidence of recurrence at 78 months post-resection.
1 remains alive and well 14 years after initial diagnosis, having undergone multiple metastectomies for symptom control. Fig 2. Pheochromocytoma found on cancer staging. An 83-year-old man followed for prostate cancer was found to have a new lung mass (A, arrow). Imaging, with positron emission tomography with 18F-fluorodeoxyglucose (FDG) showed a FDG-avid mass in the right adrenal described as ‘‘worrisome for metastatic disease,’’ and CT was also obtained (B, arrow on adrenal mass). Biopsy was undertaken, resulting in a large retroperitoneal hemorrhage which delayed treatment. He underwent a technically difficult laparoscopic adrenalectomy followed by thoracotomy and resection of his lung cancer 3 weeks later. He remains alive and well 41 months later.
Three of the 6 paraganglioma patients and 3 of 14 pheochromocytoma patients had clinical evidence of malignant disease at the time of presentation, based on locally invasive neoplasms (n = 3), distant metastases on imaging (n = 2), or both (n = 2). Of these, 2 of the 3 malignant paraganglioma patients have died (at 9 and 16 mo), and the third has slowly progressive disease (at 45 mo) despite treatment on multiple clinical trials. Of the malignant pheochromocytoma patients, 2 have died (at 8 and 81 mo), and
DISCUSSION Clinically important biopsy-related complications were encountered in our cohort, and our series likely underestimates true complication rates, because only 7 of the 20 patients were biopsied at our center. Furthermore, 6 of 7 (86%) patients biopsied at our center had documented complications, although no mortalities. Those patients that were referred after biopsy elsewhere, by definition, survived their biopsies and did not have complications that prevented their referral. In our series, biopsies resulted in clinically important inflammation and increased difficulty of resection in almost half of patients eligible for laparoscopic approach, with 1 patient requiring conversion to an open procedure. This finding brings to question the low reported rate of hematomas, because uncomplicated biopsy would not be expected to result in relevant inflammation or adhesions; however, even small areas of contained capsular disruption after multiple needle passes---even without hematoma---can result in retroperitoneal fixation, thus complicating resection.
Surgery Volume 146, Number 6
Vanderveen et al 1163
Fig 5. Paraganglioma imaging is often confused with lymphoma. A 48-year-old asymptomatic man underwent a CT scan prior to repair of a complex abdominal wall hernia (arrow = hernia). He was found to have what appeared to be periaortic ‘‘mass or adenopathy’’ (T = tumor). He relayed symptoms of weight loss, sweats, and pallor---raising a suspicion of lymphoma. His biopsy was uncomplicated, and initially interpreted as ‘‘consistent with lymphoma’’; when further staining studies were performed, the diagnosis was revised to ‘‘pheochromocytoma.’’ He remains alive and asymptomatic 47 months postoperatively.
Fig 4. Paraganglioma confused with lymphatic metastases. A 37-year-old man with a history of a prior high grade GIST was found to have a new mass in the retroperitoneum (A, arrow), that on PET with 18 F-fluorodeoxyglucose had high uptake. The radiology reports documented the mass as ‘‘compatible with regional lymph node tumor.’’ Needle biopsy was performed (B, arrow on biopsy needle entering from posterior approach, T = tumor) resulting in severe flank pain and a moderate sized retroperitoneal hemorrhage (C, arrow). Biopsy was consistent with paraganglioma. After surgical excision he is alive and without evidence of recurrence 5 years postoperatively. He was later diagnosed with Carney Syndrome.
We did not assess length of hospital stay, although other authors have seen prolonged hospitalization as a result of biopsying a pheochromocytoma.20 Hypertension was seen in several patients, but is also likely underreported as the majority of the patients underwent their biopsies elsewhere. Others have reported that biopsy results are often misleading, inaccurate, and rarely affect treatment strategy.16 Our results confirm this bias. Only 4 patients in our series had adrenal neoplasms less than 4 cm; of these 3 of the 4 were clearly functional on biochemical testing, and the last increased in size on follow-up imaging, with subsequent development of positive biochemical testing. Even if these had represented isolated solitary metastases, surgical resection would have been offered. For the patients with paragangliomas, half (3 of 6) had suggestive imaging and should have undergone case detection testing for catecholaminesecreting neoplasms, which would have obviated the need for biopsy. The other 3 patients had neoplasms with atypical appearances and clinical presentations; all were very suggestive of lymphoma. One patient’s symptoms (night sweats, weight loss) were misinterpreted as ‘‘B-symptoms’’
1164 Vanderveen et al
of lymphoma. Needle biopsy was the key diagnostic maneuver that led to the diagnosis in 2 of 3 unusual-appearing neoplasms---markedly affecting their treatment strategy (resection vs chemotherapy for lymphoma). We do not suggest that biopsy of a suspected paraganglioma is acceptable---certainly any suspicion of a catecholamine-secreting neoplasm warrants pre-biopsy biochemical testing with either plasma fractionated metanephrines or 24-hour urine fractionated metanephrines and catecholamines. A patient who is suspected strongly of having a pheochromocytoma or paraganglioma (eg, history of a familial syndrome) with negative biochemical testing may be considered for 123 I-meta-iodobenzylguanidine scintigraphy prior to biopsy. This approach may have helped confirm the diagnosis without biopsy in a few of these patients, but likely would not have affected surgical management because we now use preoperative pharmacologic blockade for all adrenal masses with an imaging phenotype suggestive of pheochromocytoma, regardless of biochemical data. Unusual presentations, negative biochemical testing, and atypical imaging were seen in several of the patients in this series, demonstrating the importance of maintaining a high degree of clinical suspicion. The majority of the patients did not relate typical symptoms of catecholamine excess, but rather presented with imaging incidentalomas or vague abdominal discomfort. Several patients had known prior or concurrent malignancies. These presentations highlight the importance of imaging phenotype (highly vascular, enhancing, and often with cystic necrosis) and reinforce the importance of case detection testing prior to biopsy of any suspicious adrenal or retroperitoneal mass, regardless of symptomatology. Those patients with negative case detection testing and classic-appearing lesions on imaging, again demonstrate the value of imaging phenotype in guiding clinical management. Although it is rare, these cases demonstrate that symptomatology, imaging characteristics, and case detection testing are all fallible. Based on our limited follow-up to date, we have not had any unanticipated recurrences in the tumor bed suggestive of needle tract seeding from the biopsy. The 1 patient with recurrent disease in the surgical bed had a 17-cm malignant neoplasm, initially thought to be within the liver. The other patient with biochemical recurrence was a patient with a paraganglioma (Fig 5), who did not have any biopsy-related complications. Longer follow-up will be necessary to address recurrence risk in these patients.
Surgery December 2009
At the Mayo Clinic, we perform an average of 25 to 30 CT-guided adrenal biopsies per year, at a cost to the payer of approximately $6,000. The cost of 24-hour urinary fractionated metanephrines and catecholamines is less than $300, and plasma fractionated metanephrines is less than $200. Measurement of plasma fractionated metanephrines is very sensitive and convenient for the patient, because it eliminates the 24-hour urine collection period. In one multicenter study, sensitivity for plasma fractionated metanephrines was 99% and the specificity was 89%. Although the concern of false positives remains, it is far outweighed by the risk of biopsy of an unanticipated chromaffin neoplasm.21,22 Our current strategy is to use plasma fractionated metanephrines as the case detection test in patients with high clinical suspicion (suggestive symptoms, family history, or imaging) and urinary fractionated catecholamines and metanephrines in patients with low clinical suspicion. Ultimately, the choice of the case detection test---plasma or urine---is matter of debate, but the need for pre-biopsy biochemical testing should not be controversial. It is difficult to calculate the overall cost of biochemical testing for pheochromocytoma in this setting because it is not known what the denominator is for patients being screened versus the number referred for CT-guided biopsy. There is also the issue of additional testing for borderline or false positive results in case detection tests. Our belief, however, is that the cost of biochemical case detection testing is minimal relative to the potential morbidity and mortality resulting from the unnecessary and risky biopsy of an undiagnosed chromaffin neoplasm. Biopsy-related hemorrhage may be life threatening, and even when occult, may increase the difficulty of the definitive surgery or result in conversion from a laparoscopic to open procedure. Furthermore, our series by its design, under-reports the incidence of the most serious biopsy-related complications. This study is about safety in clinical practice. Adrenal neoplasms are common, and pheochromocytomas comprise a larger percentage of asymptomatic adrenal masses than was thought previously. Given the risk of a catastrophic complication, and the relative lack of clinically relevant information that biopsy provides in the diagnosis of a chromaffin neoplasm, we discourage strongly image-guided biopsy of any adrenal or suspicious retroperitoneal mass without pre-biopsy biochemical case detection testing. As others have reported, language in radiology reports that ‘‘biopsy is warranted’’ or ‘‘biopsy may be performed’’ should
Surgery Volume 146, Number 6 be avoided.16 Whether the responsibility to verify normal biochemical testing lies with the referring provider, or with the radiologist performing the biopsy, is a matter that must be addressed by individual centers, and will require effective communication between providers utilizing a systems-based approach or protocol.
REFERENCES 1. Stenstro¨m G, Sva¨rdsudd K. Pheochromocytoma in Sweden 1958-1981. An analysis of the National Cancer Registry Data. Acta Med Scand 1986;220:225-32. 2. Bovio S, Cataldi A, Reimondo G, Sperone P, Novello S, Berruti A, et al. Prevalence of adrenal incidentaloma in a contemporary computerized tomography series. J Endocrinol Investig 2006;29:298-302. 3. Young WF Jr. Management approaches to adrenal incidentalomas. A view from Rochester, Minnesota. Endocrinol Metab Clin North Am 2000;29:159-85. 4. Bu¨low B, Ahre´n B. Swedish Research Council Study Group of Endocrine Abdominal Tumours. Adrenal incidentaloma--experience of a standardized diagnostic programme in the Swedish prospective study. J Intern Med 2002;252: 239-46. 5. Grumbach MM, Biller BM, Braunstein GD, Campbell KK, Carney JA, Godley PA, et al. Management of the clinically inapparent adrenal mass (‘‘incidentaloma’’). Ann Intern Med 2003;138:424-9. 6. Young WF Jr. Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 2007;356:601-10. 7. Montali G, Solbiati L, Bossi MC, De Pra L, Di Donna A, Ravetto C. Sonographically guided fine-needle aspiration biopsy of adrenal masses. AJR Am J Roentgenol 1984; 143:1081-4. 8. Paulsen SD, Nghiem HV, Korobkin M, Caoili EM, Higgins EJ. Changing role of imaging-guided percutaneous biopsy of adrenal masses: evaluation of 50 adrenal biopsies. AJR Am J Roentgenol 2004;182:1033-7. 9. Arellano RS, Harisinghani MG, Gervais DA, Hahn PF, Mueller PR. Image-guided percutaneous biopsy of the adrenal gland: review of indications, technique, and complications. Curr Probl Diagn Radiol 2003;32:3-10. 10. Baguet JP, Hammer L, Tremel F, Mangin L, Mallion JM. Metastatic phaeochromocytoma: risks of diagnostic needle puncture and treatment by arterial embolisation. J Hum Hypertens 2001;15:209-11. 11. Casola G, Nicolet V, vanSonnenberg E, Withers C, Bretagnolle M, Saba RM, Bret PM. Unsuspected pheochromocytoma: risk of blood-pressure alterations during percutaneous adrenal biopsy. Radiology 1986;159:733-5. 12. Dalal T, Maher MM, Kalra MK, Mueller PR. Extraadrenal pheochromocytoma: a rare cause of tachycardia and hypertension during percutaneous biopsy. AJR Am J Roentgenol 2005;185:554-5. 13. Ford J, Rosenberg F, Chan N. Pheochromocytoma manifesting with shock presents a clinical paradox: a case report. CMAJ 1997;157:923-5. 14. Lembke T, Greenberg H. Cystic pheochromocytoma with inadvertent needle biopsy and glucagon administration. Can Assoc Radiol J 1987;38:232-3. 15. McCorkell SJ, Niles NL. Fine-needle aspiration of catecholamine-producing adrenal masses: a possibly fatal mistake. AJR Am J Roentgenol 1985;145:113-4.
Vanderveen et al 1165
16. Quayle FJ, Spitler JA, Pierce RA, Lairmore TC, Moley JF, Brunt LM. Needle biopsy of incidentally discovered adrenal masses is rarely informative and potentially hazardous. Surgery 2007;142:497-502. 17. Sood SK, Balasubramanian SP, Harrison BJ. Percutaneous biopsy of adrenal and extra-adrenal retroperitoneal lesions: beware of catecholamine secreting tumours!. Surgeon 2007; 5:279-81. 18. Yu R, Ginsburg E. Natural history of a sporadic pheochromocytoma. Endocr Pract 2008;14:391. 19. State-of-the-Science Conference Statement on management of the clinically inapparent adrenal mass (‘‘incidentaloma’’). NIH Consens State Sci Statements. 2002 Feb 4-6; 19:1-25. ID:14768652. Available from: http://consensus. nih.gov/2002/2002AdrenalIncidentalomasos021html.htm. [Accessed 28 Feb 2009]. 20. Yu R, Nissen NN, Chopra P, Dhall D, Phillips E, Wei M. Diagnosis and treatment of pheochromocytoma in an academic hospital from 1997 to 2007. Am J Med 2009; 122:85-95. 21. Lenders JW, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, et al. Biochemical diagnosis of pheochromocytoma: which test is best? JAMA 2002;287:1427-34. 22. Eisenhofer G, Goldstein DS, Walther MM, Friberg P, Lenders JW, Keiser HR, et al. Biochemical diagnosis of pheochromocytoma: how to distinguish true- from falsepositive test results. J Clin Endocrinol Metab 2003;88: 2656-66.
DISCUSSION Dr Scott Wilhelm (Cleveland, OH): Obviously you have shown us, which I think is really important, that these things should not be routinely biopsied. Jeff Moley’s group did a similar paper last year which was nicely done as well and certainly agrees with Nancy Perrier’s studies about the pathologic adrenal METS really turn our 15% of the field, so it’s very important that this message gets out and I commend you on continuing to get the message out. My question to you is, since you are doing 30 biopsies a year there, what are you biopsying for and what patients are you screening for? Are these people who you think have metastases that maybe have other disease elsewhere, and so you are biopsying to try to rule out doing surgery on them? Is that why you are biopsying? What’s your indication? Dr Kimberly Vanderveen (Rochester, MN): Most of these patients are actually coming through medical oncology, although there are some that come through internal medicine and primary care. Any that come through the endocrine division are automatically biochemically tested as part of our protocol. But there are patients that miss us because they have other suspected disease processes and are biopsied as part of a staging workup. That would be the most common indication, and we just don’t get to see all of these patients in our endocrine division prior to biopsy. Our plan is to take this data back to our medical oncologists and again reinforce that we have got to be biochemically testing these patients prior to biopsy, even if they have a malignancy elsewhere.
1166 Vanderveen et al
Dr Kresimira M. Milas (Cleveland, OH): My question is, this is very much an audience that supports those conclusions, so what internal conversation takes place among colleagues from different specialties for the seven cases at your own institution? And then as you query literature to include in your references, have you seen the other specialties like the radiologists performing these biopsies have addressed this topic in their own societies and publications? Dr Kimberly Vanderveen (Rochester, MN): To the first question, we have been talking with the radiologists, who are finally picking up on this, but obviously the problem is if they get an order for a biopsy they perform it. There’s not a check point there to make sure that biochemical testing has been done. So we are trying to internally create some systems and discussion on how to do this properly. Again, talking with the medical oncologists in their conferences we hope will help as well. In regards to your second question, in looking at the literature, there has been a fair amount of literature in the radiology series. But their primary interest is: ‘‘What’s the safety of adrenal biopsy?’’ not ‘‘Why are we biopsying pheochromocytomas and what are the outcomes?’’ So those studies really do not address why those patients were biopsied, whether or not they were biochemically tested or what their longterm outcomes were, and specifically not what the surgical outcomes were. So I think this is the first series to really address that. Dr Christopher R. McHenry (Cleveland, OH): We have had patients who actually developed hypertensive crisis and sustained a stroke as a result of a biopsy. So there are even more serious issues other than difficulty with the operation. So my question to you is, when you present this in a quality assurance setting and your morbidity and mortality conference, this would be classified, I would think, as a system problem and it needs a system correction. And the correction here, the stop point really is the radiologist. So I wonder how are we going to deal with it---how are we going to correct this problem at the level of the radiologist? And the radiologists tend to do other things, too, when they recommend procedures. Instead of just interpreting the film, they often recommend things that are incorrect. So we have to have a way of correcting this system problem. And I wondered if you might be able to give us some insight into how we do that.
Surgery December 2009
Dr Kimberly Vanderveen (Rochester, MN): I m certainly open to suggestions, because I think it is a systems problem. I think the systems are different in different places and how those lesions get to biopsy varies from institution to institution. So again, internally we are working on it. The vast majority of adrenal lesions do at some point make it into our endocrine clinic because of the way we are structured. But again we are still missing some. And again, I don’t know how to speak to how other institutions will do it, but I think this does, as you say, point out that it is a systems issue. We have to find a way to stop it at that radiology check point. And how we all do that I think is a great discussion to have. Dr Jean-Christophe Lifante (New York, NY): Yes, we have the same experience with the neuromonitoring of the EBSLN. Sometimes, the external branch of the superior laryngeal nerve looks like a vessel. With neuromonitoring, we were able to avoid EBSLN injury. Dr Gregory Randolph (Boston, MA): I appreciate you work, and thank you for demonstrating both rates of increased visualization and improved voice quality. You can measure EMG amplitude in the muscle. You watch the twitch of the muscle, which is a very good, accurate measure. Did you have any electrode within the muscle measuring its amplitude? It can be difficult to do because the amplitude is smaller and the latency is very short. Did you measure any EMG? Second, is the upper head or laryngeal head of the sternal thyroid muscle a very reliable landmark for the external branch? Did you find that helpful in either group? Dr Jean-Christophe Lifante (New York, NY): For all patients with neuromonitoring we have the EMG of the cricothyroid muscle thanks to the paired electrodes. In our opinion, the analysis of the cricothyroid muscle EMG after the entire dissection of the upper pole of the thyroid allows us to determine the preservation of the EBSLN. The study does not allow us to answer the following question: Does the EBSLN neuromonitoring decrease the frequency of the nerve injury? Indeed, we did not perform a stroboscopy. Our study shows only that there exists a better recovery of the voice with the use of neuromonitoring. But the cricothyroid muscle EMG that we obtained at the end of the operation argues for the preservation of the nerve.