Treatment of Malignant Pheochromocytoma/Paraganglioma With Cyclophosphamide, Vincristine, and Dacarbazine

Abstract

BACKGROUND.

A long-term follow-up was conducted of 18 patients with a diagnosis of pheochromocytoma/paraganglioma treated with a combination of cyclophosphamide, vincristine, and dacarbazine (CVD).

METHODS.

The study design was a nonrandomized, single-arm trial conducted at a government medical referral center. Eighteen patients with metastatic malignant pheochromocytoma/paraganglioma were studied. After controlling symptoms of catecholamine excess, patients were treated with cyclophosphamide at 750 mg/m², vincristine at 1.4 mg/m², and dacarbazine at 600 mg/m² on Day 1 and dacarbazine at 600 mg/m² on Day 2, every 21 to 28 days.

RESULTS.

Combination chemotherapy with CVD produced a complete response rate of 11% and a partial response rate of 44%. Median survival from a landmark was 3.8 years for patients whose tumors responded to therapy and 1.8 years for patients whose tumors did not respond (P = .65). All patients with tumors scored as responding reported improvement in their symptoms related to excessive catecholamine release and had objective improvements in blood pressure. CVD was well tolerated with only grade I/II toxicities.

CONCLUSIONS.

Combination chemotherapy with CVD produced objective tumor responses in patients with advanced malignant pheochromocytoma/paraganglioma. In this 22-year follow-up there was no difference in overall survival between patients whose tumors objectively shrank and those with stable or progressive disease. However, patients reported improvement in symptoms, had objective improvements in blood pressure, and had tumor shrinkage that made surgical resection possible. The authors conclude that CVD therapy is not indicated in every patient with metastatic pheochromocytoma/paraganglioma, but should be considered in the management of patients with symptoms and where tumor shrinkage might be beneficial.

Catecholamine-secreting tumors—pheochromocytomas and paragangliomas—are rare neoplasms that arise from chromaffin tissue of the adrenal medulla and the sympathetic ganglia. Pheochromocytomas and paragangliomas occur in 2 to 8 people per million and <1% of patients with hypertension.^1–8 The peak incidence occurs during the fourth and fifth decades of life.⁹ Lacking reliable morphological or histological criteria, the diagnosis of malignancy depends on clinical behavior.^10–12 Depending on the genetic background and location, 3% to 36% of pheochromocytomas/paraganglioma are malignant.^13–16 The rate of malignancy is much higher in patients with paraganglioma who have mutations in the succinate dehydrogenase subunit B.^17–22

Most presenting signs and symptoms of pheochromocytoma/paraganglioma result from catecholamine excess, with patients often presenting with paroxysms of severe hypertension or tachycardia. However, about 20% to 30% of patients can be asymptomatic or present with only mild hypertension.^23–26 The diagnosis of pheochromocytoma/paraganglioma is confirmed by biochemical evidence of elevated catecholamine and metanephrine production, usually corroborated by radiologic studies.^10,23,24,27

The most effective method of treatment for pheochromocytoma/paraganglioma is surgical resection.^28–30 Patients with surgically resected benign pheochromocytoma/paraganglioma have an overall survival equal to an age-matched normal population. However, patients with malignant pheochromocytoma/paraganglioma have a 5-year survival <50% that of age-matched controls.^31–34

A combination of cyclophosphamide, vincristine, and dacarbazine (CVD) is often used as initial chemotherapy for malignant pheochromocytoma/paraganglioma, based on a study that reported a high response rate and improvement in symptoms.³⁵ Cognizant of the paucity of long-term follow-ups, we now report a 22-year follow-up of 18 patients with pheochromocytoma/paraganglioma, including 14 in the original report, treated with CVD at a single institution.

MATERIALS AND METHODS

Patients

From August 1983 to December 1995, 18 patients with malignant pheochromocytoma or paraganglioma confirmed by pathology and clinical course consented to receive chemotherapy. Eligibility requirements were described previously.³⁵ Fourteen patients enrolled in the original study and 4 additional patients who met eligibility criteria and were treated similarly are included in this analysis. All patients had metastatic disease. Additional criteria for entry included elevated plasma or urinary catecholamines, metanephrines, or vanillyl mandelic acid (VMA); measurable disease, no previous chemotherapy, recovery from previous radiation therapy, including targeted radiotherapy with ¹³¹I-metaiodobenzylguanidine (MIBG); adequate bone marrow function (absolute neutrophil count [ANC] >3.5 × 10⁹/L and platelet count >100 × 10⁹/L), renal function, and hepatic function, with Karnofsky performance status >30%. Patients had to be able to provide informed consent and to be accessible for follow-up.

Drug Therapy

Before starting CVD, various drugs were administered to control symptoms of catecholamine excess and maintain a normal blood pressure and heart rate. Initial treatment consisted of oral phenoxybenzamine, an α-adrenergic blocker, in doses up to 240 mg/day, usually in combination with a β-adrenergic blocker, propranolol or atenolol. If the blood pressure remained elevated, up to 2.0 g/day metyrosine, a catecholamine synthesis inhibitor, was administered orally. Some patients also received α1-adrenergic or calcium channel blockers to achieve adequate blood pressure control. The combination chemotherapy regimen consisted of intravenous cyclophosphamide (750 mg/m²), vincristine (1.4 mg/m²), and dacarbazine (600 mg/m²) on Day 1 and dacarbazine alone (600 mg/m²) on Day 2 every 21 to 28 days. The cyclophosphamide doses were adjusted according to the ANC nadir in the previous cycle, and ANC <500 was the goal to be achieved for the myelosuppression.

Treatment Evaluation

Patients had an initial complete history and physical examination with laboratory studies and received antihypertensive and analgesic therapy. At least 2 separate 24-hour urine collections were obtained for baseline catecholamines, metanephrines, and VMA determination. Initial radiologic and nuclear medicine studies included chest x-ray, computed tomography (CT) scan of the chest, abdomen, and pelvis, bone scan, and whole-body ¹³¹I-MIBG scan. Laboratory tests were repeated every 3 to 4 weeks throughout the treatment. Radiology and nuclear medicine studies were repeated every 6 to 16 weeks, if the original studies were abnormal.

Tumor response was based on the sum of the products of the perpendicular diameters of all measurable tumors and graded as complete response (complete regression of all clinical evidence of disease), partial response (at least 50% reduction), minimal response (at least 25% but not greater than 50% reduction), no change, and progression (the appearance of a new lesion or an increase in the size by 25%). Biochemical response was based upon the 24-hour urinary catecholamines, metanephrines, and VMA, as reported previously.³⁵

Statistical Methods

Survival durations were calculated in 3 different ways: 1) from date of diagnosis until date of death or last follow-up; 2) from on-study date until date of death or last follow-up; and 3) to estimate survival for patients whose tumors responded to treatment versus those whose tumors did not respond, a landmark analysis was done in which a date 3 months after starting therapy was selected as the starting point for survival, because this would permit an unbiased conditional estimate of the survival for these 2 groups of patients.³⁶ In each case, the Kaplan-Meier method was used to estimate the probability of survival as a function of time. The Mantel-Haenszel method was used to determine the statistical significance of the difference between a pair of Kaplan-Meier curves; the P value reported is from a 2-tailed test.^37,38

Patient 1

In 1991, a 42-year-old man presented with recurrent episodes of hematuria, hypertension, palpitations, diaphoresis, and headaches. CT scan revealed a 5.5 × 7.5 × 8.0-cm mass invading the urinary bladder posteriorly with right-sided hydronephrosis. Biopsy obtained via cystoscopy revealed a paraganglioma with many mitotic figures in the urinary bladder. The tumor was felt to be unresectable, and he was treated with 9 fractions of radiotherapy (RT) (1900 centigrays) without response.

The patient was referred for chemotherapy. His blood pressure was 142 over 92 mm Hg on phenoxybenzamine 20 mg twice a day, metyrosine 50 mg 3 times a day, verapamil 240 mg daily, and Maxzide 25 mg daily. Magnetic resonance imaging (MRI) before chemotherapy showed an enhancing mass in the posterior wall of the urinary bladder confluent with the prostate (Fig. 1 Pre-CVD). A 24-hour urinary test showed catecholamine at 889 μg (normal range 0–115 μg/24 h), metanephrines at 8.8 mg (normal range <0.9 mg/24 h), and VMA at 25 mg (normal range 2–10 mg/24 h). A ¹³¹I-MIBG scan showed increased activity in the bladder area.

FIGURE 1.

(A) Magnetic resonance imaging (MRI) done before the chemotherapy showed an enhancing mass in the posterior wall of the urinary bladder. (B) MRI done after the chemotherapy showed a decrease in the tumor size.

The patient began CVD chemotherapy in July 1992, and after 2 cycles an MRI scan showed a decrease in the tumor size. After 6 cycles, urinary catecholamines had decreased to 43 μg, metanephrines to 3.8 mg, and VMA to 10.1 mg, and the product of the perpendicular diameters was less than 20% of the original value (Fig. 1 Post-CVD). In February 1993 the patient underwent a radical cystectomy with an ileal conduit. Postsurgery, urinary catecholamines, metanephrines, and VMA returned to normal.

Regular follow-ups were unremarkable until March 1997, when multiple lung metastases and a 0.5-cm lesion in the right fourth rib were found on a CT scan. The patient reported increasing symptoms of catecholamine excess, and was hypertensive, with 24-hour urinary catecholamine of 298 μg, metanephrines of 1.8 mg, and VMA of 10.2 mg. He was restarted on phenoxybenzamine at 20 mg, with improvement in symptoms and normalization of blood pressure, and serial CT scans showed slowed progression of disease. In December 1999 the patient developed increasing right rib pain with an enlarging lesion at the fourth rib. He underwent radiofrequency ablation of the rib metastasis and did well until April 2001, when symptoms again increased, and CT showed new pelvic, rectum, and sigmoid masses. Therapy with CVD was restarted, but continued disease progression, especially in the liver, led to his death in August 2002.

Patient 2

A 48-year-old woman was diagnosed with pheochromocytoma in 1985 when she presented with palpitations, headache, sweating, and hypertension. Evaluation showed a left adrenal mass that was resected. She did well until 1988, when she developed bone pain, and a bone scan found increased uptake in the thoracolumbar spine. With symptomatic treatment, she did well until 1990, when multiple pulmonary metastases were found, and she was referred for chemotherapy.

At initial evaluation blood pressure was 140 over 98 mm Hg on 50 mg atenolol, 250 mg metyrosine, and 10 mg phenoxybenzamine twice daily. A 24-hour urine showed elevated catecholamines (470 μg), metanephrines (l3 mg), and VMA (15.5 mg). CT scan showed multiple pulmonary nodules and osteolytic lesions in T11–12 and L1–5. Bone scan showed metastatic lesions involving the spine, pelvis, skull, and ribs. A ¹³¹I-MIBG scan showed diffuse, increased uptake in the thoracolumbar area.

The patient was started on CVD in September 1990, and her first cycle was complicated by cardiogenic shock after the administration of the chemotherapy. Cycles 2 and 3 administered in the intensive care unit were associated with transient hypotension, and after a fourth inpatient cycle, she went on to receive 62 cycles as an outpatient. After 6 cycles, her performance status improved, and the intensity of bone scan lesions were reduced, and after 25 cycles vincristine was discontinued because of neurotoxicity. With further therapy, the patient continued to have stable disease. However, because of continued stress, chemotherapy was stopped in February 1995 when the 24-hour urinary catecholamine was 113 μg, metanephrines 5.4 mg, and VMA 8.8 mg, and ¹³¹I-MIBG and bone scan showed marked improvement. She did well off chemotherapy until March 2001, when she received palliative RT for a right shoulder bony metastasis, but otherwise proceeded with evidence of gradual disease progression until July 2002, when she died suddenly. An autopsy revealed an unusual variant of anomalous coronary circulation consisting of a single left coronary ostium giving rise to an attenuated right coronary artery passing intramurally through the anterior aortic wall, representing the predisposing cause of death. Additional autopsy findings included the multifocal bilateral lung metastases and a single liver metastasis. Although a complete survey of the skeleton was not undertaken, no bony disease was identified in the vertebral column.

RESULTS

Patient Characteristics

This report describes a total of 18 patients, 14 patients reported previously early in their course who began treatment between 1983 and 1985 and 4 who began between 1989 and 1995. All patients were evaluable for both tumor and biochemical response. Patient characteristics are shown in Tables 1 and 2. All patients had metastatic disease. Nine patients had a diagnosis of paraganglioma. Six patients had previous treatment to localized sites with radiation. Commonly used antihypertensives included phenoxybenzamine, propanolol, atenolol, prazosin, and clonidine.

TABLE 1.

Characteristics of Patients With Malignant Pheochromocytoma/Paraganglioma

Patient Characteristic
Mean age, y [range]	34 [6–64]
Men, n (%)	14 (77)
Women, n (%)	4 (22)
Primary site
Adrenal gland, n (%)	9 (50)
Extra-adrenal, n (%)	9 (50)
Karnofsky performance status [range]	76 [40–90]
Metastatic sites
Retroperitoneum, n (%)	10 (55)
Bone, n (%)	11 (61)
Spinal cord compression, n (%)	5 (27)
Lung, n (%)	9 (50)
Mediastinum, n (%)	4 (22)
Lymph nodes, n (%)	4 (22)
Other,* n (%)	8 (44)

^*Other sites of metastatic disease included the mediastinum, lymph nodes other than retroperitoneal, spleen, urinary bladder, and bone marrow.

TABLE 2.

Characteristics of Patients With Malignant Pheochromocytoma/Paraganglioma

Patient No.	Age at Diagnosis, y	Age at Time of Enrollment, y	Initial Site(s) of Disease	Sites of Metastases	Family History of Cancer	Functional Status	Presumed Mutation
1	26	27	Left retroperitoneal mass	Liver, left femoral neck	Unknown	Y	SDHB
2	64	64	Right adrenal and liver masses	Lungs, liver, multiple bones	Unknown carcinoma	N
3	30	30	Left adrenal mass, multiple lung and liver masses	Lungs and liver	No	Y
4	36	38	Para-aortic and retroperitoneal masses	Anterior mediastinal mass, lungs, multiple bones	No	Y	SDHB or SDHD
5	40	42	Mass in bladder	Urinary bladder, ribs, liver	No	Y	SDHB
6	38	39	Left adrenal mass	Cervical lymph nodes, lungs, liver, retroperitoneal masses	Unknown	Y
7	27	42	Organ of Zuckerkandl	Mediastinum, mesenteric and para-aortic masses, left supraclavicular lymph node	No	Y	SDHB or SDHD
8	32	32	Left abdominal mass and left supraclavicular node	Retroperitoneal masses, lungs, mediastinal mass, peripheral lymph nodes	No	Y	SDHB
9	44	50	Retroperitoneal mass	Right chest wall, left subpleural-based mass, multiple bones, cord compression	No	Y	Possible SDHB
10	40	46	Left adrenal mass	Multiple bones (shoulder, spine, hip, skull, pelvis, rib, thorax, T2, T10, and T12) and lung	No	Y
11	13	17	Left adrenal mass	Liver and retroperitoneal mass	No	Y
12	6	25	Left adrenal mass	Retroperitoneal mass, lungs, left skull	Several with pheochromocytoma	Y	SDHB or SDHD
13	58	61	Right adrenal mass	Lungs, retroperitoneal mass, periaortic lymph nodes, liver, spleen, multiple bones	Bladder cancer	Y
14	23	25	Right adrenal mass	Lungs, liver, retroperitoneal mass, pelvic mass, left ischium	No	Y
15	43	46	Lytic lesion in right hip and left adrenal mass	Pelvis and low lumbar and sacral spine	? Sarcoma	Y
16	19	32	Right neck mass	Right adrenal mass and abdominal mass near celiac artery	Colon cancer	Y	SDHD confirmed
17	30	30	Retroperitoneal mass	Right paratracheal mass, thoracic and lumbar spine, right supraclavicular lymph node, spinal cord compression	No	Y	SDHB or SDHD
18	34	38	Presacral mass	Paraspinal thoracic mass, multiple bones, pelvic mass	No	Y	Possible SDHB

Y indicates yes; SDHB, succinate dehydrogenase subunit B gene; N, no; SDHD, succinate dehydrogenase subunit D gene.

Tumor and Biochemical Response

As of July 2007, all patients were evaluable for tumor and biochemical responses, as shown in Table 3. Two (11%) patients had a complete tumor response, 8 (44%) had a partial tumor response, and 3 (16%) had a minimal response. Biochemical responses were seen in 13 (72%) of the patients, including all patients with complete, partial, and minimal tumor responses. Among the patients whose tumors responded to the treatment, although most responded after 2 to 4 cycles, in 1 patient a tumor response was not recorded until after Cycle 6. The median duration of response is 20 months (Table 4). The median potential follow-up is now 22 years. One patient remains alive, and 1 was lost to follow-up after 5 years. There are 16 confirmed deaths. As shown in Figure 2, the median survival for all patients from the on-study date is 3.3 years. Median survival from date of diagnosis is 6.5 years. Median survival from the landmark date (3 months after starting chemotherapy) is 3.8 years for patients whose tumors had a complete or partial response to chemotherapy and 1.8 years for the rest. The 2-tailed log-rank P value for the comparison of the 2 curves, according to response category, is .65.

TABLE 3.

Tumor and Biochemical Responses

Type of Response	No. (%)
Tumor response
Complete response	2 of 18 (11)
Partial response	8 of 18 (44)
Minimal response	3 of 18 (16)
No response	5 of 18 (28)
Biochemical response
Complete response	3 of 18 (17)
Partial response	10 of 18 (55)
No response	5 of 18 (28)

TABLE 4.

Onset and Duration of Tumor Response Among Patients Whose Tumor Responded to the Treatment

Patient No.	Total No. of Cycles Received	Cycle of Onset of Tumor Response	Duration of Response, mo
1	39	3	26
2	7	2	7
3	13	2	8
4	13	3	18
5	15	2	56
6	21	2	22
7	25	2	20
8	36	3	20
9	43	4	16
10	62	6	126

FIGURE 2.

Kaplan-Meier plots are shown for overall survival. (Left) The survival curve for all patients is shown. (Right) The survival curve of those patients whose tumors responded to cyclophosphamide, vincristine, and dacarbazine therapy versus those whose tumors did not is shown.

Treatment Tolerability and Toxicity

As one might expect, patients whose tumors responded favorably remained on study longer. Patients whose tumors were scored as complete or partial response received a mean of 27.4 cycles of CVD and a median of 23. In the patients whose tumors did not respond to CVD treatment, the mean and median numbers of treatment cycles were 8.75 and 5.5, respectively. Initially, cyclophosphamide doses were escalated; however, if a patient had a nadir ANC less than 0.5 × 10⁹/L on 3 measurements or a platelet count less than 25 × 10⁹/L, the dose in the subsequent cycle was reduced 20% to 50%. The chemotherapy dose administered as a percentage of that planned was 80.0%, 74.7%, and 80.7% for cyclophosphamide, vincristine, and dacarbazine, respectively, in those whose tumors were scored as complete or partial response, and 81.3%, 78.4%, and 93.4% in those without a tumor response or with only minimal shrinkage. The median dose of cyclophosphamide administered in the first 5 cycles compared with that in cycles beyond 5 were 100% and 76% of the planned dose for those patients whose tumors responded to therapy and 100% and 69% of the planned dose for those whose tumors were scored as not responding. The percentage for vincristine in patients with a response is similar to patients without a response. One patient received 43 cycles of vincristine at a full dose of 1.4 mg/m² every cycle, for a cumulative dose of 60 mg/m². Another patient received 36 cycles of vincristine at close to full dose without significant neurological side effects, and 2 other patients received 21 cycles at almost full dose.

As reported previously,³⁵ the common toxicities encountered were myelosuppression, peripheral neuropathy, and gastrointestinal toxicity. Four episodes of hypotension in the first 3 cycles of treatment were observed, and they occurred 3 to 5 days after treatment started. In 1 patient, described above as Patient 2, hypotension occurred in the setting of cardiogenic shock; otherwise all other episodes were easily corrected with volume replacement. Overall the treatment was well tolerated.

DISCUSSION

Up to about 36% of all catecholamine-secreting tumors are malignant,^13–16 with local invasion or distant metastasis the only reliable clues of malignancy. The most effective treatment remains surgical resection. Although prognosis in confirmed malignancy is difficult to predict, the average 5-year survival with metastases is approximately 50%.^30,33,39,40 Although pharmacological control of catecholamine excess and judicious RT provide symptomatic relief and improve well-being, these measures are eventually inadequate, underscoring the need for systemic therapies. We describe the long-term follow-up of 18 patients with pheochromocytoma/ paraganglioma treated with CVD and make recommendations for managing these patients, who are often very difficult to manage medically.

The literature suggests there are 2 subsets of patients with malignant pheochromocytoma/paraganglioma with distinct clinical courses and outcomes.^32,34,41,42 Approximately half have a relatively rapid course, dying within 4 to 5 years of diagnosis.³² The remaining half present a more indolent course and live 20 or more years without cytotoxic therapy. Twelve of our patients had an aggressive course, with early mortality and significant morbidity as a result of tumor dissemination. By comparison, 6 presented a more indolent course, living more than 10 years, including 2 for more than 20 years.

The CVD regimen was originally chosen because its component drugs were effective in neuroblastoma.^43,44 As reiterated here, the regimen is active in patients with malignant pheochromocytoma/paraganglioma. Fifty-five percent of patients had a tumor response, and 72% had a biochemical response. The median duration of response was 20 months. Four patients previously treated with radiation had objective tumor or biochemical responses, which is somewhat surprising because most in vitro and in vivo models would predict some degree of cross-resistance between radiation and both cyclophosphamide and alkylating agents, such as dacarbazine, although a lack of cross-resistance has also been noted.^45–47

The biochemical response correlated closely with tumor response determined by standard criteria and mediated objective improvements in performance status and blood pressure. Biochemical markers were useful in confirming tumor response, relapse, and progression in all the patients except 1. Pending confirmation, it can be argued that serial urine collections may be as sensitive if not more so in following patients whose tumors secrete catecholamines, obviating the need for at least some imaging studies. In addition, functional imaging studies such as ¹³¹I-MIBG scintigraphy and ¹⁸F-fluorodopamine positron emission tomography can help clarify whether abnormalities seen on CT scan retain active tumor cells. For example, an ¹⁸F-fluorodopamine scan showed residual disease in the mediastinum and abdomen of the patient who is still alive 32 years after diagnosis and 18 years after commencing chemotherapy.

As previously reported, CVD was well tolerated, as evidenced by the ability to administer a median of 23 cycles to patients with an objective response. Side effects occurred without significant sequelae and were generally not dose limiting, except in 2 patients in whom vincristine was discontinued after 3 and 25 cycles because of peripheral neuropathy. One patient had cardiogenic shock after receiving chemotherapy—a result of tachyphylaxis after massive catecholamine discharge—and was treated with intravenous phentolamine. Oncologists should be aware that several drugs (Table 5), including several commonly used in oncologic practice, are contraindicated in patients with pheochromocytoma/paraganglioma.⁴⁸ Episodes of massive spontaneous catecholamine discharge after drug administration argue for the administration of the first cycle with close monitoring, possibly in collaboration with an endocrinologist.

TABLE 5.

Drugs to Be Avoided in Pheochromocytoma/Paraganglioma Patients

ACTH

Amphetamine

Cocaine

Droperidol

Deprenyl

Glucagon

Histamine

Methamphetamine

Metoclopramide

Morphine

Phentermine

Phendimetrazine

Phenylethylamine

Saralasin

Steroids

Tricyclic antidepressants

Tyramine

Vasopressin

ACTH indicates corticotropic hormone.

Should CVD be administered routinely to patients with unresectable pheochromocytoma? The median survival from the landmark date 3 months after starting chemotherapy is 3.8 years for patients with objective tumor responses and 1.8 years for nonresponders. However, the 2-tailed log-rank P value for the comparison of the 2 curves, according to response category, was .65. Thus, based upon long-term follow-up, there is no statistically significant difference in overall survival (OS) between patients whose tumors responded to therapy compared with those whose tumors did not respond. The lack of statistical significance in OS may be attributed to the small number of patients treated with CVD. In the absence of larger numbers, we would argue that an appropriate strategy would be to treat patients with CVD if their disease is symptomatic, either because of difficulty controlling the symptoms of catecholamine excess pharmacologically or as a result of painful metastases. We would also propose one may try this therapy in a neo-adjuvant approach to render a tumor amenable to surgical resection, as was achieved in Patient 1. But we would not advocate at present that all patients receive CVD in the face of metastases. The long-term survival of many patients with metastases counsels a prudent approach guided by the clinical course. We would argue that patients with an indication for starting therapy who benefit from CVD continue therapy indefinitely, provided the patient is tolerating the therapy well and symptomatic improvement has been obtained.

As for patients previously treated with CVD, we would note anecdotally that in several patients reinstitution of CVD has been ineffective, even when therapy had been discontinued without evidence of progressive disease. Other chemotherapy regimens tested in small numbers of patients with various response rates include cisplatin and 5-fluorouracil,⁴⁹ etoposide, carboplatin, vincristine, cyclophosphamide and adriamycin,⁵⁰ a modified CVD regimen adding an anthracycline,⁵¹ and oral temozolomide.⁵² Administration of these regimens should be balanced against recommending enrollment in a clinical trial, the option we encourage.

Finally, we would note that recent advances identifying genetic factors predisposing to pheochromocytoma/paraganglioma might further clarify the clinical behavior and response to therapy. It would not be surprising to find that genetic factors underlie a better or worse prognosis or response to chemotherapy. Unfortunately, enrollment of our patients in the distant past precluded a genetic analysis.

In summary, this report shows that CVD provides effective palliation in patients with progressive, advanced malignant pheochromocytoma/paraganglioma without a clear impact on OS. Most patients with an objective response were clinically better in the opinions of treating physicians and had decreased requirements or discontinuation of blood pressure medications. We recommend that this effective, well-tolerated regimen be reserved for patients with metastases with symptoms not controlled medically, patients with unresectable disease in an attempt to render their disease resectable, and patients with rapidly growing tumors. Further investigations should identify therapies that will result in higher response rates and perhaps increased survival.