High-dose ¹³¹I-metaiodobenzylguanidine therapy for 12 patients with malignant pheochromocytoma

Abstract

BACKGROUND

¹³¹I-Metaiodobenzylguanidine (¹³¹I-MIBG) can be used systemically to treat malignant pheochromocytoma. To improve outcome, the authors used higher levels of activity of ¹³¹I-MIBG than previously reported. The authors reported the response rates and toxicity levels in patients with malignant pheochromocytoma or paraganglioma who were treated with high-dose ¹³¹I-MIBG.

METHODS

Following debulking surgery and stem cell harvest, 12 patients with malignant pheochromocytoma or paraganglioma were treated with ¹³¹I-MIBG. Five had received previous external beam radiation and/or chemotherapy. The median single treatment dose was 800 mCi (37 gigabecquerels; range, 386–866 mCi) or 11.5 mCi/kg (range, 5.6–18.3 mCi/kg). The median cumulative dose was 1015 mCi (range, 386–1690 mCi).

RESULTS

Three patients had a complete response, two of whom had soft tissue and skeletal metastases. Their median follow-up was 45 months (range, 23–101 months). Seven patients had a partial response (PR), with a median follow-up 43 months (range, 6–47 months). Two patients without a response died with progressive disease (PD) and 2 patients with an initial PR died of PD at 13 and 11 months, respectively. Grade 3 thrombocytopenia occurred after 79% (15 of 19) of treatments had been administered. Grade 3 and 4 neutropenia followed 53% (10 of 19) and 19% (4 of 19) of treatments, respectively. One patient required stem cell infusion, and one developed primary ovarian failure.

CONCLUSIONS

The single and cumulative doses of ¹³¹I-MIBG were approximately 2–3.5 times higher than those used at other centers. Unlike previous reports, two patients with both skeletal and soft tissue metastases had a complete response. Hematologic toxicity was significant but tolerable. High-dose ¹³¹I-MIBG may lead to long-term survival in patients with malignant pheochromocytoma. Cancer 2003;98:239–48. © 2003 American Cancer Society.

DOI 10.1002/cncr.11518

Malignant pheochromocytoma and paraganglioma (extraadrenal pheochromocytoma) account for 10–20% of all cases of pheochromocytoma.1, 2 Patients with malignant pheochromocytoma have a poor prognosis with an average 5-year survival of about 40%.3 Typically, major primary and metastatic lesions are debulked surgically. Antihypertensives and α-methyltyrosine block the effects or synthesis of catecholamines. However, neither treatment alters tumor growth. Systemic treatment options are limited and include chemotherapy and radioiodine-labeled metaiodobenzylguanidine (MIBG). To improve outcome, we developed a treatment protocol infusing activity of ¹³¹I-MIBG of up to 18 millicuries (mCi) per kilogram of body mass, to a maximum of about 850 mCi per treatment. These doses are larger than those previously reported.

Metaiodobenzylguanidine is a guanethidine analog that resembles norepinephrine and is concentrated similarly by adrenergic tissues. In 1981, MIBG was shown to localize pheochromocytomas.4 ¹³¹I-Metaiodobenzylguanidine therapy for malignant pheochromocytomas first was reported in 1984.5 A review of the world's experience with ¹³¹I-MIBG for the treatment of malignant pheochromocytoma reported tumor and hormonal responses of 30% and 45%, respectively.6 The single-treatment dose of ¹³¹I-MIBG ranged from 96 to 300 mCi, with a mean cumulative dose of 490 mCi (range, 96–2322 mCi). There had been only five documented cases of complete responses (CR) following ¹³¹I-MIBG therapy and no reported CRs in patients with both skeletal and soft tissue metastases. However, the effect of higher treatment doses had not been determined.

A Phase I ¹³¹I-MIBG treatment dose-escalation study of children with malignant neuroblastoma reported a higher response rate with increasing ¹³¹I-MIBG activity. In that study, 1 of 9 patients was treated with 9 mCi/kg or less, and 10 of 21 patients received doses of 12 mCi/kg or more.7 Grade 4 hematologic toxicity occurred in greater than 80% of the patients treated with doses of 15–18 mCi/kg body weight, indicating the need for stem cell harvest before such therapy. Because pheochromocytomas are similar to neuroblastomas in many respects, we decided to determine whether therapy with higher doses of ¹³¹I-MIBG would lead to an improved response rate in patients with malignant pheochromocytoma and paraganglioma. We treated 12 patients with high-dose ¹³¹I-MIBG and report their response rates and toxicity.

MATERIALS AND METHODS

Patient Eligibility

Patients were required to have metastatic or unresectable pheochromocytoma or paraganglioma. The metastases were required to concentrate ¹²³I-MIBG or ¹³¹I-MIBG to at least twice that of soft tissue background. Massive hepatic metastasis precluded treatment with high-dose ¹³¹I-MIBG due to concerns that such treatment might cause hepatic necrosis. Patients were required to be older than age 1 year, with a minimum life expectancy of at least 12 weeks. No chemotherapy or radiation treatments were given for at least 3 weeks before ¹³¹I-MIBG therapy. All the patients who received previous chemotherapy or radiation had documented progressive or persistent disease on computed tomography scan (CT) and ¹²³I-MIBG before receiving ¹³¹I-MIBG therapy. Patients were required to have normal renal and cardiac function, liver enzyme levels less than twice the upper limit of normal, as well as an absolute neutrophil count (ANC) and platelet count greater than 1 × 10⁹ and 50 × 10⁹/L, respectively. From 1991 to 2000, 12 patients with documented pheochromocytoma or paraganglioma with metastases were treated with high-dose ¹³¹I-MIBG. Five patients have been reported already.6, 8-11

Pretreatment Procedures

Pretreatment evaluation included a history and physical examination. A 24-hour urine sample was collected to evaluate total fractionated catecholamines and metanephrines. Serum specimens were assayed for glucose, calcium, albumin, electrolytes, blood urea nitrogen, and creatinine. Pulmonary function tests were performed in patients with lung metastases. Echocardiograms were obtained for patients who had significant cardiac uptake on a ¹²³I-MIBG scan. Bone marrow devoid of tumor was required before stem cell harvest. Two patients (Patients 1 and 3) had bone marrow harvests. The rest of our patients had stem cells collected by leukapheresis. Blood pressure was controlled before treatment. Patients did not receive medications that were known to interfere with MIBG uptake (e.g., labetalol, decongestants, phenothiazines, tricyclic antidepressants). All patients had their primary tumor resected or debulked before receiving ¹³¹I-MIBG therapy. The protocol was approved by the University of California–San Francisco (UCSF), Committee on Human Research. Informed consent was obtained from all patients or their parents.

¹³¹I-Metaiodobenzylguanidine Preparation

Metaiodobenzylguanidine was synthesized and exchange-labeled at UCSF as described previously.12, 13 The free iodine content at the time of dosing was less than 5%. The specific activity was 10–12 Ci/mmol MIBG.

Administration

Patients were admitted to a lead-shielded room. They were hydrated with intravenous fluids beginning 12 hours before the infusion of ¹³¹I-MIBG. The maximal single treatment activity delivered was 18 mCi/kg up to a maximal single dose of 850 mCi. The intended single treatment dose of 12–18 mCi/kg varied predominately secondary to patient's weight. Patient 1 was treated during the dose escalation phase of our protocol and therefore was treated with a smaller dose of ¹³¹I-MIBG. A bladder catheter was inserted and drained into a lead-shielded container to reduce radiation exposure to the bladder. The bladder catheter was left in place and drained continuously for about 3 days following the infusion of ¹³¹I-MIBG, or until at least 75% of the isotope was excreted. The dose was administered over 2 hours from behind an additional lead shield of 2.5 cm thickness. To block thyroid uptake of unbound ¹³¹I, potassium iodide (KI) was given at least 2 hours before the infusion (6 mg/kg load), followed by 0.88 mg/kg orally every 4 hours for 7 days and then 1 mg/kg orally for 45 days after the infusion. A loading dose of 8 mg/kg potassium perchlorate was given at least 2 hours before ¹³¹I-MIBG followed by 2 mg/kg orally every 6 hours for 5 days posttreatment. The patients remained behind a lead shield until radiation decreased to less than 2 millirads (< 0.02 mSv) per hour at 1 meter.

Posttreatment Follow-Up

Whole-body gamma scans were performed 5 days and 2–3 months after ¹³¹I-MIBG therapy. A CT or magnetic resonance imaging scan was ordered at 2–3 months following ¹³¹I-MIBG treatment and regularly thereafter. A complete blood count with differential was obtained twice per week for 6 weeks and then monthly until normal. Serum liver enzyme and creatinine levels were measured once per week for 6 weeks, and then monthly for 1 year. Urine levels of catecholamines and metanephrines were measured at 3, 6, and 12 months. Thyroid function tests were obtained at 1, 2, 3, 6, and 12 months posttreatment. An autologous bone marrow or stem cell infusion was performed per physician discretion, using the following guidelines: ANC less than 200 × 10⁹/L for more than 2 weeks or if dependent on platelet transfusions for at least 3 weeks. Platelet and red blood cell (RBC) transfusions and granulocyte–colony-stimulating factor (G-CSF) were used at the discretion of the patients' local oncologists.

Retreatment

Repeat ¹³¹I-MIBG treatments were administered on a case-by-case basis in an effort to improve the overall response. Multiple factors were considered, such as previous response and toxicity, tumor burden, and total radiation. The median interval between treatments was 4.4 months (range, 3–7 months). Patients were required to have normal hematologic, hepatic, renal, and cardiac function before repeat ¹³¹I-MIBG treatment and to be free of active infection. The need for stem cell infusion or an autologous bone marrow transplant precluded further ¹³¹I-MIBG therapy.

Definitions

All responses were evaluated with respect to measurements obtained before the initial ¹³¹I-MIBG treatment. Hormonal responses were classified by the change in urinary total levels of catecholamines or metanephrines as follows: CR, normalization; partial response (PR), reduction by ≥ 50%; no change (NC), reduction by < 50% or an increase of < 25%; progressive disease (PD), an increase of ≥ 25%. Tumor responses on CT scan were determined by the maximal tumor diameter of the predominant lesions and defined as follows: CR, complete resolution of all tumors on imaging; PR, a reduction of ≥ 50% in the measured maximal tumor diameter; stable disease (SD), no change; and PD, an increase of ≥ 25% in a single lesion or the development of new lesions. Metaiodobenzylguanidine responses were defined qualitatively with respect to the distribution of ¹²³I-MIBG uptake as follows: CR, complete resolution of ¹²³I-MIBG uptake; PR, reduction of distribution of uptake; SD, no change; and PD, an increase in the regions of uptake. Best response refers to the greatest level of response per symptom, urinary catecholamine/metanephrine levels, ¹²³I-MIBG scan, and CT scan. The best response was maintained at follow-up, unless the patient developed PD. Hematologic toxicities were classified according to the National Cancer Institute's common toxicity criteria, Version 2.0.

Statistical Methods

Follow-up was measured from the first ¹³¹I-MIBG treatment. The duration of response was measured from the initial ¹³¹I-MIBG treatment to the initial date of disease recurrence or PD. Except where otherwise stated, values are expressed as median (range). Student t tests using the Sigma Stat Version 2.0 statistical package (SPSS, Inc., Chicago, IL) were performed to compare age responders versus nonresponders, cumulative doses and mCi/kg values for responders and nonresponders, and platelet and ANC grade versus total cumulative dose and mCi/kg. Wilcoxon signed rank tests were used to compare nonparametric values. A two-tailed Fisher exact test was used to compare proportions. P < 0.05 was considered significant.

RESULTS

Patient Characteristics

Twelve patients were treated with high-dose ¹³¹I-MIBG (Table 1). Their median age was 31 years (range, 10–58 years). Five patients were younger than age 18 years. Six patients had malignant pheochromocytoma, and five had malignant paraganglioma. Patient 7 had multiple peritoneal metastases secondary to intraabdominal seeding of his tumor.11 These lesions were clearly visible on CT scan and were avid for ¹²³I-MIBG.

Table 1. Characteristics of 12 Patients with Malignant Pheochromocytoma/Paraganglioma Treated with ¹³¹I-Metaiodobenzylguanidine

Patient no.	Age (yrs)	Gender	Type	Metastatic locations before ^131I-MIBG treatment	Previous treatmenta	Mean dose/treatment (mCi/kg)	No. of treatments	Cumulative dose (mCi)	Best responseb				Duration of response (mos)c	F/U (mgs)d	Current status (mos)d
									Symptom	Hormone	MIBG	CT
1	50	F	Pheochromocytoma	Retroperitoneum/lungs/skeletal		5.6	1	386	PD	PD	PD	PD	0	11	DPD
2	42	M	Paraganglioma	Abdomen/skeletal	Rad	7.7	2	1640	PR	PRe	PR	PR	6	13	DPD
3	44	M	Pheochromocytoma	Retroperitoneum/lungs/abdomen		8.1	1	845	PR	PR	PR	SD	45	45+	AWD
4	17	F	Paraganglioma	Thigh/buttock/skeletal		9.2	1	557	CR	NS	CR	CR	101	101+	ANoD
5	58	M	Pheochromocytoma	Liver/skeletal	Chemo	10.2	2	1650	CR	CR	PR	SD	47	47+	AWD
6	28	M	Paraganglioma	Liver/abdomen/retroperitoneum/skeletal	Rad	10.6	1	770	PR	NE	PD	PD	0	8	DPD
7f	34	M	Pheochromocytoma	Abdomen/liver/retroperitoneum	Rad	11.3	1	875	CR	CR	CR	CR	24	24	ANoD
8	48	M	Pheochromocytoma	Abdomen/lungs/liver/skeletal	Rad Chemo	11.4	2	1717	PR	PR	PR	PR	7	11	DPD
9	14	M	Paraganglioma	Abdomen/retroperitoneum/skeletal		12.4	2	1605	CR	NE	PR	SD	43	43+	AWD
10	11	F	Paraganglioma	Retroperitoneum/skeletal		17.7	3	1689	CR	NS	PR	SD	44	44+	AWD
11	15	M	Pheochromocytoma	Mediastinum/skeletal		18.2	1	800	None	NS	CR	NE	45	45+	ANoD
12	10	M	Paraganglioma	Thorax/groin/axilla/skeletal	Rad	18.0	2	1155	None	NC	PR	SD	25	32+	AWPD

• mCi: millicuries (1 mCi = 0.037 gigabecquerels); MIBG: metaidobenzylguanidine; CT: computed tomography; F/U: follow-up; PD: progressive disease; DPD: died with progressive disease; Rad: external beam radiation; PR: partial response; SD: stable disease; AWD: alive with disease; CR: complete response; NS: not secretory; AnoD: alive with no disease; Chemo: chemotherapy; NE: not evaluable.
• a All patients had surgery before ¹³¹I-MIBG treatment.
• b Greatest response following ¹³¹I-MIBG treatment.
• c Interval from initial ¹³¹I-MIBG treatment to either progressive disease or follow-up.
• d Months of follow-up after the first ¹³¹I-MIBG treatment.
• e Urinary measurement not available; partial response with respect to chromogranin A.
• f Patient 7 had intraabdominal seeding of his primary pheochromocytoma.

Of the 12 patients, 10 had skeletal metastases at the time of the initial ¹³¹I-MIBG treatment. The skull was the most common site of metastasis at initial presentation; skull metastases were observed in eight patients. Other locations of metastatic disease included the vertebral spine in seven patients; the pelvis and abdominal cavity in six each; the lung in five; the liver, femur, and shoulder in four each; and the ribs and thigh soft tissue in two each. Patient 11 had reproducible ¹²³I-MIBG and posttreament ¹³¹I-MIBG uptake in the skull and vertebrae. Metaiodobenzylguanidine scanning with reproducible uptake in bone is highly specific for metastatic pheochromocytoma. The lack of observed cortical disruption per CT scan was attributed to the smallness of the metastatic lesions. The ¹²³I-Metaiodobenzylguanidine uptake in the thigh and buttock in Patient 4 was reproducible and therefore was not attributed to urine contamination. No patients had involvement of the central nervous system.

Five patients had received chemotherapy or radiotherapy before ¹³¹I-MIBG treatment. The median time from the discovery of metastases to therapy with ¹³¹I-MIBG was 9 months (range, 1.4–53 months).

The median single treatment activity of ¹³¹I-MIBG was 800 mCi (29.2 gigabecquerels [GBq]; range, 386–866 mCi) or 11.5 mCi/kg (range, 5.6–18.3 mCi/kg). Six patients received one dose, five patients received two doses, and one patient received three doses of ¹³¹I-MIBG. The activities of repeat treatments were on average within 2% of previous treatments. The median cumulative dose was 1015 mCi (37 GBq) ¹³¹I-MIBG (range, 386–1717 mCi; Table 2).

Table 2. Characteristics of 12 Patients with Malignant Pheochromocytoma/Paraganglioma and Response to Treatment with ¹³¹I-Metaiodobenzylguanidine

Characteristic	Median (range)	Frequency
Age (yrs)	31 (10–58)	—
Follow-up after initial ^131I-MIBG treatment (mos)	38 (8–101)	—
Activity of ^131I-MIBG per treatment (mCi)	800 (386–866)	—
Activity of ^131I-MIBG per treatment (mCi/kg)	11.5 (5.6–18.3)	—
Cumulative activity of ^131I-MIBG (mCi)	1015 (386–1690)	—
Duration of response (mos)
Symptomatic
PR + CR	43 (2–101)	9/10
Hormonal
PR + CR	34 (7–47)	4/6
MIBG
PR + CR	44 (6–101)	10/12
CT
PR + CR	15 (6–101)	4/11
SD	44 (25–47)	5/11

• ¹³¹I-MIBG: ¹³¹I-metaiodobenzylguanidine; mCi: millicuries (1 mCi = 0.037 gigabecquerels); PR: partial response; CR: complete response; SD: stable disease.

Response to ¹³¹I-Metaiodobenzylguanidine Therapy

After ¹³¹I-MIBG therapy, the tumor response was evaluated with respect to pretreatment symptoms, urinary levels of catecholamines and metanephrines, and tumor reduction per ¹²³I-MIBG and CT scan. An initial response was measured at a median of 2.4 months (range, 1.2–4.7 months).

Pain and other symptoms were reduced in the majority of patients after treatment with ¹³¹I-MIBG (Tables 1, 2). Ten had symptoms at baseline. Patient 1 did not respond. She was treated in the dose escalation phase of our protocol with a comparatively low level of ¹³¹I-MIBG activity. A PR occurred in 4 patients, for a median duration of 6.5 months (range, 2–45 months). Patient 6 had a 2-month interval of reduced pain and was not treated with another dose of ¹³¹I-MIBG because of a surgical infection. Five patients experienced CR, for a median duration of 44 months (range, 23–101 months). Patient 10, who was dependent on opioid analgesics before treatment, had a complete resolution of her pelvic pain and administration of opioid analgesics was discontinued within 2 weeks of her initial ¹³¹I-MIBG treatment. With the exception of Patient 6, the improvement or progression of symptoms coincided with the hormonal response in all patients with secretory malignancies.

A reduction (PR or CR) in total urinary levels of catecholamine or metanephrine excretion following ¹³¹I-MIBG treatment was observed in 4 of 6 patients (Figs. 1, 2; Table 1), for a median duration of 34 months (range, 7–47 months). The tumor did not secrete catecholamines or metabolites in three patients (Patients 4, 10, and 11). Patient 7 had a complete normalization of elevated norepinephrine levels. His metanephrine levels have remained normal. Patient 9 was not evaluable secondary to surgical reduction of an intraabdominal tumor between the first and second treatments. Patient 12 had a period of normalization; however, pretreatment values were not greatly elevated, leading to a designation of NC. Urinary levels of catecholamines and metanephrines returned to normal in Patient 8 at 7 months after therapy, before PD. Serum levels of chromogranin A (CgA) were not measured in all patients. Patient 2 had an elevated level of CgA of 137 ng/mL (1.6–5.6 ng/mL) that decreased 2 months after ¹³¹I-MIBG to 56 ng/mL, increased to 470 ng/mL at 6 months, and further progressed to 1430 ng/mL in concordance with PD on imaging studies.

Figure 1

Total urinary levels of metanephrines versus time after the initial ¹³¹I-metaiodobenzylguanidine (¹³¹I-MIBG) treatment. Patient 12 received two ¹³¹I-MIBG doses before the first data point. Patients 4, 10, and 11 had nonsecretory malignant pheochromocytomas or paragangliomas. Patients 1, 2, 6, and 9 had incomplete data. The dashed lines represent the upper and lower bounds of the normal range.

Figure 2

Total urinary levels of catecholamines versus time after the initial ¹³¹I-metaiodobenzylguanidine (¹³¹I-MIBG) treatment. Patients (Pts) 4, 10, and 11 had nonsecretory malignant pheochrmocytoma or paraganglioma. Patients 2, 3, and 6 had incomplete data. The dashed lines represent the upper and lower bounds of the normal range.

Of the 12 patients in the current study, 7 had a PR with respect to the ¹²³I-MIBG scan, with a median duration of 43 months (range, 6–47 months). Three of these seven patients subsequently developed PD. A CR as determined by ¹²³I-MIBG scanning occurred in three patients. Two patients did not respond to treatment. A CT scan of Patient 1 showed multiple pulmonary metastases that did not exhibit ¹³¹I-MIBG uptake.

CT scans were available for evaluation in 11 patients. Two patients (Patients 2 and 8) had a PR before dying of PD. Five had SD on CT scan, with a median duration of 44 months (range, 25–47 months). Per CT scan, SD was observed secondary to a lack of measurable change of skeletal lesions. Patient 12 had new skeletal metastases at 25 months of follow-up. Patients 4 and 7 had a CR per CT scan.

With respect to current status, 4 patients have died secondary to PD, after a median follow-up 11 months (range, 8–13 months). Eight patients are alive at a median follow-up of 45 months. Of these, four have clinical signs of improvement, two with minimal disease (Patients 5 and 10). One patient has PD, and 3 have experienced a CR and currently are without evidence of disease, at a median follow-up of 45 months (range, 24–101 months). The median follow-up for all patients was 38 months (range, 8–101 months). The median duration of response was 34 months (range, 0–101 months).

Patients with a CR or PR, as evidenced by urinary levels of catecholamines/metanephrines, exhibited a reduction in tumor size as observed by an MIBG scan. In addition, imaging studies showed that patients with a urinary hormonal response had PD. Two of the three patients with a CR had nonsecretory tumors. The Student t test did not show a statistical relationship when comparing the dose of ¹³¹I-MIBG (cumulative or cumulative dose per kilogram) of those with a hormonal, symptom, ¹²³I-MIBG, or CT response versus those with PD. The Fisher exact test showed a trend (P = 0.067) toward the occurrence of PD in patients who received external beam radiation before ¹³¹I-MIBG treatment. A greater proportion of patients older than age 18 years experienced PD within 12 months after initial ¹³¹I-MIBG treatment compared with patients who were younger than age 18 years (P = 0.08).

Toxicity from ¹³¹I-Metaiodobenzylguanidine Therapy

Both hematologic (Table 3) and nonhematologic toxicities were common but not life-threatening. There was no clear association between treatment dose, cumulative dose, or patient age and the grade of hematologic toxicities within the treatment range utilized in this study. Grade 3 thrombocytopenia occurred in 79% (Grade 4 thrombocytopenia was not observed). The median onset of Grade 3 thrombocytopenia was 27 days (range, 18–40 days). The median time to normal platelet levels without the need for transfusions was 16 days (range, 1–31 days).

Table 3. Grade of Hematologic Toxicity in 12 Patients after ¹³¹I-MIBG Treatment (n = 19)

	Thrombocytopenia gradea		Neutropenia gradeb
	1–2	3	1–2	3	4
Frequency (%)	4 (21)	15 (79)	5 (26)	10 (53)	4 (19)
Onset (days)
Median	28	27	34	32	30
Range	26–40	18–40	21–35	20–40	24–41
^131I-MIBG activity
Median (range; mCi/kg)	12.4 (5.6–18.2)	11.3 (7.7–18.3)	10.2 (5.6–12.4)	14.6 (8.1–18.3)	10.0 (7.7–17)
Median (range; GBq/kg)	0.46 (0.21–0.67)	0.42 (0.28–0.68)	0.38 (0.21–0.46)	0.54 (0.30–0.68)	0.37 (0.28–0.63)

• ¹³¹I-MIBG: ¹³¹I-metaiodobenzylguanidine;
• a Grade 1: > 75 × 10⁹/L (normal); Grade 2: ≥ 50 × 10⁹/L, ≤75 × 10⁹/L; Grade 3: > 10 × 10⁹/L, < 50 × 10⁹/L; Grade 4: ≤ 10 × 10⁹/L. mCi: millicuries; GBg: gigabecquerels.
• b Grade 1: > 1.5 × 10⁹/L, < 2.0 × 10⁹/L; Grade 2: > 1.0 × 10⁹/L,< 1.5 × 10⁹/L; Grade 3: > 0.5 × 10⁹/L, < 1.0 × 10⁹/L; Grade 4: ≤ 0.5 × 10⁹.

Grade 1–2 neutropenia occurred in 5 of 19 (26%) patients, and Grade 3 neutropenia occurred in 10 of 19 patients (53%). Grade 4 neutropenia occurred in 4 of 19 patients (19%). The median onset of Grade 3–4 neutropenia was 32 days (range, 20–41 days). An increase in the grade of hematologic toxicity was not observed in the six patients who had repeat treatments (data not shown). Patient 2 had stem cell replacement after his second treatment after experiencing 1 week of neutropenic fever. He had been treated with extensive external beam radiotherapy before receiving ¹³¹I-MIBG treatment.

Platelet and RBC transfusion data were available for 17 of 19 treatments. Platelets were transfused following 8 (47%) ¹³¹I-MIBG treatments. Packed RBCs were transfused following 6 of 17 (35%) ¹³¹I-MIBG treatments. Following 10 treatments (56%), G-CSF was used. There was no apparent association between treatment dose (mCi/kg) and the use of hematologic support.

The severity of nausea and emesis was recorded after 15 of 19 treatments. Following ¹³¹I-MIBG treatments, 12 of 15 patients continued to be able to eat. Of the 12 patients, 5 had 2–5 episodes of emesis per 24 hours following treatment. Transient episodes of hypertension were observed in three patients during the infusion of ¹³¹I-MIBG and were treated with nifedipine. Other complications included parotitis in three patients, herpes zoster in three patients, thrush in two patients, bacteremia in one patient, and pneumonia in one patient. Patient 10 was treated with a cumulative dose of 1680 mCi of ¹³¹I-MIBG and developed primary ovarian failure. No patients developed hypothyroidism.

DISCUSSION

Patients with malignant pheochromocytoma and paraganglioma have a median survival of 4.5 years, although tremendous variability is observed. Patients with metastases have survived for more than 10 years without tumor-reducing therapy.14, 15 Recently, a patient with skeletal metastasis was reported as having a 26-year survival period without chemotherapy or radiation treatment.16 Therefore, it is difficult for a single institution to effectively evaluate therapies for this disease.

Surgery for primary lesions and large metastatic lesions can improve morbidity resulting from local tumor and systemic catecholamine effects. However, surgery cannot treat multiple diffuse metastatic lesions. Disease remissions have been reported with chemotherapy. Of 14 patients with malignant pheochromocytoma who were treated with a combination chemotherapy regimen of cyclophosphamide, vincristine, and dacarbazine (CVD), 79% had a hormone response and 57% had a tumor PR or CR.17 However, continued cycles of chemotherapy were required to maintain disease remission, and long-term follow-up was not provided. Another study reported a median survival of 67 months (range, 12–300 months) after various regimens.18 Multiple case reports have illustrated palliation with intermittent chemotherapy with a survival period of more than 4 years.19-21 Complete long-term disease remissions with chemotherapy, to our knowledge, have not been reported.

Therapy with ¹³¹I-MIBG has been administered at multiple centers throughout the world. In 1997, Loh et al.6 published a review of the worldwide experience involving 116 patients treated with ¹³¹I-MIBG for malignant pheochromocytoma. Overall, there was a symptomatic response in 76%, a hormonal response in 43 of 96 (45%), and tumor regression in 35 of 116 patients (30%). The activity of ¹³¹I-MIBG per single dose was 96–300 mCi, and the mean cumulative activity (± standard deviation [SD]) was 490 ± 350 mCi. Only five CRs to ¹³¹I-MIBG were reported. All patients who had CR had minimal soft tissue metastases. In 2001, Mukherjee et al.22 published a report on a series of 37 patients with neuroendocrine tumors who were treated with serial doses of ¹³¹I-MIBG. Of these patients, 15 had either malignant pheochromocytoma or paraganglioma. The symptomatic response rate was 100%, and a complete or partial hormonal response occurred in 8 of 9 patients; tumor reduction was observed in 7 of 13 patients (54%). The mean single treatment dose of ¹³¹I-MIBG was 189 mCi (range, 70–300 mCi) and the mean cumulative activity was 638 mCi (range, 200–1282 mCi). No CRs were reported in their series.

The mean single treatment and cumulative activity of ¹³¹I-MIBG used in our study was about 2–3.5 times higher than that employed at other centers.6, 22, 23 Following high-dose therapy with ¹³¹I-MIBG, 3 of the 12 patients in the current study experienced a CR that has been sustained with up to 8 years of follow-up. Unlike previous reports of CR, two of these patients had both skeletal and soft tissue metastases.

In addition, two patients alive with disease are asymptomatic with minimal tumor burden. A reduction in symptoms and urinary catecholamine/metanephrine levels occurred in about 70% of patients. In contrast, lower dose protocols often call for a greater number of treatments than the one to three treatments used in our series, and CVD chemotherapy must be given in repeated cycles indefinitely to maintain a response.6, 17, 22, 23

Results for two of the patients with a CR require further clarification. Patient 7 had intraperitoneal seeding of a pheochromocytoma at the time of his original surgical resection. The natural history of this pheochromocytoma is unknown and may behave differently than that of a malignant pheochromocytoma. The patient had progression of both catecholamines and tumor burden that required a second major surgery. All tumor foci were not respectable, and further progression occurred. Although it is impossible to distinguish benign pheochromocytoma from malignant pheochromocytoma histologically, Patient 7 did not have distant metastases. However, his disease behaved aggressively and his CR following ¹³¹I-MIBG therapy was noteworthy. Patient 11 did not have skeletal metastases visible per CT scans but had reproducible uptake per ¹²³I-MIBG scans to the skull and vertebrae. Unlike Technetium 99m bone scans, MIBG scanning with reproducible uptake in bone is highly specific for metastatic pheochromocytoma. We consider the resolution of ¹²³I-MIBG scans in Patient 11 to be evidence of a CR to therapy. Malignant pheochromocytoma can dedifferentiate and lose avidity for ¹²³I-MIBG uptake. However, this would be less likely in Patient 11, because there was no evidence of disease recurrence after nearly 4 years of follow-up.

The 3 patients in the current series who experienced CR had a median follow-up of 45 months. Following primary tumor resection in these patients, a small number of metastases remained. Therefore, these patients had only a relatively small tumor burden when treated with ¹³¹I-MIBG. It is not possible to say whether the relatively large dose of ¹³¹I-MIBG therapy contributed to their CRs. Our experience suggests that skeletal metastases frequently respond to ¹³¹I-MIBG therapy. In addition, two of these three patients had nonsecretory tumors (per urine levels of catecholamines/metanephrines). Nonsecretory malignant pheochromocytomas may retain the ability to concentrate ¹³¹I-MIBG. Patients with nonsecreting tumors may respond to ¹³¹I-MIBG therapy.

The SD observed per CT scan in five patients may be representative of the described patient population with stable malignant lesions. However, all of these patients had a PR per ¹²³I-MIBG scan, improvement of symptoms, and reduction of previously elevated urinary levels of catecholamines/metanephrines. We attribute these improvements to the ¹³¹I-MIBG treatment. The lack of change per CT scan may be secondary to delayed recalcification of bone metastases.23

There was a trend toward an increased likelihood of PD for patients who had received external beam radiation before ¹³¹I-MIBG treatment. It is possible that this trend is simply the result of selection bias; patients who were treated with external beam radiation had extensive or aggressive metastatic lesions. Alternatively, previous exposure to external beam radiation may reduce the effectiveness of ¹³¹I-MIBG therapy. Patients in the current series typically had poor uptake of ¹³¹I-MIBG into areas of the tumor that previously had been treated with external beam radiation.

Serum levels of CgA are correlated with tumor weight and response to chemotherapy in patients with malignant pheochromocytoma.24, 25 Serum CgA levels were not routinely available at the start of our protocol. However, we currently are employing assays of CgA routinely. As suggested by Patient 2, serum CgA levels may be helpful in following the response of malignant pheochromocytoma to ¹³¹I-MIBG therapy. Hematologic toxicity with the higher dose of ¹³¹I-MIBG usually was transient but was much greater than that reported with lower-dose protocols. In a series of 15 patients with a mean single activity of 127 mCi ¹³¹I-MIBG and a mean cumulative activity of 624 mCi (range, 300–2322 mCi), the platelet count decreased to 60–100 × 10⁹/L in only 2 patients.23 Leukopenia at a level of 1.6–2.0 × 10⁹/L occurred in 2 patients. In another recent study, only 1 of 15 patients developed bone marrow suppression when relatively low doses of ¹³¹I-MIBG were employed.22 In our study of high-dose ¹³¹I-MIBG, we observed Grade 3–4 thrombocytopenia and neutropenia in patients after about 80% of treatments. As in other series of patients with malignant pheochromocytoma treated with ¹³¹I-MIBG, we did not observe a correlation between dose (mCi/kg) and hematologic toxicity.6, 23 However, because about 80% of patients with neuroblastoma who were treated with ≥ 12 mCi/kg developed Grade 4 hematologic toxicity,7 the lack of a correlation in the current series can be attributed to patients' being on the plateau phase of the hematologic toxicity curve. Stem cell replacement was necessary in one patient in our series; this patient previously had been treated with extensive external beam radiotherapy for widespread skeletal metastases.

Patient 10 developed primary ovarian failure. The ovarian failure likely occurred secondary to local radiation from extensive pelvic bone metastases. Ovarian failure previously was reported in a 36-year-old woman with retroperitoneal and spinal metastases after a cumulative dose of 1233 mCi.22 The boys in our series have not shown abnormal pubertal development. Secondary to the development of hypothyroidism in a past neuroblastoma protocol using only KI, we adapted by using both KI and potassium perchlorate.7 Despite using both medications, the thyroid has been visualized on posttherapy scans. We have not observed hypothyroidism.

Patients treated with high-dose ¹³¹I-MIBG may be at a greater risk for developing secondary malignancies.7, 26, 27 Leukemia has been reported in several patients following treatment with ¹³¹I-MIBG as well as after ¹³¹I treatment of thyroid carcinoma, although it is not possible to dissect the relative contributions of previous extensive chemotherapy and radiation in some of these patients.7, 28, 29 This risk must be considered before repeat ¹³¹I-MIBG treatments.

We have demonstrated the feasibility of high-dose ¹³¹I-labeled MIBG therapy. Hematologic toxicity and other morbidity have been significant but tolerable. High-dose ¹³¹I-MIBG therapy is expensive and logistically difficult. The greatest logistic problem is that high-dose ¹³¹I-MIBG is not commercially available and must be synthesized on site by a nuclear pharmacist. In addition, the patient must be hospitalized for about a week in a lead-shielded room, and bone marrow harvesting is required. Given the rarity of malignant pheochromocytoma and the logistic constraints of high-dose ¹³¹I-MIBG, treatment with this technique is likely to be performed only at a limited number of facilities.

Other methods for enhancing the uptake of ¹³¹I-MIBG into the tumor and reducing systemic toxicity must be developed. The relatively acute infusion of ¹³¹I-MIBG (plus the much larger amount of cold ¹²⁷I-MIBG in the infusate) may partially saturate the catecholamine reuptake mechanism of malignant pheochromocytomas. Prolonging the infusion time of ¹³¹I-MIBG well beyond the 2-hour infusions we used in these patients may be beneficial. However, this is logistically difficult due to the requirement for constant monitoring of the patient during a radioisotope infusion.

Improvement in the specific activity and purity of ¹³¹I-MIBG may improve uptake into metastatic pheochromocytoma tumors. Current methods for preparing ¹³¹I-MIBG rely on a catalyzed exchange reaction in which unbound ¹³¹I is incubated with a much larger molar amount of nonradioactive ¹²⁷I-MIBG. The resultant product is such that only about 1% of the iodinated MIBG is ¹³¹I-MIBG. The rest is nonradioactive ¹²⁷I-MIBG. There may be competitive inhibition of ¹³¹I-MIBG uptake by the larger amount of circulating ¹²⁷I-MIBG. Purer synthesis of ¹³¹I-MIBG has been described by the radioiododesilylation of 3-trimethylsiylbenzguanidine, resulting in no-carrier–added ¹³¹I-MIBG.30

Malignant pheochromocytomas are very rare tumors. Patients with this malignancy often have a poor prognosis. However, their prognosis is quite variable and prolonged survival may occur. Therefore, it is difficult for a single institution to run a prospective and controlled clinical trial with enough statistical power to compare different treatment modalities adequately. Patients in the current series experienced greater hematologic toxicity without attributable mortality. The overall response rate of patients in the current series to high-dose ¹³¹I-MIBG appears to be comparable with the rate reported in trials using lower doses of ¹³¹I-MIBG given repetitively. We observed a CR following high-dose ¹³¹I-MIBG therapy in two patients with pheochromocytoma metastases to the bone. This finding has not been reported in studies using lower-dose protocols.

Therapy with ¹³¹I-MIBG can lead to clinical improvements and possibly cure patients with malignant pheochromocytoma and paraganglioma. The current study demonstrates the feasibility of high-dose ¹³¹I-labeled therapy for this and other malignancies. Refinements in ¹³¹I-MIBG specific activity, purity, dosing, and patient selection ultimately may improve therapeutic results and reduce toxicity.

Acknowledgements

The authors thank Dorothy Bennett, R.N., Janet Veatch, R.N., Randall A. Hawkins, M.D., Ph.D., Lloyd Damon, M.D., Scot Merrick, M.D., Peter Carroll, M.D., Orlo Clark, M.D., Alan B. Newman, M.D., Robert S. Hattner, M.D., Morris Schambelan, M.D., and Joan Lo, M.D., University of California–San Francisco, San Francisco, CA; Jane Sande, M.D., Theodore Zwerdling, M.D., and Robert J. Wells, M.D., Children's Hospital Medical Center, Cincinnati, OH; George Fisher, M.D., Stanford University, Stanford, CA; Daniel George, M.D., Dana-Farber Cancer Institute, Boston, MA; Mark A. Creager, M.D., Brigham and Women's Hospital, Boston, MA; Jami D. Frost, M.D., University of New Mexico, Albuquerque, N.M.; Karen Cartwright, M.D., Presbyterian Hospital, Albuquerque, NM; Steven S. Ginsberg, M.D., Group Health Cooperative of Puget Sound, Seattle, WA; Valerie Peck, M.D., and Perry Cook, M.D., New York University, New York, NY; Timothy Moynihan, M.D., Scott L. Stafford, M.D., and William Young, Jr., Mayo Clinic, Rochester, MN; Edward Greeno, M.D., University of Minnessota, Minneapolis, MN; Bruce Lowe, M.D., and Jose Leis, M.D., University of Oregon, Portland, OR; Seifedin Abdulkarim, M.D., Lawrence Solberg, M.D., Herbert W. Boro, M.D., Greg Matsubara, M.D., Kaiser Permanente Medical Center, and Brent Kane, M.D., Fresno, CA; Chien-Shing Chen, M.D., Loma Linda University, Loma Linda, CA; Victor Randolph, M.D., and Blaine Morton, M.D., Kaiser Permanente Medical Center, Fontana, CA; Ernest E. Pund, Jr., M.D., Scripps Memorial Hospital, LaJolla, CA; Margo Hoover-Regan, M.D., Children's Hospital of Orange County, Orange, CA; J. William Allgood, M.D., and Robert Kaplan, M.D., Escondido, CA; Asad Bashey, M.D., University of California, San Diego, CA; and Carlos Alvarado, M.D., Egleston Children's Hospital at Emory University, Atlanta, GA.