Phase I Trial of Subcutaneous Recombinant Human Interleukin-2 in Patients with Metastatic Melanoma

127 Phase I Trial of Subcutaneous Recombinant Human Interleukin-2 in Patients with Metastatic Melanoma Omar Eton, M.D. 1 Michael G. Rosenblum, Ph.D. 2 Sewa S. Legha, M.D. 3 Wehei Zhang, M.D. 2 Mary Jo East, R.N. 1 Agop Bedikian, M.D. 1 Nicholas Papadopoulos, M.D. 1 Antonio Buzaid, M.D. Robert S. Benjamin, M.D. 1 1 Department of Melanoma/Sarcoma, The University of Texas M. D. Anderson Cancer Center, Houston, Texas. 2 Bioimmunotherapy, The University of Texas M. D. Anderson Cancer Center, Houston, Texas. 3 Department of Medical Oncology, St. Luke s Episcopal Hospital, Houston, Texas. Oncology Center, Hospital Sirio-Libanes, Sao Paulo, Brazil. Presented as preliminary clinical results by Dr. Eton at the annual meeting of the American Society of Clinical Oncology, May 15 18, 1999. The preliminary laboratory results were presented by Dr. Rosenblum at the annual meeting of the American Association of Cancer Research, San Francisco, CA, April 1 5, 2000. Address for reprints: Omar Eton, M.D., Department of Melanoma, Box 30, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; Fax: (713) 75-106; E-mail: oeton@mdanderson.org Received October 30, 2001; revision received December 29, 2001; accepted January 29, 2002. BACKGROUND. Interleukin-2 (IL-2) has activity in metastatic melanoma when given in high doses by the intravenous (IV) route, but its side effects and effectiveness when given in intermediate to high doses by the subcutaneous (SC) route have not been studied adequately. This study sought to determine the maximum tolerated dose (MTD) of IL-2 administered once daily by the SC route. METHODS. Outpatients with progressive metastatic melanoma after chemotherapy were enrolled in a Phase I trial of IL-2 administered SC daily for 5 days per week for consecutive weeks, repeated at 6-week intervals. Patients were instructed to drink at least 2Loffluid daily. IL-2 pharmacokinetic studies were performed at the two highest dose levels. Toxicity was recorded weekly using the National Cancer Institute Common Toxicity Criteria. Response was assessed at 6-week intervals. RESULTS. Three patients, 6 patients, 6 patients, and patients received a median of 2 courses of SC IL-2 at dose levels of 6 MIU/m 2, 9 MIU/m 2, 12 MIU/m 2, and 15 MIU/m 2, respectively. Failure to maintain adequate fluid intake was responsible for 2 episodes of syncope at the 9 MIU/m 2 dose level and for 2 incidents of reversible prerenal azotemia at the 15 MIU/m 2 dose level. IL-2 treatment was resumed in these patients without incident. At the 15 MIU/m 2 dose level, 2 patients had severe headaches, depression, and visual hallucinations requiring discontinuation of treatment. Cough and fluid retention at the end of the third and fourth weeks at the 15 MIU/m 2 dose level approximated the symptoms reported by inpatients treated by continuous IV infusion at 9 MIU/m 2 on the same schedule. There was a partial response and a complete response in subcutaneous disease at the 12 MIU/m 2 and 15 MIU/m 2 dose levels, respectively, each lasting 2 months. Plasma IL-2 levels after SC injection of 1000 5000 pg/ml reached maximum by 3 hours and were detectable for up to 8 hours after administration. The half-lives for SC IL-2 absorbance and clearance were 1.6 hours and 5.2 hours, respectively, and the calculated area under the curve was 30,58 pg/ml hour. CONCLUSIONS. SC IL-2 was well tolerated and had high sustained bioavailability at the higher doses studied. The MTD for a daily SC regimen was 12 MIU/m 2 and is recommended for future studies. Cancer 2002;95:127 3. 2002 American Cancer Society. DOI 10.1002/cncr.10631 KEYWORDS: clinical trial, pharmacokinetics, metastatic melanoma, cytokine, interleukin-2, biochemotherapy. In January 1998, the use of recombinant human interleukin-2 (rhil-2) was recommended for approval by a U.S. Food and Drug Administration (FDA) advisory panel as a single-agent treatment for patients with metastatic melanoma. This recommendation was based on the outcome of 8 studies of 270 patients who were treated between 1985 and 1993 with intravenous (IV) bolus infusions of IL-2 at a dose of either 600,000 IU/kg or 720,000 IU/kg. The overall response rate 2002 American Cancer Society

128 CANCER July 1, 2002 / Volume 95 / Number 1 was 16%. It is noteworthy that, of 17 patients, 6% achieved complete response (CR), and 10 patients remained in remission for more than 2 years (median not yet reached). The median duration of partial response (PR), however, was only 6 months. Toxic reactions were common; 95% of patients had Grade 3 toxicity, and 35% of patients had Grade toxicity. The most salient toxic reactions were hypotension, oliguria/anuria, and pulmonary symptoms. Evidence of cardiac toxicity included arrhythmias and myocardial infarction. There were 28 serious adverse events in 10% of the patients and 6 deaths (2.2%) related to serious treatment-related organ dysfunction. Given this suboptimal toxicity profile, the FDA advisory panel recommended the evaluation of other doses and schedules of IL-2, either alone or in combination with other agents. 1,2 Several immunologic rationales for evaluating less toxic and less costly IL-2 programs have been advanced over 16 years that have elapsed since the introduction of this cytokine in the clinic. Although even very low and single weekly doses of IL-2 have yielded biologic effects in vivo, 3, it is only at the higher doses that responses have been reported consistently in patients with melanoma. 1,2,5 IL-2 also has become an integral component of combination anticancer regimens that include interferon- 2 and chemotherapy. These biochemotherapy regimens have achieved encouraging efficacy in patients with metastatic melanoma. 6 8 The clinical significance of IL-2 in these regimens is being evaluated in the ongoing European Organization for Research and Treatment of Cancer (EORTC) 18951 trial. 9 The effect of this cytokine may not be related primarily to immunologic modulation. Instead, the biophysical effects and antiangiogenic properties of IL-2 and secondarily expressed cytokines may play an important role, especially when IL-2 is used in the high dose range. 10,11 IL-2 is associated with significant morbidity, whether it is administered IV by bolus technique or by continuous infusion. The toxicities and the high cost of hospitalization render repeated IL-2 administration over time impractical. IL-2 administered IV also is associated with a high frequency of catheter-related infections. In an effort to control toxicity and, hence, to transfer IL-2 therapy to the less costly and more flexible outpatient setting, several investigators have substituted subcutaneous (SC) IL-2 for IV bolus or continuous infusion IL-2. 12 The empirically derived dosing of SC IL-2 in these trials to date, unfortunately, may be suboptimal, because, until now, no formal Phase I study to establish a maximum tolerated dose (MTD) has been reported. There have been only modest pharmacokinetic data to support any choice of dose and schedule of SC IL-2. Consequently, we performed a Phase I study to determine the MTD of SC IL-2 beginning at a dose of 6 MIU/m 2. We chose an intensive schedule of daily SC injection 5 days per week for consecutive weeks in a manner analogous to our previous trial of continuous IV infusion IL-2. 5 This latter trial had formed the basis for IL-2 dosing in biochemotherapy regimens at The University of Texas M. D. Anderson Cancer Center. 13 To confirm that once-daily SC injections of IL-2 would result in reasonable sustained levels in the peripheral blood, we also performed a 72-hour IL-2 bioavailability study in patients who were treated at the two highest dose levels. MATERIALS AND METHODS Patients Eligibility criteria were previous unsuccessful treatment with chemotherapy for pathologically confirmed melanoma, measurable metastatic disease, a Karnofsky performance status 70%, no evidence of clinically significant major organ dysfunction, no prior IL-2 therapy, and an expected survival duration of at least 12 weeks. Patients with brain metastases were excluded. The Institutional Review Board at M. D. Anderson Cancer Center approved this protocol, and written informed consent was obtained from each patient. Treatment Il-2 (Proleukin ; Aldesleukin; Chiron Corporation, Emeryville Corporation, Berkeley, CA) was provided as a sterile, preservative free, lyophilized powder. Each vial containing 22 MIU (1.3 mg) of IL-2 was reconstituted aseptically with 1.2 ml of sterile, nonbacteriostatic water for injection, USP, while avoiding foaming, to yield a final concentration of 18 MIU/mL (1.1 mg). The entire volume was drawn up at once in a Becton-Dickinson syringe using 20-gauge, half-inch needles and was stored in a refrigerator at 2 8 C (not frozen). It was stored for no more than 96 hours and was thawed to room temperature 15 minutes before injection. One course of therapy consisted of SC IL-2 administered daily for 5 consecutive days per week for weeks followed by a 2-week rest period. The starting dose was 6 MIU/m 2. Three patients were treated at each dose level, with no escalation of dose permitted in individual patients. Dose increments were 3 MIU/m 2 over the preceding level. Using the Phase I 3 3 design, one serious adverse event (National Cancer Institute [NCI] Grade 3 or ) required the enrollment of three more patients at the current level. With

IL-2 Cytokine Therapy for Melanoma/Eton et al. 129 two or three adverse events, the previous dose level would be established as the MTD of IL-2 delivered under the tested schedule. Patient Management Patients received the first dose of IL-2 therapy in the M. D. Anderson Cancer Center outpatient treatment area so they could receive instructions on sterile SC injection and could be monitored for signs of hypotension. Optional pharmacokinetic studies were performed at that time. Subsequent therapy was selfadministered at a time of day that would permit support from a spouse or other caretaker. Injection sites were rotated among the anterior thighs, deltoid regions, and lower abdomen. Patients were required to remain in the vicinity of the M. D. Anderson Cancer Center for the first week until stable control of IL-2- related toxicities was achieved. Patients were evaluated weekly for the first month by the attending physician. Complete blood counts and serum creatinine levels were obtained weekly. Patients were encouraged to drink at least 2 L of fluids daily. Injection sites were chosen on a rotating basis to reduce local inflammation and induration. Cold compresses and divided dosing also were permitted but rarely were necessary. Eucerin cream and diphenhydramine 25 50 mg were used for dry skin and pruritus, respectively. Steroid creams were not permitted. Any orthostatic symptoms or Grade 3 prerenal azotemia were treated with IV saline. Acetaminophen was taken every 3 hours 2 doses after each IL-2 injection was administered to diminish febrile symptoms. Naprosyn 250 mg could be used sparingly (once per week) for uncomfortable symptoms that were not relieved with acetaminophen. The oral antiemetics prochlorperazine, lorazepam, and ondansetron and the antidiarrhea agents Lomotil or loperamide were taken as needed. Side effects were recorded using the NCI Common Toxicity Criteria. Treatment was withheld for patients with Grade 3 and organ toxicity until the symptoms subsided or resolved, and then treatment was resumed at the same or one lower dose level. Because most acute side effects of outpatient IL-2 resolve within 8 hours of stopping treatment, the dose of IL-2 was not reduced automatically for easily reversible toxicity, for example, toxicity that typically results from inadequate fluid intake. With proper weekly medical advice, patients were able to make adjustments that prevented second occurrences. Dosing was reduced by one level for patients with Grade 3 or toxicity that could not, by its nature, be reversed easily. Treatment was discontinued in the event of unmanageable toxicity, progression of disease, or patient refusal. Metastatic lesions were assessed before beginning the next course (every 6 weeks). A CR required the disappearance of all indicator lesions for at least 1 month. A PR required at least a 50% decrease in the sum of the products of the two greatest perpendicular dimensions of all measurable lesions for at least 1 month. Stable disease had to last at least 2 months. Progressive disease was defined as an unequivocal increase in the size of any of the measurable lesions or the appearance of new lesions. The time to disease progression and survival duration were calculated from the start of treatment. Pharmacokinetic Studies Pharmacokinetic (bioavailability) studies of SC IL-2 were examined by enzyme-linked immunosorbent assay (ELISA) at the two highest dose levels. Serum collections were obtained at the time of the first and fifth SC IL-2 dose of course 1; 10 measurements were obtained over 72 hours (the Day 2 IL-2 dose was deleted). For a control, a single, 2-hour IV infusion of IL-2 at 1.5 MIU/m 2 was given the week before the start of treatment, and 10 serum specimens were obtained over the next hours. Heparinized blood samples were collected and maintained on ice, centrifuged, and the plasma was decanted and stored at 80 C until it was analyzed. All samples were then thawed at ambient temperature and analyzed for IL-2 content by ELISA (Quantikine; R&D Systems, Minneapolis, MN). Analyses of IL-2 concentrations were then performed using nonlinear regression. RESULTS From May 199 through April 1998, 19 patients with progressive, metastatic melanoma after prior systemic chemotherapy and/or interferon treatment were enrolled in the study. Two patients had metastatic ocular melanoma. The pretreatment performance status, the number of visceral organs involved, and levels of serum lactate dehydrogenase and albumin were relatively favorable (Table 1). All patients were evaluable for both toxicity and response. Overall, 35 courses of IL-2 were administered to 19 patients, as shown in Table 2. Treatment was terminated in all but one patient because of evidence of progression of disease; that one patient, who was treated at the 12 MIU/m 2 dose level, refused a third course for nonmedical reasons. Toxicity During each month-long course of outpatient treatment, patients were unable to work full time and

130 CANCER July 1, 2002 / Volume 95 / Number 1 TABLE 1 Patient Characteristics Characteristic No. of patients Enrolled 19 Female (%) 10 (53) Median age in yrs (range) 55 (36 65) Karnofsky performance status (%) 90 12 80 7 70 1 Prior therapy Chemotherapy alone 8 Chemotherapy and interferon- 2 11 Melanoma characteristics In-transit disease 2 Advanced disease 17 No. of visceral organs 0 (subcutaneous, lymph nodes) 3 1 12 2 1 3 1 Visceral sites of disease Lung alone 7 Adrenal Liver 3 Bone, intestine, bladder 1 each Elevated serum level of LDH (%) 6 (32) Serum level of albumin 3. mg/dl (%) 0 (0) LDH: lactate dehydrogenase. TABLE 2 Interleukin-2 Dose Escalation Dose level Dose (MIU/m 2 /day) No. of patients No. of courses per patient 1 6 3 1,2,1 2 9 6 1,1,2,3,2,1 3 12 6 1, 1, 2, 2, 2, 6 15 1, 3, 2, 3 remained mostly at home. Despite this schedule, all patients completed the first course of treatment, and most were able to undergo repeated courses (Table 2) without cumulative toxicity. Of 10 anticipated weeks of therapy, 3% were modified for reversible Grade 3 toxicity (see below) by withholding IL-2 treatment for the remainder of 1 week. There was no Grade adverse event or death attributable to IL-2 in this study. By 2 hours, patients developed inflammatory changes extending 2 10 cm from the local injection site. Residual induration/nodularity persisted from 2 weeks to 2 months. Some patients experienced mild pain at the injection sites. The intensity of the local reaction was increased only modestly at the higher dose levels. Symptoms were managed successfully by adding new injection sites to the rotation and application of a cold pack in two patients. In the patients who received multiple courses, abdominal SC injections resulted in less discomfort than injections into the anterior of the thigh. Patients experienced Grade 1 fever and chills 2 hours after injection that lasted 1 2 hours. Grade 1 or 2 anorexia, nausea, emesis, diarrhea, fatigue, myalgias, and dry mouth and mucous membranes plus sinus complaints were prevalent and were more likely to be Grade 2 at the 12 MIU/m 2 and 15 MIU/m 2 dose levels. There was no discernible or cumulative pattern in these reported toxic reactions week to week, and patients invoked coping strategies that made them more resilient over time. No Grade 3 or toxicity was reported at the 6 MIU/m 2 and 12 MIU/m 2 dose levels. At the 9 MIU/m 2 dose level, two patients had one syncopal episode each at home during the third week of the first course. Neither patient reported this to the physician at the time of occurrence, and both incidents likely resulted from diminished systemic vascular resistance and mild volume depletion during IL-2 treatment. Both patients were advised again to drink 2 L of fluids daily, and both were able to resume treatment without a repeated incident. Only at the 15 MIU/m 2 dose level did cough, fluid retention, prerenal azotemia, weight gain 5 kg, and serious neurologic symptoms become evident. At the end of Week 1 at this dose level, prerenal azotemia, as indicated by serum creatinine levels of 3.1 mg/dl and.7 mg/dl in two patients, required hospitalization and overnight IV hydration with normal saline. After their discharge, both patients markedly improved compliance with the 2-L fluid intake guideline, and both resumed treatment without a second incident. The median and range values for serum levels of creatinine by dose level are listed in Table 3. Headache, blurry vision, confusion, anxiety, depression, sleep disturbances, and emotional lability were reported in the higher two dose levels, mild at the 12 MIU/m 2 dose level and moderate to severe at the 15 MIU/m 2 dose level. Two patients at the 15 MIU/m 2 dose level declined treatment for the remainder of Week 1 because of severe progressive headache, anxiety, depression, visual hallucinations, and emotional lability. These symptoms resolved within 2 8 hours after cessation of treatment, and both patients were able to resume treatment and complete the first course; however, the same severe neurologic symptoms recurred in both patients in the same pattern during a second course. Peak and nadir blood counts during the first course of weeks of SC IL-2 by dose level are outlined

IL-2 Cytokine Therapy for Melanoma/Eton et al. 131 TABLE 3 Selected Peak and Nadir Blood Counts and Serum Levels of Creatinine Measured Weekly during the First Course of Subcutaneous Interleukin-2 Peak count Nadir count Dose (MIU/m 2 ) Median ( 10 9 /L) Range ( 10 9 /L) Week Median ( 10 9 /L) Range ( 10 9 /L) Week Leukocytes 6 8.8 6.8 18.8.6.3 9.6 9 13.0 5.7 22.2 6.9.5 8.5 12 1.9 8.0 20.8 6..7 7.8 15 10.9 8.3 15.7 5.8.7 6.3 Lymphocytes 6 3.1 2.6 3.7 1.3 0.7 1.8 9.8 1.1 8.8 2.2 0.3.2 12.2 3.3 9.8 1.9 1.2.3 15 3.9 3.2 5.0 1. 0.6 3.2 Eosinophils 6 7.1 2. 7.2 9.1 0.9 6.7 12 6.5 1.9 9.3 15 3.6 0.7 7.9 Neutrophils 6 1.7 1.6 7. 1.3 0.7 2. 9.5 2.1 7.8 1.6 1.2.5 3 12.0 1.7 5.1 2.3 0.9.5 15.8 3.2 5.8 2.1 0.5 3.0 Platelets 6 172 55 296 9 181 15 27 12 167 135 291 15 126 62 19 Creatinine (mg/dl) 6 0.8 0.8 1.2 9 1.1 0.9 1.6 12 1.1 0.7 2.1 15 1.2 0.8.7 1 1 1 in Table 3. Generally, in the first week of treatment, total white blood cell, absolute lymphocyte, and platelet counts diminished to their lowest levels. By the third and fourth weeks, nadir neutrophil counts coincided with a rebound lymphocytosis and eosinophilia. Because blood tests were performed weekly, it was not possible to detect lymphopenia during the first week. Week-to-week trends in absolute lymphocyte count were not discernible readily at the 6 MIU/m 2 and 9 MIU/m 2 dose levels. In contrast, nadir and peak absolute lymphocyte counts occurred consistently in the first week and the fourth week at the two highest dose levels of 12 MIU/m 2 and 15 MIU/m 2. Overall, the blood counts did not vary widely. Table 3 shows that a neutrophil nadir 1000/mm 3 and a platelet nadir 100,000/mm 3 were observed briefly in only of 35 courses and 3 of 35 courses, respectively. In addition, the overall spread between median peak and nadir values of white blood cell, lymphocyte, and neutrophil counts was only by a factor of two, and the range of values was narrow. There were no trends in hematologic toxicity by response category (data not shown). Liver function test results were affected only mildly. The value for serum bilirubin rose above 1.0 mg/dl in only of 35 courses (11%), with 1. mg/dl the highest value reported. There was evidence that treatment induced hypothyroidism in 1 patient (among 12 patients with evaluable baseline and post-treatment thyroid function tests). One patient developed vitiligo after two courses of SC IL-2 at the 9 MIU/m 2 dose level without an objective tumor response. One patient developed a herpes zoster rash during the fourth week at the 15 MIU/m 2 dose level. Pharmacokinetic Studies After SC administration, IL-2 levels were 1000 5000 pg/ml, reaching maximal levels at approximately 3 hours, as shown in Table and Figure 1. Measurable

132 CANCER July 1, 2002 / Volume 95 / Number 1 TABLE Bioavailability of Interleukin-2 after Subcutaneous Administration Patient Dose C MAX (pg/ml) T MAX (hours) t 1/2 (ABS) (hours) t 1/2 (CL) (hours) AUC (pg/ml hour) rt (hours) 1 1 a 2288 3.6 2.5 2.5 23,27 9.1 1 b 3831.9 3.5 3.5 51,770 9.9 2 1 5207 2.6 1. 2.3 37,65 5.3 3 1 206 2. 0.7 6.6 2,25 9.9 3 1991 1.7 0.3 15.2 31,092 10.1 1 606 2.5 0.9 3.7 38,832 6.6 820 3.1 1.7 2.8 7.79 6. Mean 2973 3 1.6 5.2 30,58 8.2 Sem 603 0. 0. 1.8 5359 0.8 t 1/2 (ABS) : Half-life for absorbance; t 1/2 (CL) : half-life for clearance; AUC: area under the curve; SEM: standard error of the mean; C MAX : the maximum concentration in pg/ml of plasma IL-2 measured at time T MAX in hours after treatment with IL-2; rt: the residence time, or time that IL-2 stays in the plasma, determined without making assumptions concerning the number of pharmacologic compartments. a The first injection of subcutaneous interleukin-2 (SC IL-2) at 12 MIU/m 2 was given on Day 1 of the 1st week of Course 1. b ThefourthinjectionofSCIL-2at12MIU/m 2 was given on Day 5 of the 1st week of Course 1. FIGURE 1. The bioavailability of interleukin-2 after subcutaneous administration. Dose 1 represents the first injection of 12 MIU/m 2 on Day 1, and Dose 2 represents the fourth injection given on Day 5 during the first week of treatment. SEM: standard error of the mean. drug remained in the plasma for up to 8 hours after SC administration. The half-life (t 1/2 ) for absorbance was 1.6 0. hours, and the t 1/2 for and clearance was 5.2 1.8 hours. The area under the curve (AUC) for IL-2 after SC administration was 30,58 5359 pg/ml per hour. There did not appear to be accumulative changes in pharmacokinetics, because the Day 5 studies did not show statistically significant changes in blood levels or pharmacokinetic parameters. After the 2-hour IV infusion in the control study 1 week before the start of SC IL-2 therapy, initial plasma levels of IL-2 of 2000 7000 pg/ml declined rapidly, with an initial and terminal t 1/2 of 9.6 3.8 minutes and 55.0 6.6 minutes, respectively. IL-2 levels were undetectable 6 hours after the start of the IV test dose. The calculated AUC for the IV test dose was 323 109 pg/ml minute. Antitumor Activity There was one CR and one PR, both fleeting, in multiple subcutaneous lesions in two patients who were treated at the 15 MIU/m 2 and 12 MIU/m 2 dose levels, respectively. Some of the responding lesions were injected perilesionally. The responses were not durable, lasting 2 months each. The two patients survived 27 months and 17 months, respectively. Less than 50% diminution in pulmonary metastases was seen in another two patients at the 15 MIU/m 2 and 12 MIU/m 2

IL-2 Cytokine Therapy for Melanoma/Eton et al. 133 dose levels lasting 5 months and 30 months, respectively. The latter patient was a woman age 58 years with choroidal melanoma and pulmonary and perirenal metastases; she was the only patient who was censored for progression of disease. She received 6 courses of IL-2 at 12 MIU/m 2 and developed hypothyroidism during treatment. Fifteen patients (79%) had progressive disease within 3 months of the start of therapy. Despite prior progressive disease on chemotherapy, the median overall survival from the start of IL-2 was 10.6 months (range, from 2.3 months to 6.0 months) with two patients at the 12 MIU/m 2 dose level alive at 36 months and 6 months. Only two other patients lived for 2 months, and they both were treated at the 15 MIU/m 2 dose level. DISCUSSION Several Phase I trials of IL-2 administered by the SC route in relatively low doses have confirmed the feasibility and tolerability of this approach. 1,15 Our Phase I trial of IL-2 administered by SC injection daily for 5 consecutive days per week for weeks indicates that 12 MIU/m 2 is the MTD for this schedule. The acute reversible side effect of prerenal azotemia in two patients at the 15 MIU/m 2 dose level could be avoided by a midweek bolus IV saline infusion. However, symptoms of moderate to severe central nervous system toxicity, including headache, confusion, and hallucinations, as early as 1 week into the first course indicated that 15 MIU/m 2 was too toxic. These symptoms plus cough and fluid retention approximated the symptoms of inpatients who were treated intravenously at 9 MIU/m 2 in our previously reported Phase II trial. 5 The supportive care measures described above (see Materials and Methods) were adequate to make this outpatient regimen tolerable at home. The serious acute toxicity of hypotension reported using IV IL-2 was avoided by strict adherence to an oral intake of 2 L of electrolyte-containing fluids, such as juices, daily. Because treatment with IL-2 results in the expression of a cascade of secondary cytokines with both additive and antagonistic effects, 16 it is not surprising that the median values for circulating leukocyte, lymphocyte, and neutrophil counts varied within a narrow range, regardless of dose level. The characteristic IL-2-induced rise in the circulating eosinophil count was not dose dependent in the dose range studied. The pharmacokinetics illustrated in Figure 1 showed sustained, high levels of IL-2 after SC administration, supporting excellent bioavailability. This high dose SC IL-2 study was undertaken in part because the results of Phase I II trials dating back to the 1980s showed modest antimelanoma activity only at high IL-2 dose levels, unlike renal cell carcinoma, in which activity also is seen at lower dose levels. 1,2,5 Thus, it is not surprising that, in this study, the two major responses were observed only at the two highest dose levels studied. These responses in subcutaneous metastases were short-lived, however, precluding the scheduling of a biopsy for either patient to assess the mechanism of antitumor activity. Neither responding patient volunteered to receive high-dose IV IL-2 after their disease progressed on SC IL-2. The prolonged disease stabilization in one patient with chemotherapy and interferon-resistant metastatic choroidal melanoma supports evaluating IL-2 in the management of patients with this generally unresponsive malignancy. The response rate for SC IL-2 alone approaches 0% at low doses and is predicted to be no greater than 16% at higher doses, as reported for the high-dose IV bolus IL-2 regimen. Given this modest range of predicted antitumor activity, the most effective way to develop IL-2 is in combination with other agents. Atkins and Gollob are evaluating concurrent biochemotherapy versus chemotherapy in the Eastern Cooperative Oncology Group (ECOG) trial E-3695, in collaboration with the Southwest Oncology Group and Cancer and Leukemia Group B. 17 In the ECOG trial, biochemotherapy includes IL-2 administered by continuous IV infusion at 9 MIU/m 2, requiring hospitalization and close monitoring for hemodynamic instability and other toxicities, including catheter-related infections. 13 Using low-dose SC IL-2 at only 3 MIU/m 2, Thompson et al. reported a high CR rate of 19% in 53 patients who were treated with biochemotherapy. 12 However, in a randomized Phase II trial, the Cytokine Working Group reported that biochemotherapy using SC IL-2 at 5 MIU/m 2 had lower activity than the same regimen using IV IL-2 at 18 MIU/m 2. 18 The contribution of IL-2 to the activity of biochemotherapy, at best, is controversial and is currently being addressed in the ongoing EORTC 18951 trial. 9 Therefore, it is unreasonable at this point to design a study that would be powered sufficiently to detect differences in activity attributable to the dose or route of IL-2 administration in biochemotherapy. If IL-2 proves to be an important component of biochemotherapy, then this agent administered by the SC route may be useful in reducing morbidity and cost. However, it would be appropriate to test doses of SC IL-2 close to the MTD and associated with antimelanoma activity. At M. D. Anderson Cancer Center, we have incorporated SC IL-2 at the dose determined herein of 12 MIU/m 2 daily for 5 days immediately after combined cisplatin, vinblastine, and dacarbazine

13 CANCER July 1, 2002 / Volume 95 / Number 1 (CVD) plus interferon in a hybrid biochemotherapy regimen in which, except for dacarbazine, the other four agents are dosed 25 100% higher than previously reported. 19 One goal of this study is to at least reproduce the high frequency of durable CRs initially reported previously in patients who received for biochemotherapy, confirming that biochemotherapy is a potent response induction regimen in selected patients. 19 REFERENCES 1. Litwin SD. FDA medical reviewers report BLA supplement 97-0501 proleukin indication: treatment of metastatic melanoma [letter]. Washington, DC: U.S. Food and Drug Administration, 1998. 2. Atkins MB, Lotze MT, Dutcher JP, et al. High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol. 1999;17:2105 2116. 3. Lindemann A, Brossart P, Hoffken K, et al. Immunomodulatory effects of ultra-low-dose interleukin-2 in cancer patients: a Phase Ib study. Cancer Immunol Immunother. 1993; 37:307 315.. Maas RA, Dullens HFJ, Den Otter W. Interleukin-2 in cancer treatment: disappointing or (still) promising? A review. Cancer Immunol Immunother. 1993;36:11 18. 5. Legha SS, Gianan MA, Plager C, Eton O, Papadopoulous NEJ. Evaluation of interleukin-2 administered as continuous infusion in metastatic melanoma. Cancer. 1996;77:89 96. 6. Keilholz U, Conradt C, Legha SS, et al. Results of interleukin- 2-based treatment in advanced melanoma: a case recordbased analysis of 631 patients. J Clin Oncol. 1998;16:2921 2929. 7. Allen IE, Kupelnick B, Kumashiro M, Luo D, Ross SD, Wolin MJ. Efficacy of interleukin-2 in the treatment of metastatic melanoma. Systematic review and meta-analysis. Cancer Therapeutics. 1998;1:168 173. 8. Eton O, Legha SS, Bedikian A, et al. Phase III randomized trial of cisplatin, vinblastine, dacarbazine plus interleukin-2 and interferon-alpha-2b versus CVD in patients with metastatic melanoma [abstract]. Proc Am Soc Clin Oncol. 2000; 19:A217. 9. Keilholz U, Eggermont AM. The role of interleukin-2 in the management of Stage IV melanoma: the EORTC melanoma cooperative group program. Cancer J Sci Am. 2000;6(Suppl 1):S99 S103. 10. Kristensen CA, Nozue M, Boucher Y, Jain RK. Reduction of interstitial fluid pressure after TNF-alpha treatment of three human melanoma xenografts. Br J Cancer. 1996;7:533 536. 11. Watanabe M, McCormick KL, Volker K, et al. Regulation of local host-mediated anti-tumor mechanisms by cytokines: direct and indirect effects on leukocyte recruitment and angiogenesis. Am J Pathol. 1997;150:1869 1880. 12. Thompson JA, Gold PJ, Markowitz DR, Byrd DR, Lindgren CG, Fefer A. Updated analysis of an outpatient chemoimmunotherapy regimen for treating metastatic melanoma. Cancer J Sci Am. 1997;3(Suppl 1):S29 S3. 13. Legha SS, Ring S, Eton O, et al. Development of a biochemotherapy regimen with concurrent administration of cisplatin, vinblastine, dacarbazine, interferon alfa, and interleukin-2 for patients with metastatic melanoma. J Clin Oncol. 1998;16:1752 1759. 1. Leahy MG, Pitfield D, Popert S, Gallagher CJ, Oliver RT. Phase I study comparing continuous infusion of recombinant interleukin-2 by subcutaneous or intravenous administration. Eur J Cancer. 1992;28A:109 1051. 15. Angevin E, Valteau-Couanet D, Farace F, et al. Phase I study of prolonged low-dose subcutaneous recombinant interleukin-2 (IL-2) in patients with advanced cancer. J Immunother Emphasis Tumor Immunol. 1995;18:188 195. 16. Engelhardt M, Wirth K, Mertelsmann R, Lindemann A, Brennscheidt U. Clinical and immunomodulatory effects of repetitive 2-day cycles of high-dose continuous infusion IL-2. Eur J Cancer. 1997;33:1050 105. 17. Atkins MB, Gollob JA. Chemotherapy and cytokine-based immunotherapy for high-risk and metastatic melanoma. Adv Oncol. 1999;15:22 29. 18. Flaherty LE, Atkins M, Sosman J, et al. Outpatient biochemotherapy with interleukin-2 and interferon alfa-2b in patients with metastatic malignant melanoma: results of two Phase II cytokine working group trials. J Clin Oncol. 2001; 19:319 3202. 19. Legha SS, Ring S, Eton O, Bedikian A, Plager C, Papadopoulos N. Development and results of biochemotherapy in metastatic melanoma: the University of Texas M. D. Anderson Cancer Center experience. Cancer J Sci Am. 1997;3(Suppl 1):S7 S8.