ARTICLES
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Pilot Study of Interleukin-2 and Lymphokine-Activated Killer Cells Combined With Immunomodulatory Doses of Chemotherapy and Sequenced With Interferon Alfa-2a in Patients With Metastatic Melanoma and Renal Cell Carcinoma Mario Sznol,* Jeffrey W. Clark, John W. Smith II, Ronald G. Steis, Walter J. Urba, Lawrence V. Rubinstein, Louis A. VanderMolen, John Janik, William H. Sharfman, Robert G. Fenton, Stephen P. Creekmore, Peter Kremers, Kevin Conlon, Jean Hursey, Joy Beveridge, Dan L. Longo
Background: Experiments in animal tumor models suggest that the antitumor effects of interleukin-2 (IL-2) or IL-2 in combination with lymphokine-activated killer (LAK) cells can be enhanced by chemotherapy agents such as cyclophosphamide or doxorubicin or by the biologic agent interferon a. Purpose: We determined the toxicity and clinical response rate of an IL-2-LAK cell regimen modified by the addition of moderate, immunomodulatory doses of chemotherapy and sequenced with interferon alfa-2a (IFN a-2a) in patients with metastatic melanoma and renal cell carcinoma. Methods: IL-2 (3-6 million units/m2 per day) was administered by continuous infusion on days 0-5 and days 11-16. LAK cells were infused on days 11 and 12 or on days 11, 12, and 14. Low doses of cyclophosphamide (300 mg/ m2) and doxorubicin (25 mg/m2) were given on day 9 before the LAK cell infusions. Following the IL-2-LAK cell infusion, IFN a-2a (12 million units/m2) was administered for a total of nine doses to complete a cycle of treatment. A total of 89 patients were enrolled in the study. Results: For each histology, there were eight partial responses in 40 assessable patients, for an overall response rate of 20% (90% confidence interval = 10%-33%). The median response duration was 5 months, although two patients with renal cell carcinoma and one patient with metastatic melanoma had almost complete disappearance of tumor and are still responding after 26 + , 22 + , and 26+ months, respectively. Toxic effects were severe in patients receiving the highest dose of IL-2 administered in this study and similar to those reported with other high-dose IL-2-LAK cell regi-
Vol. 84, No. 12, June 17, 1992
mens. Although toxic effects were completely reversible in most patients, there were four treatment-related deaths. Conclusions: This regimen is active in patients with metastatic melanoma and renal cell carcinoma and produces meaningful responses in a small percentage of these patients; however, it is not clear whether cyclophosphamide, doxorubicin, and IFN a-2a as used in this protocol appreciably augmented the antitumor activity of the IL-2-LAK cell regimen. [J Natl Cancer Inst 84:929-937, 1992] Received November 21, 1991; revised February 21, 1992; accepted March 3, 1992. Supported in part by Public Health Service contract NOI-CO-74102 from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services. M. Sznol, Clinical Research Branch, Biological Response Modifiers Program, Division of Cancer Treatment, National Cancer Institute (NCI), NCIFrederick Cancer Research and Development Center, Frederick, Md.; and Investigational Drug Branch, Cancer Therapy Evaluation Program, Division of Cancer Treatment, NCI, Bethesda, Md. J. W. Clark, J. W. Smith II, R. G. Steis, L. A. VanderMolen, J. Janik, W. H. Sharfman, R. G. Fenton, S. P. Creekmore, K. Conlon, D. L. Longo, Clinical Research Branch, Biological Response Modifiers Program, Division of Cancer Treatment, NCI-Frederick Cancer Research and Development Center. W. J. Urba, J. Beveridge, Program Resources Inc./DynCorp, NCI-Frederick Cancer Research and Development Center. L. V. Rubinstein, Biometric Research Branch, Cancer Therapy EvaJuation Program, Division of Cancer Treatment, NCI, Bethesda. P. Kremers, J. Hersey, Frederick Memorial Hospital. We thank the Biological Response Modifiers Program nursing staff for providing excellent care to the patients and Terry Phillips for her outstanding work as editorial assistant. Correspondence to: Mario Sznol, M.D., Executive Plaza North, Rm. 715, National Institutes of Health, Bethesda, MD 20892.
ARTICLES
929
When first introduced into clinical trials, interleukin-2 (IL-2) and lymphokine-activated killer (LAK) cells appeared to have substantial activity in patients with renal cell carcinoma and metastatic melanoma {1,2). A number of confirmatory studies were initiated, and various attempts were made to modify the dose, schedule, and route of administration of the combination of IL-2 and LAK cells to increase efficacy and decrease toxicity (3-7). The trials produced low objective response rates in the range of 20% or less, and there was no evidence that manipulation of the original bolus IL-2-LAK cell regimen improved clinical outcome. Although most patients in these studies did not benefit from IL-2-LAK cell therapy, transient regression of tumor occurred in many of the patients classified as nonresponders. This finding raised the possibility that combining IL-2-LAK cells with agents that enhanced their antitumor immunomodulatory properties in vivo or adding other active biologic agents to this regimen may result in a greater percentage of patients reaching durable partial or complete responses.
In similar animal models, cyclophosphamide was shown to decrease the toxicity of IL-2 while enhancing its antitumor activity (10,13-15). A clinical trial in patients with melanoma indicated that low-dose cyclophosphamide could be combined with outpatient doses of IL-2, resulting in response rates comparable to the high-dose, more toxic IL-2-LAK cell regimen (16). These data suggested that low-dose cyclophosphamide may potentiate the antitumor activity of IL-2-LAK cells. Interferon-a (IFN-a) was known to be active in patients with melanoma and renal cell carcinoma, with response rates in the range of 15%-20% (17,18). Although the mechanism of action in vivo was not understood, IFN-a had direct antiproliferative effects in vitro and increased the expression of tumorassociated and histocompatibility antigens on tumor cells (19,20). With respect to antitumor activity in animal models, the combination of IFN-a and IL-2 was found to be superior to either of the agents given singly; although concurrent admin-
930
The Biological Response Modifiers Program of the National Cancer Institute (NCI) had conducted pilot studies of IL-2 administered by continuous infusion in combination with LAK cells. This regimen was tolerable, induced a high yield of LAK cells, and produced responses in patients with renal cell carcinoma (27). On the basis of the immunomodulatory properties of cyclophosphamide and doxorubicin, as well as their synergy with IL-2 and IL-2-LAK cells in animal models, we administered these agents at fixed doses 48 hours before the first LAK cell infusion. Interferon a-2a (IFN a-2a) was given as a single agent three times a week during the interval between courses of treatment with IL-2-LAK cells. This article describes the toxicity and antitumor activity of this regimen in patients with renal cell carcinoma and metastatic melanoma.
Patients and Methods All patients signed informed consent forms before enrollment in the study. The protocol was reviewed and approved by the NCI Clinical Research Subpanel, the Institutional Review Board of the NCI-Frederick Cancer Research and Development Center, and the Cancer Therapy Evaluation Program of the NCI. Eligibility
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Preclinical animal studies demonstrated that two chemotherapy agents, doxorubicin and cyclophosphamide, could increase the antitumor activity of IL-2. Salup et al. (8) examined the effects of nephrectomy followed by the administration (intravenous and intraperitoneal) of doxorubicin, IL-2, and LAK cells to mice bearing implanted renal adenocarcinomas. Following nephrectomy, only the combination of doxorubicin administered 24 hours before a 3-day course of IL-2-LAK cells markedly prolonged the survival of mice with tumors disseminated to lymph nodes. Other investigators demonstrated an increase in survival when IL-2 administration preceded and followed a single dose of doxorubicin in mice bearing the EL-4 lymphoma (9) and when doxorubicin was given before IL-2 in mice with established micrometastases of the MCA-105 sarcoma or MCA-38 adenocarcinoma (10). In addition to its direct antitumor effects in vivo, doxorubicin was found to increase allogeneic cytotoxic T-lymphocyte responses, IL-2 production by splenic T cells, macrophage cytotoxicity, the number of immature macrophages and monocytes, and interleukin-1 and tumor necrosis factor (TNF) production by macrophages (77). Preliminary evidence also suggested that doxorubicin could direct the migration of activated lymphocytes to sites of tumor (72).
istration produced the best results, sequential administration was also effective in some of the models (21-23) and obviated the inhibitory effects of IFN-a on generation of LAK cells (24). Furthermore, IFN-a was not induced in vivo during treatment with IL-2-LAK cells, unlike certain other cytokines (TNF and IFN-7) which could be detected in serum following IL-2 administration (25,26). Thus, IFN-a was considered a potential non-cross-resistant agent that could maintain or, perhaps, enhance responses to IL-2-LAK cells.
Tests to determine eligibility and tumor measurements were performed within 14 days of beginning therapy (stresstreadmill and pulmonary function tests within 42 days). These tests were similar to those in our prior studies on IL-2-LAK cells (27). Patients at least 15 years old with histologically confirmed metastatic or unresectable renal cell carcinoma or malignant melanoma were eligible for this trial. All patients were required to have a Karnofsky performance status of 70%-100%, an expected survival of greater than 3 months, bidimensionally measurable disease that was clearly progressive, normal results from a stress-treadmill examination, a diffusion capacity on pulmonary function testing of greater than 70% of predicted and a Po2 on room air of greater than 65 mm Hg, as well as normal renal and liver functions. Patients were excluded from the study for the following reasons: a history or evidence of coronary artery disease; congestive heart failure or serious cardiac arrhythmias; active infection requiring antibiotics; the presence of organ allografts or brain metastases; a requirement for corticosteroids; prior therapy with IL-2 (this exclusion criterion took effect after the initial 31 patients were accrued to the protocol); and prior chemotherapy, prior radiotherapy, or prior immunotherapy or hormonal therapy within 3 weeks (6 weeks for nitrosoureas or mitomycin C) of entering the protocol.
Journal of the National Cancer Institute
Patient Characteristics A total of 89 patients were registered for this study. Four patients did not receive IL-2 and, therefore, could not be assessed for toxic effects or response. Forty-three patients with metastatic melanoma and 42 patients with renal cell carcinoma could be assessed for toxic effects. The characteristics of these patients are shown in Tables 1 and 2, respectively. Treatment
Vol. 84, No. 12, June 17, 1992
Characteristic No.
45 43 40
Assessable for toxic effects Assessable for response Age, y Median Average Range Men/women
45 44 19-70 25:18 28/15
Time from diagnosis, d Median Average Range
828 1302 33-5027
Prior therapy No chemotherapy or immunotherapy Prior surgery Prior chemotherapy Prior radiotherapy Prior immunotherapy IFN-a or IL-2
24 (56) 40 (93) 10 (23) 7 (16) 13 (30) 10 (23)
Sites of disease Skin or lymph node only Lung ± above Liver ± above Other ± above
23 (53) 12 (28) 4 (9) 4 (9)
Performance status, 100:90
•Unless otherwise indicated, values = No. of patients (%).
Table 2. Patient characteristics—renal cell carcinoma* Characteristic No.
44
Assessable for toxic effects
42
Assessable for response
40
Age, y Median Average Range
55 54 26-71
Performance status, 100.90:80:70
22:13:4:3
Men/women
35/7
Time from diagnosis, d Median Average Range
371 624 46-2780
Prior therapy No chemotherapy or immunotherapy Prior surgery (nephrectomy) Prior chemotherapy Prior radiotherapy Prior immunotherapy, IFN-a or IL-2
34(81) 34(81) 4(10) 10 (24) 4(10)
Site of disease
Lung or pleura only Lymph node ± above Intra-abdominal ± above Liver or bone ± above
12(29) 9(21) 14 (33) 7(17)
•Unless otherwise indicated, values = No. of patients (%).
dose was re-escalated to 6 million units/m2 per day to complete 120 hours of continuous infusion. Vasopressors and furosemide were used as needed to reverse hypotension and weight gain resulting from a capillary leak ARTICLES 931
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
A central venous access device was placed in all patients. IL-2 (Hoffmann-LaRoche, Nutley, N.J.) was administered at a dose of 3-6 million units/m2 per day by continuous infusion for 120 hours on days 0-5 and days 11-16 (Fig. 1). On days 7 and 8 (and day 9 prior to chemotherapy for the first 45 patients on study) patients underwent a 5-hour leukapheresis. On day 9, patients received cyclophosphamide (300 mg/m2) and doxorubicin (25 mg/m2) intravenously by bolus. LAK cells were generated in vitro as previously described (27) and administered by a central venous access device (> 1 hour) beginning 1 hour after IL-2 was initiated on day 11 (cells from pheresis on day 7), day 12 (cells from pheresis on day 8), and day 14 (in the first 45 patients on study who underwent pheresis on day 9). On day 21, patients started IFN a-2a (HoffmannLaRoche) at a dose of 12 million units/m2 subcutaneously thrice weekly for nine doses. Patients were evaluated for response following the last dose of IFN ot-2a (day 42). Those with stable disease or partial response were eligible for repeat cycles of treatment beginning 1-4 weeks after the completion of the previous cycle. Our previous trials (27) of continuous infusion of IL-2 and LAK cells used Cetus IL-2 at a dose of 3 million Cetus units/m2 per day, which produced reversible grade 3 and grade 4 toxicity in the majority of patients and required treatment administration in a specialized monitoring unit. Since the clinical conversion factor for Cetus IL-2 units to Roche IL-2 units was not clear, we initiated the current study of Roche IL-2 at 3 million units/m2 per day. After treatment was completed in 21 melanoma patients and eight renal cell carcinoma patients, we concluded that the dose of IL-2 could be raised to 6 million units/m2 per day, based on preliminary experience from other investigators in concurrent trials and the mild toxicity seen in our patients receiving 3 million units/m2 per day. A total of 34 patients (22 melanoma patients and 12 renal cell carcinoma patients) were subsequently treated with IL-2 at a dose of 6 million units/m2 per day. Five of the first 16 patients receiving this dose experienced sufficient pulmonary toxicity during the LAK cell infusions to require omission of a LAK cell dose; all subsequent patients, therefore, had pheresis procedures reduced to two and a maximum of two LAK cell infusions. Analysis of the latter group (18 patients) continued to show excessive pulmonary toxicity during the 2 days of LAK cell infusions, and several major toxic events, including three deaths, were reported in the last 12 of the 34 patients entered at this dose of IL-2. In an attempt to reduce toxicity, the remaining patients (22 total, all renal cell carcinoma patients, since melanoma accrual had been completed) received a reduced dose of IL-2 (4.5 million units/m2 per day) only on days 11 and 12 in conjunction with LAK cell administration. In the absence of dyspnea at rest, neurologic toxicity, or hypotension supported by maximum tolerable doses of pressors on day 13, the IL-2
Table 1. Patient characteristics—melanoma*
DAY
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 lntarlcukin-2 pharasla LAK calU
XX
X
Cjciopbocptualda/doxozublcln lntarfaron-a
Criteria for Response Patients who received at least the 1 st week of IL-2 and who underwent restaging of their disease were considered eligible for response evaluation. Complete response was defined as disappearance of all measurable and assessable disease for a minimum of 4 weeks. Partial response was defined as a 50% or greater reduction in the sum of the products of the perpendicular diameters of all measured lesions without an increase in the size of any lesion or appearance of new lesions. Because the cycle was long (6-10 weeks), a partial response was determined on a single evaluation done before a new cycle of treat932
ment was begun. Progressive disease was defined as an increase of 25% in the sum of the products of the perpendicular diameters of all measured lesions from the point of maximal response or the appearance of new lesions. All patients not meeting criteria for partial or complete response (stable disease, minor response, and progressive disease) were classified as nonresponders. Response duration was measured from the date tumor measurements met the criteria for a partial response until the date progressive disease was documented (following the best response) or, for ongoing responses, until the last clinic visit in which tumor measurements were recorded. Statistical Methods
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
syndrome induced by IL-2. Other supportive therapy included cimetidine (300 mg given orally four times a day) and a nonsteroidal anti-inflammatory agent (piroxicam at a dose of 20 mg given orally every day in the first 25 patients and indomethacin at a dose of 50 mg given orally three times a day in all subsequent patients) starting 7 days before the first dose of IL-2. Cimetidine was continued without interruption while the patient remained on study. The piroxicam or indomethacin was continued until day 16. Aspirin (650 mg given orally four times a day) was administered daily beginning on day 21 and concurrently with IFN a-2a administration. A multivitamin (Vicon-C; Whitby Pharmaceuticals, Inc., Richmond, Va.) was given twice a day, orally, throughout the treatment cycle to prevent vitamin and zinc deficiencies previously documented in patients receiving IL-2-LAK cells (28). Acetaminophen and antiemetic and antidiarrheal medications were given as needed. As prophylaxis for the frequent catheter infections and malfunctions seen in our earlier IL-2-based studies, patients received cefadroxil (Duricef; Mead Johnson Pharmaceutical, Evansville, Ind.) at a dose of 500 mg given orally twice a day or cefazolin (Ancef; Smith Kline & French, Philadelphia, Pa.) at a dose of 1 g given intravenously every 12 hours and heparin at a dose of 5000 units given subcutaneously twice a day from day 0 to day 16 (29). Treatment with IL-2 was withheld in instances where toxicity could not be managed safely with intensive supportive care and monitoring. Such treatment was restarted at 75% of the original dose after toxic effects resolved. LAK cells were not administered if the blood oxygen saturation was less than 90% on low-flow oxygen (2 L/min). IFN ot-2a was held and reduced in dose level by 50% for severe fatigue, intractable vomiting or diarrhea, neurologic toxicity, or elevation in the levels of transaminases by greater than five times the normal level. Granulocytopenia was not considered a dose-limiting toxic effect for IFN a-2a. No dose modifications were made for cyclophosphamide or doxorubicin.
21-42
Fig. 1. Treatment regimen. IL-2 was administered by continuous infusion for 120 hours. For the first 45 patients in the study, pheresis was done on day 9 and LAK cells were administered on day 14. Cyclophosphamide (300 mg/m2) and doxorubicin (25 mg/m2) were given by intravenous bolus. IFN a-2a was given at a dose of 12 million units/m2 thrice weekly x 9 doses.
The 90% confidence interval (CI) on the response rates was calculated on the basis of the exact binomial distribution. Response rates and toxicity rates were compared among the various subgroups by using Fisher's Exact Test and two-sided P values. Comparisons of responders and nonresponders with respect to various potential prognostic indicators were made by means of the Wilcoxon rank-sum test, with adjustment for ties and a continuity correction using two-sided significance levels.
Results Toxicity
The percentages of treated patients with toxic effects are shown in Table 3. However, the toxic effects experienced by patients who received IL-2 at a dose of 3 million units/m2 per day and underwent dose escalation in their second or third courses are excluded from Table 3 and are discussed separately. Except for granulocytopenia, the vast majority of adverse events listed in Table 3 occurred on days 0-16, the IL-2-LAK cell phase of the treatment cycle. Twenty-nine patients received IL-2 at a dose of 3 million units/m2 per day. Only four patients (14%) required pressor support for hypotension, and only one patient (3%) required temporary discontinuation of IL-2 for dyspnea at rest. One renal cell carcinoma patient with transient chest pain was able to finish IL-2 priming (days 0-5), but this patient had positive creatinine phosphate kinase-MB bands and was subsequently proven to have severe coronary artery disease by heart catheterization, which had not been detected by pretreatment treadmill testing. Although 18 (62%) of the 29 patients developed granulocyte counts less than lOOO/p-L, this toxicity was seen primarily during IFN a-2a administration, was rapidly reversible, and was not associated with episodes of infection.
Journal of the National Cancer Institute
Table 3. Percentage of patients with toxic effects IL-2 dose* Toxic effect
6 (N = 34)
6t (N = 22)
3 0 14 0 0 3 24
32 I8§ 38 3 3 3 44
32 5 41 9 0 0 23
34 3 0 0 0 3 48 0 62 7 14 0
29 21 15 50 12 9 62 3 38 21 3 9
36 14 14 32 18 5 36 0 23 9 5 5
•Dose expressed in million units per square meter. Values in columns are percents. t IL-2 dose decreased to 4.5 million units/m2 per day on days 11 and 12 $ Requiring interruption of treatment. §Two deaths attributed to respiratory failure.
The dose of IL-2 was escalated in eight of the 29 patients in their second or third cycles, resulting in toxic effects more severe than those seen at the lower dose of IL-2. Of these eight patients, one developed pericarditis, three experienced respiratory distress requiring temporary discontinuation of IL-2 and/or omission of a LAK cell dose, one patient developed severe mucositis and a myocardial infarction, and three required pressors for hypotension. Four of eight patients also had creatinine elevations greater than three times the normal upper limit. A dose of IL-2 at 6 million units/m2 per day administered during the priming and with infusions of LAK cells was associated with an increased incidence of weight gain (>10% of baseline), nausea and vomiting, mucositis, fatigue, reversible creatinine elevations (>3 times normal), and pulmonary and cardiovascular toxic effects. Of the 34 patients, 11 (32%) experienced sufficient pulmonary toxicity to require permanent or temporary discontinuation of IL-2 and/or omission of a LAK cell dose. Pulmonary toxicity in these patients usually resolved within 48-72 hours. Six patients (18%) developed severe pulmonary dysfunction. One patient was intubated because of an adrenal hemorrhage (see below); another was intubated, with rapid resolution after IL-2 was discontinued; two patients developed adult respiratory distress syndrome, not requiring intubation, which resolved over 2 weeks; and two patients died of respiratory failure following LAK cell infusions. The autopsy on one of these patients showed diffuse alveolar damage, which was thought to be the primary cause of death, and prominent myocardial lymphocytic and eosinophilic infiltrates. Cardiovascular toxicity of IL-2 at a dose of 6 million units/m2 per day was manifest primarily as hypotension requiring pressor support in 38% of patients. One patient was found to have evidence of a myocardial infarction on a routine
Vol. 84, No. 12, June 17, 1992
electrocardiogram performed after finishing IFN ot-2a on cycle 2, and another patient developed an eight-beat run of ventricular tachycardia while on phenylephrine, which required temporary cessation of IL-2. One death, occurring suddenly several hours after IL-2 priming was finished, was related to cardiac toxicity. The autopsy revealed a diffuse lymphocytic myocarditis, although creatinine phosphokinase values obtained several hours before the fatal event were within normal limits. Other serious toxic effects were encountered in this group of 34 patients, including one case each of a bowel perforation requiring a laparotomy (not associated with tumor at the site of perforation) and a pericardial effusion requiring surgical placement of a drainage window. Another patient developed an adrenal hemorrhage in his only remaining adrenal gland early during priming, followed by abrupt onset of hypotension, creatinine elevation to 6 mg/dL, mental status changes, and respiratory insufficiency (30). The patient recovered rapidly after steroids were administered. The major toxic effects for the 22 patients who received IL-2 at a dose of 6 million units/m2 per day except during the 2 days of LAK cell infusions (4.5 million units/m2 per day) were also cardiovascular and pulmonary. Of 22 patients, nine (41%) required vasopressors for hypotension and seven (32%) required interruption of treatment for respiratory distress. Only one patient developed more severe pulmonary toxicity, but the manifestations of adult respiratory distress syndrome resolved within I week and were managed without intubation. One death occurred in this cohort. A 42-year-old woman with a massive primary renal cell carcinoma metastatic to supraclavicular lymph nodes developed hypophosphatemia during priming; this condition appeared to precipitate a severe hemolytic reaction following a blood transfusion. The immediate consequences were hypoxia (not requiring intubation), metabolic acidosis, renal failure, and encephalopathy. Although the ARTICLES 933
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Dyspnea or hypoxemiat Intubation or adult respiratory distress syndrome Hypotension requiring pressors Arrhythmias requiring therapy Pericarditis Myocardial infarction Weight gain of >IO% Creatinine level >3-6 times normal level > 6 times normal level Neurologic, grades 3/4 Nausea/vomiting precluding oral intake Diarrhea. >7 stools/d Bilirubin level >5 times normal level Transaminase level >5 times normal level Bowel perforation Granulocyte count
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Pilot Study of Interleukin-2 and Lymphokine-Activated Killer Cells Combined With Immunomodulatory Doses of Chemotherapy and Sequenced With Interferon Alfa-2a in Patients With Metastatic Melanoma and Renal Cell Carcinoma Mario Sznol,* Jeffrey W. Clark, John W. Smith II, Ronald G. Steis, Walter J. Urba, Lawrence V. Rubinstein, Louis A. VanderMolen, John Janik, William H. Sharfman, Robert G. Fenton, Stephen P. Creekmore, Peter Kremers, Kevin Conlon, Jean Hursey, Joy Beveridge, Dan L. Longo
Background: Experiments in animal tumor models suggest that the antitumor effects of interleukin-2 (IL-2) or IL-2 in combination with lymphokine-activated killer (LAK) cells can be enhanced by chemotherapy agents such as cyclophosphamide or doxorubicin or by the biologic agent interferon a. Purpose: We determined the toxicity and clinical response rate of an IL-2-LAK cell regimen modified by the addition of moderate, immunomodulatory doses of chemotherapy and sequenced with interferon alfa-2a (IFN a-2a) in patients with metastatic melanoma and renal cell carcinoma. Methods: IL-2 (3-6 million units/m2 per day) was administered by continuous infusion on days 0-5 and days 11-16. LAK cells were infused on days 11 and 12 or on days 11, 12, and 14. Low doses of cyclophosphamide (300 mg/ m2) and doxorubicin (25 mg/m2) were given on day 9 before the LAK cell infusions. Following the IL-2-LAK cell infusion, IFN a-2a (12 million units/m2) was administered for a total of nine doses to complete a cycle of treatment. A total of 89 patients were enrolled in the study. Results: For each histology, there were eight partial responses in 40 assessable patients, for an overall response rate of 20% (90% confidence interval = 10%-33%). The median response duration was 5 months, although two patients with renal cell carcinoma and one patient with metastatic melanoma had almost complete disappearance of tumor and are still responding after 26 + , 22 + , and 26+ months, respectively. Toxic effects were severe in patients receiving the highest dose of IL-2 administered in this study and similar to those reported with other high-dose IL-2-LAK cell regi-
Vol. 84, No. 12, June 17, 1992
mens. Although toxic effects were completely reversible in most patients, there were four treatment-related deaths. Conclusions: This regimen is active in patients with metastatic melanoma and renal cell carcinoma and produces meaningful responses in a small percentage of these patients; however, it is not clear whether cyclophosphamide, doxorubicin, and IFN a-2a as used in this protocol appreciably augmented the antitumor activity of the IL-2-LAK cell regimen. [J Natl Cancer Inst 84:929-937, 1992] Received November 21, 1991; revised February 21, 1992; accepted March 3, 1992. Supported in part by Public Health Service contract NOI-CO-74102 from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services. M. Sznol, Clinical Research Branch, Biological Response Modifiers Program, Division of Cancer Treatment, National Cancer Institute (NCI), NCIFrederick Cancer Research and Development Center, Frederick, Md.; and Investigational Drug Branch, Cancer Therapy Evaluation Program, Division of Cancer Treatment, NCI, Bethesda, Md. J. W. Clark, J. W. Smith II, R. G. Steis, L. A. VanderMolen, J. Janik, W. H. Sharfman, R. G. Fenton, S. P. Creekmore, K. Conlon, D. L. Longo, Clinical Research Branch, Biological Response Modifiers Program, Division of Cancer Treatment, NCI-Frederick Cancer Research and Development Center. W. J. Urba, J. Beveridge, Program Resources Inc./DynCorp, NCI-Frederick Cancer Research and Development Center. L. V. Rubinstein, Biometric Research Branch, Cancer Therapy EvaJuation Program, Division of Cancer Treatment, NCI, Bethesda. P. Kremers, J. Hersey, Frederick Memorial Hospital. We thank the Biological Response Modifiers Program nursing staff for providing excellent care to the patients and Terry Phillips for her outstanding work as editorial assistant. Correspondence to: Mario Sznol, M.D., Executive Plaza North, Rm. 715, National Institutes of Health, Bethesda, MD 20892.
ARTICLES
929
When first introduced into clinical trials, interleukin-2 (IL-2) and lymphokine-activated killer (LAK) cells appeared to have substantial activity in patients with renal cell carcinoma and metastatic melanoma {1,2). A number of confirmatory studies were initiated, and various attempts were made to modify the dose, schedule, and route of administration of the combination of IL-2 and LAK cells to increase efficacy and decrease toxicity (3-7). The trials produced low objective response rates in the range of 20% or less, and there was no evidence that manipulation of the original bolus IL-2-LAK cell regimen improved clinical outcome. Although most patients in these studies did not benefit from IL-2-LAK cell therapy, transient regression of tumor occurred in many of the patients classified as nonresponders. This finding raised the possibility that combining IL-2-LAK cells with agents that enhanced their antitumor immunomodulatory properties in vivo or adding other active biologic agents to this regimen may result in a greater percentage of patients reaching durable partial or complete responses.
In similar animal models, cyclophosphamide was shown to decrease the toxicity of IL-2 while enhancing its antitumor activity (10,13-15). A clinical trial in patients with melanoma indicated that low-dose cyclophosphamide could be combined with outpatient doses of IL-2, resulting in response rates comparable to the high-dose, more toxic IL-2-LAK cell regimen (16). These data suggested that low-dose cyclophosphamide may potentiate the antitumor activity of IL-2-LAK cells. Interferon-a (IFN-a) was known to be active in patients with melanoma and renal cell carcinoma, with response rates in the range of 15%-20% (17,18). Although the mechanism of action in vivo was not understood, IFN-a had direct antiproliferative effects in vitro and increased the expression of tumorassociated and histocompatibility antigens on tumor cells (19,20). With respect to antitumor activity in animal models, the combination of IFN-a and IL-2 was found to be superior to either of the agents given singly; although concurrent admin-
930
The Biological Response Modifiers Program of the National Cancer Institute (NCI) had conducted pilot studies of IL-2 administered by continuous infusion in combination with LAK cells. This regimen was tolerable, induced a high yield of LAK cells, and produced responses in patients with renal cell carcinoma (27). On the basis of the immunomodulatory properties of cyclophosphamide and doxorubicin, as well as their synergy with IL-2 and IL-2-LAK cells in animal models, we administered these agents at fixed doses 48 hours before the first LAK cell infusion. Interferon a-2a (IFN a-2a) was given as a single agent three times a week during the interval between courses of treatment with IL-2-LAK cells. This article describes the toxicity and antitumor activity of this regimen in patients with renal cell carcinoma and metastatic melanoma.
Patients and Methods All patients signed informed consent forms before enrollment in the study. The protocol was reviewed and approved by the NCI Clinical Research Subpanel, the Institutional Review Board of the NCI-Frederick Cancer Research and Development Center, and the Cancer Therapy Evaluation Program of the NCI. Eligibility
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Preclinical animal studies demonstrated that two chemotherapy agents, doxorubicin and cyclophosphamide, could increase the antitumor activity of IL-2. Salup et al. (8) examined the effects of nephrectomy followed by the administration (intravenous and intraperitoneal) of doxorubicin, IL-2, and LAK cells to mice bearing implanted renal adenocarcinomas. Following nephrectomy, only the combination of doxorubicin administered 24 hours before a 3-day course of IL-2-LAK cells markedly prolonged the survival of mice with tumors disseminated to lymph nodes. Other investigators demonstrated an increase in survival when IL-2 administration preceded and followed a single dose of doxorubicin in mice bearing the EL-4 lymphoma (9) and when doxorubicin was given before IL-2 in mice with established micrometastases of the MCA-105 sarcoma or MCA-38 adenocarcinoma (10). In addition to its direct antitumor effects in vivo, doxorubicin was found to increase allogeneic cytotoxic T-lymphocyte responses, IL-2 production by splenic T cells, macrophage cytotoxicity, the number of immature macrophages and monocytes, and interleukin-1 and tumor necrosis factor (TNF) production by macrophages (77). Preliminary evidence also suggested that doxorubicin could direct the migration of activated lymphocytes to sites of tumor (72).
istration produced the best results, sequential administration was also effective in some of the models (21-23) and obviated the inhibitory effects of IFN-a on generation of LAK cells (24). Furthermore, IFN-a was not induced in vivo during treatment with IL-2-LAK cells, unlike certain other cytokines (TNF and IFN-7) which could be detected in serum following IL-2 administration (25,26). Thus, IFN-a was considered a potential non-cross-resistant agent that could maintain or, perhaps, enhance responses to IL-2-LAK cells.
Tests to determine eligibility and tumor measurements were performed within 14 days of beginning therapy (stresstreadmill and pulmonary function tests within 42 days). These tests were similar to those in our prior studies on IL-2-LAK cells (27). Patients at least 15 years old with histologically confirmed metastatic or unresectable renal cell carcinoma or malignant melanoma were eligible for this trial. All patients were required to have a Karnofsky performance status of 70%-100%, an expected survival of greater than 3 months, bidimensionally measurable disease that was clearly progressive, normal results from a stress-treadmill examination, a diffusion capacity on pulmonary function testing of greater than 70% of predicted and a Po2 on room air of greater than 65 mm Hg, as well as normal renal and liver functions. Patients were excluded from the study for the following reasons: a history or evidence of coronary artery disease; congestive heart failure or serious cardiac arrhythmias; active infection requiring antibiotics; the presence of organ allografts or brain metastases; a requirement for corticosteroids; prior therapy with IL-2 (this exclusion criterion took effect after the initial 31 patients were accrued to the protocol); and prior chemotherapy, prior radiotherapy, or prior immunotherapy or hormonal therapy within 3 weeks (6 weeks for nitrosoureas or mitomycin C) of entering the protocol.
Journal of the National Cancer Institute
Patient Characteristics A total of 89 patients were registered for this study. Four patients did not receive IL-2 and, therefore, could not be assessed for toxic effects or response. Forty-three patients with metastatic melanoma and 42 patients with renal cell carcinoma could be assessed for toxic effects. The characteristics of these patients are shown in Tables 1 and 2, respectively. Treatment
Vol. 84, No. 12, June 17, 1992
Characteristic No.
45 43 40
Assessable for toxic effects Assessable for response Age, y Median Average Range Men/women
45 44 19-70 25:18 28/15
Time from diagnosis, d Median Average Range
828 1302 33-5027
Prior therapy No chemotherapy or immunotherapy Prior surgery Prior chemotherapy Prior radiotherapy Prior immunotherapy IFN-a or IL-2
24 (56) 40 (93) 10 (23) 7 (16) 13 (30) 10 (23)
Sites of disease Skin or lymph node only Lung ± above Liver ± above Other ± above
23 (53) 12 (28) 4 (9) 4 (9)
Performance status, 100:90
•Unless otherwise indicated, values = No. of patients (%).
Table 2. Patient characteristics—renal cell carcinoma* Characteristic No.
44
Assessable for toxic effects
42
Assessable for response
40
Age, y Median Average Range
55 54 26-71
Performance status, 100.90:80:70
22:13:4:3
Men/women
35/7
Time from diagnosis, d Median Average Range
371 624 46-2780
Prior therapy No chemotherapy or immunotherapy Prior surgery (nephrectomy) Prior chemotherapy Prior radiotherapy Prior immunotherapy, IFN-a or IL-2
34(81) 34(81) 4(10) 10 (24) 4(10)
Site of disease
Lung or pleura only Lymph node ± above Intra-abdominal ± above Liver or bone ± above
12(29) 9(21) 14 (33) 7(17)
•Unless otherwise indicated, values = No. of patients (%).
dose was re-escalated to 6 million units/m2 per day to complete 120 hours of continuous infusion. Vasopressors and furosemide were used as needed to reverse hypotension and weight gain resulting from a capillary leak ARTICLES 931
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
A central venous access device was placed in all patients. IL-2 (Hoffmann-LaRoche, Nutley, N.J.) was administered at a dose of 3-6 million units/m2 per day by continuous infusion for 120 hours on days 0-5 and days 11-16 (Fig. 1). On days 7 and 8 (and day 9 prior to chemotherapy for the first 45 patients on study) patients underwent a 5-hour leukapheresis. On day 9, patients received cyclophosphamide (300 mg/m2) and doxorubicin (25 mg/m2) intravenously by bolus. LAK cells were generated in vitro as previously described (27) and administered by a central venous access device (> 1 hour) beginning 1 hour after IL-2 was initiated on day 11 (cells from pheresis on day 7), day 12 (cells from pheresis on day 8), and day 14 (in the first 45 patients on study who underwent pheresis on day 9). On day 21, patients started IFN a-2a (HoffmannLaRoche) at a dose of 12 million units/m2 subcutaneously thrice weekly for nine doses. Patients were evaluated for response following the last dose of IFN ot-2a (day 42). Those with stable disease or partial response were eligible for repeat cycles of treatment beginning 1-4 weeks after the completion of the previous cycle. Our previous trials (27) of continuous infusion of IL-2 and LAK cells used Cetus IL-2 at a dose of 3 million Cetus units/m2 per day, which produced reversible grade 3 and grade 4 toxicity in the majority of patients and required treatment administration in a specialized monitoring unit. Since the clinical conversion factor for Cetus IL-2 units to Roche IL-2 units was not clear, we initiated the current study of Roche IL-2 at 3 million units/m2 per day. After treatment was completed in 21 melanoma patients and eight renal cell carcinoma patients, we concluded that the dose of IL-2 could be raised to 6 million units/m2 per day, based on preliminary experience from other investigators in concurrent trials and the mild toxicity seen in our patients receiving 3 million units/m2 per day. A total of 34 patients (22 melanoma patients and 12 renal cell carcinoma patients) were subsequently treated with IL-2 at a dose of 6 million units/m2 per day. Five of the first 16 patients receiving this dose experienced sufficient pulmonary toxicity during the LAK cell infusions to require omission of a LAK cell dose; all subsequent patients, therefore, had pheresis procedures reduced to two and a maximum of two LAK cell infusions. Analysis of the latter group (18 patients) continued to show excessive pulmonary toxicity during the 2 days of LAK cell infusions, and several major toxic events, including three deaths, were reported in the last 12 of the 34 patients entered at this dose of IL-2. In an attempt to reduce toxicity, the remaining patients (22 total, all renal cell carcinoma patients, since melanoma accrual had been completed) received a reduced dose of IL-2 (4.5 million units/m2 per day) only on days 11 and 12 in conjunction with LAK cell administration. In the absence of dyspnea at rest, neurologic toxicity, or hypotension supported by maximum tolerable doses of pressors on day 13, the IL-2
Table 1. Patient characteristics—melanoma*
DAY
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 lntarlcukin-2 pharasla LAK calU
XX
X
Cjciopbocptualda/doxozublcln lntarfaron-a
Criteria for Response Patients who received at least the 1 st week of IL-2 and who underwent restaging of their disease were considered eligible for response evaluation. Complete response was defined as disappearance of all measurable and assessable disease for a minimum of 4 weeks. Partial response was defined as a 50% or greater reduction in the sum of the products of the perpendicular diameters of all measured lesions without an increase in the size of any lesion or appearance of new lesions. Because the cycle was long (6-10 weeks), a partial response was determined on a single evaluation done before a new cycle of treat932
ment was begun. Progressive disease was defined as an increase of 25% in the sum of the products of the perpendicular diameters of all measured lesions from the point of maximal response or the appearance of new lesions. All patients not meeting criteria for partial or complete response (stable disease, minor response, and progressive disease) were classified as nonresponders. Response duration was measured from the date tumor measurements met the criteria for a partial response until the date progressive disease was documented (following the best response) or, for ongoing responses, until the last clinic visit in which tumor measurements were recorded. Statistical Methods
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
syndrome induced by IL-2. Other supportive therapy included cimetidine (300 mg given orally four times a day) and a nonsteroidal anti-inflammatory agent (piroxicam at a dose of 20 mg given orally every day in the first 25 patients and indomethacin at a dose of 50 mg given orally three times a day in all subsequent patients) starting 7 days before the first dose of IL-2. Cimetidine was continued without interruption while the patient remained on study. The piroxicam or indomethacin was continued until day 16. Aspirin (650 mg given orally four times a day) was administered daily beginning on day 21 and concurrently with IFN a-2a administration. A multivitamin (Vicon-C; Whitby Pharmaceuticals, Inc., Richmond, Va.) was given twice a day, orally, throughout the treatment cycle to prevent vitamin and zinc deficiencies previously documented in patients receiving IL-2-LAK cells (28). Acetaminophen and antiemetic and antidiarrheal medications were given as needed. As prophylaxis for the frequent catheter infections and malfunctions seen in our earlier IL-2-based studies, patients received cefadroxil (Duricef; Mead Johnson Pharmaceutical, Evansville, Ind.) at a dose of 500 mg given orally twice a day or cefazolin (Ancef; Smith Kline & French, Philadelphia, Pa.) at a dose of 1 g given intravenously every 12 hours and heparin at a dose of 5000 units given subcutaneously twice a day from day 0 to day 16 (29). Treatment with IL-2 was withheld in instances where toxicity could not be managed safely with intensive supportive care and monitoring. Such treatment was restarted at 75% of the original dose after toxic effects resolved. LAK cells were not administered if the blood oxygen saturation was less than 90% on low-flow oxygen (2 L/min). IFN ot-2a was held and reduced in dose level by 50% for severe fatigue, intractable vomiting or diarrhea, neurologic toxicity, or elevation in the levels of transaminases by greater than five times the normal level. Granulocytopenia was not considered a dose-limiting toxic effect for IFN a-2a. No dose modifications were made for cyclophosphamide or doxorubicin.
21-42
Fig. 1. Treatment regimen. IL-2 was administered by continuous infusion for 120 hours. For the first 45 patients in the study, pheresis was done on day 9 and LAK cells were administered on day 14. Cyclophosphamide (300 mg/m2) and doxorubicin (25 mg/m2) were given by intravenous bolus. IFN a-2a was given at a dose of 12 million units/m2 thrice weekly x 9 doses.
The 90% confidence interval (CI) on the response rates was calculated on the basis of the exact binomial distribution. Response rates and toxicity rates were compared among the various subgroups by using Fisher's Exact Test and two-sided P values. Comparisons of responders and nonresponders with respect to various potential prognostic indicators were made by means of the Wilcoxon rank-sum test, with adjustment for ties and a continuity correction using two-sided significance levels.
Results Toxicity
The percentages of treated patients with toxic effects are shown in Table 3. However, the toxic effects experienced by patients who received IL-2 at a dose of 3 million units/m2 per day and underwent dose escalation in their second or third courses are excluded from Table 3 and are discussed separately. Except for granulocytopenia, the vast majority of adverse events listed in Table 3 occurred on days 0-16, the IL-2-LAK cell phase of the treatment cycle. Twenty-nine patients received IL-2 at a dose of 3 million units/m2 per day. Only four patients (14%) required pressor support for hypotension, and only one patient (3%) required temporary discontinuation of IL-2 for dyspnea at rest. One renal cell carcinoma patient with transient chest pain was able to finish IL-2 priming (days 0-5), but this patient had positive creatinine phosphate kinase-MB bands and was subsequently proven to have severe coronary artery disease by heart catheterization, which had not been detected by pretreatment treadmill testing. Although 18 (62%) of the 29 patients developed granulocyte counts less than lOOO/p-L, this toxicity was seen primarily during IFN a-2a administration, was rapidly reversible, and was not associated with episodes of infection.
Journal of the National Cancer Institute
Table 3. Percentage of patients with toxic effects IL-2 dose* Toxic effect
6 (N = 34)
6t (N = 22)
3 0 14 0 0 3 24
32 I8§ 38 3 3 3 44
32 5 41 9 0 0 23
34 3 0 0 0 3 48 0 62 7 14 0
29 21 15 50 12 9 62 3 38 21 3 9
36 14 14 32 18 5 36 0 23 9 5 5
•Dose expressed in million units per square meter. Values in columns are percents. t IL-2 dose decreased to 4.5 million units/m2 per day on days 11 and 12 $ Requiring interruption of treatment. §Two deaths attributed to respiratory failure.
The dose of IL-2 was escalated in eight of the 29 patients in their second or third cycles, resulting in toxic effects more severe than those seen at the lower dose of IL-2. Of these eight patients, one developed pericarditis, three experienced respiratory distress requiring temporary discontinuation of IL-2 and/or omission of a LAK cell dose, one patient developed severe mucositis and a myocardial infarction, and three required pressors for hypotension. Four of eight patients also had creatinine elevations greater than three times the normal upper limit. A dose of IL-2 at 6 million units/m2 per day administered during the priming and with infusions of LAK cells was associated with an increased incidence of weight gain (>10% of baseline), nausea and vomiting, mucositis, fatigue, reversible creatinine elevations (>3 times normal), and pulmonary and cardiovascular toxic effects. Of the 34 patients, 11 (32%) experienced sufficient pulmonary toxicity to require permanent or temporary discontinuation of IL-2 and/or omission of a LAK cell dose. Pulmonary toxicity in these patients usually resolved within 48-72 hours. Six patients (18%) developed severe pulmonary dysfunction. One patient was intubated because of an adrenal hemorrhage (see below); another was intubated, with rapid resolution after IL-2 was discontinued; two patients developed adult respiratory distress syndrome, not requiring intubation, which resolved over 2 weeks; and two patients died of respiratory failure following LAK cell infusions. The autopsy on one of these patients showed diffuse alveolar damage, which was thought to be the primary cause of death, and prominent myocardial lymphocytic and eosinophilic infiltrates. Cardiovascular toxicity of IL-2 at a dose of 6 million units/m2 per day was manifest primarily as hypotension requiring pressor support in 38% of patients. One patient was found to have evidence of a myocardial infarction on a routine
Vol. 84, No. 12, June 17, 1992
electrocardiogram performed after finishing IFN ot-2a on cycle 2, and another patient developed an eight-beat run of ventricular tachycardia while on phenylephrine, which required temporary cessation of IL-2. One death, occurring suddenly several hours after IL-2 priming was finished, was related to cardiac toxicity. The autopsy revealed a diffuse lymphocytic myocarditis, although creatinine phosphokinase values obtained several hours before the fatal event were within normal limits. Other serious toxic effects were encountered in this group of 34 patients, including one case each of a bowel perforation requiring a laparotomy (not associated with tumor at the site of perforation) and a pericardial effusion requiring surgical placement of a drainage window. Another patient developed an adrenal hemorrhage in his only remaining adrenal gland early during priming, followed by abrupt onset of hypotension, creatinine elevation to 6 mg/dL, mental status changes, and respiratory insufficiency (30). The patient recovered rapidly after steroids were administered. The major toxic effects for the 22 patients who received IL-2 at a dose of 6 million units/m2 per day except during the 2 days of LAK cell infusions (4.5 million units/m2 per day) were also cardiovascular and pulmonary. Of 22 patients, nine (41%) required vasopressors for hypotension and seven (32%) required interruption of treatment for respiratory distress. Only one patient developed more severe pulmonary toxicity, but the manifestations of adult respiratory distress syndrome resolved within I week and were managed without intubation. One death occurred in this cohort. A 42-year-old woman with a massive primary renal cell carcinoma metastatic to supraclavicular lymph nodes developed hypophosphatemia during priming; this condition appeared to precipitate a severe hemolytic reaction following a blood transfusion. The immediate consequences were hypoxia (not requiring intubation), metabolic acidosis, renal failure, and encephalopathy. Although the ARTICLES 933
Downloaded from http://jnci.oxfordjournals.org/ at Fudan University on May 16, 2015
Dyspnea or hypoxemiat Intubation or adult respiratory distress syndrome Hypotension requiring pressors Arrhythmias requiring therapy Pericarditis Myocardial infarction Weight gain of >IO% Creatinine level >3-6 times normal level > 6 times normal level Neurologic, grades 3/4 Nausea/vomiting precluding oral intake Diarrhea. >7 stools/d Bilirubin level >5 times normal level Transaminase level >5 times normal level Bowel perforation Granulocyte count
