- © 2003 by American Society of Clinical Oncology
Interferon Alfa-2b Three Times Daily and Thalidomide in the Treatment of Metastatic Renal Cell Carcinoma
- Address reprint requests to Micaela Hernberg, MD, PhD, Department of Oncology, Helsinki University Central Hospital, PO Box 180, FIN-00029 Helsinki, Finland; e-mail: micaela.hernberg{at}hus.fi.
Abstract
Purpose: The antiangiogenic effect of interferon (IFN) may improve with frequent dosing and by combination with other agents with antiangiogenic activity. To evaluate this potential, we treated patients with metastatic renal cell carcinoma (RCC) with frequently dosed IFN and thalidomide.
Patients and Methods: Thirty patients were given IFN-α-2b 0.9 MU subcutaneously three times daily for 1 month and subsequently 1.2 MU tid unless serious toxicity was encountered. Thalidomide was first given 100 mg/d for 1 week and 300 mg/d thereafter. Sera were collected before and during treatment for serum vascular endothelial growth factor (S-VEGF) analyses performed using enzyme-linked immunosorbent assay.
Results: The intention-to-treat response rate was 20% (95% CI, 6% to 34%) and response rate for assessable patients (n = 27) was 22% (95% CI, 6% to 38%). All responses were partial. In addition, 17 patients (63%; 95% CI, 45% to 81%) had stable disease for 3 months or longer. The median time to treatment failure was 7.7 months, and median survival time was 14.9 months. The most common cause of thalidomide discontinuation was neuropathy. S-VEGF levels decreased more in patients who responded to therapy compared with those in patients whose condition had stabilized or who had progressive disease (P = .036).
Conclusion: The combination of frequently dosed IFN-α-2b and low-dose thalidomide is feasible and active in advanced RCC, but the clinical benefit may remain small compared with that of IFN alone. Results from an ongoing phase III trial comparing IFN-α with or without thalidomide need to be analyzed before this combination can be recommended for use outside clinical studies.
INTERLEUKIN-2 (IL-2) and interferon (IFN) are associated with a response rate of approximately 12% to 20% in the treatment of renal cell carcinoma (RCC) and have been considered the basis of standard immunotherapy.1,2 The combination of IL-2 and IFN may not provide additional benefit, and RCC is generally considered relatively resistant to chemotherapy.2–,5 Overtly metastatic RCC cannot be cured with any known systemic therapy, although a few patients with advanced disease may survive more than 10 years after initiation of therapy.1 Hence, novel and more effective therapeutic options are needed for treatment of advanced RCC.
Inhibition of tumor angiogenesis is potentially an attractive approach in the treatment of RCC. Von Hippel-Lindau tumor suppressor gene (VHL) alterations are present in up to 80% of sporadic RCCs.6,7 The VHL protein suppresses transcription of the vascular endothelial growth factor (VEGF) gene, and VHL inactivation may thus result in VEGF overproduction and promotion of angiogenesis in RCC.6,8 Hypoxia-inducible factor alpha (HIF-α) protein is upregulated in the majority of RCCs, and HIF-α protein expression is associated with upregulation of VEGF protein and increased microvessel density.9 The VHL tumor suppressor protein acts as the substrate recognition component of an ubiquitin E3 ligase that targets HIF-α subunits for proteolysis. Serum VEGF (S-VEGF) levels are elevated in RCC patients, and high levels have been associated with poor outcome in some studies.10,11
IFN has been regarded as a weakly antiangiogenic agent, and some of its activity in the treatment of RCC might result from prevention of blood vessel growth.12 The frequency and the mode of IFN administration may be of importance regarding inhibition of angiogenesis. IFN-α is often given as three subcutaneous injections weekly, which may not be optimal for this purpose. IFN-α administered daily was more effective than IFN-α given at longer intervals in inhibition of tumor vascularization and growth in a nude mouse human bladder cancer xenograft model, which suggests that frequent administration and relatively flat serum concentration levels of IFN might improve its antiangiogenic efficiency.12
Hypothetically, concomitant use of even relatively weak antiangiogenic agents might result in a more marked clinical antiangiogenic activity than the use of the same agents as monotherapy.13,14 Several lines of evidence suggest that thalidomide has some antiangiogenic activity. The teratogenic effects of thalidomide have been attributed to inhibition of blood vessel growth in the developing fetal limb bud,15 and it has been used with success in the treatment of the vascular tumor Kaposi’s sarcoma.16 Thalidomide increases degradation of the fibroblast growth factor and the tumor necrosis factor alpha mRNA,17,18 and induces apoptosis of established neovasculature in some experimental models.17,19 Thalidomide has shown limited activity as monotherapy in the treatment of advanced RCC but with considerable toxicity at high daily doses.20–,24
The aim of this study was to evaluate the safety and efficacy of the combination of IFN-α and thalidomide in the treatment of metastatic RCC. We administered IFN-α as small subcutaneous doses three times daily and thalidomide twice daily to achieve a relatively stable blood concentration of each drug, and we studied whether this might improve the anticancer efficiency, as suggested by some experimental findings. To our knowledge, this approach has not been tested previously in the treatment of patients with advanced RCC.
PATIENTS AND METHODS
Patients
Thirty consecutive patients with metastatic RCC were accrued in a trial evaluating the combination of IFN-α-2b and thalidomide as the first-line systemic treatment for metastatic RCC. The study was carried out in the Helsinki University Central Hospital (Helsinki, Finland) between January 1999 and January 2002. Eligible patients had histologically verified RCC and measurable, progressive metastatic lesions confirmed either by histologic biopsy, cytology, or imaging. Patients who had ever had an invasive cancer other than RCC were ineligible, as were those with WHO performance status higher than 2, known hypersensitivity to IFN-α, severe insufficiency of the liver or the kidney, or deficiency in CNS function. Patients were also excluded from the protocol if they had received prior chemotherapy or immunotherapy for metastatic RCC, or if subcutaneous IFN therapy was not possible for logistic reasons. Premenopausal women and women who had menstrual bleeding during the 12 months before accrual were ineligible unless hysterectomy had been performed and the serum follicle-stimulating hormone level was more than 40 U/L. Written informed consent was obtained from all participating patients. The study was approved by an institutional review committee before accrual of any study participants.
Patient characteristics are listed in Table 1⇓. The median age at enrollment was 61 years and 80% of patients were male. Eighty percent had undergone prior nephrectomy, one patient had prior arterial embolization, and four patients had received local radiotherapy. Nineteen patients (63%) had metastatic disease at the time of diagnosis, and the median distant disease-free interval in the other patients (n = 11) before study enrollment was 24.1 months (range, 5.1 to 50.9 months). Seventy-three percent (n = 22) had two or more metastatic organ sites. Six patients belonged to the low-risk group, 23 to the intermediate-risk group, and one to the high-risk group, as defined by Motzer et al.1 Except for one patient, who had chromophobe carcinoma, all patients had clear-cell renal carcinomas.
Patient Characteristics
Pretreatment Evaluation
An isotope bone scan, a computed tomography (CT) scan or a chest x-ray of the mediastinum and the lungs, and a CT scan of the abdomen were performed within 4 weeks before the initiation of the study treatment. The sites with abnormal uptake in the isotope bone scan were further evaluated using x-ray or CT. Palpable metastases were measured in two perpendicular dimensions. A CBC, a blood chemistry profile, and an ECG were also evaluated before study accrual.
Therapy
The starting dose of thalidomide was 100 mg/d orally, taken in the evening to reduce drug-related daytime somnolence. The dose was gradually increased to 300 mg/d unless grade 3 or greater toxicity occurred. The 300 mg/d dose was divided into two doses: 100 mg taken in the morning and 200 mg taken in the evening. The drug was purchased from Laphal Laboratories (Allauch, France). Thalidomide dose was reduced to 200 mg/d whenever the patient developed grade 2 sensory neuropathy, and the drug was discontinued if grade 3 sensory neuropathy was encountered. When thalidomide was discontinued because of thalidomide-related grade 3 or 4 toxicity, patients were allowed to be treated with IFN alone.
IFN-α-2b 18MU/mL was given subcutaneously using an injection pen (Intron A, Schering-Plough [Brinny] Co, Innishannon, Ireland), and patients were instructed to administer the IFN injections at home. The starting dose was 0.9 MU tid (ie, 2.7 MU/d; 18.9 MU/wk). IFN therapy was started on study day 8, 1 week after the start of thalidomide treatment, in an attempt to avoid concomitant occurrence of thalidomide-related lethargy and flu-like symptoms associated with initiation of IFN therapy. If no severe IFN-related adverse events were encountered, the dose of IFN-α-2b was escalated after the first month of treatment to 1.2 MU tid (3.6 MU/d; 25.2 MU/wk). Thus, the total IFN doses used in this study were comparable to those commonly used in the treatment of metastatic RCC in which IFN was administered three times per week. Whenever grade 3 or 4 adverse events considered to be related to IFN occurred, the IFN dose was reduced, and in some cases treatment was interrupted for a few days. If grade 3 or more serious adverse events considered to be related to IFN persisted or recurred, IFN treatment was discontinued. When IFN-α was discontinued because of toxicity considered to be related to IFN, the patients were allowed to receive single-agent thalidomide.
Evaluation of Response and Toxicity
Clinical evaluation was performed before treatment initiation, on study days 8, 15, and 29, then monthly for the first 6 months of treatment, and thereafter at 2- to 3-month intervals. At every visit the clinical status including the performance status was evaluated, treatment-associated toxicity was recorded, and the blood cell counts and blood chemistry profile were analyzed. Treatment response was evaluated with CT (n = 29) or with chest x-ray and CT of the abdomen (n = 1). Examinations were performed 1, 2, 3, 5, and 7 months after initiation of treatment, and thereafter every 2 months. Objective responses were confirmed with CT or chest x-ray 1 month after the response was first detected. Evaluation of response was assessed according to the Response Evaluation Criteria in Solid Tumors Group guidelines by a radiologist (P.V.).25 Adverse events were registered and classified according to the WHO toxicity criteria.26
S-VEGF Analysis
Blood (≥ 1 mL) was collected in ordinary test tubes, centrifuged at 2,000 × g for 10 minutes at 4°C, and then stored in aliquots at −20°C or colder. The baseline serum samples were taken a median of 10 days before starting the treatment, and the follow-up samples were taken 0.8 months (range, 0.1 to 1.2 months), 2.6 months (range, 1.3 to 3.2 months), 4.2 months (range, 3.3 to 5.5 months), and 7.4 months (range, 5.7 to 13.9 months) after the first treatment day. S-VEGF levels were determined using a commercially available VEGF enzyme-linked immunosorbent assay kit (R & D Systems, Minneapolis, MN) according to the manufacturer’s instructions. The kit uses antihuman VEGF–coated microwells and horseradish peroxidase-conjugated anti-VEGF antibody as the secondary antibody, and it detects VEGF isoforms VEGF121 and VEGF165. For each analysis, 100 mL of serum was used, and all analyses and calibrations were performed in duplicate. The blank was subtracted from the duplicate readings for each standard and sample. A standard curve was created by plotting the logarithm of the mean absorbance of each standard versus the logarithm of the VEGF concentration. S-VEGF levels were determined without prior knowledge of the clinical data. Intra-assay variability was assessed by analyzing two replicates of 15 serum samples and the interassay variation by assaying 10 samples twice in two separate assays. Both intra- and interassay variations were less than 10%. Concentrations are reported as picograms per milliliter.
Statistical Analyses
Analyses were performed using SAS Version 8.2 (SAS, Cary, NC) and the Statview version 4.53 (Abacus Concepts Inc, Berkeley, CA). Groups with abnormal distributions were compared with the Kruskal-Wallis analysis of variance (ANOVA). ANOVA for repeated measures was used to study associations between S-VEGF levels and the time factor, treatment response, and time-response interaction. Baseline measurements were used as covariates. In the multivariate analysis, sex, age at diagnosis, WHO performance status, number of metastatic sites, change in the S-VEGF levels, and nephrectomy status were entered as covariates, and 95% CIs were calculated for treatment response. Survival distributions were estimated using the Kaplan-Meier method. A two-way significance level of 5% was considered to be statistically significant in all analyses.
RESULTS
Treatment Duration and Efficacy
The median duration of thalidomide treatment was 6.5 months (range, 0.5 to 9.1 months) and that of IFN-α-2b therapy 7.2 months (range, 0.4 to 29.6 months). Eleven patients (37%) continued receiving IFN-α monotherapy for a median of 4.7 months (range, 2.8 to 26.8 months) after discontinuation of thalidomide because of toxicity, and two patients (7%) received thalidomide monotherapy for 7.2 and 9.9 months, respectively, after discontinuation of IFN-α because of IFN-related toxicity. Twenty-seven patients (90%) were assessable for treatment response. The reasons for ineligibility of response assessment were short treatment duration (n = 1; 15 days), inadequate pretreatment radiologic evaluation (n = 1), and presence of metastatic lesions too small (< 10 mm in diameter) for reliable response evaluation (n = 1). The overall response rate by an intention-to-treat analysis was 20% (six of 30; 95% CI, 6% to 34%) and overall response for assessable patients was 22% (six of 27; 95% CI, 6% to 38%). All responses (n = 6) were partial, with a median duration of 5.7 months (range, 4.5 to 18.9 months). Seventeen patients (63%; 95% CI, 45% to 81%) had stable disease lasting at least 3 months.
The median time to treatment failure was 7.8 months (95% CI, 7.3 to 13.0 months). At the time of analysis, 10 patients were alive, three of them without progression. The median progression-free survival (PFS) time was 6.9 months (95% CI, 6.6 to 14.4 months), and the median overall survival time was 15.5 months (95% CI, 14.0 to 24.2 months; Fig 1⇓). The median times to treatment failure were 15.2, 7.7, and 6.0 months in the subsets of patients who responded to treatment, who had stabilized disease during treatment, and who had progressive disease, respectively. The respective median overall survival times in these groups were 28.9, 16.2, and 7.5 months, respectively.
Cumulative overall survival (-•-) and time to treatment failure (-▪-) among 30 patients with metastatic renal cell cancer treated with frequently dosed interferon alfa and thalidomide. The proportions of patients alive and without treatment failure at 12 and 24 months after treatment initiation are provided.
Adverse Events
The most commonly observed adverse events were lethargy (100%), constipation (97%), and peripheral sensory neuropathy (80%; Table 2⇓). Lethargy was usually most severe in the beginning of therapy. Constipation generally responded well to laxatives. Thalidomide dose was reduced in 12 patients because of neuropathy. Nineteen patients (63%) discontinued thalidomide because of either the presence of grade 3 neuropathy (n = 7) or progressive grade 2 neuropathy (n = 12), although thalidomide discontinuation was not required by the protocol in case of grade 2 neuropathy. Grade 3 neuropathy developed in four patients within the first 3 study months (range, 0.5 to 5.2 months), and in the rest of the patients (n = 3) within the first 10 months on study. Progressive grade 2 neuropathy that resulted in thalidomide discontinuation took place 7.1 to 29.6 months after thalidomide initiation (median, 9.1 months), except for one patient who had thalidomide discontinued 2 months after enrollment because of grade 2 neuropathy and other concomitant adverse events. After thalidomide discontinuation, the neuropathy symptoms usually persisted, although some improvement was seen in a few patients with longitudinal follow-up.
Adverse Events
Hematologic toxicity was generally mild. Only a minority of the patients had nausea that needed intervention. During the first weeks of IFN-α-2b therapy, 47% of the patients had mild fever (grade 1 or 2). Three patients had deep vein thrombosis detected during therapy, and one patient developed thyrotoxicosis.
S-VEGF Levels and Treatment Response
S-VEGF levels decreased during therapy in responding patients, whereas the decrease was more short-lived in patients who were judged to have either stabilized or progressive disease during the study treatment (P = .036; Fig 2⇓). There was no statistical difference in the pretreatment S-VEGF levels between patients who responded to therapy (median, 248 pg/mL; range, 165 to 1,580 pg/mL; P = .68), those who had stabilized disease (median, 255 pg/mL, range, 46 to 1,150 pg/mL; P = .68), or those who progressed during treatment (median, 410 pg/mL; range, 78 to 825 pg/mL; P = .68). In an ANOVA, treatment response was the only factor that correlated significantly with the degree of change in the S-VEGF levels during treatment (P = .0008), whereas the number of metastatic sites (P = .073), sex (P = .46), age at diagnosis (P = .54), prior nephrectomy (P = .20), and WHO performance status did not (P = .13).
Median serum vascular endothelial growth factor levels during therapy grouped by response to treatment. SD, stable disease; PD, progressive disease; PR, partial response.
DISCUSSION
We evaluated an experimental regimen consisting of frequently dosed IFN-α and thalidomide in an attempt to improve treatment efficacy in metastatic RCC. An overall response rate of 11% was obtained among 463 RCC patients treated with IFN-α in six series,1 for which response rate was within the 95% CI for response (20%; 95% CI, 6% to 34%) obtained in this study. Thalidomide has only modest activity in advanced RCC when used as a single agent. Twelve (6%) of 185 assessable RCC patients treated with thalidomide in eight recent series achieved a partial response, and 73 (39%) had stable disease lasting at least for a few months (Table 3⇓). However, most patients treated with thalidomide in these studies had received prior systemic therapies for metastatic RCC, which may have influenced thalidomide efficacy adversely.
Efficacy of Thalidomide in Advanced Renal Cell Cancer in Nine Studies
The median overall survival time was 15.5 months (95% CI, 14.0 to 24.2 months) in the present series, which is similar to the median survival time of 13.0 months (95% CI, 12 to 15 months) of 463 patients with advanced RCC treated with IFN-α alone in six prospective clinical trials.1 Direct comparisons of response rates and survival times between different series is notoriously difficult because patient selection and other confounding factors may have considerable influence on the reported efficacy parameters. Motzer et al1 found that five factors, consisting of Karnofsky performance status less than 80% at diagnosis, a high serum lactate dehydrogenase level (> 1.5 × upper limit of normal), a low serum hemoglobin level, a high corrected serum calcium level (> 10 mg/dL), and time from diagnosis to IFN-α therapy of less than 1 year, were independently associated with poor outcome in RCC carcinoma patients treated with IFN-α. These authors devised a prognostic score using the five factors to help comparisons of treatment results between different series. In their study the median PFS times of patients with low risk (no risk factors), intermediate risk (one or two risk factors), and high risk (> two risk factors) were 8.3, 5.1, and 2.5 months, respectively. In the present series, 77% of the patients were placed in the intermediate-risk category using this scoring, and the median PFS in the entire series (6.9 months; 95% CI, 6.6 to 14.4 months) is not markedly superior to that of the intermediate-risk group when RCC patients are treated with IFN-α alone (5.1 months).
Because antiangiogenesis agents may work by causing disease stabilization rather than tumor regression as a result of their mechanism of action, it has been suggested that long-lasting periods of stabilized disease should be considered a favorable treatment outcome when the efficacy of antiangiogenic agents is evaluated. Long-lasting disease stabilizations have been observed in prior studies of single-agent thalidomide21,24 and IFN-α,31 but whether such periods of disease stabilization can be interpreted as a useful therapeutic effect remains controversial. The clinical course of RCC is often unpredictable, ranging from rapidly progressing disease to tumors that grow only slowly over several years without any anticancer therapy.
The most difficult adverse events in this series were usually neurological. Sensory neuropathy was common and long lasting. Because the maintenance thalidomide dose was only 300 mg/d, the neuropathy symptoms were somewhat greater than we expected on the basis of reports from earlier studies that used single-agent thalidomide at higher doses.22–,24 Of note, one recent study that evaluated the use of thalidomide and IFN-α in advanced RCC was closed prematurely because of excessive neurotoxicity,32 but in that study, larger single doses of IFN-α (9 MU three times a week) were administered than in the present series. Other adverse neurologic effects, such as visual symptoms and vertigo and dizziness, were not graded as severe, and skin dryness, erythema, hematologic toxicity, constitutional symptoms, hair loss, and nausea were usually mild.
The optimal dose of the IFN-α and thalidomide combination when used to achieve an antiangiogenic effect in the treatment of RCC is not known. In the present study, we used a dose of 0.9 MU given three times daily, which is at the lower end of the range of doses commonly used for treatment of advanced RCC when one of the three-times-weekly regimens is used (15 to 50 MU/wk). In studies that have used single-agent thalidomide the highest response rate (17%) was reported in a study where the smallest daily dose (100 mg/d) was used (Table 3⇑). The higher doses are more toxic, and at present there is little evidence to suggest that the use of high daily thalidomide doses will result in a larger clinical benefit as compared with smaller doses in the treatment of advanced RCC.
IFN administration by the patients with an IFN pen was well accepted, and none of the patients discontinued treatment because of the frequent subcutaneous injections required by the protocol. Sustained serum levels of IFN-α can also be achieved in a practical manner with pegylated IFN, which was not available to us at the time of the study. Experience from phase I and II trials suggests that the adverse effect profile of pegylated IFN-α-2b resembles that of unpegylated IFN-α, but weekly subcutaneous administration of pegylated IFN may be associated with a shorter duration of constitutional symptoms as compared with daily administration.33–,35 This is in line with our present experience, which suggests that patients treated with the regimen three times a day may have less severe constitutional symptoms than those treated with comparable doses of subcutaneous IFN three times weekly.
S-VEGF levels have been suggested to function as a possible predictive marker for treatment response in RCC,10 and in this series, a more sustained decrease in the S-VEGF levels was seen in patients who responded to the therapy as compared with those who did not. No association has usually been found between S-VEGF levels and treatment response in studies in which RCCs have been treated with thalidomide alone. Because most VEGF in the blood circulation is carried in platelets and the WBCs, handling of the blood samples might thus influence S-VEGF levels.36 Therefore, serum VEGF levels need to be interpreted with caution. Yet serum VEGF levels may be considered as a reasonable surrogate parameter for the whole-blood VEGF, given that high pretreatment serum VEGF levels are generally associated with high whole-blood VEGF concentrations in the same cancer patients (unpublished data).
We conclude that the combination of frequently dosed IFN-α-2b and thalidomide is feasible and active in the treatment of advanced RCC. These results suggest that improvement in treatment efficacy appears to remain small at best when IFN-α and thalidomide are used in a combination as compared with IFN-α monotherapy, and the combination is neurotoxic, requiring careful follow-up of patients. Results of the ongoing phase III trial comparing IFN-α with or without thalidomide conducted by Eastern Cooperative Oncology Group are awaited with interest and should be available before this combination can be recommended for use outside clinical trials.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The authors indicated no potential conflicts of interest.
Footnotes
-
Supported by grants from the Cancer Society of Finland, and Research Funds of the Helsinki University Central Hospital.
- Received January 22, 2003.
- Accepted July 30, 2003.
