|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Interferon Adjuvant to Radical Nephrectomy in Robson Stages II and III Renal Cell Carcinoma: A Multicentric Randomized StudyFrom the Division of Urology, Department of Surgery, Division of Medical Statistics and Biometry, and Division of Human Tumors Immunotherapy, Department of Experimental Oncology, Istituto Nazionale per lo Studio e la Cura dei Tumori; and Institute of Medical Statistics and Biometry, University of Milan, Milan, Italy. Address reprint requests to Giorgio Pizzocaro, MD, Urology Division, Department of Surgery, Istituto Nazionale Tumori, Via Venezian 1, 20133 Milan, Italy; email pizzocaro{at}istitutotumori.mi.it
PURPOSE: Because interferon gave promising results in the management of metastatic renal cell carcinoma in the 1980s, a multicentric randomized controlled trial was planned to compare adjuvant recombinant interferon alfa-2b (rIFN  2b) with observation after radical nephrectomy in patients with Robson stages II and III renal cell carcinoma. Overall and event-free survival were to be evaluated together with prognostic factors. PATIENTS AND METHODS: Overall and event-free survival curves for 247 patients (124 controls and 123 treated) were estimated by the Kaplan-Meier method and compared using the log-rank test. Cox’s multiple regression models were adopted to perform a joint analysis of treatment and prognostic factors.
RESULTS: The 5-year overall and event-free survival probabilities were 0.665 and 0.671, respectively, for controls and 0.660 and 0.567, respectively, for the treated group; the differences were not statistically significant (2P = .861 for overall and 2P = .107 for event-free survival with the log-rank test). Regarding prognostic factors, only grade, pT, and pN demonstrated a significant prognostic role. First-order interactions of treatment with pT and pN category were investigated; a significant interaction was found between pN and treatment. A harmful effect of rIFN
CONCLUSION: Adjuvant rIFN
INTERFERON ALFA (IFN ) gave promising results in the management of metastatic renal cell carcinoma (RCC) in the 1980s: objective responses were reported in approximately 15% of patients, and the major responders were patients with lung metastases after tumor nephrectomy.1-3 Two large cooperative randomized studies were carried out in the United States4 and in Germany,5 using lymphoblastoid IFN and recombinant IFN2a, respectively, in radically resected Robson stages II (perinephric fat involved) and III (tumor extension into renal vein or inferior vena cava; resected lymph node metastases) RCC.6
A multicentric randomized controlled trial was planned to compare adjuvant recombinant IFN
The study was initiated on February 1, 1990, after approval was received from the Scientific and Ethics Committees of the Istituto Nazionale per lo Studio e la Cura dei Tumori (INT) in Milan, Italy. All consecutive patients submitted to radical nephrectomy for RCC in the participating centers were to be registered at the secretariat of the Division of Urology at the INT. Unilateral para-aortic lymph node dissection was recommended by protocol, and we relied on the pathologic report to verify that lymphadenectomy was performed. Patients with pathologic stages II and III RCC (1987 tumor-node-metastasis categories T3aN0M0 and T3bN0M0 or T2/3N1-3M0) were eligible for the study. Stratification by participating centers was accomplished according to a scheme of balanced randomized blocks of eight patients each. Randomization within each center was performed by the secretary (S.F.) of the Division of Urology at the INT under the supervision of the study coordinator (L.P.) after the pathologic stage was verified and the signature of informed consent was obtained and in the absence of any exclusion criteria (ie, lymph nodes not described in the pathologic report; age > 70 years; Eastern Cooperative Oncology Group validity index of 2 or more; any serious cardiovascular disease or renal, hepatic, or hematologic disorders; a second tumor, with the exclusion of squamous or basal cell carcinoma of the skin and carcinoma-in-situ of the uterine cervix; a pregnancy status or a risk of pregnancy during IFN therapy; any treatment with biologic response modifiers during the last 6 months; no availability for a regular follow-up for at least 5 years). Patients randomized to the observation (control) arm were to receive no adjuvant therapy and to be followed-up every 6 months for 5 years and yearly thereafter with repetition of full clinical examination, chest x-rays, and abdominal ultrasound at each visit. Bone scan and computed tomography or magnetic resonance imaging were to be performed only on request. In the event of relapse, the patients were to be treated with IFN 10 million IU intramuscularly 3 times per week or with the best available therapy. Patients randomized to adjuvant therapy were to receive IFN 6 million IU intramuscularly 3 times per week for 6 months starting within 1 month of surgery. Only paracetamol could be given (up to 2 g/d) to control the flu syndrome. Patients were to be seen on an outpatient basis every month during therapy to check toxicity. In the event of objective signs of toxicity (WBC < 3,000/µL, platelets < 100.000/µL, serum creatinine > 2 mg/100 mL, AST and ALT increase > 50% of normal values, intractable flu syndrome), the IFN dosage was to be reduced by 50% for 1 month and the normal dose restored after resolution of toxic signs. In the event of persistent toxicity after dose reduction of IFN, therapy was to be suspended for 1 month and the treatment restarted with the reduced dosage. Normal dosage could be restored if no toxicity appeared after 1 month of reduced IFN. After the adjuvant therapy, the follow-up of patients in the treatment arm was to be the same as for patients in the observation arm. In the event of relapse, patients were to be treated with the best available treatment. According to a previous study,8 the probability of event-free survival at 5 years for the control group was estimated to be approximately 50%. By postulating a clinically relevant difference of 20% in the event-free survival probability between the two arms of the study (two-sided test, alpha = 0.05, beta = 0.10), it was determined that a total of 256 patients were needed.9 The follow-up was closed in May 1998. Time to death or to the last observation was computed from the date of surgery; survival curves were estimated using the Kaplan-Meier method and compared using the log-rank test. Local-regional relapses and distant metastases were the events considered to estimate event-free survival. Time to the occurrence of the first of these was computed from the date of surgery. Deaths without evidence of RCC were censored. Deaths from other malignancies were also censored because of possible interference with the second tumor treatment and the difficulty in ascribing an incidental relapse to the first or the second tumor. Cox’s multiple regression models were adopted to investigate the role of prognostic factors, which were inserted into the initial models together with their first-order interaction with adjuvant treatment. Backward procedures were applied to obtain the final parsimonious models. It is well known that if subsets of patients are identified by important prognostic factors, differences between treatment effects are expected to vary from one subset to another, thereby enabling the investigator to anticipate quantitative interactions. In the analysis, attention was therefore focused on qualitative interaction, as only it may be considered relevant from a clinical viewpoint (see Marubini and Valsecchi10). The Gail and Simon11 test was used to discriminate between the two aforementioned types of interaction. To evaluate the appropriateness of the model including the interaction terms, a bootstrap procedure according to Altman and Andersen12 and to Sauerbrei and Schumacher13 was adopted. Univariate analysis was performed on the following prognostic factors: sex (male v female), age (four categories: 32 to 52 years [reference category], 53 to 59 years, 60 to 64 years, and 65 to 70 years), grade (two categories: G1 + G2 [reference category] v G3 + G4), pathologic T category (three categories: pT2, pT3a [reference category], and pT3b), metastatic lymph nodes (three categories: pN0 [reference category], pN1, and pN2/pN3), and Robson stage (two categories: II [reference category] v III). In the multivariate analysis, the two variables pN and pT were separately inserted into the model to test their independent main effects together with their first-order interaction with treatment.
Recruitment was closed in October 1995, and a total of 1,914 radical nephrectomies were registered by 21 participating urological units in Lombardy, Italy. The flowchart of the trial is shown in Fig 1. Overall, 264 patients (13.8% of those registered and 53.7% of the Robson II or III stage patients) were randomized. A total of 226 Robson stage II/III patients were considered ineligible because of age greater than 70 years (42%), lymph nodes not described in the pathologic report (27.9%), patient refusal (16%), or other reasons (14.1%). After analysis of the clinical forms, 17 randomized patients had to be excluded from results evaluation: 15 because they were treated at six centers that were dropped from the study because of poor cooperation, and two (randomized to IFN therapy) because they were lost to follow-up immediately after randomization. Of the 1,914 registered patients who underwent radical nephrectomy, 247 (12.9%) were assessable for the study: 123 treated and 124 observed.
The two treatment groups were well matched for baseline patient and tumor features (Table 1). Overall, 68 (55.3%) of the 123 treated patients developed signs of toxicity (Table 2), and 35 (28%) had a dose reduction and/or suspension of IFN therapy (Table 3). Thus 88 patients completed all planned interferon therapy.
After a median follow-up period of 62 months (range, 5 to 99 months), 89 patients relapsed: four in the retroperitoneum and 85 in distant sites. Relapses occurred in 38 of the 124 controls and in 51 of the 123 treated patients. Four of the 38 relapsed controls received IFN as secondary therapy. Eighty patients died: 39 in the control group and 41 in the IFN group. Seventy renal cancer–related deaths occurred: 33 in the control group and 37 in the IFN group. Ten patients died of unrelated cause without any evidence of RCC: six in the control group and four in the treated arm (three died of unrelated malignancies, five of cardiovascular diseases, one of injuries sustained in a car accident, and one of pneumonia). The overall survival probability at 5 years from surgery as estimated by the Kaplan-Meier method (Fig 2) was .665 for controls and .660 for the treated group. The difference between the two curves was not statistically significant (P = .861 with the log-rank test; hazards ratio [IFN/control] = 1.040, with 95% confidence interval, 0.671 to 1.613). The corresponding probabilities for event-free survival were .671 and .567, respectively (Fig 3); the difference between the two curves was not statistically significant (P = .107 with the log-rank test; hazards ratio [IFN/control] = 1.412, with 95% confidence interval, 0.927 to 2.149).
The results of the univariate analysis on prognostic factors of event-free survival are listed in Table 4. The effects of sex and age were negligible, whereas those of the tumor characteristics were all statistically significant. It is to be pointed out that all of the 16 pT2 patients who were admitted to the study had metastatic nodes (Robson stage III) compared with 8.2% of nodal metastases in the 147 pT3a and 17.8% in the 84 pT3b patients. Such a distribution explains the worst prognosis of the pT2 patients in the present series (1,267 pT1 or pT2 pN0 patients were excluded from the study because they were Robson stage I).
As can be seen from Table 1, 13 tumors were classified as Gx; the patients were excluded from the multivariate analysis. The latter investigated the independent effects of pathologic T categories and involvement of regional lymph nodes. The results of the final model, in terms of hazards ratio, are listed in Table 5. Tumor grading (P = .002) and pT category (P = .006) were highly significant, as was interaction treatment x pN category (P = .0001 with the likelihood ratio test). Attention should be focused here on the interaction terms more than on the main effects. Progressing from pN0 to pN1/2/3 categories, the hazards ratios of IFN to control decreased progressively: from significantly greater than 1.0 in pN0 to significantly less than 1.0 in pN2/pN3.
The three panels of Fig 4 report the Kaplan-Meier event-free survival curves for the three pathologic N categories. The curves, together with the aforementioned comments to Table 5, indicate a harmful effect of IFN in the 97 treated pN0 patients (A) and a protective effect in the 13 treated pN2/pN3 patients (C). The bootstrap model selection included the first-order treatment x pathologic N category interaction 289 times of 300 (96%). According to the Gail and Simon test,11 the interaction was qualitative (min [Q+,Q-] = 7.86 > 4.23, the pertinent reference value with alpha = 0.05). The treatment effect on pN patient categories was also evidenced by the cumulative probability of death: at 3 years of follow-up, there was no relevant difference between treated and controls in pN0 (0.16 for IFN v 0.10 for controls) and pN1 patients (0.25 for IFN v 0.25 for controls), whereas in pN2/pN3 patients, the greater difference was statistically significant (0.39 for IFN v 0.92 for control) with the Wald’s test.
The overall results of the present randomized clinical phase III trial carried out on 247 assessable patients with radically resected Robson stage II and III RCC showed no advantage of adjuvant rIFN 2b therapy over observation in terms of overall and event-free survival. Regarding mortality and relapse, these results are in agreement with those reported by the Eastern Cooperative Study Group4 in a large randomized study on the same categories of patients using lymphoblastoid IFN. Similarly, the Delta-P Study Group phase III trial5 on the same patient population found no significant difference between observation and adjuvant treatment with IFN2a in terms of time to treatment failure or survival. It could be argued that these negative results with IFN in the adjuvant setting were unforeseen in view of the objective response rate reported in metastatic RCC: nearly 15% in the 1980s,1-3 with a 10% response rate reported in the early 1990s.14
Two recent large randomized trials suggested a small but significant improvement in survival for patients with metastatic RCC treated with IFN
Therefore, it seems that IFN From a biologic point of view, one could argue that metastases to regional lymph nodes in RCC can elicit an RCC antigen-specific T-cell response, as well as an activation of macrophages and natural killer lymphocytes, which might be further stimulated and expanded by the subsequent administration of IFN after radical nephrectomy. In fact, RCC-specific antigens that can trigger a T-cell response and that persist for a long time in vivo have been recently described.21
In conclusion, adjuvant rIFN
APPENDIX
Supported in part by grant no. 95.00531.PF39 from the Italian National Research Council (C.N.R.), Rome, Italy, and grant no. 198.672 from the Italian Association for Cancer Research (A.I.R.C.), Milan, Italy. We thank Nadia Crose for data management and B. Johnston for editing the manuscript.
1. Quesada JR, Swanson DA, Gutterman JU: Phase II study of interferon alpha in metastatic renal-cell carcinoma: A progress report. J Clin Oncol 3: 1086-1092, 1985 2. Horoszewicz JS, Murphy GG: An assessment of the current use of human interferons in therapy of urological cancers. J Urol 142: 1173-1180, 1989[Medline] 3. Williams RD: Immunotherapy of advanced renal cancer. Eur Urol 18: 46-47, 1990 (suppl 2) 4. Trump DL, Elson P, Propert K, et al: Randomized, controlled trial of adjuvant therapy with lymphoblastoid interferon in resected, high risk renal cell carcinoma: An ECOG study. Proc Am Soc Clin Oncol 15: 253, 1996 (abstr 648) 5. Porzsolt F: Adjuvant therapy of renal cell cancer with interferon alpha-2a. Proc Am Soc Clin Oncol 11: 202, 1992 (abstr 622) 6. Robson CJ, Churchill BM, Anderson W: The results of radical nephrectomy for renal cell carcinoma. J Urol 101: 297-301, 1969[Medline] 7. Pizzocaro G, Piva L, Faustini M, et al: Adjuvant interferon alpha in renal carcinoma with a high risk of recurrence: Multicenter pilot study. Arch It Urol Androl 65: 173-176, 1993 8. Pizzocaro G, Piva L, Di Fronzo G, et al: Adjuvant medroxyprogesterone acetate to radical nephrectomy in renal cell carcinoma: 5 year results of a prospective randomized study. J Urol 138: 1379-1381, 1987[Medline] 9. Machin D, Campbell MJ: Statistical Tables for the Design of Clinical Trials. Chichester, United Kingdom, Blackwell Scientific, 1987 10. Marubini E, Valsecchi MG: Analyzing Survival Data From Clinical Trials and Observational Studies. New York, NY, John Wiley & Sons, 1995 11. Gail M, Simon R: Testing for qualitative interactions between treatment effects and patient subsets. Biometrics 41: 361-372, 1985[Medline] 12. Altman D, Andersen PK: Bootstrap investigation of the stability of a Cox regression model. Stat Med 8: 771-783, 1989[Medline] 13. Sauerbrei W, Schumacher M: A bootstrap resampling procedure for model building: Application to the Cox regression model. Stat Med 11: 2093-2109, 1992[Medline]
14.
Minasian LM, Motzer RJ, Gluck L, et al: Interferon alfa-2a in advanced renal cell carcinoma: treatment results and survival in 159 patients with long-term follow-up. J Clin Oncol 11: 1368-1375, 1993 15. Medical Research Council Renal Cancer Collaborators: Interferon-alfa and survival in metastatic renal carcinoma: Early results of a randomized controlled trial. Lancet 353:14-17, 1999
16.
Pyrhönen S, Salminen E, Ruutu M, et al: Prospective randomized trial of interferon alfa-2a plus vinblastine versus vinblastine alone in patients with advanced renal cell cancer. J Clin Oncol 17: 2859-2867, 1999
17.
Negrier S, Escudier B, Lasset C, et al: Recombinant human interleukin-2, recombinant human interferon alfa-2a or both in metastatic renal cell carcinoma. N Engl J Med 338: 1272-1278, 1998 18. Prümmer O: Interferon-alpha antibodies in patients with renal cell carcinoma treated with recombinant interferon-alpha-2a in an adjuvant multicenter trial. Cancer 71: 1828-1834, 1993[Medline] 19. Pizzocaro G, Piva L, Costa A, et al: Adjuvant interferon to radical nephrectomy in Robson’s stages II and III renal cell cancer, a multicentric randomized study with some biological evaluations. Proc Am Soc Clin Oncol 16: 318, 1997 (abstr 1132) 20. Blom JHM, van Poppel H, Marechal JM, et al: Radical nephrectomy with and without lymph node dissection: Preliminary results of EORTC randomized phase III protocol 30881. Eur Urol 36: 570-575, 1999[Medline]
21.
Jantzer P, Schendel DJ: Human renal cell carcinoma antigen-specific CTLs: Antigen-driven selection and long-term persistence in vivo. Cancer Res 58: 3078-3086, 1998 Submitted December 20, 1999; accepted September 1, 2000.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
