Prognostic factors of survival and rapid progression in 782 patients with metastatic renal carcinomas treated by cytokines: a report from the Groupe Français d’Immunothérapie

S. Négrier1,+, B. Escudier2, F. Gomez1, J.-Y. Douillard3, A. Ravaud4, C. Chevreau5, M. Buclon1, D. Pérol1 and C. Lasset1,§

1 Departments of Biostatistics and Medicine, Centre Léon Bérard, Lyon; 2 Department of Medicine, Institut Gustave Roussy, Villejuif; 3 Department of Medicine, Centre René Gauducheau, Nantes; 4 Department of Medicine, Institut Bergonié, Bordeaux; 5 Department of Medicine, Centre Claudius Regaud, Toulouse, France

Received 19 November 2001; revised 2 April 2002; accepted 11 April 2002


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Interleukin-2 (IL-2) or/and interferon (IFN) are routinely used for treating patients with metastatic renal cell cancer. However, results have been disappointing, with a majority of treatment failure. Over 6 years, the Groupe Français d’Immunothérapie enrolled 782 patients in successive multicenter trials using cytokine regimens. Univariate and multivariate analyses were performed on this large prospective database to identify prognostic factors for survival. The presence of biological signs of inflammation, short time interval from renal tumor to metastases (<1 year), elevated neutrophil counts, liver metastases, bone metastases, patient performance status (PS), the number of metastatic sites, alkaline phosphatases and hemoglobin levels were predictive of survival outcome. When compared with previous results, our study showed that PS, number of metastatic sites, disease-free interval, biological signs of inflammation and hemoglobin levels can be considered as validated prognostic factors. We also identified four independent factors predictive of rapid progression under cytokine treatment: presence of hepatic metastases, short interval from renal tumor to metastases (<1 year), more than one metastatic site and elevated neutrophil counts. Patients who combined at least three of these factors have >80% probability of rapid progression despite treatment. We think that these results must be taken into account when making the decision to treat with cytokine.

Key words: interferon, interleukin-2, metastatic renal carcinoma, prognostic factor, rapid progression, survival


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Renal cancer represents 3% of adult malignancies, but its incidence increased by 35% from 1973 to 1991 [1]. About one-third of patients diagnosed with renal cancer will die from the disease after the occurrence of metastases. Metastases from renal cancer have proved resistant to all forms of chemotherapy tested until now [2, 3]. A few durable and complete remissions have been achieved by immunotherapy using mainly {alpha}-interferon (IFN) or interleukin-2 (IL-2) and, more recently, by use of allogeneic lymphoid transplants [4, 5]. However, only a small minority of patients experienced therapeutic benefit, whereas most patients were concerned with the important adverse effects of the treatments. In addition, the impact of treatment on survival is controversial, despite two recent phase III trials that provided evidence of a limited gain in survival associated with IFN compared with medroxyprogesterone or vinblastine [6, 7].

Determining prognostic factors of survival or of rapid progression under treatment would be of help for selecting patients for cytokine treatments that are routinely used. Some previously published analyses identified a number of survival prognostic factors [813], but, to our knowledge, no author outside our group has tried to identify predictive factors of rapid progression.

The Groupe Français d’Immunothérapie conducted several large multicenter trials between 1992 and 1998: two randomized trials and three open phase II trials, to evaluate the interest of IL-2 and/or IFN treatments in patients with metastatic renal cell carcinoma (MRCC) [1417]. Within 6 years, 782 patients were enrolled by the same group of investigators. This prospective database was used to identify pretreatment features that were predictive for survival or for rapid progression under cytokine treatment. The results were analyzed by multivariate analysis, and patients were stratified according to the risk of rapid progression.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patients
All patients were enrolled in five different multicenter prospective trials, which were approved by the relevant ethics committees according to French law. Three trials have been reported previously [1416], the other two are submitted. Every patient gave written informed consent prior to trial entry. Inclusion criteria were histologically confirmed and measurable progressive MRCC, age between 18 and 80 years (or limited to 65 or 70 years in some trials), and adequate cardiac, renal, hepatic and hematological functions. Patients with untreated and/or progressive brain metastases were excluded, as were patients requiring corticosteroids. Patients must have adequate Eastern Cooperative Oncology Group (ECOG) performance status (PS) (0 or 1, except in one trial that admitted patients with an ECOG score of 2). Pretreatment tumor status was evaluated by computed tomography scans (CT scan) of the chest, abdomen and brain, and radioisotope bone scans. Tumor status was re-evaluated 10–12 weeks after start of treatment, using at least chest and abdominal computed tomography (CT) scans, as well as all other valuable controls if required. Tumor response was classified according to the WHO criteria. All cases of response were reviewed by an independent committee. The databases were constituted with similar procedures and identical quality controls. Additional tests were performed on the pooled database to verify the consistency of the data.

Treatment
Treatment consisted of IL-2 or IFN, or both. In the large randomized CRECY trial, IL-2 was given as a continuous infusion, as described previously [14]. In one further trial, IL-2 combined with IFN was administered with identical doses and schedules but by the subcutaneous route [17]. In the remaining three trials, a subcutaneous combination of IL-2 and IFN was given using an intermittent schedule. It was associated with 5-fluorouracil (5-FU) in some patients for one phase II trial and for one group of a randomized trial [15, 16].

Prognostic factor analysis
The data from the five clinical trials were pooled and then analyzed with SPSS 10.07 software (SPSS Inc., Chicago, IL, USA). The end-points of the study were overall survival and rapid progression under cytokine treatment. Survival was measured from treatment initiation to the date of death or last follow-up. Rapid progression was considered in the case of progression of the disease at tumor evaluation (between weeks 10 and 12 after start of treatment) or at any time between treatment start and the planned date of tumor re-assessment. All potential prognostic features were transformed to be on a categorical form. Clinical features examined included the number (1 versus >1) and sites (lung, mediastinum, bone, liver) of metastases, ECOG PS (0 versus >=1), age (<=60 versus >60 years), initial Robson score (1–2 versus 3–4), prior nephrectomy (yes versus no), time from renal tumor to occurrence of metastases (<=12 versus >12 months), recent weight loss (<10% versus >=10% of body weight) and some baseline biochemical features previously identified as potential predictive factors of survival [12, 13]: hemoglobin [<115 g/l female (F); <130 g/l male (M) versus >=115 g/l (F); >= 130 g/l (M)], alkaline phosphatase (<=100 versus >100 UI/l) , number of circulating neutrophils (<=7.5 versus >7.5 103/l), lymphocytes (<=1.5 versus >1.5 103/l) and platelets (<=400 versus >400 103/l), and signs of inflammation defined by a sedimentation rate (SR) of at least 100 or a C-reactive protein (CRP) level of >=50. This last parameter was generated because data for SR and CRP were missing, and because a high degree of correlation had previously been identified between SR and CRP level [18]. The two chosen cut-off points, 100 and 50, have previously been identified as significant threshold values [9, 11, 19]. The lactate dehydrogenase level, which was recently shown to be relevant in this setting [12, 13], was not included in the analysis because it had not been registered in the 425 patients enrolled in the CRECY trial.

Survival distributions were estimated using the Kaplan–Meier method [20]. The relationship between survival and each parameter was analyzed with the log-rank test [21]. Correlations between rapid progression under treatment and potential prognostic features were assessed using {chi}2 tests.

Multivariate model
Fifteen parameters were included in a stepwise Cox’s model on the basis that they were significantly connected with survival (P <0.10) [21]. Thirteen variables were entered in a stepwise logistic regression model on the basis that they were significantly connected with rapid progression (P <0.10) [21].

Variables were entered in these models if their P value was <0.05 and rejected if their P value was >0.05. There was a limited number of missing values (i.e. maximum 13% for inflammation markers). Case deletion was used to control missing values in all analyses.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient characteristics and treatment results
As shown in Table 1, 563 patients were male, median age was 58 years (range 21–79 years), 76.1% had two or more metastatic sites, 69.1% had an interval from diagnosis of renal tumor to metastases of 1 year or less, and a large majority (93.1%) had undergone prior nephrectomy. With regards to cytokine treatments, 138 patients received i.v. IL-2 alone, 147 s.c. IFN alone, 140 a combination of both treatments; 357 received combinations of IFN and s.c. IL-2, and this treatment was associated with 5-FU in 172 patients. Tumor response was assessed using the WHO criteria. The overall response rate was 10%, with 11 complete and 67 partial responses; 24.2% patients had stable disease at time of tumor evaluation, 54% had a progression of their disease at time of evaluation or before. In 48 (6%) patients, tumor response was not evaluable.


View this table:
[in this window]
[in a new window]
 
Table 1. Patient characteristics
 
The median follow-up is 77 months. The median overall survival is 12.8 months [95% confidence interval (CI) 11.5–14 months] (Figure 1).



View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. Overall survival of the 782 patients with metastatic renal cell carcinoma.

 
Univariate analysis
Factors significantly (P <0.05) associated with a limited survival were initial Robson score (3–4), prior nephrectomy (no), time interval between diagnosis of renal tumor and occurrence of metastases (<=12 months), presence of metastases in liver, bone or mediastinum, weight loss (>=10%), number of metastatic sites (>1), ECOG PS (>=1), number of circulating lymphocytes (<=1.5 x 103/l), number of neutrophils (>7.5 x 103/l), hemoglobin level [<115 g/l (F); <130 g/l (M)], number of platelets (>400 x 103/l), alkaline phosphatase level (>100 UI/l), and signs of inflammation (SR >=100 or CRP >=50). Table 2 indicates parameter estimates and corresponding risk ratio and P values.


View this table:
[in this window]
[in a new window]
 
Table 2. Univariate analysis of the different parameters correlated with limited overall survival
 
Factors significantly (P <0.05) correlated with rapid progression under treatment were: prior nephrectomy (no), time from renal tumor to metastases (<=12 months), presence of liver metastases, number of metastatic sites (>1), ECOG PS (>= 1), number of circulating neutrophils (>7.5 x 103/l), number of platelets (>400 x 103/l), hemoglobin level [<115 g/l (F); <130 g/l (M)], level of alkaline phosphatase (>100 UI/l), signs of inflammation (SR >=100 or CRP >=50), presence of bone metastases, weight loss (>=10%) and mediastinum involvement. Table 3 indicates parameter estimates and corresponding risk ratio and P values.


View this table:
[in this window]
[in a new window]
 
Table 3. Univariate analysis of the different parameters correlated with rapid progression under treatment
 
Multivariate analysis
Parameters identified as being correlated with survival with a degree of significance <0.10 were included in the multivariate analysis. Nine parameters with P values <0.01 were considered as independent prognostic factors: signs of inflammation, time interval from renal tumor to metastases, number of circulating neutrophils, presence of liver metastases, presence of bone metastases, ECOG PS, number of metastatic sites, elevated level of alkaline phosphatase, number of metastatic sites and hemoglobin level. Table 4 displays parameters, together with their risk ratio and P value, the number of patients and the median survival estimates for each corresponding subgroup. Mediastinum involvement also appears to be predictive, although to a lesser degree (P value 0.01–0.05).


View this table:
[in this window]
[in a new window]
 
Table 4. Multivariate analysis of the different factors correlated with overall survival
 
The results of the literature survey permitted to identify five different prognostic factors in at least three independent series, including ours: PS, number of metastatic sites, time interval from renal tumor to metastases (or to treatment of metastases), signs of inflammation and hemoglobin level. Based on a combination of these factors, we generated two possible risk models, shown in Figures 2 and 3. Finally, we analyzed the variables identified as prognostic factors for survival in the subgroup of responding patients (n = 78). Two parameters only were found to have significant correlation with survival: PS (0 versus >=1) and number of metastatic sites (1 versus >=2).



View larger version (22K):
[in this window]
[in a new window]
 
Figure 2. Possible risk model according to the combination of the five different validated prognostic factors.

 


View larger version (16K):
[in this window]
[in a new window]
 
Figure 3. Possible risk model according to the presence of one, two or three, and four or five prognostic factors.

 
The parameters that were correlated with rapid progression of the disease with a degree of significance <0.05 were included in the multivariate analysis. Four parameters were considered as independent prognostic factors with a degree of significance <0.01, whereas three additional factors were also predictive, although to a lesser degree (P values 0.01–0.05). The highest relative risks of rapid progression are either linked to the presence of liver metastases, to elevated numbers of neutrophils, to the time interval from renal tumor to metastases, or to the number of metastatic sites. These results, as well as the relative risk and the corresponding P values, are indicated in Table 5.


View this table:
[in this window]
[in a new window]
 
Table 5. Multivariate analysis of the different factors correlated with rapid progression under treatment
 
The probability rates of rapid progression under treatment according to a combination of the four different factors with a relative risk >1.8 are shown on Figure 4. It should be noted that patients who have a combination of at least three of the four factors have a probability of progression under treatment of >80% (Figure 4). These patients represent 22.3% of the population studied. They have a short survival (i.e. median survival 5.5 months, 95% CI 4–7 months). Of note, this poor outcome was observed for all types of cytokine treatments, even in the subgroup of patients treated with IFN alone (147 patients enrolled in the CRECY trial).



View larger version (17K):
[in this window]
[in a new window]
 
Figure 4. Rapid progression under cytokine treatment according to four different factors. Probability of rapid progression varies with the number of predictive factors. Patients with a combination of at least three of the four factors have >80% risk of progressing under treatment.

 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The prognostic factors identified in this study were discriminated from the largest prospective database ever reported, collected over a rather short period of time (within 6 years). Nine independent prognostic factors considered as capable of predicting survival with a high degree of significance (P <=0.01) were identified: biological signs of inflammation, time interval from renal tumor to metastases, liver metastases, bone metastases, ECOG PS, elevated neutrophil count, elevated level of alkaline phosphatase, number of metastatic sites and hemoglobin level. All but one (level of alkaline phosphatase) had previously been identified on at least one other cohort of patients [813]. The results of previous prognostic factor analyses reported in the last 15 years are summarized in Table 6. It must be noted that only the most recent reports include some biological factors, usually measured as routine tests in the patients. As emphasized by Motzer et al. [13], the populations concerned by these studies varied over time; indeed, patients who were treated before the 1990s have a poorer outcome than patients who were treated more recently. This is clearly due to recent improvements in the medical management of the patients, the earlier detection of metastases and, probably, the use of more restrictive inclusion criteria in most recent trials. For example, patients selected for the present study have a good PS (90% have PS of 0 or 1). Since they have been identified in at least three independent series, PS, number of metastatic sites, time interval from renal tumor to metastases (or to treatment of metastases), signs of inflammation (mostly identified by the SR) and hemoglobin level appear as the most consistent prognostic factors of survival. We note, however, that each factor, either reported in the literature or in our study, is usually associated with a rather limited increase in the risk of death, since most relative risks are rated between 1 and 2. Because each study has analyzed different parameters and none of these factors has been associated with a very high relative risk of death, risk stratification groups often differ from one model to another. For this reason we will not propose yet another model based on the factors we have identified. More consistently, we prefer to develop risk models based on the five validated factors (see Figures 2 and 3). These models can be used to stratify subgroups of patients included in prospective randomized trials; however, this selection would obviously be unsuitable for routine treatment decisions in individual patients.


View this table:
[in this window]
[in a new window]
 
Table 6. Independent prognostic factors identified in previous studies including different populations of patients with metastatic renal cell carcinoma
 
Using a more pragmatic approach, we chose to identify factors of rapid progression under cytokine treatment or, in other words, of early cytokine failure. Early treatment failure corresponds to the progression of the disease despite cytokine treatment within 3 months of treatment initiation. The progression rate in our large series of patients was 54%, which is consistent with other large reports of cytokine-treated patients [11, 22, 23]. Four independent prognostic factors capable of predicting rapid progression with a high degree of significance (P <=0.005) were identified: presence of hepatic metastases, elevated neutrophil count, <1 year from renal tumor to metastases, and two or more metastatic organ sites. Three of these factors had previously been identified in a study including part of the patient population [14]. They were confirmed by our analysis comparing a number of additional factors on a much larger number of patients. Combining these different factors, as indicated in Figure 4, showed that patients who present with at least three adverse prognosis factors have >80% probability of progressing rapidly under cytokine treatment. This subgroup of patients with a limited survival (median time 5.5 months) represents 22.3% of the patients enrolled in our trials.

Since cytokine treatment induces adverse effects in almost every patient, it is necessary to avoid treating patients who will not, in the end, benefit from the treatment. As a consequence, the subgroup of patients presenting at least three factors predictive of rapid progression should no longer be considered candidates for cytokine trials in France. In addition, we believe it important that physicians in charge of these patients should be informed of the high risk of rapid progression despite cytokine treatment. We think it reasonable, however, to make new therapeutic agents available to these patients in the frame of clinical trials. Indeed, in strict observance of the informed consent process, these patients may prefer this approach to best supportive care. In addition, the efficacy of any new agent that might become available will be more easily demonstrated in this group of patients that in the general population including patients with spontaneous, very slow-growing tumors.

In conclusion, a number of independent prognostic factors for survival have been identified in this large cohort of cytokine-treated patients with good PS. Performance status, number of metastatic sites, disease-free interval, signs of inflammation and hemoglobin levels are the most consistent with previous results. They can be considered as validated prognostic factors for survival. We have also identified factors predictive of rapid disease progression, and calculated that patients combining at least three of these factors have >80% probability of disease progression despite treatment. We think that cytokine treatment cannot be recommended in this subgroup of patients.


    Acknowledgements
 
We thank M. D. Reynaud for skillful editorial assistance.

Groupe Français d’Immunothérapie investigators: Armelle Caty, Centre Oscar Lambret, Lille; Bruno Coudert, Pierre Fargeot, Centre Georges Francois Leclerc, Dijon; Stéphane Culine, Michel Fabbro, Centre Val D’aurelle, Montpellier; Rémy Delva, Centre Paul Papin, Angers; Lionnel Geoffrois, Centre Alexis Vautrin, Vandoeuvre-Les-Nancy; Thierry Lesimple, Centre Eugene Marquis, Rennes; Jean Peny, Centre Francois Baclesse, Caen; Patrice Viens, Institut Paoli-Calmettes, Marseille; Bruno Audhuy, F. Khoser, Hopital Louis Pasteur, Colmar; Jaafar Benouna, Centre Rene Gauducheau, St-Herblain; Michel Bittard, CHRU Hopital St-Jacques, Besancon; Olivier Boulat, Hopital Henri Duffaut, Avignon; Thierry Dorval, Institut Curie, Paris; Jean-Christophe Eymard, Institut Jean Godinot, Reims; Jean-Marc Ferrero, Antoine Thyss, Centre Antoine Lacassagne, Nice; Joël Fleury, Centre Jean Perrin, Clermont-Ferrand; Alain Goupil, Centre Rene Huguenin, St-Cloud; Eric Legouffe, Jean-François Rossi, CHRU Lapeyronie, Montpellier; Claude Linassier, CHU Tours, Tours; Yacine Merrouche, CHRU Minjoz, Besancon; Laurent Mignot, Hopital Foch, Suresnes; Mireille Mousseau, CHRU Michallon, Grenoble; Xavier Murraciole, CHRU La Timone, Marseille; Hubert Orfeuvre, Hopital Fleyriat, Bourg En Bresse.


    Footnotes
 
+ Correspondence to: Dr S. Négrier, Department of Oncology, Centre Léon Bérard, 28 rue Laënnec, 69373 Lyon cedex 08, France. Tel: +33-478-78-27-51; Fax: +33-478-78-27-16; E-mail: negrier{at}lyon.fnclcc.fr Back

§ Groupe Français d’Immunothérapie investigators are listed in the Acknowledgements. Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Motzer RJ, Bander NH, Nanus DM. Renal-cell carcinoma. N Engl J Med 1996; 335: 865–875.[Free Full Text]

2. Yagoda A, Abi-Rached B, Petrylak D. Chemotherapy for advanced renal-cell carcinoma: 1983–1993. Semin Oncol 1995; 22: 42–60.

3. Rini BI, Vogelzang NJ, Dumas MC et al. Phase II trial of weekly intravenous gemcitabine with continuous infusion fluorouracil in patients with metastatic renal cell cancer. J Clin Oncol 2000; 18: 2419–2426.[Abstract/Free Full Text]

4. Bukowski RM. Natural history and therapy of metastatic renal cell carcinoma: the role of interleukin-2. Cancer 1997; 80: 1198–1220.[ISI][Medline]

5. Childs R, Chernoff A, Contentin N et al. Regression of metastatic renal-cell carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation. N Engl J Med 2000; 343: 750–758.[Abstract/Free Full Text]

6. Medical Research Council Renal Cancer Collaborators. Interferon-{alpha} and survival in metastatic renal carcinoma: early results of a randomised controlled trial. Lancet 1999; 353: 14–17.[ISI][Medline]

7. Pyrhonen S, Salminen E, Ruutu M et al. Prospective randomized trial of interferon {alpha}-2a plus vinblastine versus vinblastine alone in patients with advanced renal cell cancer. J Clin Oncol 1999; 17: 2859–2867.[Abstract/Free Full Text]

8. Elson PJ, Witte RS, Trump DL. Prognostic factors for survival in patients with recurrent or metastatic renal cell carcinoma. Cancer Res 1988; 48: 7310–7313.[Abstract]

9. De Forges A, Rey A, Klink M et al. Prognostic factors of adult metastatic renal carcinoma: a multivariate analysis. Semin Surg Oncol 1988; 4: 149–154.[ISI]

10. Palmer PA, Vinke J, Philip T et al. Prognostic factors for survival in patients with advanced renal cell carcinoma treated with recombinant interleukin-2. Ann Oncol 1992; 3: 475–480.[Abstract]

11. Fossa SD, Kramar A, Droz JP. Prognostic factors and survival in patients with metastatic renal cell carcinoma treated with chemotherapy or interferon-{alpha}. Eur J Cancer 1994; 30: 1310–1314.

12. Lopez Hänninen E, Kirchner H, Atzpodien J. Interleukin-2 based home therapy of metastatic renal cell carcinoma: risks and benefits in 215 consecutive single institution patients. J Urol 1996; 155: 19–25.[ISI][Medline]

13. Motzer RJ, Mazumdar M, Bacik J et al. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol 1999; 17: 2530–2540.[Abstract/Free Full Text]

14. Négrier S, Escudier B, Lasset C et al. Recombinant human interleukin-2, recombinant human interferon {alpha}-2a, or both in metastatic renal-cell carcinoma. N Engl J Med 1998; 338: 1272–1278.[Abstract/Free Full Text]

15. Ravaud A, Audhuy B, Gomez B et al. Subcutaneous interleukin-2, interferon {alpha}-2a, and continuous infusion of fluorouracil in metastatic renal cell carcinoma: a multicenter phase II trial. Groupe Francais d’Immunotherapie. J Clin Oncol 1998; 16: 2728–2732.[Abstract]

16. Négrier S, Caty A, Lesimple T et al. Treatment of patients with metastatic renal cell carcinoma with a combination of subcutaneous interleukin-2 and interferon {alpha} with or without fluorouracil. J Clin Oncol 2000; 18: 4009–4015.[Abstract/Free Full Text]

17. Négrier S, Ravaud A, Delva R et al. Combination of cytokines in metastatic renal cell carcinoma (MRCC). Is the subcutaneous (sc) route less active than the intravenous (iv) route? Proc Am Soc Clin Oncol 1999; 18: 331a (Abstr 1273).

18. Ljungberg B, Grankvist K, Rasmuson T. Serum interleukin-6 in relation to acute-phase reactants and survival in patients with renal cell carcinoma. Eur J Cancer 1997; 33: 1794–1798.[ISI][Medline]

19. Blay JY, Négrier S, Combaret V et al. Serum level of interleukin 6 as a prognosis factor in metastatic renal cell carcinoma. Cancer Res 1992; 52: 3317–3322.[Abstract]

20. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.[ISI]

21. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50: 163–170.[Medline]

22. Motzer RJ, Mazumdar M, Bacik J et al. Effect of cytokine therapy on survival for patients with advanced renal cell carcinoma. J Clin Oncol 2000; 18: 1928–1935.[Abstract/Free Full Text]

23. Gold PJ, Thompson JA, Markowitz DR et al. Metastatic renal cell carcinoma: long-term survival after therapy with high-dose continuous-infusion interleukin-2. Cancer J Sci Am 1997; 3: S85–S91.[ISI][Medline]