Original article. 5-Fluorouracil versus 5-fluorouracil plus a-interferon as treatment of metastatic colorectal carcinoma. A randomized study*

Transcription

1 Annals of Oncology 7: , O 1996 Kluwer Academic Publishers. Printed in the Netherlands. Original article 5-Fluorouracil versus 5-fluorouracil plus a-interferon as treatment of metastatic colorectal carcinoma. A randomized study* P. Dufour, 1 F. Husseini, B. Dreyfus, 3 H. Cure, 4 C. Martin, 5 G. Prevost, 6 J.-P. Olivier, 7 F. Dumas, 8 B. Duclos, 1 R. Olivares, 9 A. Leszler, 9 J.-P. Bergerat, 1 B. Audhuy, L. Thill 9 & F. Oberling 1 1 Departement d'oncologie, Hopitaux Univcrsitaires de Strasbourg, Strasbourg; Service Onco Hematologie, Hopital Pasteur, Colmar; i Service Oncologie, Centre Hospitalo Universitaire Poitiers; ^Centre J. Perrin, Clermont Ferrand;! Service Oncologie, Centre Hospitalier d'annecy; 6 Service Oncologie, Centre Hospitalier de Mulhouse; ''Service Oncologie, Centre Hospitalo Universitaire de Limoges; *Hopital Haut-Leveque, Pessac; 9 Produits Roche, France Summary Background: In 1989, S. Wadler reported very promising results (76% response rate) with a combination of 5-fluorouracil (5-FU) plus a-a interferon (IFN) in the treatment of metastatic colorectal carcinoma (MCRC). In vitro, there are several potential explanations for synergism between the two agents. We therefore decided in 1989 to start a randomized study comparing 5-FU alone with 5-FU plus IFN. Patients and methods: 15 non-pretreated patients with measurable metastatic colorectal carcinoma entered into this study. The patients were randomly allocated either in arm A (n - 49) with 5-FU: 75 mg/m i.v. Cl dl-d5 fouowed by 75 mg/m i.v. bolus once a week, or in arm B (n 56) with 5-FU as in arm A plus IFN 9 x 1 6 IU sub-cutaneously three times a week. Results: After two months of treatment we observed 1 CR and PR in arm A (response rate 6.1%), 3 CR and 8 PR in arm B (response rate 19.6%), i.e., a significant difference (P =.5). Event-free survival was significantly higher in arm B (6 months) than in arm A ( months) (P <.1), while median survival was slightly higher in arm B (1 months) than in arm A (1 months) (P <.5). For overall survival the difference was not significant after adjustment on center treatment and baseline Karnofsky status (/>-.13). Toxicity was also greater in arm B. Sixteen percent of patients in arm A and 36% in arm B experienced certain grade 3-4 side effects (P<.5). Conclusion: 5-FU plus IFN is more effective than 5-FU alone in terms of response rate, event free survival but not of overall survival. 5-FU plus IFN is more toxic. As IFN has no demonstrated efficacy in MCRC as a single agent, this study suggests that IFN is acting as a 5-FU modulatory agent. The response rate observed (19.6%) is similar to the results obtained elsewhere with 5-FU plus leucovorin. Key words: metastatic, colorectal carcinoma, 5-FU, a-interferon Introduction In 1989, S. Wadler published promising results with a combination of 5-FU and interferon alpha-a (IFN) in the treatment of metastatic colorectal carcinoma (MCRC) [1]. The reported response rate was 76% in a subgroup of non-pretreated patients. 5-FU is at present the sole registered cytotoxic agent efficacious in colorectal carcinoma. The response rate with 5-FU alone is approximately 1% in large multicentric phase EQ studies []. Several approaches have been adopted in an attempt to improve these results. The most promising is the adjunction of a modulatory agent, leucovorin, which interferes with the mechanisms of 5-FU cytotoxicity. Such a combination increases approximately two fold the response rate with the same dose of 5-FU alone []. In vitro, EFN enhances the cytotoxicity of 5-FU, for which several potential explanations have * Presented in part at the 19th ESMO Congress in Lisbon. been reported, including inhibition of thymydilate synthetase and thymidilate kinase activities and the concurrent cytotoxic effect of interferon and 5-FU [3,4]. In order to assess the potential benefit of a-ifn combined with 5-FU in the treatment of metastatic colorectal carcinoma, we started a randomized study in 1989 comparing 5-FU alone and 5-FU plus a-interferon according to the scheme proposed by Wadler et al. The end points of this study were response rate, event free survival and overall survival. Patients and methods Inclusion criteria were age between 18 and 7, Karnofsky index >7%, metastatic colorectal carcinoma with bidimensional measurable lesions, no previous chemotherapy and/or immunotherapy (adjuvant or curative), life expectancy > 6 months. One hundred five patients from 9 centers entered into this study. Their characteristics are summarized in Table 1. Both groups were well balanced for major prognostic factors except for Karnofsky Downloaded from on 17 July 18

2 576 Table I. Patients' characteristics. No. Sex(M/F) Age (median) Range Karnofsky (>9%) Colon carcinoma Rectal carcinoma Histologic grade Well differentiated Moderately differentiated Undifferentiated Not done Synchronous metastases Metachronous metastases Time from diagnosis to metastases (months) Standard deviation Hemoglobin (mean value) Liver metastases Lung metastases Lymph node metastases Locoregional recurrence (5-FU) 49 31/18 61 (3-7) 3 (61%) 36 (73.5%) 13 (6.5%) 6 (53%) 17(35%) 3 (6%) 3 (6%) (41%) 9 (59%) (8%) 14 (8.5%) 6 (1%) 1 (%) ArraB (5-FU + IFN) 56 3/4 59 (36-7) 45 (8%) 41 (73%) 15 (7%) 3 (54%) 18 (3%) 3 (5%) 5 (9%) 31 (55%) 5 (45%) (75%) 17 (3%) 1 (18%) 1 (%) index (>9%). Significantly more patients in arm B had a Kamofsky index >9% (Table 1). Since the Karnofsky index is a variable of high prognostic value, response rate, event free survival and overall survival are given following adjustment on it. All eligible patients were randomized either in arm A with 5-FU: 75 mg/m /d continuous infusion dl-d5, then one week's rest, followed by 5-FU 75 mg/m i.v. bolus weekly or in arm B with 5-FU as in arm A, plus a-a interferon 9 x 1 6 IU administered subcutaneously, in the late afternoon, three times a week. In the event of grade hematological toxicity, the 5-FU dose was reduced to 5 mg/m /week; in the case of grade 3 hematological toxicity, both 5-FU and IFN were stopped until recovery. In the case of grade diarrhea or stomatitis, 5-FU was stopped until complete resolution, then 5-FU was resumed at a dose of 5 mg/m /week. Initial work-up consisted of clinical examination, liver and renal biological function tests, gamma GT, CEA, CA 19.9, thoracic X-ray and abdominal CT scan. Target lesions, CEA and CA 19.9 were re-evaluated at, 4, 6, 9 and 1 months of treatment Tumor response was strictly assessed according to the WHO criteria. Patients with a complete or partial remission, or stabilisation, received the same treatment until progression. Patients in progression after two months or any time later went off the study. This study was approved by the local ethical committee and written informed consent was obtained from patients before randomization. Statistical methods Characteristics of patients were compared by the chi-square test for qualitative variables and by Student's f-test for quantitative variables. This trial was designed to demonstrate an increase of 35% in response rate for patients receiving 5-FU + interferon. The estimated response rate of 5-FU alone was 15%. In a two-sided test with a power of 95% and a type I error of 5% we needed 5 patients per arm to test this hypothesis. Event free survival O^FS) and overall survival (OS) were calculated according to the Kaplan-Meier method and the curves compared with the log-rank test. EFS was the time between the date of randomization and the start of progressive disease. Survival was calculated from the date of randomization. End-point of this study was 1 December In the multivariate analysis, a Cox proportional hazard model was used to evaluate the treatment effect adjusted on center treatment and baseline Karnofsky status. The center effect, center treatment and Karnofsky status treatment interactions were tested in the Cox model, they were not significant. Results All results are given based on intent to treat, regardless of the number of cycles received. Five patients in arm A and three patients in arm B received two or less cycles as they decided to stop the treatment. They were considered non evaluable but are all included in the analysis of event free survival, survival and response rate. Response to therapy (Table ) The response rate, in 15 eligible patients, was significantly higher in arm B (5-FU plus a-interferon) than in arm A (5-FU alone) with 19.6% versus 6.1% respectively (P =.4). There was no difference in response rate according to the metastatic sites, but most patients with lung metastases had concomitant liver metastases. The difference was just significant after adjustment on center treatment and baseline Karnofsky status (P -.5) in a regression logistic model. Event free survival (EFS) (Figure 1) was significantly higher in arm B (median: 6 months) than in arm A (median: months) (P <.1). This difference was also significant after adjustment on center treatment and baseline Karnofsky status (P <.1). Overall survival (OS) (Figure ) was slightly higher in arm B (median duration: 1 months) than in arm A (median duration: 9 months) (P<.5). This difference was not significant after adjustment on center treatment and baseline Karnofsky status (P -.13). Tolerance (Table 3) In arm B, we observed a significantly higher incidence of grade 3-4 leucopenia and asthenia. There was a trend to more frequent anemia and thrombocytopenia in arm B. We observed a significantly higher incidence Table. Response rate. Response (intent to treat) CR PR ST PD NE (n-49) 1() (4.1) 6 (1.) 35 (71.4) 5 (1.) Arm B («- 56) 3 (5.4) 8 (14.3) 16 (8.5) 6 (46.4) 3 (5.4) Overall response rate: 6.1 versus 19.6% (P -.5). Downloaded from on 17 July 18

3 577 Table 3. Tolerance to treatment (according to the WHO classification). 8 (% of patients) 5-FU Arm B (% of patients) 5-FU + IFN \ \ Fever Myalgia Headache Mucositis Nausea/vomiting Diarrhea Cardiac Leucopenia Anemia Thrombocytopenia.5 -i,. I.1 j III Figure 1. Event free survival. 5-FU alone (dotted line), Ann B - 5-FU + IFN (solid line). The difference is significant () 1. c.9 ; u t.7 Compliance to treatment In arm A 73.5% of patients received the projected dose of 5-FU versus 53.6% in arm B (P =.3). Due to the higher rate of non progression, patients received significantly more cycles in arm B than in arm A (16 versus 8 cycles). Treatment was stopped in four patients (8.1%) in arm A and in 14 patients (5%) in arm B (P <.5) because of side effects. There was no dose reduction of interferon. When there was major intolerance the treatment was stopped and the patient went off the study. "..6 ". 5 Discussion In the present study, the response rate, event free survival but not overall survival were significantly im- 3proved by the combination of 5-FU ct-interferon compared with 5-FU alone. As the projected 5-FU dose ; was the same in both arms (the received 5-FU dose was actually lower in the interferon arm) and remembering o i that a-interferon alone has a minimal effect in colorectal carcinoma [5], this study suggests that interferon is oo{_ acting as a 5-FU modulatory agent. Hlltt! Several previous phase II studies [6-13] reported the results obtained with a combination of 5-FU and afigure. Overall survival. - 5-FU alone (dotted line), Arm B - 5-FU + IFN (solid line). Arm B was slightly better than interferon in MCRC using the same schedule. In these (P <.5), but the difference was not significant after adjustment on monocentric studies, the number of patients was small center effect and baseline Karnofsky status (P -.13). (average 3) and the range of response rate was wide (4% to 63%), with a mean value of 38%. The response of flu-like symptoms (fever, myalgia) due to interferon rate we observed in the present study is lower, but this in arm B. study is multicentric and the results concerned all eliin all, eight patients (16%) in arm A and patients gible patients in an unselected patient population and (36%) in arm B displayed a grade 3-4 side effect (P < not only evaluable patients..5). Four other broad ranging phase HI studies have We did not observe any sign of cumulative toxicity in been published: York compared in 45 patients 5-FU either arm. alone with 5-FU plus a-interferon according to the There was no toxic death in either arm. same scheme as the present study. The response rate ".4 Downloaded from on 17 July 18

4 578 was 19% with 5-FU alone and 31% with 5-FU plus a- interferon.this difference was not significant [14]. The study of the Royal Marsden group is similar to the York study but interferon a-b was used instead of interferon a-a. 16 patients were randomized and the authors observed a 19% response rate with 5-FU plus IFN and 3% with 5-FU alone. This difference was not significant and the 5-FU plus IFN arm was more toxic [15]. By comparison with these two studies, the observed response rate of 5-FU alone in our study is low (6%). The range of 5-FU response rate reported in the literature is wide (3% to 45%) and highly dependent on dose and schedule, but our response rate is close to that reported in several multicentric phase HI studies using 5-FU alone in one arm. In a series of 179 patients, Lokich reported a 7% response rate with 5-FU alone (5 mg/m i.v. dl-5 every 5 weeks) [16]. Ehrlichman also observed a 7% response rate with 5-FU alone [17]. In 3 patients, Martoni with a weekly 5-FU regimen (6 mg/m i.v.) observed a 3% response rate [18]. Piedbois in a meta-analysis found a mean response rate of 11% for 5-FU alone [19]. Hill also reported a phase HI study of 5-FU versus 5-FU plus IFN, but the design of this study was quite different. 5-FU was administered as protracted venous infusion (3 mg/m /d for 1 weeks) and the dose of IFN was lower (5 x 1 6 IU) than in other studies. In 155 evaluable patients, he observed a 33% response rate with 5-FU alone and 8% with 5-FU plus IFN. There was no significant difference for response rate, failure free survival and overall survival between the two arms []. The Corfu-A study group compared 5-FU plus a- interferon with 5-FU plus leucovorin in 496 patients. The observed response rates were 1 and 19%, respectively, not a significant difference. In the 5-FU plus a- interferon arm significantly more hematological toxicity was observed [1]. Ragnhammar explored different dose regimens of 5-FU and a-interferon in patients with metastatic colorectal carcinoma. The doses of IFN were in a range of 3 to 9 x 1 6 IU three times a week and 5-FU given either as in the present study or as continuous infusion over five days every three weeks. In the subgroup of non pretreated patients, he observed a high response rate regardless of doses and schedules (38% to 64%), but the differences were not significant. The number of patients in this study was however small (less than in each group) []. 5-FU + leucovorin (LV) is the most commonly used combination in MCRC. Piedbois in a meta-analysis reports a mean response rate of 3% with such a combination [19]. If we pool the results of 5-FU plus IFN observed in the five reported phase HI studies comparing 5-FU +/- LV with 5-FU plus IFN (York, Royal Marsden study, Hill study, Corfu-A study group and the present study), the response rate of 5-FU plus IFN in 549 patients is 3.5% (19/549). This response rate is very close to that observed with the 5-FU plus leucovorin combination, but 5-FU plus IFN is clearly more toxic and more expensive than the 5-FU + LV combination, which makes 5-FU + LV still the preferred treatment for metastatic colorectal carcinoma. In conclusion, the present study in MCRC provides evidence of an increased response rate and event free survival, but not overall survival, with a-interferon plus 5-FU compared to 5-FU alone (according to this scheme), but at the expense of increased toxicity. A similar response rate could possibly be obtained with 5-FU alone at equitoxic doses. As the modulatory effects of IFN on chemotherapy in experimental models are highly dependent on the timing and the dose of IFN, alternative schedules of 5-FU and IFN combinations are worth exploring in clinical trials in order to increase the response rate and/or to reduce toxicity. This combination is also worth evaluating at an early stage of the disease with a low burden of tumor, i.e., in an adjuvant setting; at this point, the immunomodulatory properties of IFN could be helpful. Acknowledgements Supported by Produits Roche, Neuilly, France. We are indebted to Marjorie Bettison for help in the preparation of the manuscript. References 1. Wadler S, Schwartz EL, Goldman M et al. Fluorouracil and recombinant alpha-a interferon: An active regimen against advanced colorectal carcinoma. J Clin Oncol 1989; 7: Bruckner HW, Motwani BT. Chemotherapy of advanced cancer of the colon and rectum. Semin Oncol 1991; 18: Houghton JA, Cheshire PJ, Morton CL et al. Potentiation of 5-fluorouracil-leucovorin activity by a-interferon in colon adenocarcinoma xenografts. Clin Cancer Res 1995; 1: Wadler S, Schwartz EL. Antineoplastic activity of the combination of interferon and cytotoxic agents against experimental and human malignancies: A review. Cancer Res 199; 5: Kirkwood JM, Ernstoff MS. Interferons in the treatment of human cancer. J Clin Oncol 1984; : John WJ, Neefe JR, MacDonald JS et al. 5-Fluorouracil and interferon-a a in advanced colorectal carcinoma. Results of two treatment schedules. Cancer 1993; 7: Pazdur R, Ajani JA, Patt YZ et al. Phase II evaluation of recombinant alpha-a-interferon and continuous infusion fluorouracil in previously untreated metastatic colorectal adenocarcinoma. Cancer 1993; 71: Wadler S, Lembersky B, Atkins M et al. Phase II trial of fluorouracil and recombinant interferon alpha-a in patients with advanced colorectal carcinoma; An eastern cooperative oncology group study. J Clin Oncol 1991; 9: Huberman M, McClay E, Atkins M et al. Phase n trial of 5-fluorouracil (5-FU) and recombinant alpha-a interferon in advanced colorectal cancer. Proc Am Soc Clin Oncol 1991; 1: 153(Abstr478). 1. Kemeny N, Younes A, Seiter K. Interferon alpha-a and 5-fluorouraci] for advanced colorectal carcinoma. Assessment of activity and toxicity. Cancer 199; 66: Downloaded from on 17 July 18

5 Douillard JY, Leborgene J, Danielou JY et al. Phase II trial of 5-fluorouracil and recombinant a-interferon in metastatic previously untreated colorectal cancer. Proc Am Soc Clin Oncol 1991;1:139(Abstr4). 1. Hansen RM, Ritch PS, Libnoch JA et al. Continuous 5-fluorouracil infusion and a-interferon in advanced cancers: A report of initial treatment results. Am J Med Sci 1991; 31: Raderer M, Scheithauer W. Treatment of advanced colorectal cancer with 5-fluorouracil and interferon-a: An overview of clinical trials. Eur J Cancer 1995; 31 A: York M, Greco FA, Einhorn L et al. A randomized phase HI trial comparing 5-FU with or without interferon alfa-a for advanced colorectal cancer. Proc Am Soc Clin Oncol 1993; 1:, (Abstr 59). 15. Hill M, Norman A, Cunningham D et al. Royal Marsden phase HJ trial of fluorouracil with or without interferon alfa-b in advanced colorectal cancer. J Clin Oncol 1995; 13: Lokich JJ, Ahlgren JD, Gullo JJ et al. A prospective randomized comparison of continuous infusion fluorouracil with a conventional bolus schedule in metastatic colorectal carcinoma: A midatlantic oncology program study. J Clin Oncol 1989; 7: Ehrlichman C, Fines S, Wong. A randomized trial of fluorouracil with folinic acid in patients with metastatic colorectal carcinoma. J Clin Oncol 1988; 6: Martoni A, Cricca A, Guaraldi M et al. Randomized clinical trial with a weekly regimen of 5-FU vs. 5-FU + intermediatedose folinic acid in the treatment of advanced colorectal cancer. Ann Oncol 199, 3: Advanced colorectal cancer meta-analysis project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer Evidence in terms of response rate. J Clin Oncol 199; 1: Hill M, Norman A, Cunningham D et al. Impact of protracted venous infusion fluorouracil with or without interferon alfa-b on tumor response, survival and quality of life in advanced colorectal cancer. J Clin Oncol 1995; 13: Corfu-A study group. Phase HI randomized study of two fluorouracil combinations with either interferon alfa-a or leucovorin for advanced colorectal cancer. J Clin Oncol 1995; 13: Ragnhammar P, Blomgren H, Edler D et al. Different dose regimens of 5-fluorouracil and interferon-a in patients with metastatic colorectal carcinoma. Eur J Cancer 1995; 31 A: 315. Received 9 January 1996; accepted 7 June Correspondence to: Prof. Patrick Dufour Service Onco Hematologie Hopitaux Universitaires de Strasbourg 6798 Strasbourg Cedex France Downloaded from on 17 July 18