Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Irinotecan Plus Fluorouracil/Leucovorin for Metastatic Colorectal Cancer: A New Survival Standard

Irinotecan Plus Fluorouracil/Leucovorin for Metastatic Colorectal Cancer: A New Survival Standard Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 The Oncologist Promising New Drugs and Combinations Irinotecan Plus Fluorouracil/Leucovorin for Metastatic Colorectal Cancer: A New Survival Standard a b c d LEONARD B. SALTZ, JEAN-YVES DOUILLARD, NICOLETTA PIROTTA, MAY ALAKL, d d c c GABRIELA GRUIA, LUCILE AWAD, GARY L. ELFRING, PAULA K. LOCKER, LANGDON L. MILLER a b Memorial Sloan-Kettering Hospital, New York, New York, USA; Centre Rene Gauducheau, Nantes, France; c d Pharmacia Corporation, Peapack, New Jersey, USA; Aventis, SA, Antony, France Key Words. Irinotecan · Fluorouracil · Leucovorin · Camptothecin · Antineoplastic agents · Colorectal neoplasms · Quality of life · Randomized controlled trials · Human · Prognosis ABSTRACT Background. Irinotecan is a topoisomerase I inhibi- study 2, response rates for irinotecan/5-FU/LV versus tor that prolongs survival in patients with colorectal can- 5-FU/LV alone were 35% and 22% (p = .005); median cer refractory to fluorouracil (5-FU) and leucovorin TTPs were 6.7 months and 4.4 months, respectively (p (LV). This demonstrated activity of irinotecan as effec- < .001). Survival time increased significantly with tive second-line therapy for colorectal cancer led to eval- irinotecan/5-FU/LV versus 5-FU/LV alone in both uation of combination irinotecan/5-FU/LV as first-line studies (study 1: median 14.8 months versus 12.6 therapy for patients with metastatic disease. The results months, p = .042; study 2: median 17.4 months versus of two prospective phase III randomized, controlled, 14.1 months, p = .032). The combined analysis of the multicenter, multinational clinical trials in patients with data from the two studies showed median survivals of previously untreated metastatic colorectal cancer served 15.9 months versus 13.3 months, favoring the irinote- as the basis for U.S. and European approval of irinote- can-containing combinations (stratified-by-study p = can/5-FU/LV for this indication. An overview of the find- .003). Patients in study 1 had a 36% lower risk of ings of these two pivotal studies provides insights tumor progression and a 20% lower risk of death with regarding the application of this new combination in the irinotecan combination than with 5-FU/LV alone; clinical practice. comparable risk reduction values in study 2 were 42% Methods. Patients were randomly assigned to receive and 23%. While grade 3 diarrhea and vomiting were 5-FU/LV, either alone, or with concurrent irinotecan. The more common with irinotecan/5-FU/LV, grade 4 neu- study conducted primarily in North America (study 1), tropenia, neutropenic fever, and mucositis were less employed bolus 5-FU/LV schedules, while the study per- common with irinotecan/5-FU/LV than with the Mayo formed primarily in Europe (study 2), employed infu- Clinic 5-FU/LV regimen. sional 5-FU/LV regimens. Major endpoints included Conclusion. The combination of irinotecan/5-FU/LV tumor response rate, time to tumor progression (TTP), is superior to 5-FU/LV alone as first-line therapy for overall survival, quality of life, and safety. patients with metastatic colorectal cancer, offering consis- Results. In study 1, the respective confirmed tently improved tumor control and prolonged survival. response rates for irinotecan/5-FU/LV versus 5-FU/LV Irinotecan-based combination therapy sets a new survival were 39% and 21% (p < .001); median TTPs were 7.0 standard for the treatment of this life-threatening disease. months and 4.3 months, respectively (p = .004). In The Oncologist 2001;6:81-91 Correspondence: Leonard B. Saltz, M.D., Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021, USA. Telephone: 212-639-2501; Fax: 212-794-7186; e-mail: saltzl@mskcc.org Accepted for publication December 5, 2000. ©AlphaMed Press 1083-7159/2001/$5.00/0 The Oncologist 2001;6:81-91 www.TheOncologist.com Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 82 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer INTRODUCTION PATIENTS AND METHODS Until recently, patients with colorectal cancer had few available chemotherapeutic options. In the absence of any Patient Demographics and Study Design Both studies were large multinational, multicenter, ran- more effective agent, 5-fluorouracil (5-FU) served as the domized, open-label, controlled trials with similar inclusion mainstay of treatment for patients with advanced colorectal and exclusion criteria. Male or female patients were eligible cancer for nearly 45 years [1]. A fluorinated pyrimidine, 5- for enrollment if they were aged 18 years or older (but not FU acts by inhibiting thymidylate synthase, an enzyme nec- more than 75 years for study 2), had a histologic diagnosis of essary for the production of thymidine nucleotides required colorectal cancer with unresectable measurable metastases for DNA synthesis. 5-FU is usually given in combination and a performance status of 0-2, and had adequate hemato- with leucovorin (LV), a biomodulating agent that increases logic, renal, and hepatic function. They could not have the binding of 5-FU to thymidylate synthase, thereby increas- received prior chemotherapy for metastatic disease, or 5-FU- ing the inhibition of DNA synthesis and enhancing the based adjuvant chemotherapy during the 12 months (study 1) antitumor effect of 5-FU. This approach has increased or six months (study 2) preceding study entry. Patients who response rates from 11% with 5-FU alone to 23% with 5- had previously undergone pelvic radiotherapy were excluded FU/LV but has provided no meaningful survival benefit from study 1, but could enter study 2. All patients provided (median survival 11.0 months with 5-FU alone versus 11.5 written informed consent before being enrolled. months with 5-FU/LV) [2]. Attempts to improve the efficacy In study 1, patients were prospectively stratified according of 5-FU by administration of protracted continuous infusions to age (<65 or ≥65 years), time from initial diagnosis (<6 or ≥6 also showed a significant increase in response rate to 22%, but months), performance status (0 or 1-2), and prior adjuvant 5-FU- again, without improvement in survival beyond one year [3]. based therapy (yes or no). The patients were then randomized to Irinotecan (CPT-11, CAMPTOSAR , Pharmacia one of three treatment regimens: irinotecan/bolus 5-FU/LV, Corporation; Peapack, NJ; CAMPTO , Aventis SA, Antony, bolus 5-FU/LV, or irinotecan alone. The treatment scheme for France), a topoisomerase I inhibitor, offers a mechanism of study 1 is depicted in Figure 1. In study 2, each study site chose action completely different from that of 5-FU in the treat- one of two infusional regimens for administration of 5-FU/LV ment of colorectal cancer. Irinotecan and its metabolites according to local clinical practice or preference, i.e., either bind to a complex of DNA and topoisomerase I (an enzyme required for unwinding of DNA during replication), induc- ARM ing DNA strand breaks and consequent tumor cell death CPT-11: 125 mg/m /wk x 4 wks, q 6 wks 5-FU: 500 mg/m /wk x 4 wks, q 6 wks n = 231 [4]. Irinotecan has shown antitumor activity in patients with LV: 20 mg/m /wk x 4 wks, q 6 wks colorectal cancer when administered alone as first-line ther- RAMDOMIZATION apy [5-8] or as second-line therapy after 5-FU failure [5, 7- 5-FU: 425 mg/m /d × 5 d, q 4 wks STUDY 1 n = 226 LV: 20 mg/m /d × 5 d, q 4 wks 9]. In two randomized phase III studies in patients who experienced failure of first-line 5-FU, irinotecan was com- pared with either best supportive care or with intensive 5-FU- C 2 CPT-11: 125 mg/m /wk × 4 wks, q 6 wks n = 226 based infusional therapy. Both studies showed a statistically significant survival benefit for patients treated with irinote- can [10, 11]. The activity of irinotecan against untreated ARM and 5-FU-resistant colorectal cancer led to studies of this 2 CPT-11: 80 mg/m /wk × 6 wks, q 7 wks A1* 5-FU: 2.3 gm/m /wk × 6 wks, q 7 wks agent in combination with 5-FU/LV as first-line therapy for n = 54 2 LV: 500 mg/m /wk × 6 wks, q 7 wks or this disease. CPT-11: 180 mg/m d1 q 2 wks A2 5-FU: 400 IV/600 CI mg/m d1, 2 q 2 wks n = 145 Recently Douillard et al. and Saltz et al. reported the LV: 200 mg/m d1, 2 q 2 wks RAMDOMIZATION results of two studies conducted in parallel at a combined total STUDY 2 of 154 institutions located primarily in North America and Europe [12, 13]. These studies evaluated irinotecan combined 5-FU: 2.6 gm/m /wk × 6 wks, q 7 wks B1* n = 43 LV: 500 mg/m /wk × 6 wks, q 7 wks with either bolus (study 1: Saltz) or infusional 5-FU/LV (study or 5-FU: 400 IV/600 CI mg/m d1, 2 q 2 wks B2 2: Douillard) in previously untreated patients with metastatic n = 143 LV: 200 mg/m d1, 2 q 2 wks *AIO colorectal cancer. Both studies demonstrated statistically sig- de Gramont CI = continous infusion nificant clinical benefits with the irinotecan/5-FU/LV combi- CPT-11 = irinotecan nations, including improved tumor control and prolonged survival. The major efficacy and safety results of these studies are summarized in this review. Figure 1. Treatment regimens for studies 1 and 2. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 83 the once-weekly regimen of the Arbeitsgemeinschaft Safety was evaluated in terms of adverse events, labora- Internische Onkologie (AIO) cooperative German group for tory abnormalities, and deaths. Adverse-event assessments oncology [14] (designated A1 and B1 in Fig. 1), or the every- and complete blood counts were performed weekly, and two-week regimen of de Gramont [15] (designated as A2 and chemistry studies were performed at the beginning of each B2 in Fig. 1). Once a site selected its preferred regimen (either chemotherapy cycle. AIO or de Gramont), patients at the site were randomized to receive that method of 5-FU/LV administration, with or with- STATISTICAL METHODS out irinotecan. After initial treatment, doses in all treatment In study 1, a sample size of 220 patients per treatment arms of both studies could be adjusted according to specified arm was considered necessary to detect a 40% improve- guidelines to accommodate individual patient tolerance to the ment in TTP (i.e., median TTP from five to seven months) study drugs [12, 13]. Treatment was continued until tumor in the irinotecan/5-FU/LV arm versus the 5-FU/LV arm progression, the occurrence of unacceptable toxicity, or the with a power of 85% (two-tailed unstratified log-rank test, patient withdrew consent. alpha of .05). In study 2, assuming a response rate of 35% In both studies, atropine was provided for the treatment for the 5-FU/LV arm and 50% for the irinotecan/5-FU/LV of cholinergic symptoms, and loperamide for treatment of arm, a total of 338 evaluable patients (169 per arm) was delayed diarrhea. Antiemetic agents were administered as considered necessary to provide 80% power to detect a sig- prophylaxis for nausea and/or vomiting. Granulocyte colony- nificant difference in response rate (two-tailed chi-square stimulating factor could be given for prolonged neutropenia test, alpha of .05). In the statistical comparison of irinote- or infectious complications during neutropenic episodes. can/5-FU/LV versus 5-FU/LV, results for patients who received regimens A1 and A2 were combined, as were EVALUATIONS results from patients treated with B1 and B2. The major clinical efficacy endpoints were similar in the For both studies, time-to-event endpoints (e.g., duration of two studies; both evaluated objective tumor response rates, response, TTP, survival) were analyzed by Kaplan-Meier time to tumor progression (TTP), and survival. In study 1, curves and unstratified log-rank tests. Response rates were tumor measurements were obtained every six weeks until compared using chi-square tests. The influence of stratification week 24, and then every 12 weeks until the end of study treat- factors and other baseline characteristics on confirmed objec- ment. In study 2, tumor measurements were obtained after tive tumor response rates were assessed using multiple logistic each chemotherapy cycle (every six to seven weeks) and at the regression modeling. Similarly, TTP and survival were end of study treatment. Objective tumor responses were con- assessed using Cox proportional hazard regression modeling. firmed at least four weeks after the first documentation of Changes in QOL subscale scores were assessed using analysis response. In both studies, quality of life (QOL) was measured of variance for repeated measures. In study 1, worst changes at the start of each treatment cycle using the European from baseline were compared with Student’s t-tests. Organization for the Research and Treatment of Cancer Irinotecan alone was included as a third treatment arm in Quality of Life Questionnaire (EORTC QLQ-C30). Following study 1 only to document the efficacy and safety associated completion of study drug treatment, data were collected on with its first-line single-agent use; consequently no hypotheses post-study treatment for colorectal cancer and on survival. regarding its comparative efficacy were tested. Table 1. Patient disposition Study 1 Study 2 Irinotecan Irinotecan Patients 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B a a Randomized (n) 231 226 226 199 188 Untreated (n)4 8 4 1 1 a a Full-analysis (n) —— — 198 187 b, c d e b Treated with a different arm (n)2 1 1 01 f f f f f As-treated (n) 225 219 223 199 186 a b c Population analyzed for efficacy; Randomized to irinotecan/5-FU/LV but received 5-FU/LV; Randomized to irinotecan/5-FU/LV but received d e f irinotecan alone; Randomized to 5-FU/LV but received irinotecan alone; Randomized to irinotecan alone but received 5-FU/LV; Population analyzed for safety Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 84 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Table 2. Baseline patient characteristics Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B Characteristic (n = 231) (n = 226) (n = 226) (n = 198) (n = 187) Age (years) Median 62 61 61 62 59 (Range) (25-85) (19-85) (30-87) (27-75) (24-75) Sex (%) b c Male 65 54 64 67 53 Female 34 45 35 33 47 Performance Status (%) 039 41 46 5151 146 45 46 4241 215 13 8 7 8 Site, primary tumor (%) Colon 81 85 84 55 65 Rectum 17 14 15 45 35 No. of involved organ sites (%) 164 66 62 6263 226 23 28 2328 >2 10 10 9 15 9 Prior adjuvant 5-FU (%) 11 8 10 26 24 Prior radiotherapy (%) Any 3 2 1 20 16 Pelvis/abdomen 2 1 1 –– Other sites 1 1 0 –– Baseline laboratory abnormalities (%) CEA ≥100 ng/ml 40 39 37 35 32 Hemoglobin <11 g/dl 26 25 26 16 21 3 3 WBC ≥8 × 10 /mm 52 53 51 47 38 LDH >UNL 60 56 53 43 45 Total bilirubin >UNL 7 4 10 7 7 Abbreviations: CEA = carcinoembryonic antigen; LDH = lactate dehydrogenase; UNL = upper normal limit; WBC = white blood cell count a b c Data not available for some patients who were randomized but not treated; p = .019 Arm A versus Arm B; p = .006 Arm A versus Arm B; Percentage based on patients with baseline data RESULTS Colonic primary tumors predominated in both studies, as might be expected given the epidemiology of the disease. Patients The proportion of patients with rectal tumors in study 1 The studies enrolled a combined total of 1,070 patients (~15%) was lower than that in study 2 (35% to 45%), pre- treated at 154 multinational sites. A small number of patients sumably because patients with prior pelvic irradiation were were untreated or received a treatment regimen to which they excluded from study 1 but not from study 2. Because most were not originally randomized. Table 1 summarizes the patients had metastatic disease at initial diagnosis, relatively populations for the two studies. few patients (about 10% in study 1 and 25% in study 2) had Baseline characteristics are summarized in Table 2. Across received prior adjuvant 5-FU therapy. all treatment arms, the populations were generally divided sim- With respect to baseline laboratory values, carcinoem- ilarly between patients with a performance status of 0 and those bryonic antigen (CEA) and total bilirubin levels were gener- with a performance status of 1 or 2. However, the proportions ally similar across the two studies. However, there were of patients with a performance status of 2 in the irinotecan/5- notably greater proportions of patients with a depressed FU/LV and 5-FU/LV groups of study 1 (15% and 13%, respec- hemoglobin level, elevated WBC, or abnormal serum lactate tively) were about twice those of patients with a performance dehydrogenase (LDH) in study 1 than in study 2. These lab- status of 2 in either arm of study 2 (7% and 8%, respectively). oratory abnormalities are consistent with the performance Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 85 Table 3. Efficacy results Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B Efficacy endpoint (n = 231) (n = 226) (n = 226) (n = 198) (n = 187) Overall objective response rate (%) 50 28 29 49% 31% p < .0001 p < .001 Confirmed tumor response rate (%) 39 21 18 35 22 b b p < .0001 p < .005 CEA response rate (≥50% decrease) (%) 55 38 44 —— p < .001 — Median TTP (months) 7.0 4.3 4.2 6.7 4.4 c c p = .004 p < .001 Median survival (months) 14.8 12.6 12.0 17.4 14.1 c c p = .042 p = .032 Abbreviations: CEA = carcinoembryonic antigen; TTP = time to tumor progression a b c Confirmed ≥4 weeks after first evidence of objective response; Chi-square test; Unstratified log-rank test status findings, possibly indicating greater tumor burden or CEA during treatment) was evaluated in patients with ele- tumor-related organ dysfunction among the patients in study 1. vated baseline CEA values and at least one CEA assessment while on treatment. The CEA response rate with irinotecan/ EFFICACY 5-FU/LV was substantially higher than with 5-FU/LV; this Efficacy data for studies 1 and 2 are summarized in difference was significant (p < .001) (Table 3). Table 3. Remarkable consistency between the studies was observed when examining the efficacy outcome measures. TTP TTP was significantly longer in both studies for patients RESPONSE RATES who received irinotecan/5-FU/LV than for those who received In study 1, the overall response rate (response defined as 5-FU/LV (study 1: median 7.0 months versus 4.3 months, p = shrinkage of measured tumors by ≥50% on at least one occa- .004; study 2: median 6.7 months versus 4.4 months, p < .001). sion) was significantly higher for patients treated with irinote- As in the case of response rate, TTP was universally improved can/5-FU/LV than for those treated with 5-FU/LV (50% versus in all patient subgroups [16]. Kaplan-Meier TTP curves for 28%, p < .0001); comparable rates were also significantly both studies are shown in Figure 2. higher in study 2 (49% versus 31%, p < .001). The confirmed response rates (based on responses confirmed by imaging tests Survival ≥4 weeks later) were also significantly (p < .005 to p < .001) Comparison of survival with irinotecan/5-FU/LV versus and consistently higher among patients who received irinote- 5-FU/LV in study 1 showed a significant (p = .042) differ- can/5-FU/LV compared with those who received 5-FU/LV in ence favoring the irinotecan combination (Table 3); at any both studies (Table 3). The median duration of confirmed given time on study, patients treated with irinotecan/5-FU/LV objective tumor response from time of randomization was had a 19% reduction in the risk of death relative to those about nine months across all treatment arms. Examination of treated with 5-FU/LV alone (hazard ratio 0.81, 95% CI = confirmed response rates by subgroup analyses showed that, for 0.65-0.99). Similarly, in study 2, survival was significantly (p every subgroup evaluated, including patients with poor perfor- = .032) longer with irinotecan/5-FU/LV therapy than with 5- mance status, extensive metastatic disease, prior adjuvant ther- FU/LV; the risk of death at any time on study was decreased apy, or abnormal baseline laboratory values, the response rate by 23% with the irinotecan combination relative to 5-FU/LV with irinotecan/5-FU/LV was approximately double that with alone (hazard ratio 0.77, 95% CI = 0.60-0.98). The combined 5-FU/LV alone [16]. data from the two studies showed median survivals of 15.9 months for irinotecan/5-FU/LV versus 13.3 months for 5- CEA FU/LV alone; a stratified-by-study analysis indicated a sig- Serum CEA was assessed systematically during study nificant (p = .003) reduction in the risk of death at any time on 1. CEA response (≥50% decrease from baseline in serum study of 21% with the irinotecan/5-FU/LV arms relative to Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 86 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Study 1 Study 1 1.0 1.0 Irinotecan/5-FU/LV (n = 231) Irinotecan/5-FU/LV (n = 231) 0.9 0.9 5-FU/LV (n = 226) 5-FU/LV (n = 226) 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 p = .042* p = .004* 0.1 0.1 0.0 0.0 03 61 9 2 15 18 21 24 27 30 0 3 6 912 15 Months Months *Log-rank test *Log-rank test Study 2 Study 2 1.0 1.0 Irinotecan/5-FU/LV (n = 198) 0.9 Irinotecan/5-FU/LV (n = 198) 0.9 5-FU/LV (n = 187) 0.8 5-FU/LV (n = 187) 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 p < .001* p = .032* 0.1 0.1 0.0 0.0 03 61 9 2 15 18 21 24 27 30 0 6 912 15 Months Months *Log-rank test *Log-rank test Figure 2. TTP—Kaplan-Meier estimates from two phase III studies. Figure 3. Survival—Kaplan-Meier estimates from two phase III studies. the 5-FU/LV control arms (hazard ratio 0.67, 95% CI = 0.57- Efficacy results in this arm of the study appeared generally 0.79) [17]. Kaplan-Meier survival curves for studies 1 and 2 similar to those observed with 5-FU/LV alone (Table 3). are shown in Figure 3. It is apparent that the irinotecan/5-FU/LV curve sepa- Post-Study Treatment rates from the 5-FU/LV curve later in study 1 than in study Over 90% of the patients had follow-up information 2. This finding may possibly be explained by the greater regarding post-study anticancer therapy. In study 1, 52% of proportion of patients with performance status 2 in study 1; these patients in the irinotecan/5-FU/LV-treated arm received such patients generally had survival times <6 months with some form of second-line therapy, compared with 70% in the either therapy. If only patients with performance status 0-1 5-FU/LV-treated arm. For 56% of the patients in the 5-FU/LV are considered, survival results for the irinotecan/5-FU/LV arm, this second-line regimen consisted of irinotecan as a sin- arms are quite similar (medians of 17.2 months and 17.4 gle agent or in combination. In study 2, 49% of patients in the months in the 195 patients of study 1 and the 185 patients of irinotecan/5-FU/LV arm received post-study therapy com- study 2, respectively) and continue to contrast with the 5- pared with 65% of those in the 5-FU/LV arm. In this study, FU/LV arms (medians of 13.8 months and 15.5 months in 34% of the 5-FU/LV-treated patients given post-study the 195 patients of study 1 and the 173 patients of study 2, chemotherapy received second-line irinotecan (Table 4). respectively). Proportional Hazards Modeling Single-Agent Irinotecan Results In study 1, Cox regression techniques were used to evalu- A single-agent irinotecan arm was included in study 1 ate the effects of the study treatments on TTP and overall sur- only to document the activity of this agent as monotherapy. vival as influenced by the four predefined stratification factors Probability Probability Probability Probability Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 87 Table 4. Post-study therapy Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B b c Efficacy endpoint (n = 205) (n = 203) (n = 195) (n = 167) (n = 171) Total patients with post-study therapy 52% 70% 79% 49% 65% Irinotecan-based 1% 38% 3% 2% 28% Irinotecan/5-FU-based 13% 18% 9% 4% 6% 5-FU-based 30% 10% 64% 32% 21% Other therapy 8% 4% 3% 11% 10% a b Based on patients with follow-up information (approximately 90% of patients); Altogether, 56% of patients in this arm ultimately received a second-line regimen containing irinotecan as a single agent or in combination; Altogether, 34% of patients in this arm ultimately received a second-line regimen containing irinotecan as a single agent or in combination. Table 5. Cox regression results TTP Survival Hazard Hazard Factors Values Ratio 95% CI p Ratio 95% CI p Study 1 Serum LDH ≤UNL versus >UNL 0.60 0.47-0.76 .0001 0.47 0.37-0.60 .0001 Performance status 0 versus ≥1 0.74 0.59-0.93 .0088 0.57 0.45-0.71 .0001 No. of organ sites 1 versus ≥2 sites 0.63 0.50-0.80 .0001 0.67 0.54-0.84 .0004 Bilirubin ≤UNL versus >UNL 0.56 0.35-0.89 .0132 0.55 0.35-0.86 .0051 3 3 WBC <8 versus ≥8 × 10 /mm —— — 0.64 0.51-0.80 .0001 Hemoglobin ≥11 versus <11 0.74 0.58-0.95 .0157 —— — Age ≥65 versus <65 0.78 0.63-0.98 .0315 —— — Treatment Irinotecan/5-FU/LV versus 5-FU/LV 0.64 0.51-0.79 .0001 0.80 0.64-0.99 .0372 Study 2 Serum LDH ≤UNL versus >UNL 0.61 0.46-0.80 .0012 0.55 0.42-0.72 .0001 Performance status 0 versus ≥1 —— — 0.52 0.41-0.67 .0001 Time from metastatic ≥1 mo versus <1 mo 0.62 0.48-0.80 .0003 0.63 0.49-0.82 .0005 diagnosis No. of organ sites 1 versus ≥2 sites 0.71 0.55-0.91 .0070 0.73 0.57-0.94 .0127 Treatment Irinotecan/5-FU/LV versus 5-FU/LV 0.58 0.45-0.75 .0001 0.77 0.61-0.98 .0365 Abbreviations: CI = confidence interval; LDH = lactate dehydrogenase; TTP = time to tumor progression; UNL = upper normal limit; WBC = white blood cell count (age, performance status, time from initial diagnosis, and prior both improved TTP and survival in study 1 were normal LDH, adjuvant therapy), as well as other baseline factors prospec- good performance status, fewer involved organ sites, and tively considered to have potential prognostic significance normal bilirubin. Higher hemoglobin level and normal (e.g., number of involved organ sites, liver involvement, serum WBC were significantly predictive of longer TTP and sur- lactic dehydrogenase [LDH], hemoglobin level) [10, 11, 18- vival time, respectively. Surprisingly, older age also 20]. As shown in Table 5, factors significantly predictive for appeared to be significantly associated with a longer TTP. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 88 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Treatment with combination irinotecan/5-FU/LV remained a SAFETY significant independent predictor of longer TTP (p < .001) The most frequently occurring adverse effects noted in and survival (p = .037) when other significant baseline studies 1 and 2 are shown in Table 6. For study 2, this review patient characteristics were taken into account. In this focuses on the safety data from the FDA-approved de adjusted analysis, irinotecan/5-FU/LV combination therapy Gramont irinotecan/5-FU/LV regimen and the corresponding was associated with a 36% lower risk of tumor progression 5-FU/LV control. In both studies, approximately 23% of and a 20% lower risk of death, than with 5-FU/LV therapy. patients treated with the combination irinotecan/5-FU/LV These findings were corroborated by application of the regimens experienced grade 3-4 diarrhea compared with model developed in study 1 to the data from study 2 (Table 5). approximately 10% to 14% of patients receiving 5-FU/LV In this study, normal baseline serum LDH and fewer involved alone. This difference was primarily in the incidence of grade organs were found to be significant predictors for improved 3 diarrhea; grade 4 diarrhea was comparably infrequent in the TTP and survival; better performance status was also signifi- treatment and control arms of the two trials; for example, in cantly associated with longer survival. In this study, a longer study 1, the incidence was 8% in the irinotecan/5-FU/LV arm time from diagnosis of metastatic disease to randomization and 7% in the 5-FU/LV arm. As expected, grade 3-4 vomit- was predictive for better outcomes with respect to both TTP ing was somewhat more common with irinotecan-based ther- and survival. As in study 1, treatment with irinotecan/5- apy but occurred in <10% of patients in any of the FU/LV in study 2 remained a significant predictor of improved combination arms. survival and was associated with a 42% lower risk of tumor Of note, grade 3-4 mucositis was quite infrequent with progression and a 23% lower risk of death relative to treatment irinotecan-based therapy, occurring in <4% of patients receiv- with 5-FU/LV when adjusted for baseline prognostic factors. ing combination therapies. By contrast, the Mayo Clinic Table 6. Incidence of adverse events Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B a a Adverse event (n = 225) (n = 219) (n = 223) (n = 145) (n = 143) Diarrhea (%) Grade 3/4 23 13 31 14 6 Grade 3 15 6 18 10 4 Grade 4 8 7 13 4 2 Vomiting (%) Grade 3/4 10 4 12 4 2 Grade 3 5 3 6 3 1 Grade 4 4 1 6 1 1 Mucositis (%) Grade 3/4 2 17 2 4 3 Grade 3 2 15 2 4 3 Grade 4 0 2 0 0 0 Neutropenia (%) Grade 3/4 54 67 31 46 14 Grade 3 30 24 19 36 13 Grade 4 24 43 12 10 1 Neutropenic complications (%) Neutropenic fever 7 15 6 3 1 Neutropenic infection 2 0 2 2 0 Discontinuations due to adverse events and drug-related deaths (%) Discontinuations 8 6 12 6 1 Drug-related deaths 1 1 1 1 0 de Gramont regimen only Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 89 schedule of 5-FU/LV used in the control arm of study 1 and metastatic colorectal cancer. The studies were similar in that commonly employed as first-line therapy in North America they enrolled analogous patient populations, evaluated similar was associated with a much higher frequency of severe, grade efficacy and safety endpoints, and applied standardized meth- 3-4 mucositis (17%). ods of analysis. The trials were complementary in that they For neutropenia, only grade 4 events are usually associ- assessed the use of irinotecan in combination with two differ- ated with clinical consequences. Of interest, the frequency of ent methods of 5-FU/LV administration (bolus and infu- grade 4 neutropenia with combination irinotecan/5-FU/LV sional therapy). Study 1 compared a new combination of therapy (24%) in study 1 was essentially half that observed in irinotecan/bolus 5-FU/LV with that of the Mayo Clinic reg- patients receiving 5-FU/LV in the control group (43%) and imen of bolus 5-FU/LV that had been most widely employed proportionately fewer patients experienced neutropenic fever in North America. Study 2 symmetrically determined the when contrasting irinotecan/5-FU/LV (7%) with 5-FU/LV therapeutic ratio associated with adding irinotecan to two (15%). In study 2, where schedules of chemotherapy admin- different infusional 5-FU/LV regimens that were widely used istration were similar, grade 4 neutropenia was more fre- in Europe. Because the comparator regimens have been com- quently seen when irinotecan was added to 5-FU/LV (10%) monly used in oncology practices worldwide, the two studies than with 5-FU/LV alone (1%). The incidence of neutropenic provide clinicians with insights into the efficacy and safety of fever was also higher in the irinotecan/5-FU/LV arm (5.0%) the new combinations relative to familiar standards. than in the 5-FU/LV arm (1%); however, these rates of neu- The efficacy results of the two studies were remarkably tropenic complications are quite low when contrasted with consistent and show that the combination of irinotecan with 5- many other chemotherapy regimens [21]. FU provides patients with significant reductions in tumor size Discontinuations due to adverse events were acceptably in conjunction with prolonged suppression of tumor growth. low across all arms of both studies. The incidence of treat- The confirmed objective response rates with the irinotecan/5- ment-related death ranged from 0% to 1% in all treatment FU/LV arms (39% and 35%) were 1.5 to 2 times those in the groups. 5-FU/LV arms (21% and 22%), and these differences were statistically significant. In studies 1 and 2, respectively, TTP QOL was significantly improved with combination treatment (medi- The primary repeated measures analyses of variance of ans, 7.0 and 6.7 months) relative to treatment with 5-FU/LV changes from baseline in QOL in both studies showed no sig- alone (medians, 4.3 and 4.4 months). Assessment of response nificant differences between the treatment arms. In study 1, rates and TTP across demographic and disease-related sub- results were significantly better with irinotecan/5-FU/LV than groups showed improvements with irinotecan-based combina- with 5-FU/LV when comparing worst changes from baseline tion therapy in all prospectively defined subgroups [16]. These for the subscales of role functioning, fatigue, appetite loss, results indicate that irinotecan/5-FU/LV has the potential to and pain [22]. In study 2, deterioration of performance status offer better tumor control to all patients who are eligible for occurred later in patients treated with the irinotecan combina- first-line combination chemotherapy. tion than in those treated with 5-FU/LV alone (median 11.2 The most important finding of the studies was that first- months versus 9.9 months, p = .046). line irinotecan/5-FU/LV combination treatment provided a statistically significant survival advantage. It is particularly DISCUSSION noteworthy that this advantage occurred even though most Attempts to improve outcome in patients with metasta- control patients received second-line irinotecan therapy after tic colorectal cancer with modified 5-FU-based regimens, on-study failure of 5-FU/LV. These results suggest that early while generally resulting in higher response rates, have combination irinotecan/5-FU/LV may be superior to sequen- failed to significantly improve survival [3, 4]. However, the tial administration of first-line 5-FU followed by second-line results of recent studies of the topoisomerase I inhibitor irinotecan. The combination therapy survival outcomes noted irinotecan have improved the outlook for patients with this dis- among patients with good performance status accentuate this ease. In these studies, irinotecan was shown to improve sur- observation [16]. Such findings were presaged by the results vival, first as single-agent second-line therapy [10, 11], and of a prior study in metastatic colorectal cancer in which early now as a component of first-line treatment with 5-FU/LV administration of chemotherapy before symptom develop- [12, 13, 17]. ment was compared with delayed treatment after symptom The two phase III randomized multicenter, multinational, development; as in the current experience, early therapy in controlled studies reviewed here compared the efficacy and that study resulted in significantly better survival [23]. safety of new combinations of irinotecan/5-FU/LV to that of Cox regression modeling, which assessed treatment traditional regimens of 5-FU/LV given as first-line therapy of effect adjusted for significant prognostic factors, provided Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 90 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer further evidence that combination therapy with irinotecan/5- and now as a component of first-line combination treatment. FU improves TTP and survival. This analysis indicated that The studies discussed here are the first trials to document irinotecan/5-FU/LV treatment resulted in an approximate that the combination of a new agent with 5-FU/LV can 40% reduction in the relative risk of tumor progression and safely benefit patients with metastatic colorectal cancer by a 20% decrease in the relative risk of death. inducing tumor shrinkage, extending tumor control, and sig- In both studies, gastrointestinal toxicities were more com- nificantly prolonging life without an impairment of QOL. mon with combination treatment, but grade 4 diarrhea— The strength and consistency of the data from these large largely defined by the need for hospitalization for supportive complementary studies led the FDA’s Oncology Drug care—was infrequent (<8%). In study 1, grade 4 neutropenia, Advisory Committee (ODAC) to conclude that irinotecan/5- neutropenic fever, and mucositis were observed less often with FU/LV should be the reference standard against which weekly irinotecan/5-FU/LV than with Mayo Clinic bolus 5- future first-line therapies for metastatic colorectal cancer are FU/LV alone; data from other studies that compared weekly compared. versus monthly 5-FU/LV therapy [24, 25] suggest that this Given the positive findings in patients with metastatic reduced toxicity with combination therapy is most likely due disease, several trials in the U.S. and Europe are comparing to the differences in 5-FU/LV scheduling between the arms. In irinotecan/5-FU/LV with 5-FU/LV as adjuvant therapy for study 2, clinically relevant grade 3 and 4 events, e.g., vomit- patients with stage III colon cancer; it is hoped that the ing, mucositis, and neutropenic fever, were infrequent with the addition of irinotecan to 5-FU/LV will offer an increased de Gramont irinotecan/5-FU/LV regimen. The safety findings opportunity for cure in patients with early-stage disease. were supported by the results of the QOL analyses, which showed that administration of irinotecan in combination with ACKNOWLEDGMENTS 5-FU/LV did not result in significant worsening of QOL. The work described in this publication was supported Attempts to improve outcomes in patients with metasta- by grants from Pharmacia Corporation, Peapack, NJ, and tic colorectal cancer with the limited tools of 5-FU and LV Aventis SA, Antony, France. have been the source of decades of frustration and disap- L.S. and J-Y.D. have received major research support and pointment. Irinotecan has now broken the barrier to honoraria as consultants for both Pharmacia Corporation and improved survival, first as single-agent second-line therapy, Aventis Corporation. REFERENCES 1 Mayer RJ. Moving beyond fluorouracil for colorectal cancer. 8 Rougier P, Bugat R, Douillard JY et al. Phase II study of N Engl J Med 2000;343:963-964. irinotecan in the treatment of advanced colorectal cancer in chemotherapy-naive patients and patients pretreated with fluo- 2 Advanced Colorectal Cancer Meta-analysis Project. Modulation rouracil-based chemotherapy. J Clin Oncol 1997;15:251-260. of fluorouracil by leucovorin in patients with advanced colorec- 9 Rothenberg ML, Cox JV, DeVore RF et al. A multicenter, tal cancer: evidence in terms of response rate. J Clin Oncol phase II trial of weekly irinotecan (CPT-11) in patients with pre- 1992;10:896-903. viously treated colorectal carcinoma. Cancer 1999;85:786-795. 3 Meta-analysis Group in Cancer. Efficacy of IV continuous 10 Cunningham D, Pyrhonen S, James RD et al. Randomised infusion of fluorouracil compared with bolus administration trial of irinotecan plus supportive care versus supportive care in advanced colorectal cancer. J Clin Oncol 1998;16:301-308. alone after fluorouracil failure for patients with metastatic 4 Pommier Y, Tanizawa A, Kohn KW. Mechanisms of topoiso- colorectal cancer. Lancet 1998;352:1413-1418. merase I inhibition by anticancer drugs. In: Liu LF, ed. Advances 11 Rougier P, Van Cutsem E, Bajetta E et al. Randomized trial in Pharmacology. New York: Academic Press, 1994;29B:73-92. of irinotecan versus fluorouracil by continuous infusion after 5 Shimada Y, Yoshino M, Wakui A et al. Phase II study of CPT- fluorouracil failure in patients with metastatic colorectal can- 11, a new camptothecin derivative, in metastatic colorectal can- cer. Lancet 1998;352:1407-1412. cer. CPT-11 Gastrointestinal Cancer Study Group. J Clin Oncol 12 Saltz LB, Cox JV, Blanke C et al. Irinotecan plus fluorouracil 1993;11:909-913. and leucovorin for metastatic colorectal cancer. N Engl J Med 6 Conti JA, Kemeny NE, Saltz LB et al. Irinotecan is an active 2000;343:905-914. agent in untreated patients with metastatic colorectal cancer. 13 Douillard JY, Cunningham D, Roth AD et al. Irinotecan com- J Clin Oncol 1996;14:709-715. bined with fluorouracil compared with fluorouracil alone as 7 Pitot HC, Wender DB, O’Connell MJ et al. Phase II trial of first-line treatment for metastatic colorectal cancer: a multicen- irinotecan in patients with metastatic colorectal carcinoma. tre randomised trial. Lancet 2000;355:1041-1047 [Erratum, J Clin Oncol 1997;15:2910-2919. Lancet 2000;355:1372]. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 91 14 Köhne CH, Schöffski P, Wilke H et al. Effective biomodula- metastases from colorectal cancer. Br J Surg 1995;82:1397- tion by leucovorin of high-dose infusion fluorouracil given as 1400. a weekly 24-hour infusion: results of a randomized trial in 20 Miller LL, Petit RG, Elfring GL. Evaluation of carcinoembry- patients with advanced colorectal cancer. J Clin Oncol onic antigen (CEA) levels during CPT-11 treatment of patients 1998;16:418-426. with previously treated colorectal cancer. Proc Am Soc Clin Oncol 1998;17:297a. 15 de Gramont A, Bosset JF, Milan C et al. Randomized trial comparing monthly low-dose leucovorin and fluorouracil 21 American Society of Clinical Oncology. American Society of bolus with bimonthly high-dose leucovorin and fluorouracil Clinical Oncology recommendations for the use of hematopoi- bolus plus continuous infusion for advanced colorectal cancer: etic colony-stimulating factors: evidence-based clinical practice a French intergroup study. J Clin Oncol 1997;15:808-815. guidelines. J Clin Oncol 1994;12:2471-2508. 16 Knight RD, Miller LL, Pirotta N et al. First-line irinotecan 22 Locker PK, Miller LL, Pirotta N et al. Weekly irinotecan (C) (C), fluorouracil (F), leucovorin (L) especially improves sur- combined with 5-fluorouracil (F) and leucovorin (L) in patients vival (OS) in metastatic colorectal cancer (MCRC) patients with untreated metastatic colorectal cancer (MCRC): impact on (PT) with favorable prognostic indicators. Proc Am Soc Clin quality of life. Proc Am Soc Clin Oncol 2000;19:620a. Oncol 2000;19:991a. 23 Nordic Gastrointestinal Tumor Adjuvant Therapy Group. 17 Saltz LB, Douillard J, Pirotta N et al. Combined analysis of Expectancy or primary chemotherapy in patients with advanced two phase III randomized trials comparing irinotecan (C), asymptomatic colorectal cancer: a randomized trial. J Clin fluorouracil (F), leucovorin (L) versus F alone as first-line Oncol 1992;10:904-911. therapy of previously untreated metastatic colorectal cancer 24 Buroker TR, O’Connell MJ, Wieand HS et al. Randomized (MCRC). Proc Am Soc Clin Oncol 2000;19:242a. comparison of two schedules of fluorouracil and leucovorin 18 Kemeny N, Niedzwiecki D, Shurgot B et al. Prognostic vari- in the treatment of advanced colorectal cancer. J Clin Oncol ables in patients with hepatic metastases from colorectal can- 1994;12:14-20. cer. Importance of medical assessment of liver involvement. 25 Leichman CG, Fleming TR, Muggia FM et al. Phase II study Cancer 1989;63:742-747. of fluorouracil and its modulation in advanced colorectal can- 19 Rougier P, Milan C, Lazorthes F et al. Prospective study of cer: a Southwest Oncology Group Study. J Clin Oncol prognostic factors in patients with unresected hepatic 1995;13:1303-1311. Related articles in The Oncologist: Irinotecan: A New Agent Comes of Age. Saltz LB. The Oncologist 2001;6:65. Irinotecan (CPT-11): Recent Developments and Future Directions–Colorectal Cancer and Beyond. Rothenberg ML. The Oncologist 2001;6:66-80. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Oncologist Oxford University Press

Irinotecan Plus Fluorouracil/Leucovorin for Metastatic Colorectal Cancer: A New Survival Standard

Loading next page...
 
/lp/oxford-university-press/irinotecan-plus-fluorouracil-leucovorin-for-metastatic-colorectal-0sNnfbP0bc

References (31)

Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
ISSN
1083-7159
eISSN
1549-490X
DOI
10.1634/theoncologist.6-1-81
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 The Oncologist Promising New Drugs and Combinations Irinotecan Plus Fluorouracil/Leucovorin for Metastatic Colorectal Cancer: A New Survival Standard a b c d LEONARD B. SALTZ, JEAN-YVES DOUILLARD, NICOLETTA PIROTTA, MAY ALAKL, d d c c GABRIELA GRUIA, LUCILE AWAD, GARY L. ELFRING, PAULA K. LOCKER, LANGDON L. MILLER a b Memorial Sloan-Kettering Hospital, New York, New York, USA; Centre Rene Gauducheau, Nantes, France; c d Pharmacia Corporation, Peapack, New Jersey, USA; Aventis, SA, Antony, France Key Words. Irinotecan · Fluorouracil · Leucovorin · Camptothecin · Antineoplastic agents · Colorectal neoplasms · Quality of life · Randomized controlled trials · Human · Prognosis ABSTRACT Background. Irinotecan is a topoisomerase I inhibi- study 2, response rates for irinotecan/5-FU/LV versus tor that prolongs survival in patients with colorectal can- 5-FU/LV alone were 35% and 22% (p = .005); median cer refractory to fluorouracil (5-FU) and leucovorin TTPs were 6.7 months and 4.4 months, respectively (p (LV). This demonstrated activity of irinotecan as effec- < .001). Survival time increased significantly with tive second-line therapy for colorectal cancer led to eval- irinotecan/5-FU/LV versus 5-FU/LV alone in both uation of combination irinotecan/5-FU/LV as first-line studies (study 1: median 14.8 months versus 12.6 therapy for patients with metastatic disease. The results months, p = .042; study 2: median 17.4 months versus of two prospective phase III randomized, controlled, 14.1 months, p = .032). The combined analysis of the multicenter, multinational clinical trials in patients with data from the two studies showed median survivals of previously untreated metastatic colorectal cancer served 15.9 months versus 13.3 months, favoring the irinote- as the basis for U.S. and European approval of irinote- can-containing combinations (stratified-by-study p = can/5-FU/LV for this indication. An overview of the find- .003). Patients in study 1 had a 36% lower risk of ings of these two pivotal studies provides insights tumor progression and a 20% lower risk of death with regarding the application of this new combination in the irinotecan combination than with 5-FU/LV alone; clinical practice. comparable risk reduction values in study 2 were 42% Methods. Patients were randomly assigned to receive and 23%. While grade 3 diarrhea and vomiting were 5-FU/LV, either alone, or with concurrent irinotecan. The more common with irinotecan/5-FU/LV, grade 4 neu- study conducted primarily in North America (study 1), tropenia, neutropenic fever, and mucositis were less employed bolus 5-FU/LV schedules, while the study per- common with irinotecan/5-FU/LV than with the Mayo formed primarily in Europe (study 2), employed infu- Clinic 5-FU/LV regimen. sional 5-FU/LV regimens. Major endpoints included Conclusion. The combination of irinotecan/5-FU/LV tumor response rate, time to tumor progression (TTP), is superior to 5-FU/LV alone as first-line therapy for overall survival, quality of life, and safety. patients with metastatic colorectal cancer, offering consis- Results. In study 1, the respective confirmed tently improved tumor control and prolonged survival. response rates for irinotecan/5-FU/LV versus 5-FU/LV Irinotecan-based combination therapy sets a new survival were 39% and 21% (p < .001); median TTPs were 7.0 standard for the treatment of this life-threatening disease. months and 4.3 months, respectively (p = .004). In The Oncologist 2001;6:81-91 Correspondence: Leonard B. Saltz, M.D., Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021, USA. Telephone: 212-639-2501; Fax: 212-794-7186; e-mail: saltzl@mskcc.org Accepted for publication December 5, 2000. ©AlphaMed Press 1083-7159/2001/$5.00/0 The Oncologist 2001;6:81-91 www.TheOncologist.com Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 82 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer INTRODUCTION PATIENTS AND METHODS Until recently, patients with colorectal cancer had few available chemotherapeutic options. In the absence of any Patient Demographics and Study Design Both studies were large multinational, multicenter, ran- more effective agent, 5-fluorouracil (5-FU) served as the domized, open-label, controlled trials with similar inclusion mainstay of treatment for patients with advanced colorectal and exclusion criteria. Male or female patients were eligible cancer for nearly 45 years [1]. A fluorinated pyrimidine, 5- for enrollment if they were aged 18 years or older (but not FU acts by inhibiting thymidylate synthase, an enzyme nec- more than 75 years for study 2), had a histologic diagnosis of essary for the production of thymidine nucleotides required colorectal cancer with unresectable measurable metastases for DNA synthesis. 5-FU is usually given in combination and a performance status of 0-2, and had adequate hemato- with leucovorin (LV), a biomodulating agent that increases logic, renal, and hepatic function. They could not have the binding of 5-FU to thymidylate synthase, thereby increas- received prior chemotherapy for metastatic disease, or 5-FU- ing the inhibition of DNA synthesis and enhancing the based adjuvant chemotherapy during the 12 months (study 1) antitumor effect of 5-FU. This approach has increased or six months (study 2) preceding study entry. Patients who response rates from 11% with 5-FU alone to 23% with 5- had previously undergone pelvic radiotherapy were excluded FU/LV but has provided no meaningful survival benefit from study 1, but could enter study 2. All patients provided (median survival 11.0 months with 5-FU alone versus 11.5 written informed consent before being enrolled. months with 5-FU/LV) [2]. Attempts to improve the efficacy In study 1, patients were prospectively stratified according of 5-FU by administration of protracted continuous infusions to age (<65 or ≥65 years), time from initial diagnosis (<6 or ≥6 also showed a significant increase in response rate to 22%, but months), performance status (0 or 1-2), and prior adjuvant 5-FU- again, without improvement in survival beyond one year [3]. based therapy (yes or no). The patients were then randomized to Irinotecan (CPT-11, CAMPTOSAR , Pharmacia one of three treatment regimens: irinotecan/bolus 5-FU/LV, Corporation; Peapack, NJ; CAMPTO , Aventis SA, Antony, bolus 5-FU/LV, or irinotecan alone. The treatment scheme for France), a topoisomerase I inhibitor, offers a mechanism of study 1 is depicted in Figure 1. In study 2, each study site chose action completely different from that of 5-FU in the treat- one of two infusional regimens for administration of 5-FU/LV ment of colorectal cancer. Irinotecan and its metabolites according to local clinical practice or preference, i.e., either bind to a complex of DNA and topoisomerase I (an enzyme required for unwinding of DNA during replication), induc- ARM ing DNA strand breaks and consequent tumor cell death CPT-11: 125 mg/m /wk x 4 wks, q 6 wks 5-FU: 500 mg/m /wk x 4 wks, q 6 wks n = 231 [4]. Irinotecan has shown antitumor activity in patients with LV: 20 mg/m /wk x 4 wks, q 6 wks colorectal cancer when administered alone as first-line ther- RAMDOMIZATION apy [5-8] or as second-line therapy after 5-FU failure [5, 7- 5-FU: 425 mg/m /d × 5 d, q 4 wks STUDY 1 n = 226 LV: 20 mg/m /d × 5 d, q 4 wks 9]. In two randomized phase III studies in patients who experienced failure of first-line 5-FU, irinotecan was com- pared with either best supportive care or with intensive 5-FU- C 2 CPT-11: 125 mg/m /wk × 4 wks, q 6 wks n = 226 based infusional therapy. Both studies showed a statistically significant survival benefit for patients treated with irinote- can [10, 11]. The activity of irinotecan against untreated ARM and 5-FU-resistant colorectal cancer led to studies of this 2 CPT-11: 80 mg/m /wk × 6 wks, q 7 wks A1* 5-FU: 2.3 gm/m /wk × 6 wks, q 7 wks agent in combination with 5-FU/LV as first-line therapy for n = 54 2 LV: 500 mg/m /wk × 6 wks, q 7 wks or this disease. CPT-11: 180 mg/m d1 q 2 wks A2 5-FU: 400 IV/600 CI mg/m d1, 2 q 2 wks n = 145 Recently Douillard et al. and Saltz et al. reported the LV: 200 mg/m d1, 2 q 2 wks RAMDOMIZATION results of two studies conducted in parallel at a combined total STUDY 2 of 154 institutions located primarily in North America and Europe [12, 13]. These studies evaluated irinotecan combined 5-FU: 2.6 gm/m /wk × 6 wks, q 7 wks B1* n = 43 LV: 500 mg/m /wk × 6 wks, q 7 wks with either bolus (study 1: Saltz) or infusional 5-FU/LV (study or 5-FU: 400 IV/600 CI mg/m d1, 2 q 2 wks B2 2: Douillard) in previously untreated patients with metastatic n = 143 LV: 200 mg/m d1, 2 q 2 wks *AIO colorectal cancer. Both studies demonstrated statistically sig- de Gramont CI = continous infusion nificant clinical benefits with the irinotecan/5-FU/LV combi- CPT-11 = irinotecan nations, including improved tumor control and prolonged survival. The major efficacy and safety results of these studies are summarized in this review. Figure 1. Treatment regimens for studies 1 and 2. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 83 the once-weekly regimen of the Arbeitsgemeinschaft Safety was evaluated in terms of adverse events, labora- Internische Onkologie (AIO) cooperative German group for tory abnormalities, and deaths. Adverse-event assessments oncology [14] (designated A1 and B1 in Fig. 1), or the every- and complete blood counts were performed weekly, and two-week regimen of de Gramont [15] (designated as A2 and chemistry studies were performed at the beginning of each B2 in Fig. 1). Once a site selected its preferred regimen (either chemotherapy cycle. AIO or de Gramont), patients at the site were randomized to receive that method of 5-FU/LV administration, with or with- STATISTICAL METHODS out irinotecan. After initial treatment, doses in all treatment In study 1, a sample size of 220 patients per treatment arms of both studies could be adjusted according to specified arm was considered necessary to detect a 40% improve- guidelines to accommodate individual patient tolerance to the ment in TTP (i.e., median TTP from five to seven months) study drugs [12, 13]. Treatment was continued until tumor in the irinotecan/5-FU/LV arm versus the 5-FU/LV arm progression, the occurrence of unacceptable toxicity, or the with a power of 85% (two-tailed unstratified log-rank test, patient withdrew consent. alpha of .05). In study 2, assuming a response rate of 35% In both studies, atropine was provided for the treatment for the 5-FU/LV arm and 50% for the irinotecan/5-FU/LV of cholinergic symptoms, and loperamide for treatment of arm, a total of 338 evaluable patients (169 per arm) was delayed diarrhea. Antiemetic agents were administered as considered necessary to provide 80% power to detect a sig- prophylaxis for nausea and/or vomiting. Granulocyte colony- nificant difference in response rate (two-tailed chi-square stimulating factor could be given for prolonged neutropenia test, alpha of .05). In the statistical comparison of irinote- or infectious complications during neutropenic episodes. can/5-FU/LV versus 5-FU/LV, results for patients who received regimens A1 and A2 were combined, as were EVALUATIONS results from patients treated with B1 and B2. The major clinical efficacy endpoints were similar in the For both studies, time-to-event endpoints (e.g., duration of two studies; both evaluated objective tumor response rates, response, TTP, survival) were analyzed by Kaplan-Meier time to tumor progression (TTP), and survival. In study 1, curves and unstratified log-rank tests. Response rates were tumor measurements were obtained every six weeks until compared using chi-square tests. The influence of stratification week 24, and then every 12 weeks until the end of study treat- factors and other baseline characteristics on confirmed objec- ment. In study 2, tumor measurements were obtained after tive tumor response rates were assessed using multiple logistic each chemotherapy cycle (every six to seven weeks) and at the regression modeling. Similarly, TTP and survival were end of study treatment. Objective tumor responses were con- assessed using Cox proportional hazard regression modeling. firmed at least four weeks after the first documentation of Changes in QOL subscale scores were assessed using analysis response. In both studies, quality of life (QOL) was measured of variance for repeated measures. In study 1, worst changes at the start of each treatment cycle using the European from baseline were compared with Student’s t-tests. Organization for the Research and Treatment of Cancer Irinotecan alone was included as a third treatment arm in Quality of Life Questionnaire (EORTC QLQ-C30). Following study 1 only to document the efficacy and safety associated completion of study drug treatment, data were collected on with its first-line single-agent use; consequently no hypotheses post-study treatment for colorectal cancer and on survival. regarding its comparative efficacy were tested. Table 1. Patient disposition Study 1 Study 2 Irinotecan Irinotecan Patients 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B a a Randomized (n) 231 226 226 199 188 Untreated (n)4 8 4 1 1 a a Full-analysis (n) —— — 198 187 b, c d e b Treated with a different arm (n)2 1 1 01 f f f f f As-treated (n) 225 219 223 199 186 a b c Population analyzed for efficacy; Randomized to irinotecan/5-FU/LV but received 5-FU/LV; Randomized to irinotecan/5-FU/LV but received d e f irinotecan alone; Randomized to 5-FU/LV but received irinotecan alone; Randomized to irinotecan alone but received 5-FU/LV; Population analyzed for safety Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 84 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Table 2. Baseline patient characteristics Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B Characteristic (n = 231) (n = 226) (n = 226) (n = 198) (n = 187) Age (years) Median 62 61 61 62 59 (Range) (25-85) (19-85) (30-87) (27-75) (24-75) Sex (%) b c Male 65 54 64 67 53 Female 34 45 35 33 47 Performance Status (%) 039 41 46 5151 146 45 46 4241 215 13 8 7 8 Site, primary tumor (%) Colon 81 85 84 55 65 Rectum 17 14 15 45 35 No. of involved organ sites (%) 164 66 62 6263 226 23 28 2328 >2 10 10 9 15 9 Prior adjuvant 5-FU (%) 11 8 10 26 24 Prior radiotherapy (%) Any 3 2 1 20 16 Pelvis/abdomen 2 1 1 –– Other sites 1 1 0 –– Baseline laboratory abnormalities (%) CEA ≥100 ng/ml 40 39 37 35 32 Hemoglobin <11 g/dl 26 25 26 16 21 3 3 WBC ≥8 × 10 /mm 52 53 51 47 38 LDH >UNL 60 56 53 43 45 Total bilirubin >UNL 7 4 10 7 7 Abbreviations: CEA = carcinoembryonic antigen; LDH = lactate dehydrogenase; UNL = upper normal limit; WBC = white blood cell count a b c Data not available for some patients who were randomized but not treated; p = .019 Arm A versus Arm B; p = .006 Arm A versus Arm B; Percentage based on patients with baseline data RESULTS Colonic primary tumors predominated in both studies, as might be expected given the epidemiology of the disease. Patients The proportion of patients with rectal tumors in study 1 The studies enrolled a combined total of 1,070 patients (~15%) was lower than that in study 2 (35% to 45%), pre- treated at 154 multinational sites. A small number of patients sumably because patients with prior pelvic irradiation were were untreated or received a treatment regimen to which they excluded from study 1 but not from study 2. Because most were not originally randomized. Table 1 summarizes the patients had metastatic disease at initial diagnosis, relatively populations for the two studies. few patients (about 10% in study 1 and 25% in study 2) had Baseline characteristics are summarized in Table 2. Across received prior adjuvant 5-FU therapy. all treatment arms, the populations were generally divided sim- With respect to baseline laboratory values, carcinoem- ilarly between patients with a performance status of 0 and those bryonic antigen (CEA) and total bilirubin levels were gener- with a performance status of 1 or 2. However, the proportions ally similar across the two studies. However, there were of patients with a performance status of 2 in the irinotecan/5- notably greater proportions of patients with a depressed FU/LV and 5-FU/LV groups of study 1 (15% and 13%, respec- hemoglobin level, elevated WBC, or abnormal serum lactate tively) were about twice those of patients with a performance dehydrogenase (LDH) in study 1 than in study 2. These lab- status of 2 in either arm of study 2 (7% and 8%, respectively). oratory abnormalities are consistent with the performance Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 85 Table 3. Efficacy results Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B Efficacy endpoint (n = 231) (n = 226) (n = 226) (n = 198) (n = 187) Overall objective response rate (%) 50 28 29 49% 31% p < .0001 p < .001 Confirmed tumor response rate (%) 39 21 18 35 22 b b p < .0001 p < .005 CEA response rate (≥50% decrease) (%) 55 38 44 —— p < .001 — Median TTP (months) 7.0 4.3 4.2 6.7 4.4 c c p = .004 p < .001 Median survival (months) 14.8 12.6 12.0 17.4 14.1 c c p = .042 p = .032 Abbreviations: CEA = carcinoembryonic antigen; TTP = time to tumor progression a b c Confirmed ≥4 weeks after first evidence of objective response; Chi-square test; Unstratified log-rank test status findings, possibly indicating greater tumor burden or CEA during treatment) was evaluated in patients with ele- tumor-related organ dysfunction among the patients in study 1. vated baseline CEA values and at least one CEA assessment while on treatment. The CEA response rate with irinotecan/ EFFICACY 5-FU/LV was substantially higher than with 5-FU/LV; this Efficacy data for studies 1 and 2 are summarized in difference was significant (p < .001) (Table 3). Table 3. Remarkable consistency between the studies was observed when examining the efficacy outcome measures. TTP TTP was significantly longer in both studies for patients RESPONSE RATES who received irinotecan/5-FU/LV than for those who received In study 1, the overall response rate (response defined as 5-FU/LV (study 1: median 7.0 months versus 4.3 months, p = shrinkage of measured tumors by ≥50% on at least one occa- .004; study 2: median 6.7 months versus 4.4 months, p < .001). sion) was significantly higher for patients treated with irinote- As in the case of response rate, TTP was universally improved can/5-FU/LV than for those treated with 5-FU/LV (50% versus in all patient subgroups [16]. Kaplan-Meier TTP curves for 28%, p < .0001); comparable rates were also significantly both studies are shown in Figure 2. higher in study 2 (49% versus 31%, p < .001). The confirmed response rates (based on responses confirmed by imaging tests Survival ≥4 weeks later) were also significantly (p < .005 to p < .001) Comparison of survival with irinotecan/5-FU/LV versus and consistently higher among patients who received irinote- 5-FU/LV in study 1 showed a significant (p = .042) differ- can/5-FU/LV compared with those who received 5-FU/LV in ence favoring the irinotecan combination (Table 3); at any both studies (Table 3). The median duration of confirmed given time on study, patients treated with irinotecan/5-FU/LV objective tumor response from time of randomization was had a 19% reduction in the risk of death relative to those about nine months across all treatment arms. Examination of treated with 5-FU/LV alone (hazard ratio 0.81, 95% CI = confirmed response rates by subgroup analyses showed that, for 0.65-0.99). Similarly, in study 2, survival was significantly (p every subgroup evaluated, including patients with poor perfor- = .032) longer with irinotecan/5-FU/LV therapy than with 5- mance status, extensive metastatic disease, prior adjuvant ther- FU/LV; the risk of death at any time on study was decreased apy, or abnormal baseline laboratory values, the response rate by 23% with the irinotecan combination relative to 5-FU/LV with irinotecan/5-FU/LV was approximately double that with alone (hazard ratio 0.77, 95% CI = 0.60-0.98). The combined 5-FU/LV alone [16]. data from the two studies showed median survivals of 15.9 months for irinotecan/5-FU/LV versus 13.3 months for 5- CEA FU/LV alone; a stratified-by-study analysis indicated a sig- Serum CEA was assessed systematically during study nificant (p = .003) reduction in the risk of death at any time on 1. CEA response (≥50% decrease from baseline in serum study of 21% with the irinotecan/5-FU/LV arms relative to Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 86 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Study 1 Study 1 1.0 1.0 Irinotecan/5-FU/LV (n = 231) Irinotecan/5-FU/LV (n = 231) 0.9 0.9 5-FU/LV (n = 226) 5-FU/LV (n = 226) 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 p = .042* p = .004* 0.1 0.1 0.0 0.0 03 61 9 2 15 18 21 24 27 30 0 3 6 912 15 Months Months *Log-rank test *Log-rank test Study 2 Study 2 1.0 1.0 Irinotecan/5-FU/LV (n = 198) 0.9 Irinotecan/5-FU/LV (n = 198) 0.9 5-FU/LV (n = 187) 0.8 5-FU/LV (n = 187) 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 p < .001* p = .032* 0.1 0.1 0.0 0.0 03 61 9 2 15 18 21 24 27 30 0 6 912 15 Months Months *Log-rank test *Log-rank test Figure 2. TTP—Kaplan-Meier estimates from two phase III studies. Figure 3. Survival—Kaplan-Meier estimates from two phase III studies. the 5-FU/LV control arms (hazard ratio 0.67, 95% CI = 0.57- Efficacy results in this arm of the study appeared generally 0.79) [17]. Kaplan-Meier survival curves for studies 1 and 2 similar to those observed with 5-FU/LV alone (Table 3). are shown in Figure 3. It is apparent that the irinotecan/5-FU/LV curve sepa- Post-Study Treatment rates from the 5-FU/LV curve later in study 1 than in study Over 90% of the patients had follow-up information 2. This finding may possibly be explained by the greater regarding post-study anticancer therapy. In study 1, 52% of proportion of patients with performance status 2 in study 1; these patients in the irinotecan/5-FU/LV-treated arm received such patients generally had survival times <6 months with some form of second-line therapy, compared with 70% in the either therapy. If only patients with performance status 0-1 5-FU/LV-treated arm. For 56% of the patients in the 5-FU/LV are considered, survival results for the irinotecan/5-FU/LV arm, this second-line regimen consisted of irinotecan as a sin- arms are quite similar (medians of 17.2 months and 17.4 gle agent or in combination. In study 2, 49% of patients in the months in the 195 patients of study 1 and the 185 patients of irinotecan/5-FU/LV arm received post-study therapy com- study 2, respectively) and continue to contrast with the 5- pared with 65% of those in the 5-FU/LV arm. In this study, FU/LV arms (medians of 13.8 months and 15.5 months in 34% of the 5-FU/LV-treated patients given post-study the 195 patients of study 1 and the 173 patients of study 2, chemotherapy received second-line irinotecan (Table 4). respectively). Proportional Hazards Modeling Single-Agent Irinotecan Results In study 1, Cox regression techniques were used to evalu- A single-agent irinotecan arm was included in study 1 ate the effects of the study treatments on TTP and overall sur- only to document the activity of this agent as monotherapy. vival as influenced by the four predefined stratification factors Probability Probability Probability Probability Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 87 Table 4. Post-study therapy Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B b c Efficacy endpoint (n = 205) (n = 203) (n = 195) (n = 167) (n = 171) Total patients with post-study therapy 52% 70% 79% 49% 65% Irinotecan-based 1% 38% 3% 2% 28% Irinotecan/5-FU-based 13% 18% 9% 4% 6% 5-FU-based 30% 10% 64% 32% 21% Other therapy 8% 4% 3% 11% 10% a b Based on patients with follow-up information (approximately 90% of patients); Altogether, 56% of patients in this arm ultimately received a second-line regimen containing irinotecan as a single agent or in combination; Altogether, 34% of patients in this arm ultimately received a second-line regimen containing irinotecan as a single agent or in combination. Table 5. Cox regression results TTP Survival Hazard Hazard Factors Values Ratio 95% CI p Ratio 95% CI p Study 1 Serum LDH ≤UNL versus >UNL 0.60 0.47-0.76 .0001 0.47 0.37-0.60 .0001 Performance status 0 versus ≥1 0.74 0.59-0.93 .0088 0.57 0.45-0.71 .0001 No. of organ sites 1 versus ≥2 sites 0.63 0.50-0.80 .0001 0.67 0.54-0.84 .0004 Bilirubin ≤UNL versus >UNL 0.56 0.35-0.89 .0132 0.55 0.35-0.86 .0051 3 3 WBC <8 versus ≥8 × 10 /mm —— — 0.64 0.51-0.80 .0001 Hemoglobin ≥11 versus <11 0.74 0.58-0.95 .0157 —— — Age ≥65 versus <65 0.78 0.63-0.98 .0315 —— — Treatment Irinotecan/5-FU/LV versus 5-FU/LV 0.64 0.51-0.79 .0001 0.80 0.64-0.99 .0372 Study 2 Serum LDH ≤UNL versus >UNL 0.61 0.46-0.80 .0012 0.55 0.42-0.72 .0001 Performance status 0 versus ≥1 —— — 0.52 0.41-0.67 .0001 Time from metastatic ≥1 mo versus <1 mo 0.62 0.48-0.80 .0003 0.63 0.49-0.82 .0005 diagnosis No. of organ sites 1 versus ≥2 sites 0.71 0.55-0.91 .0070 0.73 0.57-0.94 .0127 Treatment Irinotecan/5-FU/LV versus 5-FU/LV 0.58 0.45-0.75 .0001 0.77 0.61-0.98 .0365 Abbreviations: CI = confidence interval; LDH = lactate dehydrogenase; TTP = time to tumor progression; UNL = upper normal limit; WBC = white blood cell count (age, performance status, time from initial diagnosis, and prior both improved TTP and survival in study 1 were normal LDH, adjuvant therapy), as well as other baseline factors prospec- good performance status, fewer involved organ sites, and tively considered to have potential prognostic significance normal bilirubin. Higher hemoglobin level and normal (e.g., number of involved organ sites, liver involvement, serum WBC were significantly predictive of longer TTP and sur- lactic dehydrogenase [LDH], hemoglobin level) [10, 11, 18- vival time, respectively. Surprisingly, older age also 20]. As shown in Table 5, factors significantly predictive for appeared to be significantly associated with a longer TTP. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 88 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer Treatment with combination irinotecan/5-FU/LV remained a SAFETY significant independent predictor of longer TTP (p < .001) The most frequently occurring adverse effects noted in and survival (p = .037) when other significant baseline studies 1 and 2 are shown in Table 6. For study 2, this review patient characteristics were taken into account. In this focuses on the safety data from the FDA-approved de adjusted analysis, irinotecan/5-FU/LV combination therapy Gramont irinotecan/5-FU/LV regimen and the corresponding was associated with a 36% lower risk of tumor progression 5-FU/LV control. In both studies, approximately 23% of and a 20% lower risk of death, than with 5-FU/LV therapy. patients treated with the combination irinotecan/5-FU/LV These findings were corroborated by application of the regimens experienced grade 3-4 diarrhea compared with model developed in study 1 to the data from study 2 (Table 5). approximately 10% to 14% of patients receiving 5-FU/LV In this study, normal baseline serum LDH and fewer involved alone. This difference was primarily in the incidence of grade organs were found to be significant predictors for improved 3 diarrhea; grade 4 diarrhea was comparably infrequent in the TTP and survival; better performance status was also signifi- treatment and control arms of the two trials; for example, in cantly associated with longer survival. In this study, a longer study 1, the incidence was 8% in the irinotecan/5-FU/LV arm time from diagnosis of metastatic disease to randomization and 7% in the 5-FU/LV arm. As expected, grade 3-4 vomit- was predictive for better outcomes with respect to both TTP ing was somewhat more common with irinotecan-based ther- and survival. As in study 1, treatment with irinotecan/5- apy but occurred in <10% of patients in any of the FU/LV in study 2 remained a significant predictor of improved combination arms. survival and was associated with a 42% lower risk of tumor Of note, grade 3-4 mucositis was quite infrequent with progression and a 23% lower risk of death relative to treatment irinotecan-based therapy, occurring in <4% of patients receiv- with 5-FU/LV when adjusted for baseline prognostic factors. ing combination therapies. By contrast, the Mayo Clinic Table 6. Incidence of adverse events Study 1 Study 2 Irinotecan Irinotecan 5-FU/LV 5-FU/LV Irinotecan 5-FU/LV 5-FU/LV Arm A Arm B Arm C Arm A Arm B a a Adverse event (n = 225) (n = 219) (n = 223) (n = 145) (n = 143) Diarrhea (%) Grade 3/4 23 13 31 14 6 Grade 3 15 6 18 10 4 Grade 4 8 7 13 4 2 Vomiting (%) Grade 3/4 10 4 12 4 2 Grade 3 5 3 6 3 1 Grade 4 4 1 6 1 1 Mucositis (%) Grade 3/4 2 17 2 4 3 Grade 3 2 15 2 4 3 Grade 4 0 2 0 0 0 Neutropenia (%) Grade 3/4 54 67 31 46 14 Grade 3 30 24 19 36 13 Grade 4 24 43 12 10 1 Neutropenic complications (%) Neutropenic fever 7 15 6 3 1 Neutropenic infection 2 0 2 2 0 Discontinuations due to adverse events and drug-related deaths (%) Discontinuations 8 6 12 6 1 Drug-related deaths 1 1 1 1 0 de Gramont regimen only Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 89 schedule of 5-FU/LV used in the control arm of study 1 and metastatic colorectal cancer. The studies were similar in that commonly employed as first-line therapy in North America they enrolled analogous patient populations, evaluated similar was associated with a much higher frequency of severe, grade efficacy and safety endpoints, and applied standardized meth- 3-4 mucositis (17%). ods of analysis. The trials were complementary in that they For neutropenia, only grade 4 events are usually associ- assessed the use of irinotecan in combination with two differ- ated with clinical consequences. Of interest, the frequency of ent methods of 5-FU/LV administration (bolus and infu- grade 4 neutropenia with combination irinotecan/5-FU/LV sional therapy). Study 1 compared a new combination of therapy (24%) in study 1 was essentially half that observed in irinotecan/bolus 5-FU/LV with that of the Mayo Clinic reg- patients receiving 5-FU/LV in the control group (43%) and imen of bolus 5-FU/LV that had been most widely employed proportionately fewer patients experienced neutropenic fever in North America. Study 2 symmetrically determined the when contrasting irinotecan/5-FU/LV (7%) with 5-FU/LV therapeutic ratio associated with adding irinotecan to two (15%). In study 2, where schedules of chemotherapy admin- different infusional 5-FU/LV regimens that were widely used istration were similar, grade 4 neutropenia was more fre- in Europe. Because the comparator regimens have been com- quently seen when irinotecan was added to 5-FU/LV (10%) monly used in oncology practices worldwide, the two studies than with 5-FU/LV alone (1%). The incidence of neutropenic provide clinicians with insights into the efficacy and safety of fever was also higher in the irinotecan/5-FU/LV arm (5.0%) the new combinations relative to familiar standards. than in the 5-FU/LV arm (1%); however, these rates of neu- The efficacy results of the two studies were remarkably tropenic complications are quite low when contrasted with consistent and show that the combination of irinotecan with 5- many other chemotherapy regimens [21]. FU provides patients with significant reductions in tumor size Discontinuations due to adverse events were acceptably in conjunction with prolonged suppression of tumor growth. low across all arms of both studies. The incidence of treat- The confirmed objective response rates with the irinotecan/5- ment-related death ranged from 0% to 1% in all treatment FU/LV arms (39% and 35%) were 1.5 to 2 times those in the groups. 5-FU/LV arms (21% and 22%), and these differences were statistically significant. In studies 1 and 2, respectively, TTP QOL was significantly improved with combination treatment (medi- The primary repeated measures analyses of variance of ans, 7.0 and 6.7 months) relative to treatment with 5-FU/LV changes from baseline in QOL in both studies showed no sig- alone (medians, 4.3 and 4.4 months). Assessment of response nificant differences between the treatment arms. In study 1, rates and TTP across demographic and disease-related sub- results were significantly better with irinotecan/5-FU/LV than groups showed improvements with irinotecan-based combina- with 5-FU/LV when comparing worst changes from baseline tion therapy in all prospectively defined subgroups [16]. These for the subscales of role functioning, fatigue, appetite loss, results indicate that irinotecan/5-FU/LV has the potential to and pain [22]. In study 2, deterioration of performance status offer better tumor control to all patients who are eligible for occurred later in patients treated with the irinotecan combina- first-line combination chemotherapy. tion than in those treated with 5-FU/LV alone (median 11.2 The most important finding of the studies was that first- months versus 9.9 months, p = .046). line irinotecan/5-FU/LV combination treatment provided a statistically significant survival advantage. It is particularly DISCUSSION noteworthy that this advantage occurred even though most Attempts to improve outcome in patients with metasta- control patients received second-line irinotecan therapy after tic colorectal cancer with modified 5-FU-based regimens, on-study failure of 5-FU/LV. These results suggest that early while generally resulting in higher response rates, have combination irinotecan/5-FU/LV may be superior to sequen- failed to significantly improve survival [3, 4]. However, the tial administration of first-line 5-FU followed by second-line results of recent studies of the topoisomerase I inhibitor irinotecan. The combination therapy survival outcomes noted irinotecan have improved the outlook for patients with this dis- among patients with good performance status accentuate this ease. In these studies, irinotecan was shown to improve sur- observation [16]. Such findings were presaged by the results vival, first as single-agent second-line therapy [10, 11], and of a prior study in metastatic colorectal cancer in which early now as a component of first-line treatment with 5-FU/LV administration of chemotherapy before symptom develop- [12, 13, 17]. ment was compared with delayed treatment after symptom The two phase III randomized multicenter, multinational, development; as in the current experience, early therapy in controlled studies reviewed here compared the efficacy and that study resulted in significantly better survival [23]. safety of new combinations of irinotecan/5-FU/LV to that of Cox regression modeling, which assessed treatment traditional regimens of 5-FU/LV given as first-line therapy of effect adjusted for significant prognostic factors, provided Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 90 Irinotecan/5-FU/LV in Metastatic Colorectal Cancer further evidence that combination therapy with irinotecan/5- and now as a component of first-line combination treatment. FU improves TTP and survival. This analysis indicated that The studies discussed here are the first trials to document irinotecan/5-FU/LV treatment resulted in an approximate that the combination of a new agent with 5-FU/LV can 40% reduction in the relative risk of tumor progression and safely benefit patients with metastatic colorectal cancer by a 20% decrease in the relative risk of death. inducing tumor shrinkage, extending tumor control, and sig- In both studies, gastrointestinal toxicities were more com- nificantly prolonging life without an impairment of QOL. mon with combination treatment, but grade 4 diarrhea— The strength and consistency of the data from these large largely defined by the need for hospitalization for supportive complementary studies led the FDA’s Oncology Drug care—was infrequent (<8%). In study 1, grade 4 neutropenia, Advisory Committee (ODAC) to conclude that irinotecan/5- neutropenic fever, and mucositis were observed less often with FU/LV should be the reference standard against which weekly irinotecan/5-FU/LV than with Mayo Clinic bolus 5- future first-line therapies for metastatic colorectal cancer are FU/LV alone; data from other studies that compared weekly compared. versus monthly 5-FU/LV therapy [24, 25] suggest that this Given the positive findings in patients with metastatic reduced toxicity with combination therapy is most likely due disease, several trials in the U.S. and Europe are comparing to the differences in 5-FU/LV scheduling between the arms. In irinotecan/5-FU/LV with 5-FU/LV as adjuvant therapy for study 2, clinically relevant grade 3 and 4 events, e.g., vomit- patients with stage III colon cancer; it is hoped that the ing, mucositis, and neutropenic fever, were infrequent with the addition of irinotecan to 5-FU/LV will offer an increased de Gramont irinotecan/5-FU/LV regimen. The safety findings opportunity for cure in patients with early-stage disease. were supported by the results of the QOL analyses, which showed that administration of irinotecan in combination with ACKNOWLEDGMENTS 5-FU/LV did not result in significant worsening of QOL. The work described in this publication was supported Attempts to improve outcomes in patients with metasta- by grants from Pharmacia Corporation, Peapack, NJ, and tic colorectal cancer with the limited tools of 5-FU and LV Aventis SA, Antony, France. have been the source of decades of frustration and disap- L.S. and J-Y.D. have received major research support and pointment. Irinotecan has now broken the barrier to honoraria as consultants for both Pharmacia Corporation and improved survival, first as single-agent second-line therapy, Aventis Corporation. REFERENCES 1 Mayer RJ. Moving beyond fluorouracil for colorectal cancer. 8 Rougier P, Bugat R, Douillard JY et al. Phase II study of N Engl J Med 2000;343:963-964. irinotecan in the treatment of advanced colorectal cancer in chemotherapy-naive patients and patients pretreated with fluo- 2 Advanced Colorectal Cancer Meta-analysis Project. Modulation rouracil-based chemotherapy. J Clin Oncol 1997;15:251-260. of fluorouracil by leucovorin in patients with advanced colorec- 9 Rothenberg ML, Cox JV, DeVore RF et al. A multicenter, tal cancer: evidence in terms of response rate. J Clin Oncol phase II trial of weekly irinotecan (CPT-11) in patients with pre- 1992;10:896-903. viously treated colorectal carcinoma. Cancer 1999;85:786-795. 3 Meta-analysis Group in Cancer. Efficacy of IV continuous 10 Cunningham D, Pyrhonen S, James RD et al. Randomised infusion of fluorouracil compared with bolus administration trial of irinotecan plus supportive care versus supportive care in advanced colorectal cancer. J Clin Oncol 1998;16:301-308. alone after fluorouracil failure for patients with metastatic 4 Pommier Y, Tanizawa A, Kohn KW. Mechanisms of topoiso- colorectal cancer. Lancet 1998;352:1413-1418. merase I inhibition by anticancer drugs. In: Liu LF, ed. Advances 11 Rougier P, Van Cutsem E, Bajetta E et al. Randomized trial in Pharmacology. New York: Academic Press, 1994;29B:73-92. of irinotecan versus fluorouracil by continuous infusion after 5 Shimada Y, Yoshino M, Wakui A et al. Phase II study of CPT- fluorouracil failure in patients with metastatic colorectal can- 11, a new camptothecin derivative, in metastatic colorectal can- cer. Lancet 1998;352:1407-1412. cer. CPT-11 Gastrointestinal Cancer Study Group. J Clin Oncol 12 Saltz LB, Cox JV, Blanke C et al. Irinotecan plus fluorouracil 1993;11:909-913. and leucovorin for metastatic colorectal cancer. N Engl J Med 6 Conti JA, Kemeny NE, Saltz LB et al. Irinotecan is an active 2000;343:905-914. agent in untreated patients with metastatic colorectal cancer. 13 Douillard JY, Cunningham D, Roth AD et al. Irinotecan com- J Clin Oncol 1996;14:709-715. bined with fluorouracil compared with fluorouracil alone as 7 Pitot HC, Wender DB, O’Connell MJ et al. Phase II trial of first-line treatment for metastatic colorectal cancer: a multicen- irinotecan in patients with metastatic colorectal carcinoma. tre randomised trial. Lancet 2000;355:1041-1047 [Erratum, J Clin Oncol 1997;15:2910-2919. Lancet 2000;355:1372]. Downloaded from https://academic.oup.com/oncolo/article/6/1/81/6387678 by DeepDyve user on 31 January 2022 Saltz, Douillard, Pirotta et al. 91 14 Köhne CH, Schöffski P, Wilke H et al. Effective biomodula- metastases from colorectal cancer. Br J Surg 1995;82:1397- tion by leucovorin of high-dose infusion fluorouracil given as 1400. a weekly 24-hour infusion: results of a randomized trial in 20 Miller LL, Petit RG, Elfring GL. Evaluation of carcinoembry- patients with advanced colorectal cancer. J Clin Oncol onic antigen (CEA) levels during CPT-11 treatment of patients 1998;16:418-426. with previously treated colorectal cancer. Proc Am Soc Clin Oncol 1998;17:297a. 15 de Gramont A, Bosset JF, Milan C et al. Randomized trial comparing monthly low-dose leucovorin and fluorouracil 21 American Society of Clinical Oncology. American Society of bolus with bimonthly high-dose leucovorin and fluorouracil Clinical Oncology recommendations for the use of hematopoi- bolus plus continuous infusion for advanced colorectal cancer: etic colony-stimulating factors: evidence-based clinical practice a French intergroup study. J Clin Oncol 1997;15:808-815. guidelines. J Clin Oncol 1994;12:2471-2508. 16 Knight RD, Miller LL, Pirotta N et al. First-line irinotecan 22 Locker PK, Miller LL, Pirotta N et al. Weekly irinotecan (C) (C), fluorouracil (F), leucovorin (L) especially improves sur- combined with 5-fluorouracil (F) and leucovorin (L) in patients vival (OS) in metastatic colorectal cancer (MCRC) patients with untreated metastatic colorectal cancer (MCRC): impact on (PT) with favorable prognostic indicators. Proc Am Soc Clin quality of life. Proc Am Soc Clin Oncol 2000;19:620a. Oncol 2000;19:991a. 23 Nordic Gastrointestinal Tumor Adjuvant Therapy Group. 17 Saltz LB, Douillard J, Pirotta N et al. Combined analysis of Expectancy or primary chemotherapy in patients with advanced two phase III randomized trials comparing irinotecan (C), asymptomatic colorectal cancer: a randomized trial. J Clin fluorouracil (F), leucovorin (L) versus F alone as first-line Oncol 1992;10:904-911. therapy of previously untreated metastatic colorectal cancer 24 Buroker TR, O’Connell MJ, Wieand HS et al. Randomized (MCRC). Proc Am Soc Clin Oncol 2000;19:242a. comparison of two schedules of fluorouracil and leucovorin 18 Kemeny N, Niedzwiecki D, Shurgot B et al. Prognostic vari- in the treatment of advanced colorectal cancer. J Clin Oncol ables in patients with hepatic metastases from colorectal can- 1994;12:14-20. cer. Importance of medical assessment of liver involvement. 25 Leichman CG, Fleming TR, Muggia FM et al. Phase II study Cancer 1989;63:742-747. of fluorouracil and its modulation in advanced colorectal can- 19 Rougier P, Milan C, Lazorthes F et al. Prospective study of cer: a Southwest Oncology Group Study. J Clin Oncol prognostic factors in patients with unresected hepatic 1995;13:1303-1311. Related articles in The Oncologist: Irinotecan: A New Agent Comes of Age. Saltz LB. The Oncologist 2001;6:65. Irinotecan (CPT-11): Recent Developments and Future Directions–Colorectal Cancer and Beyond. Rothenberg ML. The Oncologist 2001;6:66-80.

Journal

The OncologistOxford University Press

Published: Feb 1, 2001

There are no references for this article.