Presented in part at the 43rd Annual Meeting of the American Society of Clinical Oncology, Chicago, Illinois, June 1-5, 2007.
Irinotecan-based chemotherapy regimens are 1 option for treatment of metastatic colorectal cancer (mCRC). The authors report the safety and efficacy of such regimens in elderly patients using a large phase III trial (bolus, infusional, or capecitabine with camptostar-celecoxib &#91;BICC-C&#93;) cohort.
In period 1, 430 previously untreated patients with mCRC were randomized in a 3-by-2 design to receive irinotecan plus infusional 5-fluorouracil, and leucovorin (FOLFIRI), irinotecan plus bolus 5-fluorouracil/leucovorin (mIFL), and irinotecan plus oral capecitabine (CapeIRI). In period 2, an additional 117 patients were randomized to receive FOLFIRI or mIFL and bevacizumab. In both periods patients were also randomized to a double-blind treatment with celecoxib or placebo. A secondary analysis was conducted examining the safety and efficacy of these regimens in elderly (age >70 years) versus nonelderly (age ≤70 years) patients.
In period 1, 19.5% of patients were elderly, compared with 24.8% in period 2. Rates of grade 3 and higher toxicity did not differ significantly between age groups in either period by treatment arm, with the exception of asthenia in the FOLFIRI and CapeIRI arms (P = .05 and P = .03, respectively) and dehydration in the CapeIRI arm in period 1 (P = .02). Overall progression-free survival for FOLFIRI in both periods was not statistically different by age. Objective responses and overall survival did not differ by patient age within treatment arms and periods.
An estimated 153,000 persons were diagnosed with colorectal cancer and more than 50,000 died from colorectal cancer (CRC) in the United States in 2007.1 Worldwide, over 1 million individuals are diagnosed with CRC and just over 500,000 die from CRC annually.1, 2 Elderly individuals are a rapidly growing segment of the US population. In fact, individuals aged 70 years and over will comprise 18% of the United States population by the year 2030.3 They are disproportionately affected by colorectal cancer. By 2025, approximately 75% of new cases and 85% of cancer deaths from colorectal cancer will occur in patients older than 65 years.4 The median ages at colorectal cancer diagnosis and death are 71 and 75 years, respectively.1 In the United States, the expected median life expectancy1 for a 70- to 75-year-old without metastatic cancer ranges from 9 to 16 years for men, and from 9 to 12 years for women.5 Unfortunately, there is a paucity of information concerning the factors that modify treatment effect, tolerance, and toxicity of elderly patients with gastrointestinal malignancies. Given the increasing number of older patients with colorectal cancer, and their expected life expectancy, identifying safe and efficacious therapies for elderly patients is imperative.
The treatment for metastatic colorectal cancer has changed rapidly in the past 10 years, with combination chemotherapy being standard first-line treatment for most patients. Although more aggressive therapy regimens have increased median overall survival, they have also increased the likelihood of side effects from treatment.6 There is a growing body of literature suggesting similar safety and efficacy of combination regimens for elderly patients with mCRC compared with nonelderly patients.7-16 However, the bulk of this literature is limited to oxaliplatin-based therapy. As irinotecan-based regimens appear to have similar efficacy overall compared with oxaliplatin-based regimens when the fluoropyrimidine backbone is similar,17-23 we sought to test if efficacy and safety differed by age for irinotecan-based regimens.
The bolus, infusional, or capecitabine with camptostar-celecoxib (BICC-C) study was a large, multicenter, randomized phase III trial in previously untreated patients with metastatic colorectal cancer, comparing combination regimens of irinotecan and various fluoropyrimidine strategies.24 From that cohort, we now report on the safety and efficacy of irinotecan-based chemotherapy in first-line metastatic colorectal cancer patients aged >70 years (elderly) compared with those aged ≤70 years (nonelderly).
MATERIALS AND METHODS
Study Design and Patient Selection
The BICC-C study was initiated in 2003 in previously untreated patients with metastatic colorectal cancer. Full details of inclusion and exclusion criteria have been recently reported.24 Briefly, patients had histologically confirmed, metastatic adenocarcinoma of the colon or rectum with measurable disease; were ≥18 years old; had Eastern Cooperative Oncology Group (ECOG) performance status (PS) 0 to 1; and had adequate bone marrow, hepatic, and renal function. Prior adjuvant chemotherapy must have been completed at least 12 months before inclusion; any major surgery must have been completed ≥4 weeks before the first bevacizumab dose. Exclusion criteria included inadequately controlled hypertension, serious cardiovascular disease, Gilbert disease, active use of cyclooxygenase-2 inhibitors or nonsteroidal anti-inflammatory drugs, or chronic use of >325 mg/d of aspirin.
Chemotherapy Regimens and Randomization
Patients were randomized to 1 of the 3 open-label chemotherapy arms: infusional 5-flourouracil (5-FU), leucovorin, and irinotecan (FOLFIRI); bolus 5-FU, leucovorin, and irinotecan (mIFL); or oral capecitabine with irinotecan (CapeIRI), as well as to a double-blind treatment celecoxib or placebo (period 1) using a 3-by-2 factorial design (Fig. 1). In April 2004, after US Food and Drug Administration approval of bevacizumab, the trial was amended to compare FOLFIRI and bevacizumab (FOLFIRI + Bev) with mIFL and bevacizumab (mIFL + Bev) using a 2-by-2 factorial design (period 2); the CapeIRI arm was discontinued because of greater toxicity and limited safety data for the addition of bevacizumab to this arm (Fig. 1). Following the amendment, patients randomized to FOLFIRI or mIFL during period 1 who were still on study had the option of adding bevacizumab to their current regimen; 16 patients on the FOLFIRI arm and 7 patients on the mIFL arm added bevacizumab to their regimen. For the treatment trial, predefined stratification factors for randomization were age (≤70 or >70 years old), ECOG PS (0 or 1), and low-dose aspirin use (≤325 mg daily; yes or no); thus, there was balanced distribution by age to the treatment arms.
Patients were treated with chemotherapy (in conjunction with twice daily celecoxib or placebo) until progressive disease (PD), unacceptable toxicities caused by chemotherapy, or withdrawal of patient consent. Celecoxib or placebo administration for all patients remaining on study was permanently discontinued on January 19, 2005 after an announcement of the potential association of celecoxib with increased risk of major cardiovascular events in a long-term, colorectal adenoma prevention study (the adenoma prevention with celecoxib [APC] trial).25 Starting doses of all therapy were the same regardless of the age of the patient. Standard intra- and intercycle dose modifications for 5-FU, capecitabine, and irinotecan (based on the package insert for irinotecan26) were prescribed if grade 2 to 4 toxicities occurred (based on National Cancer Institute Common Toxicity Criteria, version 2.0). Leucovorin, bevacizumab, and celecoxib doses were not altered.
The protocol was reviewed and approved by the institutional review board of each participating institution. Each patient provided written informed consent. The study was sponsored by Pfizer Global Pharmaceuticals.
Baseline evaluation of patients before randomization and throughout the study has been recently reported.24 Baseline radiographic tumor assessments were performed within 21 days before the first study medication dose. During chemotherapy, follow-up radiology assessments were to be performed every 6 weeks until PD or on chemotherapy discontinuation. For the purpose of this study, PD was defined as any of the following criteria having been met: definition of PD based on the Response Evaluation Criteria in Solid Tumors (RECIST) guidelines, investigator's clinical judgment of PD, or death by any cause in the absence of previously documented PD. If a patient's disease was in response or stable at the time of treatment withdrawal, the patient was followed every 6 weeks until PD. After PD, the patient was followed every 3 months for survival. Tumor response classification was based on RECIST guidelines.27
In this secondary analysis of the data, we defined elderly as aged >70 years old at the time of randomization onto the treatment trial. Thus, 2 subgroups were designated: aged <70 years (nonelderly) and aged >70 years (elderly). The primary efficacy endpoint was progression-free survival (PFS), defined as the time from randomization to the first documentation of PD. For patients without documented PD, data were censored on the date of the last tumor assessment with nonprogression status, or, for patients who started a second-line therapy, at the date of the start of the new therapy. The primary efficacy analysis for this analysis compared PFS by age category for each study period. Secondary efficacy analyses included PFS, response rate, and overall survival within each treatment arm by age and by period. Overall survival was defined as the time from the date of randomization to the date of death; in the absence of confirmation of death, data were censored at the last date the patient was known to be alive.
The Kaplan-Meier method was used to calculate PFS and overall survival24; differences by age categories were tested by the log-rank test. The hazard ratio (HR) and 95% confidence interval (CI) for age comparisons were obtained from a Cox proportional hazards model, adjusted for ECOG performance status, low-dose aspirin use at randomization, and randomization to celecoxib treatment.28 All statistical tests were 2-sided, using a significance level of 5%. No adjustments for multiple comparisons were made.
Efficacy analyses included all patients randomized on an intent-to-treat basis. Safety analyses included all treated patients and were summarized descriptively. All efficacy and safety analyses reported were performed using a data collection cutoff date of November 15, 2007.
Patient Enrollment and Characteristics
During period 1 (February 2003 through March 2004), 430 patients were randomized to FOLFIRI, mIFL, or CapeIRI. Of these patients, 346 were aged ≤70 years, and 84 were aged >70 years (Fig. 2). During period 2 (April 2004 through December 2004), 117 patients were randomized to FOLFIRI/Bev or mIFL/Bev. Of these patients, 88 were aged ≤70 years, and 29 were aged >70 years. Baseline characteristics were fairly well balanced by age (Table 1). There are no statistically significant differences in the duration of treatment across age groups for either period 1 (P = .33) or period 2 (P = .69).
Mean duration of disease from initial diagnosis, mo (SE)
Prior therapy for CRC, No. (%)
Low-dose aspirin use, No. (%)
Rates of grade 3 or higher toxicity did not differ significantly between age groups when combining all treatment arms in period 1: leukopenia (27% for age >70 years, 21% for age ≤70 years, P = .29), diarrhea (30% vs 26%, P = .47), nausea (11% vs 12%, P = .82), deep vein thrombosis (7% vs 8%, P = .72), and fatigue (10% vs 11%, P = .81). Despite experiencing modestly higher grade 3 or higher neutropenia (46% for older patients vs 37% for younger patients, P = .13), rates of febrile neutropenia were comparable between older and younger patients (6% vs 8%, P = .52). Older patients experienced significantly more dehydration than their younger counterparts (18% vs 9%, P = .02) as well as asthenia (11% vs 3%, P = .001), although this was primarily driven by differences in the tolerance to CapeIRI (Table 2).
Table 2. Percentage of Common Grade 3-4 Adverse Events, by Treatment Arm (as Treated Population)
Similarly, rates of febrile neutropenia, diarrhea, nausea, vomiting, and deep venous thrombosis were similar within each treatment arm for elderly and nonelderly patients. Elderly patients did experience significantly more asthenia in the FOLFIRI arm (P = .05) and the CapeIRI arm (P = .03). Point estimates for bone marrow suppression and gastrointestinal complications suggest increased frequency in elderly patients, although these differences were not statistically significant (Table 2).
Older patients had similar rates of dose reductions (16% vs 17%), but more dose delays (41% vs 31%) and dose delays with reductions (24% vs 18%) than younger patients (Table 3).
P values are generated using a Cochran-Mantel-Hanszel test stratified by treatment arm and period. P<.05 indicates age group is statistically associated with dose reduction, dose delay, and dose reduction and delay, respectively.
Dose reduction and delay
As in period 1, older and younger patients experienced similar rates of grade 3 or higher leukopenia (24% for age >70 years, 20% for age ≤70 years, P = .62), diarrhea (10% vs 12%, P = .85), and nausea (7% vs 8%, P = .83) (Table 2). Elderly patients did experience a nearly 20% increased rate of neutropenia (55% vs 36%, P = .07), with more episodes of febrile neutropenia (7% vs <5%, P = .25).
Older patients experienced fewer reductions and delays in therapy than younger patients in period 2. None of the elderly patients experienced dose delays with reduction (0% vs 7%) (Table 3).
PFS did not differ significantly between age groups during the median follow-up time of 34 months (Fig. 3). Comparing subjects aged >70 years to those aged <70 years, the HR for PFS was 0.98 (95% CI, 0.74-1.29) (Table 4). Consistent with the overall results of the trial,24 within each age group, there was a trend toward a superior PFS for patients treated with FOLFIRI when compared with either mIFL or CapeIRI. Median PFS was similar for those who received FOLFIRI regardless of age category (7.9 months for age ≤70 years, 7.5 months for age >70 years; HR, 1.1; 95% CI, 0.69-1.76). In addition, no significant differences in median PFS were noted between age groups for mIFL (HR, 0.82; 95% CI, 0.51-1.32) or CapeIRI (HR, 0.98; 95% CI, 0.59-1.64).
Table 4. Progression-free Survival by Age Categories (Intent-to-treat Population)
Hazard ratio adjusted for Eastern Cooperative Oncology Group performance status (0 or 1), low-dose aspirin use at randomization (yes or no), and randomization to celecoxib treatment (celecoxib or placebo).
Objective response, as defined by RECIST criteria, did not differ significantly by age by treatment arm. In the FOLFIRI arm, nearly half of subjects aged ≤70 years (47%) had a response, compared with 50% of those aged >70 years. Rates of response were lower in the other treatment arms, but similar between the 2 age groups.
Median overall survival (OS) did not differ significantly between age groups in all treatments combined (19.0 months for age ≤70 years, 21.2 months for age >70 years) (Fig. 3). Although patients aged ≤70 years had a 4.1 month longer median OS time in the FOLFIRI arm, this difference was not statistically significant compared with those aged >70 years (Table 5).
Table 5. Overall Survival by Age Categories (Intent-to-treat Population)
Hazard ratio adjusted for Eastern Cooperative Oncology Group performance status (0 or 1), low-dose aspirin use at randomization (yes or no), and randomization to celecoxib treatment (celecoxib or placebo).
Only 29 of 117 patients were aged >70 in period 2; thus, statistical power was more limited. PFS did not differ significantly between age groups in all treatments combined once bevacizumab was added to FOLFIRI or mIFL during a median 23-month follow-up period (Fig. 3). Patients aged ≤70 years experienced a 3-month increase in median PFS compared with older patients (10.6 compared with 7.6 months, P = .14) (Table 4). As with period 1, response rate did not differ by age group between treatment arms (FOLFIRI/Bev: 58% age ≤70 years, 57% age >70 years; mIFL/Bev: 58% age ≤70 years, 40% age >70 years). Similarly, OS did not differ significantly between the 2 age groups for the overall cohort or within each treatment arm (Table 5, Fig. 3). Consistent with the overall results of the trial,24 within each age group, there was a trend toward a superior PFS and OS for patients treated with FOLFIRI/Bev when compared with mIFL/Bev.
In this secondary analysis of the BICC-C trial, elderly patients with metastatic colorectal cancer generally tolerated irinotecan and fluoropyrimidine combination regimens similarly to nonelderly patients, with the exception of modest increases in rates of neutropenia and asthenia. Furthermore, elderly patients derived similar benefit from combination therapy, with no significant differences in PFS and OS compared with nonelderly patients.
Irinotecan, a topoisomerase I inhibitor, is an active drug in metastatic colorectal cancer that is converted by the liver into 2 active metabolites, SN-38 and SN-38 glucuronide. Irinotecan monotherapy has been shown to be well tolerated in the general metastatic colorectal cancer population, inclusive of the elderly.29-31 Although plasma levels of its active metabolites may predict for toxicity, pharmacokinetic studies have revealed no significant differences in the plasma levels of these irinotecan metabolites in older and younger patients.32 Thus, the pharmacological properties of the drug would suggest a similar expected safety and efficacy profile in the elderly and nonelderly.
Our findings are congruent with the existing literature regarding irinotecan in elderly metastatic colorectal cancer patients. In a subset analysis of a multi-institutional phase II study of weekly single agent irinotecan, Rothenberg and colleagues found increased rates of grade III/IV diarrhea in patients aged 65 years or older compared with those younger than 65 years (39% vs 19%, P < .0008), although rates of first-course diarrhea were not statistically different (25% vs 15%, P = .1).33 In a trial comparing the weekly and the triweekly dosing regimens of irinotecan monotherapy as second-line therapy for metastatic disease, >⅓ of 291 patients were at least 70 years of age.31 Age >70 years did not affect survival or PFS, but was associated with an increased risk of grade III/IV neutropenia and grade III/IV diarrhea compared with patients <70 years of age. In contrast, Chau and colleagues found that elderly patients (at least 70 years old) who were treated on a trial of irinotecan monotherapy every 3 weeks (starting dose 350 mg/m2) had similar rates of grade III/IV toxicity compared with nonelderly patients (46% vs 38%, P = .2) and derived similar rates of objective response and OS.29
In regard to combination irinotecan-based regimens, Folprecht and colleagues recently published a pooled analysis of 599 patients aged ≥70 years who received fluorouracil-based therapy for metastatic colorectal cancer, 220 of whom also received irinotecan.16 This pooled analysis of 4 clinical trials noted similar tolerance of irinotecan/fluoropyrimidine combinations between older and younger subjects. With the exception of hepatic toxicity, older and younger subjects experienced similar rates of grade 3 and higher toxicity. PFS was increased by 1 month for older compared with younger subjects (9.2 vs 8.2 months, P = .04), although OS and response rate did not differ between age groups.
Mitry and colleagues recently reported interim results in abstract form of a phase III trial of infusional 5-FU and leucovorin or infusional 5-FU, leucovorin, and irinotecan in elderly patients (at least 75 years old) with metastatic colorectal cancer.34 The objective response rate was higher for those who received irinotecan and infusional 5-FU therapy (31%) compared with those treated with infusional 5-FU and leucovorin alone (18%). Rates of grade 3 or 4 neutropenia were lower in that trial (28%) compared with the BICC-C elderly population (54%), although rates of diarrhea were similar between the 2 trials (16% vs 11%).
The advantage of this analysis is that BICC-C was a large, phase III trial that stratified by age with protocol-specified collection of efficacy and toxicity. Furthermore, prior reports of irinotecan in the elderly have not included bevacizumab-containing regimens.35 In this analysis of the BICC-C trial, rates of dose reduction and delay for elderly and nonelderly patients were similar. This compares favorably to several studies.34, 36-38 The recently reported interim analysis reported by Mitry demonstrated that 67% (for combination with bolus FU) and 87% (for combination with continuous infusion FU) of patients at least 75 years old received ≥75% planned irinotecan doses during cycles 1 through 4.34 These results suggest that maintaining dose intensity is feasible in elderly patients.
Although our findings are supportive of aggressive irinotecan-based combination treatment of elderly patients with metastatic colorectal cancer, there are limitations to our analyses. First, the use of a population eligible for a clinical trial may limit generalizability to the entire metastatic colorectal cancer population. The present study is limited in representation of older patients in clinical trials. Only 20% (period 1) and 25% (period 2) of the study population was aged ≥70 years, in contrast to 50% of the population diagnosed with colorectal cancer annually in the United States (median age of diagnosis, 71 years).1 In addition to eligibility restrictions, disproportionately few elderly patients are enrolled in clinical trials, and trial populations are generally not representative of the typical older cancer patient.39-43 Possible methods to overcome this barrier in clinical trial enrollment for older patients is to 1) develop risk-stratified treatment protocols based on comprehensive geriatric assessment, inclusive of assessment of functional status, comorbid conditions, and polypharmacy; 2) develop less restrictive inclusion criteria for older patients to allow generalizability to the general cancer population; and 3) increase funding for additional studies evaluating the impact of the biology of aging, comorbid conditions, and social support of older cancer patients on treatment outcomes.44-50 A final limitation of the present analysis is that information on comorbidities was not collected as part of study enrollment, and thus adjustments for other concurrent medical problems was not possible in these analyses.
In conclusion, within the context of this multicenter randomized clinical trial, patients aged >70 years derive similar benefit from aggressive irinotecan-based combination chemotherapy as younger patients, with acceptable rates of toxicity. To continue developing treatment strategies for the growing population of older colorectal cancer patients, oncologists must continue to evaluate the safety and efficacy of novel treatment regimens in this unique population. Increasing enrollment of elderly patients in clinical trials is 1 way to accomplish this imperative and determine feasible treatment strategies.
We thank Dr. Charles Fuchs, Director of the Department of Gastrointestinal Oncology at the Dana-Farber Cancer Institute, for his contributions to the treatment trial and thoughtful suggestions on this article.
Conflict of Interest Disclosures
This study was supported by Pfizer Inc, New York, New York.
Xiaoxi Zhang is an employee of Pfizer.
José Barrueco is an employee of Pfizer.
Edith Mitchell receives research funds and speaker's honoraria from Pfizer.
John Marshall has financial interests in Pfizer, Roche, Sanofi, BMS, Amgen, and Genentech.