Cytotoxic Chemotherapy for Advanced Colorectal Cancer
Cytotoxic Chemotherapy for Advanced Colorectal Cancer
Several developments in the past few years have incrementally progressed the field and provided additional insights into the management of advanced colorectal cancer. This review discusses the components of current cytotoxic chemotherapy regimens for advanced colorectal cancer: fluorouracil (5-FU), capecitabine (Xeloda), irinotecan (Camptosar), and oxaliplatin (Eloxatin). The equivalence of several front-line regimens has provided opportunities for increased tailoring of therapies for individual patients. Preliminary data on pharmacogenomics provides hope that we will be able to better match patients with regimens and doses on the basis of individualized predictions of toxicity and response. The importance of second-line therapy in overall survival has again been highlighted; the best outcomes have occurred in patients treated with 5-FU, oxaliplatin, and irinotecan in combination with targeted therapies during the course of their disease. Elderly patients are no exception to this finding. Combination regimens and second-line therapy should be offered to elderly patients who have adequate performance status and no contraindicated comorbid conditions, without regard for their chronological age.
Colorectal cancer is the second leading cause of cancer death in the United States. Up to onethird of patients have metastatic disease at the time of initial presentation. Although metastatic colorectal cancer remains a terminal diagnosis for the majority of patients, recent advances have improved outcomes while providing more treatment options for clinicians. The median survival of patients with metastatic disease has gone from 12 to 24 months with increased utilization of combination chemotherapeutic regimens. These new developments have provided opportunities for increased tailoring of therapies to match the characteristics of individual cases and patients. Research presented over the past few years has changed our approach to several important considerations in the treatment of advanced colorectal cancer. Although much attention has been paid to the rapid developments in targeted agents, significant advances have been made in conventional chemotherapy as well. This review discusses these new developments in cytotoxic chemotherapy regimens and describes the changing approach to these treatments, with a separate emphasis on issues in the elderly patient. The addition of targeted agents to these regimens is discussed in an accompanying article by Drs. Marshall and Hwang. Fluorouracil/Leucovorin For many years fluorouracil (5-FU) was the sole chemotherapeutic agent for metastatic colon cancer. Fluorouracil and its metabolites have at least two mechanisms of action: binding thymidylate synthetase, which depletes uracil stores necessary for DNA replication, and incorporation into RNA, which subsequently disrupts protein synthesis. Recent data from the adjuvant setting suggest that below normal levels of thymidylate synthetase, as assessed by immunohistochemical studies, may be associated with prolonged survival. The addition of leucovorin to the regimen improves efficacy by increasing the amount of reduced folate cofactor needed to bind thymidylate synthetase. The addition of leucovorin increased the rate of response to 5-FU from 11% to 21% in metastatic colorectal cancer. Fluorouracil has traditionally been given by bolus infusion in the United States. However, several trials have demonstrated that infusional regimens offer greater benefit, possibly due to more prolonged inhibition of thymidylate synthetase. A meta-analysis of 1,219 patients in six randomized trials of 5-FU administered by bolus or continuous infusion demonstrated a higher response rate in patients who received the drug by continuous infusion (22% vs 14%). The sideeffect profile of continuous infusion therapy included less neutropenia and an increased incidence of hand-foot syndrome. The combination of leucovorin and infusional 5-FU has subsequently been shown to produce better outcomes than the infusional regimen alone. The most commonly used infusional regimens are the German AIO regimen of 5-FU (2,000 to 2,600 mg/m2) over 24 hours with leucovorin (500 mg/ m2) weekly for 6 weeks, followed by 1 or 2 weeks rest, and the de Gramont or LV5FU2 regimen consisting of leucovorin (200 mg/m2 IV over 2 hours) and 5-FU (bolus 400 mg/m2 and infusion 600 mg/m2) on days 1 and 2, repeated every 2 weeks. Capecitabine The oral fluoropyrimidine capecitabine (Xeloda) is a prodrug of 5-FU that has activity similar to that of infusional 5-FU. Two large randomized phase III studies demonstrated that capecitabine had a higher response rate (26% vs 17%) and a better safety profile than bolus 5-FU. Side effects are similar to those of infusional 5-FU, although the incidence of handfoot syndrome is higher in patients who receive capecitabine. Thymidine phosphorylase, one of the enzymes responsible for conversion of capecitabine to the active metabolite, is increased in tumor cells, suggesting that the oral prodrug has increased selectivity over infusional 5-FU. Capecitabine is being investigated as a substitute for 5-FU in most frontline regimens. Several recent studies have demonstrated the safety and tolerability of these combinations. Most studies have focused on combinations with oxaliplatin (Eloxatin), five of which were reported at the 2005 American Society of Clinical Oncology Annual Meeting (Table 1). These trials are briefly reviewed here. The CAPOX regimen was compared to FUFOX, a weekly infusional 5-FU plus oxaliplatin regimen. The toxicity profiles were similar, aside from a higher incidence of hand-foot syndrome with CAPOX. The response rates for CAPOX and FUFOX were similar (47% vs 49%, P = .70). Although the difference in progressionfree survival did not reach statistical significance (7.0 vs 8.0 months, P = .11), the trend favored the intravenous arm; the difference suggests caution when analyzing this study. The TREE-1 study randomized patients to receive modified FOLFOX6, bFOL, or CapeOx. With the approval of bevacizumab (Avastin), enrollment in TREE-1 was discontinued in favor of a similar design with bevacizumab added to each arm (TREE-2). The high rate of diarrhea and dehydration seen with the TREE-1 capecitabine dose of 1,000 mg/m2 twice daily led to a dose reduction in the TREE-2 trial. Bevacizumab added to the efficacy of all regimens, with a clear increase in response rates. The bFOL had the lowest response rate, and it was statistically inferior to the response rate seen with the FOLFOX6 regimen. Preliminary results from a third phase III trial of capecitabine/oxaliplatin and infusional 5-FU/oxaliplatin in a slightly different schedule produced comparable toxicity profiles and response rates in the two arms (47% vs 54%, respectively). Interim results were recently presented of another study comparing XELOX to the popular FOLFOX6 regimen after enrollment of 177 of 304 planned patients. This initial safety analysis demonstrated much lower rates of neutropenia in patients who received the XELOX regimen than in those who received FOLFOX6 at an oxaliplatin dose of 100 mg/m2. A strategy of prolonged continuous infusion 5-FU with oxaliplatin was compared to XELOX in a randomized phase II study. Survival data are still maturing, but preliminary results demonstrate similar response rates. Surgical resection of liver metastases was attempted more frequently in the XELOX arm (9.6% vs 4.1%). A report has been recently published of a more dose-intense capecitabine regimen in 89 patients. In this study, capecitabine doses of 3,500 mg/m2 were given on days 1-7 and 14-21, with oxaliplatin at 85 mg/m2 days 1 and 14, repeated every 4 weeks. Response rates were higher than those of the standard XELOX regimen (55% vs 42%), and toxicity profiles were similar. Median progression-free survival also favored the dose-intense regimen (10.5 vs 6.0 months, P < .01). Although capecitabine's potential as a substitute for infusional 5-FU was confirmed in these studies, larger phase III studies are needed to demonstrate equivalence to FOLFOX. A trial comparing FOLFOX plus bevacizumab to XELOX plus bevacizumab is ongoing and will address the interchangeability of capecitabine for 5-FU. Until these data are presented, combination capecitabine regimens should be reserved for patients understanding and willing to accept that there is a small potential risk of a difference in efficacy, for the convenience of an oral fluoropyrimidine. The principle of combining capecitabine with irinotecan (Camptosar) was proven by the combination's effectiveness in phase II studies. However, recent larger series combining capecitabine and irinotecan have been less successful. A large European Organisation for Research and Treatment of Cancer trial compared CAPE/IRI (capecitabine at 1,000 mg/m2 twice a day for days 1-14 and irinotecan at 250 mg/m2 day 1, on a 3-week cycle) to FOLFIRI (leucovorin at 200 mg/m2 day 1, 5-FU at 400 mg/m2 bolus, days 1 and 2, 600-mg infusion over 22 hours, days 1 and 2; irinotecan at 180 mg/m2 on day 1, on a 2-week cycle). The trial was stopped early owing to increased mortality in the CAPE/ IRI arm (6 deaths in 44 patients). The deaths were the result of severe diarrhea or thromboembolic events. Dose reduction was required in 61% of patients on the CAPE/IRI arm. Patients also were randomized to receive celecoxib (Celebrex) or placebo, although this did not have an apparent impact on the toxicity. These poor results were similar to those of a smaller phase II trial of the same regimen. In that small study, dose reductions were required in 80% of the patients owing to diarrhea and neutropenia. Further studies of this combination are planned with a reduced dose of irinotecan. An alternate regimen employs a 2-week cycle of irinotecan at 175 mg/m2 on day 1 and capecitabine at 1,000 mg/m2 bid days 2-8, with upfront dose reductions to 140 mg/m2 and 750 mg/m2 bid, respectively, for patients older than 65. The overall response rate was 50%. Grade 3 or 4 diarrhea or neutropenia occurred in less than 2% of cycles. This approach may be an alternative to the earlier, more toxic, regimens. Irinotecan Irinotecan has been used successfully as a single agent and in combination regimens. Irinotecan is a prodrug that is metabolized in the liver by uridine diphosphate glucuronosyl transferases (UGT) into SN-38, an active topoisomerase inhibitor. The toxicity profile is notable for significant diarrhea, which can be fatal in severe cases. Combination of irinotecan with bolus 5-FU and leucovorin in the IFL regimen resulted in high rates of toxicity. In the initial trial of IFL, grade 3 or 4 diarrhea occurred in 23% of patients. Efficacy was increased; the combination yielded a response rate of 39% and a median progression-free survival of 7.0 months. Several alternative combinations have been investigated in attempts to improve response rates and decrease toxicity. The AIO infusional 5-FU regimen was combined with irinotecan, yielding a response rate of 62% and time to progression of 8.5 months. However, the rate of grade 3 or 4 diarrhea remained high, at 36%. The FOLFIRI regimen had a similar response rate (56%) and median progression- free survival (11 months) but an improved toxicity profile; the rate of grade 3 or 4 diarrhea was only 14%. Given its high response rate and tolerability, this regimen is a reasonable first-line therapy for metastatic colorectal cancer. Several recent studies have investigated the importance of UGT promoter polymorphisms in predicting irinotecan-induced toxicity and response to therapy. In one study of a group of 81 patients receiving frontline irinotecan-containing regimens, dose-limiting toxic effects occurred in 18% of patients with mutations in UGT1A1. In contrast, none of those without the mutation discontinued the treatment because of toxicity. Response rates were lower in patients with the mutation (57% vs 43%). A similar study of 67 patients receiving irinotecan with capecitabine demonstrated that those with mutations in UGT1A7 had a higher response rate and lower toxicity rate than patients with a wild-type genotype.[ 20] The response rate was 85% (11 of 13 patients) in patients with mutations associated with lower UGT1A7 activity and 44% (19 of 43) in patients with other genotypes. Similarly, none of the patients with these mutations had grade 3 or 4 diarrhea or neutropenia, whereas 23% of patients with other genotypes experienced at least one of these toxic effects. Oxaliplatin As a single agent, oxaliplatin has minimal efficacy in metastatic colorectal cancer. In combination with 5-FU/leucovorin, however, oxaliplatin has significant synergy and yields high response rates. The most common oxaliplatin-containing combination regimen, FOLFOX, has undergone several iterations. The FOLFOX4 regimen uses bolus and continuous infusion 5-FU on days 1 and 2 with oxaliplatin at 85 mg/m2. This regimen has been shown to be superior to irinotecan with bolus 5-FU/leucovorin (IFL) in a large trial of 795 patients, which included a comparison of IFL to FOLFOX4 and found a higher response rate and overall survival duration with FOLFOX4 (45% vs 31%, 19.5 vs 15.0 months, respectively).[ 21] Although FOLFOX4 was better tolerated than IFL, 20% of the patients who received FOLFOX4 developed grade 3 neuropathy. FOLFOX6 increased the dose of oxaliplatin to 100 mg/m2 and attempted to improve patient convenience by removing the 5-FU bolus on the second day. This resulted in an increased rate of neuropathy, however, and the oxaliplatin dose was reduced back to 85 mg/m2 in the popular modified FOLFOX6 regimen. In frontline therapy, this regimen produced a response rate of 54% and a median progression-free survival duration of 8.0 months; it is now the most widely used regimen in the United States. The neuropathy associated with oxaliplatin is usually dose-dependent; most patients experience dose-limiting neuropathy after a cumulative dose of 700 to 800 mg/m2 (8 to 10 cycles). There is a subset of patients who develop early-onset neuropathy. Recent data suggest that a glutathione S-transferase polymorphism may predict this early toxicity. The 13% of patients with the GSTP1 I105V polymorphism were more likely to stop treatment because of neuropathy (23% vs 10%) and had a higher rate of early-onset neuropathy. These results await further validation before this approach can move into the clinic. The combination of irinotecan and oxaliplatin (IROX) has been investigated in the front-line setting in several studies, with disappointing results. When compared to FOLFOX4, IROX had a trend for lower response rate and shorter overall survival interval. A similar combination was recently compared to FOLFIRI, revealing a trend toward worse progression-free survival for IROX (8.7 vs 7.3 months, P = .15). Toxicity rates also were higher in the IROX arm. This combination does not appear to improve the results seen with FOLFOX or FOLFIRI and should not be used routinely in front-line therapy. Selecting a Combination Modified FOLFOX6 and FOLFIRI have shown similar efficacy and tolerability. The different side-effect profiles allow tailoring of front-line regimens to individual patients. Patients with preexisting neuropathies or underlying conditions that put them at risk for early neuropathy from oxaliplatin, such as poorly controlled diabetes, could receive FOLFIRI. Similarly, patients with symptoms of irritable bowel syndrome or poor tolerance of diarrhea may better tolerate modified FOLFOX6. It is our hope that pharmacogenomic studies, once validated, may help guide the selection and dosing of an appropriate regimen for each individual. The question of which combination regimen to use in the front-line setting was addressed by a trial that randomized patients to receive frontline FOLFOX6 or FOLFIRI, followed by planned treatment with the alternate regimen on disease progression.[ 18] Front-line response rates, progression-free survival, and overall survival were similar between the two arms. Several limitations of this study complicate interpretation, including lower rates of second-line therapy per protocol in the front-line FOLFIRI arm and a mismatch in the rates of subsequent metastasectomy in the two groups. Nevertheless, this long overall survival duration demonstrates the importance of treatment with an active second-line agent in combination with 5-FU/leucovorin. The most important determinant of overall survival duration appears to be the number of active agents that a patient receives. A retrospective study of seven phase III trials demonstrated that overall survival duration was greatest in patients who had received 5-FU/leucovorin, irinotecan, and oxaliplatin during the course of their treatment (Figure 1). A large study from the United Kingdom, MRC-FOCUS, evaluated the appropriate sequence of chemotherapy.[ 25] The trial has only been presented in preliminary form and full results are not available.Up-front combination therapy, when compared to infusional 5-FU followed by the addition of irinotecan or oxaliplatin on disease progression, did not show a significant difference in overall survival, despite a higher response rate in patients who initially received combination regimens. However, the median survivals for all arms were shorter than expected, and the relatively small number of patients receiving all three active agents may have played a significant role in these results.