|
|
|
| HOME | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | HELP |
|
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||
When the role of systemic treatment for colorectal cancer was last reviewed in the Journal, in 1994,3 fluorouracil was the only effective chemotherapeutic drug for this cancer; much exciting progress has occurred since then. Accordingly, in this review, we will consider newer cytotoxic chemotherapies and biologic agents effective against colorectal cancer (Table 1) and will assess their uses for the treatment of metastatic disease and as components of adjuvant therapy. A discussion of combined therapeutic approaches (surgery, chemotherapy, and radiation) for patients with rectal cancer5 is outside the scope of this review.
|
Pathological stage at the time of presentation remains the most important prognostic indicator in colorectal cancer. Although the original Dukes' staging system6 has been modified, the depth of disease invasion through the bowel wall and the extent of regional lymph-node involvement remain the core of the staging system. The tumor–node–metastasis (TNM) system of the American Joint Committee on Cancer is now the most commonly used system for staging colorectal cancer and serves as a benchmark for predicting the likelihood of five-year survival (Table 2).7
|
Fluoropyrimidines
Mechanisms of Action and Dose Scheduling of Fluorouracil
The backbone of treatment for colorectal cancer is fluorouracil, a fluorinated pyrimidine, which is thought to act primarily by inhibiting thymidylate synthase, the rate-limiting enzyme in pyrimidine nucleotide synthesis.13 Fluorouracil is usually administered with leucovorin, a reduced folate, which stabilizes the binding of fluorouracil to thymidylate synthase, thereby enhancing the inhibition of DNA synthesis.14 In patients with advanced colorectal cancer, treatment with fluorouracil and leucovorin reduces tumor size by 50 percent or more in approximately 20 percent of patients (the "objective-response rate") and prolongs median survival from approximately 6 months (without treatment) to about 11 months.8,9,10,11,15
The major side effects associated with fluorouracil depend on the method of administration. When the drug is given according to a "loading" schedule of bolus treatments on five consecutive days every four to five weeks, neutropenia and stomatitis are the most common toxic effects. In contrast, with weekly bolus doses, diarrhea is more frequent. Regimens involving fluorouracil administered as a continuous intravenous infusion (with a portable infusion pump) are associated with less hematologic and gastrointestinal toxicity, but palmar–plantar erythrodysesthesia ("hand–foot syndrome") is more common.16,17,18 Although regimens involving continuous intravenous infusion were previously perceived as being more expensive and less convenient than bolus regimens, recent analyses suggest that differences in cost and quality of life between the bolus and prolonged-infusion schedules are marginal.19,20,21 Furthermore, continuous infusion appears to be moderately more effective than a rapid bolus approach.22
Oral Fluoropyrimidines
Early attempts to administer fluorouracil orally fell into disfavor after the results of a double-blind, placebo-controlled, randomized study showed intravenous fluorouracil to be more effective than the oral form.23 Pharmacokinetic plasma assays suggested that this discrepancy was due to the erratic intestinal absorption of oral fluorouracil, which may in turn have been a result of the variable mucosal concentrations of dihydropyrimidine dehydrogenase, a major catabolic enzyme of the drug. Strategies developed to overcome this problem include the administration of fluorouracil prodrugs that are absorbed intact and metabolically activated after intestinal absorption24 and the coadministration of oral fluorouracil with drugs that inhibit the action of dihydropyrimidine dehydrogenase.25,26,27
Capecitabine (Xeloda) is a prodrug that undergoes a three-step enzymatic conversion to fluorouracil.24 The side-effect profile of capecitabine is similar to that observed when fluorouracil is given as a protracted infusion. The hand–foot syndrome is a prominent toxic effect; other adverse reactions include diarrhea, nausea, vomiting, and bone marrow suppression.28,29,30 Two randomized clinical trials comparing capecitabine to the monthly schedule of fluorouracil and leucovorin28,29 reported that the rate of objective response in patients treated with capecitabine was moderately improved (19 to 25 percent, as compared with 15 percent); median overall survival, however, was similar between the two regimens (12 to 13 months). Mouth sores and bone marrow suppression were more likely to develop in patients receiving a loading regimen of intravenous fluorouracil, whereas patients assigned to receive capecitabine had an increased incidence of the hand–foot syndrome.
An example of an oral fluorouracil combination that inhibits dihydropyrimidine dehydrogenase is uracil plus tegafur, which has been approved by regulatory agencies outside the United States. Tegafur, a prodrug of fluorouracil, is combined with uracil, which is a competitive blocker of dihydropyrimidine dehydrogenase, to improve the absorption and bioavailability of tegafur.26 The combination is usually administered with oral leucovorin. In two randomized studies, this therapy resulted in a response rate and median overall survival similar to those obtained with parenteral fluorouracil and leucovorin.31,32
As monotherapy, the oral fluoropyrimidines appear to have favorable safety, convenience, and cost-effectiveness profiles when compared with bolus intravenous fluorouracil administered on five consecutive days every four to five weeks.33 Whether such benefits would remain if the oral agents were compared with a less toxic method of administering intravenous fluorouracil (e.g., as a weekly bolus) or if capecitabine were combined with parenteral chemotherapies remains to be determined.
Adjuvant Therapy with Fluorouracil
For many years, the use of fluorouracil as adjuvant treatment after complete resection of stage II cancers (T3 or T4 with negative lymph nodes, according to the TNM system) or stage III cancers (any T stage with positive lymph nodes) was thought to be ineffective.34 In retrospect, previous randomized studies were limited both by insufficient numbers of patients as well as by suboptimal adherence to chemotherapy. Better-conducted randomized trials in patients with stage III disease showed that intravenous fluorouracil35 or fluorouracil plus leucovorin36,37,38 did improve outcomes. A pooled analysis of patients with stage III disease who were participating in seven clinical trials of adjuvant therapy demonstrated that adjuvant chemotherapy increased the probability of remaining free of tumor recurrence after five years from 42 percent to 58 percent and the likelihood of five-year overall survival from 51 percent to 64 percent.39 Two recent randomized trials that compared either capecitabine or uracil plus tegafur to parenteral fluorouracil showed that oral and intravenous fluoropyrimidine therapies appear to offer equivalent efficacy.40,41 Adjuvant therapy has been found to be as beneficial in elderly patients as it is in younger patients.39,42
The value of postoperative fluorouracil-based therapy after the resection of stage II colon cancers has remained controversial, however.43,44,45 No randomized clinical trial has demonstrated a survival benefit for this group of patients, and an analysis of data pooled from seven such studies showed five-year survival probabilities in the range of 80 percent, with or without treatment.39 An expert panel convened by the American Society of Clinical Oncology recently concluded that adjuvant chemotherapy should not be routinely recommended for all patients with stage II disease. According to the available data, the absolute survival advantage associated with such treatment could not exceed 5 percent, and the inclusion of at least 4000 patients would be required to detect a potentially smaller benefit.46 A retrospective subgroup analysis44 has suggested that patients with prognostic factors associated with increased rates of recurrence (e.g., adherence of the tumor to an adjacent organ or bowel perforation or obstruction) may benefit from adjuvant treatment; however, no prospective study has yet validated this observation. According to insurance-claims data, 30 percent of Medicare beneficiaries (i.e., Americans older than 65 years of age) with stage II disease receive such treatment.
Regional Therapy with Fluoropyrimidines
The rationale for administering chemotherapeutic agents directly into the liver in the presence of hepatic metastases is based on the dual blood supply of the liver: tumors derive most of their blood from the hepatic artery, while the portal circulation supplies the normal liver parenchyma. Infusions of fluorouracil or an analogue compound, floxuridine, into the hepatic artery results in a doubling of the response rates achieved with systemic administration of fluorouracil. However, no survival advantage has been demonstrated for patients with more advanced disease, most often because of the presence of metastases outside the liver.47 Furthermore, toxic effects that include chemical hepatitis, cholangitis, and catheter-related complications, as well as the high overall costs of this therapy, limit its practical value.47,48,49
To date, randomized studies assessing hepatic arterial infusion of fluorouracil or floxuridine as adjuvant therapy after the resection of liver metastases, as compared with either surgery alone50,51 or systemic chemotherapy,52 have failed to demonstrate an improvement in the likelihood of a cure with this type of adjuvant therapy, although one of these trials reported a significant difference in a predefined, two-year survival end point between hepatic infusion and systemic chemotherapy alone.52
Irinotecan
Mechanisms of Action and Toxic Effects
Irinotecan (Camptosar, also known as CPT-11) is a semisynthetic derivative of the natural alkaloid camptothecin, which exerts a cytotoxic effect through its interaction with the enzyme topoisomerase I.53 This enzyme is involved in the uncoiling of DNA for replication and transcription, and it causes single-stranded DNA breaks. Such breaks are normally transient and repaired; however, camptothecin stabilizes these breaks, leading to DNA fragmentation and cell death through interaction with the replication fork.
Irinotecan is a prodrug that is hydrolyzed to its active metabolite, SN-38, by hepatic carboxylesterases. SN-38 is detoxified to an inactive, glucuronidated form by uridine diphosphate glucuronosyltransferase isoform 1A1 (UGT1A1) and is excreted in the urine and bile.54 Several additional inactive metabolites of irinotecan are formed through oxidative metabolism by the cytochrome P-450 enzymes CYP3A4 and CYP3A5.55
The toxic effects of irinotecan include diarrhea, bone marrow suppression, nausea, vomiting, and alopecia. Polymorphisms of UGT1A1 appear to correlate with the severity of the gastrointestinal effects and bone marrow suppression,56,57,58 and clinical trials are currently under development to individualize drug dosages on the basis of a given patient's pharmacogenomic profile.
Treatment of Metastatic Disease with Irinotecan
Two randomized trials of single-agent irinotecan as a second-line therapy in patients with advanced colorectal cancer who had previously received bolus fluorouracil showed a two-to-three-month improvement in median overall survival as compared with either best supportive care alone9 or fluorouracil given by continuous infusion,59 accompanied by similar or improved quality of life. Irinotecan was subsequently examined in combination with fluorouracil and leucovorin as initial therapy for metastatic colorectal cancer; a bolus schedule (known as IFL) was tested in North America60 and a 48-hour infusion program was tested in Europe.4 Both of these multicenter trials indicated that the three-drug combination was twice as likely as fluorouracil and leucovorin alone to result in a 50 percent or greater shrinkage of tumor dimensions, resulting in a two-month extension in median survival.
Adding irinotecan to bolus fluorouracil and leucovorin increases the likelihood of clinically significant myelosuppression and diarrhea.4,60 This toxicity pattern led to a higher-than-anticipated number of treatment-related deaths when the combination of irinotecan, fluorouracil, and leucovorin was first used61; however, prompt attention to and treatment of toxic effects as well as adjustments in dosing and scheduling of the regimen have rendered this combination more tolerable and safer.61,62,63
Oxaliplatin
Mechanisms of Action and Toxic Effects
Oxaliplatin (Eloxatin) is a third-generation platinum derivative that forms bulky DNA adducts and induces cellular apoptosis.64 Despite the ineffectiveness of other platinum-based drugs (such as cisplatin and carboplatin) in the treatment of colorectal cancer, preclinical data from human cell lines suggested that oxaliplatin held promise in treating this disease.65 Furthermore, oxaliplatin and fluorouracil were shown to be highly synergistic, not only in preclinical models66 but also in subsequent clinical trials.67 A potential mechanism for this synergy is the down-regulation of thymidylate synthase by oxaliplatin, which thereby potentiates the efficacy of fluorouracil.68
The toxicity profile of oxaliplatin differs from those of cisplatin and carboplatin. Renal dysfunction, alopecia, and ototoxic effects are uncommon, but neuropathy is more frequent.69 Two types of neuropathies have been described. Most patients experience transient dysesthesias, manifested as numbness or tingling of the hands and feet and the oral or perioral regions, which are exacerbated by exposure to low temperatures. After months of therapy, patients may have a cumulative, dose-dependent sensory neuropathy in which peripheral dysesthesias and paresthesias persist between cycles of therapy; these effects usually diminish after the cessation of treatment.70
Treatment of Metastatic Disease with Oxaliplatin
Although oxaliplatin as a single agent has limited efficacy when administered as first-line71,72,73 or second-line74 treatment for patients with metastatic colorectal cancer, clinical benefit has been shown when it is combined with bolus fluorouracil and leucovorin followed by a 46-hour infusion of fluorouracil — a treatment regimen known as FOLFOX.67,70,75,76 Randomized clinical trials have consistently shown that FOLFOX results in response rates and times to disease progression that are superior to those achieved with fluorouracil and leucovorin when given as first-line70,75,76 or second-line67 treatment for advanced colorectal cancer. There was a trend toward improvement in overall survival, but it did not reach statistical significance in these initial studies.
Is There an Optimal First-Line Therapy for Metastatic Disease?
Comparisons of irinotecan, fluorouracil, and leucovorin regimens with oxaliplatin, fluorouracil, and leucovorin combinations for the initial treatment of metastatic colorectal cancer have recently been reported (Table 3). In a multicenter trial conducted in North America, 795 patients were randomly assigned to receive IFL, FOLFOX, or a combination of irinotecan with oxaliplatin (IROX).77 The patients treated with FOLFOX had a response rate, time to disease progression, and overall survival time that were superior to those observed with either IFL or IROX. However, the apparent superiority of FOLFOX may have been influenced by an imbalanced availability of second-line agents at the time the study was conducted in that 60 percent of the patients initially treated with FOLFOX subsequently received irinotecan, whereas only 24 percent in the IFL group were subsequently given oxaliplatin. In addition, the IFL regimen is based on a bolus-fluorouracil schedule that may be inferior to the two-day infusion of fluorouracil included in FOLFOX.22 Nonetheless, these data support the option of oxaliplatin-based therapy as first-line combination therapy for patients with metastatic disease.
|
Thus, the optimal sequence of these chemotherapy agents is currently unclear. The choice of initial therapy could depend on a given patient's coexisting conditions at baseline. For example, for patients who have an underlying neuropathy, irinotecan-based regimens may be more appropriate than oxaliplatin-based regimens, which may have neurotoxic effects, whereas for patients with underlying bowel dysfunction, oxaliplatin-based therapy may be more appropriate than irinotecan-based therapy, which may have toxic effects on the gastrointestinal system. Despite the choice of initial therapy, exposure to each of these cytotoxic agents at some time over the course of a patient's disease has been associated with prolonged survival.80
Incorporation of Oxaliplatin and Irinotecan into Adjuvant Therapy
Attempts to integrate irinotecan and oxaliplatin into adjuvant-treatment programs have been the focus of at least four clinical trials. The initial results of two of these studies have been reported thus far.
Saltz and colleagues assigned 1264 patients with stage III disease to receive either IFL or bolus fluorouracil and leucovorin. After a median follow-up period of 2.6 years, IFL therapy did not improve either the probability of recurrence or overall survival, but it significantly increased the risks of diarrhea and myelosuppression.81 This unanticipated outcome emphasizes the need for caution in extrapolating positive results from the setting of metastatic disease to adjuvant treatment. The outcome of a European trial of infusional fluorouracil in combination with irinotecan has not yet been reported.
In contrast, the initial results of a European trial in which 2200 patients with stage II and stage III colon cancer were randomly assigned to receive either FOLFOX or infusional fluorouracil and leucovorin showed that the FOLFOX-treated cohort had a greater likelihood of remaining free of recurrence after four years (76 percent vs. 69 percent, P<0.001).82,83 This difference was far more evident among patients with stage III disease (70 percent in the FOLFOX group vs. 61 percent in the fluorouracil-and-leucovorin group, P=0.002) than among those with stage II disease (85 percent vs. 81 percent, difference not significant). FOLFOX treatment has not yet been found to confer a statistically significant advantage in terms of overall survival. Peripheral neuropathy was the major side effect of the FOLFOX regimen, occurring in 92 percent of the patients and classified as grade 3 (i.e., limiting the activities of daily living) in 12 percent. The neurotoxic effects were generally reversible; 18 months after the completion of therapy, 76 percent of patients reported no neurologic impairment, and only 4 percent had residual grade 2 or 3 symptoms. These data have expanded the options for treating patients with early-stage disease; however, the risks and benefits of this more toxic regimen should be assessed in individual patients.
Targeted Therapies
Laboratory studies have identified molecular sites in tumor tissue that may serve as specific targets for treatment. The goal of such a therapeutic strategy is the interruption of cellular pathways essential for tumor growth, survival, and metastasis and, potentially, a reduction in the toxic effects associated with less specific cytotoxic chemotherapies. Currently, two promising classes of targeted compounds have been introduced into the clinical management of advanced colorectal cancer: epidermal growth factor receptor antagonists and angiogenesis inhibitors.
Cetuximab
The epidermal growth factor receptor is a transmembrane glycoprotein that is involved in signaling pathways affecting cellular growth, differentiation, proliferation, and programmed cell death.84 The receptor is present on the surface of normal epithelium and is overexpressed in certain tumors. Such overexpression has been associated with a poorer prognosis in colorectal cancer.85,86 Inhibition of this target can be achieved by antibodies directed against the extracellular domain or the soluble ligands of the receptor, inhibitors of the required dimerization of the receptor, or small molecules that prevent phosphorylation of the receptor by its intracellular tyrosine kinase. Cetuximab (Erbitux, also known as C-225) is a monoclonal antibody against the extracellular binding domain of the receptor and recently became the first such inhibitor to be approved in the United States for the treatment of metastatic colorectal cancer.
Preclinical studies have shown not only that therapeutic synergy exists between cetuximab and chemotherapeutic agents, but also that such synergy can occur in tumor cells already resistant to irinotecan — a finding suggesting that the inhibitor may overcome cellular resistance to irinotecan.87 As such, Saltz and colleagues gave a combination of cetuximab and irinotecan to 121 patients with advanced colorectal cancer whose tumor had been found to be unresponsive to irinotecan; 19 percent of the patients had radiographically objective tumor shrinkage88 (Table 4). To determine whether this antitumor effect was due to synergy between the two drugs or due to the independent activity of cetuximab, 60 similar patients were treated with the antibody alone; 10 percent of them had radiographically significant tumor regression.89
|
The side effects of cetuximab are fairly mild, with an acne-like rash and drying and fissuring of the skin the most common; hypersensitivity infusion reactions are less frequent (occurring in 3 percent of patients, with death in fewer than 1 in 1000). Although some degree of acneiform rash occurs in most patients, severe eruptions resulting in significant pain, pruritus, or infectious sequelae are rare. Of note, the development and severity of the rash have been correlated with an increased likelihood of an objective response; the mechanism underlying this correlation is currently unclear.94
These data suggest that cetuximab is effective in a subgroup of patients with advanced colorectal cancer. The trials reported to date have included only patients with immunohistochemical evidence of epidermal growth factor receptor expression. However, the degree of such expression appears to be unrelated to the likelihood of disease regression,90 raising questions as to whether receptor overexpression should be a prerequisite for cetuximab treatment and whether the drug is interacting with additional molecular targets.
Bevacizumab
The appreciation that tumors induce blood-vessel formation, allowing extension beyond a few millimeters in size, stimulated efforts at inhibiting this type of angiogenesis as a means of controlling the growth and spread of cancer cells.95 The most successful of these efforts to date has focused on neutralizing the vascular endothelial growth factor, which is a soluble protein instrumental in angiogenesis.96 Bevacizumab (Avastin), a humanized antibody directed against the vascular endothelial growth factor, has been examined in combination with chemotherapeutic agents in several clinical trials in patients with advanced colorectal cancer (Table 4).91,93,97 A small, randomized, phase 2 trial in patients who had received no prior treatment for their metastatic disease showed that bevacizumab, as compared with fluorouracil and leucovorin alone, improved the likelihood of a tumor response.91 This effort led to two concurrent randomized, phase 3 trials.
Hurwitz and colleagues assigned 815 patients to receive either IFL with bevacizumab or IFL with placebo.93 The addition of bevacizumab led to an impressive, statistically significant increase in the rate of response and a 4.7-month prolongation in median overall survival (to 20.3 months, vs. 15.6 months with IFL and placebo). In a study involving patients considered unable to tolerate irinotecan, Kabbinavar et al.92 found that bevacizumab added to fluorouracil and leucovorin improved response rates and extended the time to tumor progression but did not significantly prolong median survival. In both studies,92,93 bevacizumab was associated with reversible hypertension and proteinuria and was relatively well tolerated. Recently, a statistically significant prolongation in median survival was reported with the addition of bevacizumab to FOLFOX, as compared with FOLFOX alone, in patients with advanced colorectal cancer who had previously been treated with irinotecan-based therapy.98
The Food and Drug Administration approved the use of bevacizumab in combination with any intravenous fluorouracil-containing regimen as initial therapy for patients with advanced colorectal cancer. Although the addition of this drug clearly represents a therapeutic advance against this disease, further studies are required to define the extent of its clinical usefulness. Bevacizumab appears to be inactive as a single agent, and its efficacy when added to nonintravenous fluorouracil regimens is currently unknown. It is also unclear whether its activity is generated primarily by an antiangiogenic mechanism or whether it exerts its effect by altering tumor vasculature, thereby enhancing the intracellular access of chemotherapeutic agents.93,99
Future Directions and Challenges
The recent introduction of cytotoxic drugs in addition to fluorouracil as well as the development of targeted agents has resulted in significant progress in the treatment of advanced colorectal cancer. Whereas the median life expectancy for people with this condition is in the range of 6 months without the use of any form of treatment and is extended to 10 to 12 months when either fluorouracil alone or fluorouracil combined with leucovorin is administered, the median survival is increased to 14 to 16 months when either irinotecan or oxaliplatin is added to a fluorouracil-based treatment regimen. Survival appears to be further prolonged, to more than 20 months, when all three drugs are used at some point in the care of the patient or if a form of targeted therapy (e.g., bevacizumab) is combined with a cytotoxic-drug combination (Figure 1). Data suggest that the introduction of at least some of these combination regimens, when administered as adjuvant treatment after surgery,82,83 may further enhance the likelihood of cure.
|
Efforts are under way, however, to develop ways in which therapy can be tailored for individual patients. Intratumoral levels of various enzymes involved in fluorouracil activation and metabolism, such as thymidylate synthase, dihydropyrimidine dehydrogenase, and thymidine phosphorylase, have been examined to identify patients with advanced colorectal cancer who are likely to benefit from fluorouracil treatment. Unfortunately, these studies have yielded conflicting outcomes.100,101,102,103,104,105 The efficacy and toxicity of specific drugs in individual patients will be more readily assessable once the genetic polymorphisms and mutations affecting the metabolic pathways of various chemotherapeutic agents as well as the biologic pathways of targeted agents have been fully studied.106,107
Substantial progress is being made in the identification of new forms of treatment for colorectal cancer. Patients with metastatic disease are living twice as long as they were one decade ago. The use of adjuvant chemotherapy has increased the likelihood of cure by 30 percent among patients with stage III disease. In 1994, Moertel concluded his review of chemotherapy for colorectal cancer by noting that "these [new] treatments will stimulate continued research in the treatment of colorectal cancer."3 The same statement can be made today. Unresectable, advanced colorectal cancer remains incurable. As in the past, further progress can take place only through the completion of well-designed, randomized clinical trials.
Supported in part by a K07 award from the National Cancer
Institute (1K07CA097992-01A1, to Dr. Meyerhardt) and an American
Society of Clinical Oncology career development award.
Dr. Meyerhardt reports having received a consulting fee from
Bristol-Myers Squibb and a lecture fee and grant support from
Sanofi-Synthelabo and reports having given lectures as part of
speakers' bureaus (Thompson Professional, which is supported by
Pfizer; Health Sciences, which is supported by Sanofi-Synthelabo; and
Health Sciences, which is supported by Genentech). He is the lead
investigator in a multidrug trial involving capecitabine (Roche),
oxaliplatin (Sanofi-Synthelabo), and erlotinib (Genentech) but does
not receive salary support for this work. Dr. Mayer reports having
received a consulting fee from Pfizer.
We are indebted to Drs. Charles S. Fuchs and Matthew H. Kulke for
their critical review of the manuscript.
Source Information
From the Department of Medical Oncology, Dana–Farber Cancer Institute; the Department of Medicine, Brigham and Women's Hospital; and the Department of Medicine, Harvard Medical School — all in Boston.
Address reprint requests to Dr. Meyerhardt at the Dana–Farber Cancer Institute, 44 Binney St., Boston, MA 02115.
References
| ||||||||||
Previous
.gif){kind=spacer}
