Systemic Therapy for Metastatic Colorectal Cancer

Leonard B. Saltz

Introduction

The chemotherapy options available for metastatic colorectal cancer (CRC) patients have expanded dramatically since the 1990s. However, treatment for patients with metastatic disease should be approached with a balance of optimism and caution. Well-motivated patients with adequate performance status, bone marrow reserve, liver function, and renal function have a substantial potential to benefit from treatment, whereas patients with poor performance status and/or significant comorbidities should be considered for either less aggressive therapies or supportive care only. The therapeutic options for patients with metastatic CRC are reviewed in this chapter, and general treatment recommendations are given. The reader is cautioned that this is a rapidly evolving field and that many changes in practice can be anticipated in the near future.

Cytotoxic Chemotherapy

5-Fluorouracil

5-Fluorouracil (5-FU) was a rationally designed agent patented almost 50 years ago (1). Despite its venerability, this agent remains at the center of most CRC chemotherapy regimens. 5-FU is a prodrug that must be metabolized before it can become biologically active. The chemistry of this activation process has been well described, and the reader is referred elsewhere for a detailed description (2,3,4,5).

Of the biomodulation strategies explored with 5-FU, two in particular—leucovorin (LV) and protracted venous infusion (PVI)—have gained traction over the years. LV (folinic acid, citrovorum factor) is the reduced folate, 5-formyl-tetrahydrofolate. In the presence of reduced folates, the active metabolite of 5-FU binds more tightly to thymidylate synthase (TS), its primary target enzyme (6,7). Although this preclinical rationale for LV as a biomodulator of 5-FU is sound, the issue of clinical relevance, whether it improves the therapeutic index of 5-FU, remains unresolved.

Initial uncontrolled pilot trials of 5-FU plus LV showed high response rates (RRs) in comparison to historical controls of 5-FU alone. Substantial toxicity was also seen, however (8,9,10,11). Multiple treatment schedules were developed using a variety of 5-FU and LV doses, and these regimens, now not widely used, have been well described elsewhere (12,13,14). The Advanced Colorectal Cancer Meta-Analysis Project initially performed a meta-analysis of nine randomized studies that compared 5-FU/LV with 5-FU alone (15). An update of this meta-analysis was reported with longer follow-up and 10 additional trials (16). This analysis now contains 3,300 patients from 19 trials, some of which had multiple comparisons, and thus a total of 21 pairwise comparisons were analyzed. In 10 of the comparisons, the 5-FU doses were similar in both arms, with LV being added to one arm. In these comparisons, significant response and survival advantages were seen in the 5-FU/LV arms, albeit with more toxicity. An analysis of the 11 trials in which 5-FU/LV was compared to a higher dose of 5-FU failed to show benefits for 5-FU/LV over 5-FU alone. Taken as a whole, the evidence would suggest that LV adds little to the therapeutic index of 5-FU and that higher doses of 5-FU without LV would be a reasonable alternative. LV is so widely accepted, however, that it is unlikely that 5-FU–based regimens without LV will ever gain significant popularity.

Many trials have attempted to define the “optimal” schedule of LV administration. It would seem from available data that lower-dose LV has some advantages because it has been associated with less diarrhea, and once-weekly regimens have been associated with less neutropenia and stomatitis than daily ×5 regimens (17,18,19). Of note, relatively protracted LV infusion times of up to 3 hours are often used in various regimens. There are no clinical data to support the practice of prolonging these infusions, and shorter LV infusion times of 15 to 30 minutes would appear to be both clinically defensible and reasonable.

Preclinical evidence indicated that an increased duration of exposure to low-dose 5-FU could improve efficacy (20). Because the plasma half-life of 5-FU is in the range of 8 to 20 minutes, protracted infusional 5-FU schedules were explored. A PVI of 5-FU at a dose of 300 mg/m²/day was compared to bolus 5-FU (21). The PVI regimen yielded a substantially higher RR (30%) than the bolus regimen (7%); however, there was no difference in survival between the two treatment arms. The Eastern Cooperative Oncology Group (ECOG) performed a similar trial with similar results (22). In a meta-analysis of 1,219 patients in six trials comparing PVI 5-FU to bolus, RR was improved (22% vs. 14%, respectively, P = 0.0002) (23). A survival advantage of less than 1 month was seen for PVI 5-FU.

High-dose intermittent infusion schedules differ from PVI schedules in that patients receive 5-FU over 24 to 48 hours on a weekly or every other week schedule. An early pilot trial of 5-FU 2,600 mg/m² weekly over 24 hours with LV 500 mg/m² reported seven responses in 12 chemotherapy-naive patients and three responses out of 10 patients who were previously treated (24). A large randomized trial reported by Kohne et al. confirmed the activity of this regimen, with a major objective RR of 44% in 91 patients (25). This trial also had an interferon-α (IFN-α)–containing arm, which was found to have substantial toxicity but no benefit. A phase III confirmatory trial comparing weekly 5-FU 24-hour infusions of 2,600 mg/m², either alone or with 500 mg/m² of LV, to the Mayo Clinic bolus daily ×5 schedule of 5-FU was less encouraging, however. No
overall survival (OS) differences were seen between the arms, and the RRs were 12% for the Mayo Clinic bolus schedule, 10% for the infusional 5-FU, and 17% for infusional 5-FU plus LV (P = NS). Progressionfree survival (PFS) was increased in the infusion plus LV arm (P = 0.029), but diarrhea was substantially increased (26).

Exploiting the different mechanisms of cytotoxicity of bolus and infusional 5-FU, de Gramont et al. piloted a regimen using both strategies simultaneously. This LV5FU2 regimen was administered as a 2-day treatment every other week. Patients receive LV 200 mg/m² over 2 hours, followed by a 5-FU bolus of 400 mg/m², followed by 5-FU 600 mg/m² by 22-hour infusion, with all drugs given on days 1 and 2, repeated every 14 days. In a randomized comparison of this LV5FU2 schedule to the Mayo Clinic bolus schedule, the RR was superior for LV5FU2 versus Mayo Clinic 5-FU (33% vs. 14%, P = 0.0004), as was the PFS (P = 0.0012) (27). OS for LV5FU2 was approximately 5 weeks longer than the Mayo Clinic group, a difference that trended toward, but just barely missed, statistical significance (P = 0.067). LV5FU2 patients experienced less overall toxicity than the patients on the Mayo Clinic arm.

Other biomodulation strategies that have been explored unsuccessfully include the use of methotrexate (13,28,29,30,31,32,33), trimetrexate (34,35,36,37,38,39,40,41), and IFN-α (42,43,44,45,46,47,48,49). These do not have a role in the current management of CRC, and the reader is referred elsewhere if further information on these agents is desired.

Capecitabine

Absorption of 5-FU from the gut is unreliable, and inactivation of orally absorbed 5-FU by dihydropyrimidine dehydrogenase in a first-pass clearance through the liver is highly variable between patients. Capecitabine is an orally available 5-FU precursor that is absorbed intact through the gut and then activated by a series of enzymatic alterations, the last of which is conversion by thymidine phosphorylase (TP) into 5-FU. Some evidence suggests that TP levels may be higher in tumor than in normal tissue, thus suggesting a preferential activation of capecitabine within the tumor (50). A phase II trial showed activity in CRC (51). The addition of LV did not appear to provide any benefit, and toxicity was increased. Randomized phase III trials showed that oral capecitabine was at least as effective as intravenous (IV) bolus 5-FU/LV, and the side effect profile of capecitabine was superior to the Mayo Clinic 5-FU schedule (52,53,54).

The major side effects of capecitabine in these trials were palmar-plantar erythrodysesthesia (hand–foot syndrome) and diarrhea. The hand–foot syndrome is frequently the dose-limiting side effect (55). Although the approved starting dose in the United States is 1,250 mg/m² twice daily, many clinicians, especially in North America, choose to initiate therapy at a lower dose and escalate in those rare circumstances when no toxicity is seen. A retrospective review of results from two large trials suggests that efficacy was not inferior in those patients who required dose reductions (56). Whether efficacy is maintained when patients are prospectively and routinely started at a lower dose of capecitabine has not been addressed. It should be noted that despite the claim that chronic oral capecitabine approximates the pharmacokinetics of a protracted infusion IV 5-FU schedule, no randomized comparison of capecitabine to infusional schedules of 5-FU/LV has been reported. The equivalence of capecitabine to infusional 5-FU, or lack thereof, is therefore not an issue that can be definitively answered at this time. Also, although some clinicians have expressed a preference for use of capecitabine as a salvage regimen after 5-FU–based regimens have failed, data do not support this approach (57).

UFT + Leucovorin

Uracil is a competitive inhibitor of DPD, the rate-limiting enzyme in 5-FU catabolism. UFT is a combination of uracil and the 5-FU prodrug tegafur (ftorafur) in a fixed molar ratio of 4:1. Tegafur is orally absorbed and is converted in the body to 5-FU. In its early development, tegafur was found to have some activity against CRC. A metabolite of this drug was found to be neurotoxic, however, and this toxicity limited the development of tegafur. By inhibiting DPD, uracil allows for small amounts of tegafur to produce 5-FU that persists in circulation, thereby reducing the amount of neurotoxic metabolite produced. The inhibition of DPD also reduces interpatient differences in DPD activity levels, making dosing more predictable (58).

UFT has been developed with oral LV on a three times daily schedule. Phase II showed acceptable tolerability, with activity comparable to what can be achieved with IV 5-FU bolus schedules (59,60). Two large randomized studies compared oral UFT plus LV to the IV Mayo Clinic 5-FU/LV schedule. Both trials showed equivalence in terms of RRs, time to tumor progression, and OS (61,62). The trials did not, however, fulfill the U.S. regulatory requirements for noninferiority of UFT/LV, and the issue of demonstration of the contribution of uracil to the activity of the compound was also not addressed adequately for regulatory approval. Therefore, this agent is not available in the United States.

Raltitrexed

Raltitrexed is a TS inhibitor that is not related to the fluoropyrimidines. Randomized trials have shown raltitrexed 3 mg/m² given once every 3 weeks to have similar activity to bolus 5-FU/LV (63,64,65). In one trial, however, survival for the raltitrexed arm versus the 5-FU/LV arm was statistically significantly worse (9.7 vs. 12.7 months, P = 0.01). TS levels may predict for response to this agent (66). Raltitrexed is not approved for use in the United States, but it is available in many other countries.

Irinotecan

Camptothecin was identified as an agent with preclinical antitumor activity as early as 1966. Its insolubility hindered early attempts at clinical development, until the identification of camptothecin’s mechanism of action (inhibition of topoisomerase I) led to renewed efforts to develop soluble derivatives. Irinotecan, or CPT-11 (CPT is an abbreviation for camptothecin), was one such soluble derivative. CPT-11 possesses a bulky dipiperidino side chain joined to the camptothecin molecule by a carboxyl-ester bond. This side chain confers solubility but substantially decreases cytotoxic activity. Carboxylesterase, predominantly in the liver, cleaves the carboxyl-ester bond to form the more active metabolite, 7-ethyl-10-hydroxycamptothecin (SN-38) (67). SN-38 has been shown to be as much as 1,000-fold more potent in inhibiting topoisomerase I than CPT-11 and is thus the predominant active form of the drug.

CPT-11 and SN-38 function via inhibition of the enzyme topoisomerase I (topo I). Topo I facilitates the uncoiling of DNA for replication and transcription. In binding to DNA, topo I causes reversible single-stranded DNA breaks. The topo I–DNA complex allows the intact strand to pass through the break, thereby relieving torsional stress in the coiled helix. Topo I then reseals the break. CPT-11 and SN-38 stabilize these single-stranded breaks. These stabilized breaks are reversible; however, the collision of replication forks with open
single-stranded breaks result in double-stranded breaks, which result in irreversible DNA fragmentation.

In the initial phase I trials of CPT-11, some antitumor activity was observed in several CRC patients (68,69,70,71,72). Subsequently, there was a phase II trial in which a 22% RR was seen in a population of previously treated CRC patients (73). A confirmatory trial reported a 23% RR and 31% stable disease rate in 43 patients with 5-FU–refractory CRC (74). An analysis of three identical trials together using a weekly treatment for 4 weeks every 6 weeks in 5-FU–refractory CRC showed a RR of 13% (75). Using a once-every-21-day schedule of CPT-11 at a dose of 350 mg/m² an 18% RR was seen, both in the 48 chemotherapy-naive patients and in the 165 patients who had previously progressed through a 5-FU–based regimen (76). Trials in the front-line setting reported 32% and 26% RRs, respectively (77,78).

The first randomized trial to confirm the benefits of CPT-11 was a phase III comparison of CPT-11 once every 3 weeks at 350 mg/m² (300 mg/m² for patients age 70 and older) versus supportive care only in CRC patients who had progressed on 5-FU (79). The patients receiving CPT-11 had a 1-year survival that was 2.5 times greater than the control group (36% vs. 14%). Quality-of-life parameters for the CPT-11–treated patients, as measured by the EORTC QLQ-C30 questionnaire, were as good or better in all major indices than the control group. Another phase III trial compared CPT-11 to infusional 5-FU after front-line 5-FU failure and found the 1-year survival of the CPT-11 group to be 1.4 times better (76).

Diarrhea was the major dose-limiting toxicity in early trials. Two different diarrheal syndromes were identified: early onset and late onset. The early onset diarrhea, which occurs during or immediately after CPT-11 administration, is a cholinergic effect and is readily controlled by use of atropine (80). In those patients who experience this symptom (and who do not have a contraindication to atropine administration), 0.5 to 1 mg of atropine gives rapid resolution, and subsequent CPT-11 doses can then be given with atropine as a premedication. Late onset diarrhea is a far more significant clinical issue, and this has been managed by the use of intensive loperamide at the onset of late onset diarrhea.

A randomized comparison of weekly versus every-3-week CPT-11 showed similar efficacy; however, the once-every-3-week schedule showed less diarrhea in this trial (81). Subsequent studies have shown that changing the weekly schedule to a 2 week on, 1 week off schedule substantially reduces the risk of toxicity (82). A direct comparison of efficacy and safety on these two weekly schedules has not been done, however.

Camptothecin/5-Fluorouracil/Leucovorin Combinations

Building on the 4 week on, 2 week off CPT-11 schedule that had been selected for development in North America, Saltz et al. added a low dose (20 mg/m²) of LV given weekly, in order to reduce the potential for LV-exacerbated diarrhea. The phase I trial of this schedule showed that the full single-agent dose of CPT-11 (125 mg/m²) could be given with 500 mg/m² of 5-FU and 20 mg/m² LV (83).

In a large-scale randomized phase III trial, this combination of irinotecan, fluorouracil, and leucovorin (IFL) was compared to the Mayo Clinic schedule of 5-FU/LV (84). For regulatory reasons, a third arm using front-line single-agent CPT-11 was also included. This trial showed that IFL was superior to the Mayo Clinic 5-FU/LV arm in terms of RR, time to tumor progression, and OS. The CPT-11 alone arm appeared to be comparable in efficacy to the 5-FU/LV arm. Total toxicity incidence was similar in all arms of this trial. A greater amount of grades 3 and 4 diarrhea and vomiting were seen with IFL, whereas more dose-limiting neutropenia, neutropenic fever, and stomatitis were seen with 5-FU/LV. Treatment-related deaths occurred in 1% of patients in each arm of this trial.

Other investigators combined CPT-11 with infusional 5-FU (85). In France, an every other week (biweekly) 5-FU infusion for 2 consecutive days was developed, whereas in Germany investigators explored weekly 24-hour high-dose infusions of 5-FU combined with weekly CPT-11. A randomized phase III trial randomized patients to one of these 5-FU/LV schedules alone, or with CPT-11. RR, PFS, and OS were superior in the CPT-11–containing arm of this trial as well.

More recently, the biweekly schedule of LV5FU2 plus irinotecan has been studied with a simplified LV5FU2 infusion schedule, a regimen now widely known as FOLFIRI (FOL, folinic acid; F, 5-FU; IRI, irinotecan) (86). This regimen has now gained widespread acceptance as one of the preferred irinotecan/5-FU/LV administration schedules.

Oxaliplatin

The diaminocyclohexane (DACH) platinum compounds are a group of agents that demonstrated preclinical activity in some cisplatin-resistant cell lines and xenografts (87,88). One important member of this DACH platinum group is oxaliplatin, which demonstrated some preclinical activity against CRC (89). The size of the DACH carrier ligand results in bulkier platinum-DNA adducts than are created by cisplatin. This putatively results in greater resistance to repair mechanisms (90,91,92).

Phase I studies showed evidence of antitumor activity at tolerable doses, with nausea, vomiting, leucopenia, and thrombocytopenia being the major dose-limiting toxicities. Nephrotoxicity was not observed. Significant neurotoxicity, including pharyngolaryngeal dysesthesia, a sensation of choking without overt airway blockage, were also noted (93,94). A phase II trial of oxaliplatin monotherapy in previously untreated CRC showed a confirmed RR of 12% (95). A second in a similar population reported a 24% RR, with 13% grade 3 neurotoxicity (96). A trial of monotherapy in second-line treatment yielded a RR of 10% (97).

Although the single-agent activity was marginal, studies of oxaliplatin plus 5-FU/LV appeared far more promising. Based on a series of phase II trials by Levi et al. (98,99), Giachetti et al. from the same group reported a phase III trial of chronomodulated 5-FU/LV alone or with oxaliplatin (100). The group receiving oxaliplatin had a superior RR (53% vs. 16%, P <0. 001). PFS was also superior, just reaching statistical significance (8.7 vs. 7.4 months, P = 0.048). There were no differences in median OS (19.4 and 19.9 months, respectively).

FOLFOX is an acronym that denotes a series of biweekly, nonchronomodulated combinations of LV, 5-FU, and oxaliplatin (FOL, folinic acid [LV]; F, fluorouracil; OX, oxaliplatin) (27). Numerous permutations of this combination (i.e., FOLFOX 1, FOLFOX 2), involving modifications in doses and scheduling, have been evaluated. In a randomized comparison of LV5FU2 versus FOLFOX 4 in 420 previously untreated metastatic CRC patients, the FOLFOX 4 arm showed a significantly superior RR (51% vs. 22%, P = 0.001) and PFS (9.0 vs. 6.2 months, P = 0.0003) (101). The OS trended in favor of FOLFOX, but the difference was not statistically significant (16.2 vs. 14.7 months, P = 0.12). The number of patients experiencing grade 3–4 neutropenia was increased with FOLFOX 4 over LV5FU2 (42% vs. 5% of patients). Grade 3–4 diarrhea (12% vs. 5%) was also increased in the FOLFOX arm. Neurotoxicity, virtually absent in the LV5FU2 arm, was frequent in the FOLFOX arm, with 18% of patients experiencing grade 3 neurosensory toxicity.

The same FOLFOX 4 regimen was also studied in the second-line setting (102). Patients were randomized to FOLFOX 4, LV5FU2, or single-agent oxaliplatin. RRs were 10% for FOLFOX 4, 0% for LV5FU2, and 1% for oxaliplatin alone (P <0.0001 for FOLFOX vs. LV5FU2). Time to tumor progression was also superior for FOLFOX 4 (4.6 months) versus LV5FU2 (2.7 months) and oxaliplatin alone (1.6 months), while OS trended toward, but did not reach, statistical significance (P = 0.07) (103).

In FOLFOX 5, the oxaliplatin dose was increased from 85 to 100 mg/m²; however, before FOLFOX 5 was ever tested clinically, FOLFOX 6 was developed. This regimen maintained the 100 mg/m² oxaliplatin dose but used a simplified 5-FU/LV schedule (104). Oxaliplatin 100 mg/m² was given over 2 hours, with LV 400 mg/m² given concurrently via a “T” connector. These are then followed by a 400 mg/m² bolus of 5-FU, and then a 46-hour infusion of 5-FU at 2,400 to 3,000 mg/m². More recently, the FOLFOX 7 regimen has been reported, using a 130 mg/m² dose of oxaliplatin every 14 days. The simplified LV and 5-FU administration of FOLFOX 6 is maintained, with deletion of the bolus 5-FU. In the FOLFOX 7 schedule, oxaliplatin is discontinued after 3 months and is planned for reintroduction after 12 weeks or sooner if clinical progression occurs. Although reintroduction of oxaliplatin was less frequent than had been intended in the protocol, the results of this trial showed the acceptability of a planned early stopping of oxaliplatin, with efficacy parameters being similar in both arms of the trial (105). This rationale of stopping oxaliplatin at an early time point, before the development of prohibitive neurotoxicity, with the potential for reintroduction at a later date, has now become widely accepted in standard practice, regardless of which FOLFOX schedule is used.

There has not been, and never will be, a randomized trial comparing FOLFOX 4 to FOLFOX 6. Most investigators have accepted that the LV5FU2 and simplified LV5FU2 schedules are comparable in efficacy and toxicity, and the simplified schedule has de facto replaced the original LV5FU2 in many practices. Most ongoing National Cancer Institute (NCI) cooperative group studies are using a modified FOLFOX 6 (mFOLFOX 6) that contains the simplified LV5FU2 doses from FOLFOX 6 with the lower 85 mg/m² starting dose of oxaliplatin from FOLFOX 4. Currently, this mFOLFOX 6 appears to be a reasonable schedule for routine clinical use.

One other practical modification that has often been made in the infusional 5-FU schedules of simplified LV5FU2 is to administer the infusion over a full 2 days (48 hours) as opposed to the original 46 hours in the published regimens. The current NCI cooperative protocols call for the dose to be given over 46 to 48 hours. The reason for this is one of medication safety. A 48-hour infusion has the advantage in that it can be written as two consecutive 24-hour infusions, thereby minimizing the risk that the total 2-day dose could be inadvertently written in a subsequent cycle as a daily dose, which would lead to a serious overdose. In the interest of minimizing such potentially catastrophic errors, a common and advisable practice is to avoid ever writing more than a 24-hour dose of a chemotherapy agent in the medication record. As such, a 46- to 48-hour infusion of 2,400 mg/m² is more safely written as 1,200 mg/m²/day × 2 days. The 4% difference in infusion time is trivial, but the potential for improved safety is considerable.

A regimen of bolus 5-FU, oxaliplatin, and leucovorin (bFOL) has also been studied in a phase II trial (106); however, the results of a randomized phase II trial comparing this regimen to FOLFOX and to a capecitabine plus oxaliplatin combination do not support the routine use of the bFOL regimen in the metastatic setting (107).

Oxaliplatin versus Irinotecan in the First-Line Setting

First-line development of irinotecan and oxaliplatin occurred in parallel. It was not until after each had established a front-line role that head-to-head comparisons were performed. One of the more important trials to address the comparison of these two agents was the NCI intergroup study N9741. Although this study underwent many iterations, in its final form it was a three-arm study using the weekly bolus IFL regimen as the control arm, compared to FOLFOX 4 and to oxaliplatin plus irinotecan (IROX).

The results of N9741 showed superior outcome for the patients randomized to FOLFOX 4, as compared to those randomized to either IFL or IROX, in terms of RR, time to tumor progression, and OS (Table 45.1) (Table 45.2) (108). Toxicity for FOLFOX 4 was also superior for virtually all parameters, except, as would be expected, neurotoxicity. The results of the IROX arm did not differ significantly from those of the IFL arm in terms of toxicity, response, or time to tumor progression; however, survival was borderline significantly better in the IROX arm than the IFL arm (P = 0.04).

Although FOLFOX 4 was superior to IFL in both RR and time to tumor progression in this trial, interpretation of the OS results of N9741 is complicated by a number of issues. First, there were major imbalances between treatment arms in terms of availability of effective second-line therapy. Oxaliplatin was not commercially available in the United States during the course of N9741, and only a small percentage of patients on the IFL arm received second-line oxaliplatin. Furthermore, it is unknown what percentage of those who did receive oxaliplatin received it as part of the FOLFOX regimen versus as a single agent. The details of poststudy therapy were not recorded, and it was not yet known at the time of the trial that only FOLFOX, and not single-agent oxaliplatin, is active in the salvage setting (102). Second-line irinotecan, which has been shown to offer a survival benefit, was readily available to all patients who had received FOLFOX 4. To what degree these imbalances in second-line therapy may have influenced the survival results is not known. Another important point is that IFL uses bolus 5-FU, while FOLFOX 4 contains infusional LV5FU2. It is therefore not possible to isolate the irinotecan versus oxaliplatin comparison from the 5-FU bolus versus 5-FU infusion comparison in the relative efficacy of these regimens.

Tournigand et al. conducted a small phase III trial of FOLFOX 6 versus FOLFIRI in which the only variable was oxaliplatin versus irinotecan; identical simplified LV5FU2 schedules were used in each arm, and all patients were planned to cross over to the other regimen at time of progression (Table 45.3) (109). Although this study is underpowered, having a total of only 226 patients, the results are compelling, with first-line response, time to tumor progression, and OS being extremely similar in the two arms. Diarrhea and alopecia were somewhat more common with the irinotecan-based regimen; the oxaliplatin-based regimen had considerable neurotoxicity. A similar and somewhat larger trial of 360 patients that compared FOLFOX 4 to the equivalent FOLFIRI schedule also showed virtually no difference in efficacy, with more gastrointestinal toxicity and alopecia with FOLFIRI and more neurotoxicity and thrombocytopenia in the FOLFOX arm (110). A third trial using a bolus 5-FU schedule in both arms, as well as comparing irinotecan and oxaliplatin with identical 5-FU schedules and same access to effective second-line therapy, shows no difference in efficacy parameters between the two regimens (111). Another randomized trial using bolus 5-FU in each arm showed a better result for the group receiving first-line oxaliplatin; however, the oxaliplatin regimen initially had a higher dose of 5-FU, a factor that confounds interpretation of this trial (112).

Only gold members can continue reading. Log In or Register to continue