Second-Line Therapy for Advanced Colorectal Cancer

Abstract

The availability of irinotecan and oxaliplatin has dramatically altered both first- and second-line treatment of advanced colorectal cancer (CRC) compared with the era in which the sole treatment option in advanced disease was 5-fluorouracil (5-FU). Treatment options and strategies are becoming ever more enriched and complex with the recent availability of biologic agents such as bevacizumab and cetuximab. This article reviews randomized clinical trials assessing second-line treatment after failure of first-line single-agent 5-FU, irinotecan-based treatment, or oxaliplatin-based treatment. A number of conclusions can be drawn based on available data. (1) Second-line chemotherapy with active agents is superior to best supportive care alone. (2) Following 5-FU failure, active regimens include irinotecan, FOLFIRI, FOLFOX, and IROX, with IROX appearing to be superior to FOLFIRI. (3) Following irinotecanbased first-line treatment, FOLFOX is in general the best choice, and the combination of FOLFOX plus bevacizumab appears to be superior to FOLFOX alone. (4) Following first-line FOLFOX, irinotecan and FOLFIRI are currently the most appropriate options. Irinotecan plus cetuximab should emerge as an effective regimen. Further study is needed to define the best options in second-line therapy following the FOLFOXIRI triplet or the combination of bevacizumab plus chemotherapy. Studies are ongoing to determine optimal use of biologic agents, both in terms of when they should be used and in what combinations.

Prior to the relatively recent availability of irinotecan and oxaliplatin, chemotherapy for advanced colorectal cancer (CRC) was fairly straightforward and generally limited to 5-fluorouracil (5-FU). The addition of these two agents, as well as the availability of the oral fluoropyrimidine capecitabine, has permitted the development and investigation of a number of active regimens in both first-line and second-line treatment. Importantly, these developments have allowed investigation of strategies regarding sequence and timing of initial and subsequent therapy to maximize treatment response, with some of these newer approaches blurring the traditional distinction between “first-line” and “second-line” treatment. Moreover, the advent of biologic agents, such as bevacizumab and cetuximab, that augment the activity of chemotherapy has opened the door to even more complex treatment strategies and enriched treatment possibilities.

In light of recent advances and evolving treatment strategies, a note on the definition of “second-line” chemotherapy, vis-à-vis the context of this review, is in order. Traditionally, “second-line” therapy of advanced CRC describes any subsequent intervention with a new regimen after treatment failure (disease progression or discontinuing of treatment due to toxicity) has occurred with initial chemotherapy. It bears noting, however, that many situations currently exist in clinical practice in which “failure” of first-line treatment or the distinction between first- and second-line therapies is not so clear. Indeed, treatment may be stopped despite absence of progression or intolerance/toxicity. For example, chemotherapy is often discontinued in the absence of further tumor shrinkage, even when treatment is well tolerated, or simply because the patient has grown tired of treatment. Numerous clinical circumstances have arisen recently that represent entirely new challenges to the oncologist as they present unique or seemingly ambiguous situations that cannot necessarily be considered “second-line” settings as commonly understood.

General scenarios that challenge the traditional definition of second-line therapy include those in which initial chemotherapy is resumed after a prolonged pause. For example, the recently reported NO 16966 trial in more than 2000 patients revealed that the median time to treatment discontinuation was 6 months; regardless of disease progression.¹ (This confirms the long held notion that, in general, European oncologists treat advanced CRC for approximately 6 months and then stop.) If a patient remains progression free for an additional 3 to 6 months (median, 5.7 months in those trials permitting pauses)^2,3 then progresses, subsequent chemotherapy may not be considered truly “second-line.” In fact, it is likely that most such patients resume treatment with their first-line regimen in clinical practice. To illustrate an emerging model, evidence exists to suggest that rechallenge with FOLFOX (5-FU/oxaliplatin/leucovorin) after a chemotherapy- free interval following initial FOLFOX is feasible and effective.⁴ In that scenario, treatment is more analogous to intermittent first-line therapy rather than second-line therapy.

In most randomized trials that have established the superiority of doublets over single-agent 5-FU, the eligibility criteria included patients relapsing after longer than 6–12 months since completing adjuvant 5-FU. Now that FOLFOX has become a standard treatment regimen, this consideration must apply to the doublet as well. This is particularly relevant considering the potential activity of rechallenge with FOLFOX, as mentioned above, after an adequate chemotherapy-free period.⁴ In general, this situation is comparable to that of patients treated in the adjuvant setting with single-agent 5-FU. Thus, if a minimum of 12 months has passed since completing adjuvant FOLFOX, it would seem appropriate to consider it ongoing first-line treatment. Otherwise, the scenario might seem to fall somewhere in between first- and second-line therapy.

The traditional distinction between first- and second-line therapy is also blurred in the potential setting in which, for example, a patient receiving bevacizumab plus a doublet as first-line therapy discontinues the doublet after 6 months and subsequently progresses while receiving bevacizumab alone. Resumption of the initial doublet is likely in this situation, since no progression was observed during the initial 6 months of treatment, and the setting hardly constitutes a second-line strategy.

Other issues are raised by potential bevacizumab resistance—including whether it exists in the clinical setting. At present, it is unknown whether bevacizumab should be continued in subsequent treatment after failure of first-line treatment with this agent in combination with an active doublet. With the potential utility of chemotherapy-free periods and/or intermittent therapy in enhancing response to particular regimens, increased focus on strategies for optimizing timing and sequence of administration of active regimens, and the potential carryover of biologics such as bevacizumab from initial to subsequent treatment, conventional views of what constitutes first- vs. second-line therapies will continue to be challenged.

TRADITIONAL SECOND-LINE THERAPY SETTINGS

For purposes of clarity, this article reviews findings in randomized studies examining three classic “second-line” therapy scenarios—treatment administered subsequent to failure of single-agent 5-FU, irinotecan-based doublets, and oxaliplatinbased doublets. Settings including patients who fail irinotecan- or oxaliplatin-based triplet regimens and combinations involving targeted molecular therapies (eg, bevacizumab and cetuximab) are also addressed.

Second-Line Treatment After Single-Agent 5-FU

Before the introduction of irinotecan and oxaliplatin, the only available options for second-line treatment in advanced CRC involved variations in 5-FU treatment schedule and biochemical modulation.^5,6 Such strategies are now of marginal clinical relevance. Studies examining irinotecan and oxaliplatin regimens in second-line treatment are summarized in Table 1.^7–17

Table 1.

Second-line treatment after failure of first-line 5-FU–based chemotherapy: results of phase II and III randomized (except as noted) trials.

Author	Phase	Second-Line Treatment	No. Pts	RR%	Median PFS m ( o)	Median OS (mo)
Cunningham7	III	BSC	90	NR	NR	6.5
		IRI + BSC	189	NR	NR	9.2*
Rougier8	III	IRI	133	NR	NR	10.8 *
		CI 5-FU	134	NR	NR	8.5
Fuchs9	III	IRI weekly	95	NR	4.0	9.9
		IRI q21d	196	NR	3.0	9.9
Machover 10	II Nonrandomized	OXA 130 q3wks	109	10.0	6.0	NR
Andrè11	II Nonrandomized	FOLFOX3	40	20.6	4.6	10.6
		FOLFOX4	57	21.6	5.1	11.1
Rougier12	II Randomized Noncomparative	FOLFIRI	35	11.4	3.2	12.2
		FOLFOX	33	21.2	4.7	11.5
		IROX	33	15.2	4.2	11.0
Pitot13 (N 9841)	III	IRI	245	15.0	4.0	14.7
		FOLFOX4	246	27.0	5.2	13.5
Seymour14 (FOCUS)	III	IRI	259	11.0	10.5 †	13.9
		FOLFIRI	141	21.0	11.5 †	14.8
		FOLFOX	155	23.0	11.6 †	15.2
Pluzanska 15 (LIFE Study)	III	IRI	178	NR	NR	15.2
Haller16	III	IRI	628 total	7.3	2.8	11.1
		IROX		23.1*	5.3*	13.4*
Becouarn17	II Randomized Noncomparative	FOLFIRI/FOLFOX	32	6.2	8.2	9.8
		IROX	30	23.3	8.5	12.3

Abbreviations: RR = response rate; PFS = progression-free survival; OS = overall survival; BSC = best supportive care; IRI = irinotecan; CI 5-FU = continuousinfusion 5-fluorouracil (5-FU); OXA = oxaliplatin; FOLFOX = biweekly infusional plus bolus fluorouracil/leucovorin plus oxaliplatin; FOLFIRI, biweekly infusional plus bolus fluorouracil/leucovorin plus irinotecan; IROX; irinotecan plus oxaliplatin; NR = not reported.

^*Statistically significant

^†PFS of the first two lines

Irinotecan was approved for use in the late 1990s based on demonstrated activity as a single agent and in combination with 5-FU in 5-FU–resistant disease. The first study in this setting compared irinotecan with best supportive care in 279 patients and demonstrated an overall survival benefit for the irinotecan arm (9.2 vs. 6.5 months, P < .0001), with a 1-year survival of 32% in the irinotecan arm vs. 13.8% in the control arm.⁷ Irinotecan was also associated with benefits in quality of life and in symptom control, especially asthenia and pain.

A second study randomized 260 patients to receive irinotecan or infusional 5-FU.⁸ Patients receiving irinotecan lived significantly longer than those treated with 5-FU (10.8 vs. 8.5 months, P = .035). Survival at 1 year was 45% in the irinotecan group compared to 32% in the 5-FU group. Both treatments were equally well tolerated. Trends toward benefits for irinotecan relative to baseline were also seen for pain-free survival (P = .059), median time to performance status (PS) deterioration (P = .18), and median survival without weight loss of more than 5% (P = .22). Fuchs et al clarified that irinotecan given on a weekly basis and irinotecan given every 3 weeks have similar efficacy, with the latter regimen being associated with a lower incidence of severe diarrhea.⁹

Similar randomized trials for oxaliplatin alone are not available at this time because activity in the second-line setting after 5-FU failure has not been pursued as a registration strategy for this agent and because single-agent activity of this drug has been minimal vs. that of irinotecan.¹⁰ To date, only two phase II studies have been published with oxaliplatin plus 5-FU/leucovorin (LV) in second-line treatment, reporting response rates of 20%–40% with median survival durations of 11–17 months.^11,18

Rougier et al performed a multicenter randomized phase II trial examining three strategies in 101 patients with progressive disease after 5-FU–based treatment.¹³ The following regimens were compared: irinotecan 180 mg/m² on day 1 followed by infusional LV 200 mg/m², then 5-FU 400 mg/m² IV bolus followed by 5-FU 600 mg/m² (LV5FU2 regimen) on days 1 and 2 every 2 weeks (FOLFIRI); oxaliplatin 85 mg/m² on day 1 followed by the LV5FU2 regimen every 2 weeks (FOLFOX); and oxaliplatin 85 mg/m² followed by irinotecan 200 mg/m², both on day 1 every 3 weeks (IROX). Response rates were 11.4%, 21.2%, and 15.2% for FOLFIRI, FOLFOX, and IROX, respectively. Tumor growth control was ≥ 60% for all three regimens and the median overall survival (OS) durations were 12.2 months, 11.5 months, and 11 months for FOLFIRI, FOLFOX, and IROX, respectively. The authors concluded that more than 50% of patients failing 5-FU–based firstline chemotherapy benefit from any of these second-line combinations.

The phase III Intergroup N9841 trial confirmed the efficacy of both FOLFOX and irinotecan after failure of first-line 5- FU–based chemotherapy.¹⁴ Of 491 patients initially randomized, 220 were crossed over to FOLFOX as third-line therapy after progression on second-line irinotecan. In general, these patients experienced better outcomes than those who received irinotecan; response rates were 16% vs. 4%, median progression-free survival (PFS) durations were 5 vs. 2.7months, andmedian OS durations were 10 vs. 8.7 months.¹⁹

Three studies have provided indirect evidence of the activity of second-line chemotherapy after failure of initial therapy with a fluoropyrimidine. The FOCUS,¹⁴ LIFE,¹⁵ and CAIRO studies (efficacy data are pending) were designed to investigate the value of starting single-agent first-line therapy with a fluoropyrimidine then giving a doublet upon progression vs. giving doublet regimens up front. These trials represent a shift from the classic first-line/second-line model to a strategy-oriented trial design in which multiple stages of therapy are preplanned.

The FOCUS study assessed three strategies: (1) “staged single agent” consisting of 5-FU followed by irinotecan; (2) “staged combination” consisting of 5-FU followed by the addition of irinotecan or oxaliplatin as separate arms; and (3) “firstline combination” consisting of FOLFIRI or FOLFOX. Salvage therapy with the third drug (among 5-FU, irinotecan, and oxaliplatin) not used in the individual study regimens was not designed into the protocol but was permitted. The results show no significant OS differences among the treatment arms. The “staged single agent” strategy tended toward inferiority compared to any other plan, while the “staged combination” strategy was not inferior to the “first-line combination” strategy.¹⁴ These outcomes may be related, in part, to the fact that large numbers of patients in the “staged single agent” and “staged combination” groups did not receive salvage therapy with the third drug due to performance status deterioration.

A similar inference can be drawn from the LIFE study.¹⁵ Patients were randomized to receive 5-FU alone (continuous infusion or de Gramont regimen) or combined with oxaliplatin (FOLFOX4 or oxaliplatin plus continuous-infusion 5-FU). Upon disease progression, patients were to be treated with irinotecan. This study confirmed that the oxaliplatin/5-FU combination improves response rate and PFS in first-line treatment, but no difference in OS was observed. The study was plagued by low compliance rates throughout the entire treatment program, however, which makes it difficult to interpret the outcomes. A subgroup analysis showed that in centers where more than 50% of patients received second-line therapy, OS was greater in the doublet-treated group. These findings are consistent with those of a recent review suggesting that patients receiving all three available agents do better than those receiving only one or two.²⁰

Another important study in patients failing first-line 5-FU was that of Haller et al in which standard irinotecan was compared to the IROX doublet.¹⁶ Results showed that IROX was superior in terms of survival, making this combination a candidate for standard chemotherapy treatment in patients failing 5-FU.

Second-Line Treatment After Irinotecan-Based Treatment

In 2002, the US Food and Drug Administration approved oxaliplatin in combination with 5-FU/LV for the treatment of advanced colorectal cancer that has progressed or recurred following first-line therapy with irinotecan/5-FU/LV. Approval in this setting was based on the results of a large-scale phase III study reported by Rothenberg in which patients who failed first-line treatment with irinotecan, bolus 5-FU, and LV (IFL) were randomized to receive singleagent 5-FU, single-agent oxaliplatin, or FOLFOX.²¹ No responses were observed with oxaliplatin alone, indicating that this agent needs to be combined with a fluoro-pyrimidine. FOLFOX was associated with a significantly higher response rate and longer PFS compared with the other two arms.

The recently reported ECOG E3200 study investigated irinotecan resistance in light of the availability of bevacizumab.²² FOLFOX plus bevacizumab produced significantly longer PFS and OS compared with the control arm of FOLFOX alone. Outcomes with FOLFOX in this study were similar to those reported by Rothenberg.²¹ Similar results were reported for the FOLFOX control arm in another large phase III trial investigating the value of vatalanib (PTK787/ZK) in second-line treatment.²² The available data suggest that FOLFOX in patients failing irinotecanbased chemotherapy produces a response rate of approximately 10% with a PFS of approximately 5 months, which is superior to PFS times achieved with first-line 5-FU/LV in the pre-doublet era.

The strategy-oriented trials reported by Tournigand²⁴and Grothey²⁵assessed whether FOLFIRI followed by FOLFOX and CAPIRI (capecitabine/irinotecan) followed by CAPOX (capecitabine/oxaliplatin) are more efficacious than the reverse sequences. The results obtained with FOLFOX and CAPOX as second-line treatment confirm the activity data reported for the fluoropyrimidine/oxaliplatin doublets in the other trials mentioned above. A summary of trials investigating second-line treatment following irinotecan-based initial treatment is presented in Table 2.^{14, 21–27}

Table 2.

Second-line treatment after failure of an irinotecan-based first-line chemotherapy regimen: results of phase II and III randomized trials.

Author	Phase	Second-Line Treatment	No. Pts	RR%	Median PFS(mo)	Median OS(mo)
Seymour14(FOCUS)*	III	OXA+5-FU	163	NR	9.0†	16.3
Rothenberg21	III	OXA	174	1.3	1.6	8.1
		FOLFOX4	270	9.9‡	4.6‡	9.8
		LV5FU2	272	0.0	2.7	8.7
Giantonio22 (E3200)	III	FOLFOX4+BEVA	289	21.8	7.4‡	12.5‡
		FOLFOX4	290	9.2	5.5	10.7
		BEVA	243	3.0	3.5	10.2
Kohne23(CONFIRM2)	III	FOLFOX4+vatalanib	425	18.5	5.5	12.1
		FOLFOX4	430	17.5	4.1	11.8
Tournigand24	III	FOLFOX6	81	15.0	4.2	21.5§
Grothey25	II randomized Noncomparative	CAPOX	31	13.0	4.3	18.8§
Cunningham26 (BOND)	III	CET	27	17.0	NR	NR
		CET+IRI	41	19.0	NR	NR
Pfeiffer27	II	XELOX30	70	17.0	5.4	9.5

Abbreviations: RR = response rate; PFS = progression-free survival; OS = overall survival; OXA = oxaliplatin; 5-FU = 5-fluorouracil; FOLFOX = biweekly infusional plus bolus 5-FU/leucovorin plus oxaliplatin; LV5FU2 = biweekly infusional plus bolus 5-FU /leucovorin; BEVA = bevacizumab; CAPOX = oxaliplatin plus capecitabine; CET = cetuximab; IRI = irinotecan; XELOX₃₀ = oxaliplatin infused over 30 minutes plus oral capecitabine; NR = not reported.

^*The FOCUS study assessed three strategies: “staged single agent” (Plan A, consisting of 5-FU followed by irinotecan); “staged combination” (Plan B, consisting of 5-FU followed by the addition of irinotecan [B_ir] or oxaliplatin [B_ox] as separate arms); and “first-line combination” (Plan C, consisting of FOLFIRI [C_ir] or FOLFOX [C_ox]).

^†PFS of the first two lines

^‡Statistically significant

^§OS of first plus second line

Second-Line Therapy After Oxaliplatin-Based Treatment

Data relevant to the efficacy of second-line chemotherapy in patients failing first-line treatment with FOLFOX were derived from the Intergroup N9741 trial in which 795 patients were randomly assigned to receive IFL, FOLFOX, or IROX.²⁸ Response rates, PFS, and OS were superior for patients treated with FOLFOX compared with those receiving IFL or IROX. Superior survival associated with FOLFOX, however, might have been influenced by disparity in the use of second-line agents due to availability; 60% of patients receiving FOLFOX were treated with second-line irinotecan, which was FDA approved at the time, whereas only 24% of the IFL group received oxaliplatin, which had not yet received approval. These data have contributed considerably to the growing body of evidence showing the activity of newly emerging second-line treatment strategies.

Other evidence in support of the efficacy of chemotherapy following FOLFOX resistance is provided by the Tournigand study, which investigated the optimal sequencing of doublets (FOLFOX→FOLFIRI vs. FOLFIRI→FOLFOX).²³ The authors concluded that the two sequences were similar in efficacy, with both achieving impressive survival rates. After disease progression, however, FOLFOX was significantly more active than FOLFIRI, producing a longer PFS (4.2 vs. 2.5 months) and higher response rate (15% vs. 4%). This suggests that irinotecan following FOLFOX failure is less active than oxaliplatin after FOLFIRI failure, though the sequence of administration does not appear to affect OS. A study performed by Grothey et al comparing CAPOX→CAPIRI vs. CAPIRI→CAPOX supported the efficacy of CAPIRI as second-line treatment after CAPOX failure.²⁴

Results of studies evaluating secondline therapy after failure of oxaliplatinbased initial treatment are summarized in Table 3, which also reports data from the FOCUS¹⁴ and LIFE¹⁵ studies. Because these two studies tested strategies rather than activity of each line of treatment, no specific conclusions can be drawn as to the efficacy of second-line chemotherapy following oxaliplatin.

Table 3.

Second-line treatment after failure of an oxaliplatin-based first-line chemotherapy regimen: results of phase III randomized trials.

Author	Phase	Second-Line Treatment	No. Pts	RR%	Median PFS (mo)	Median OS (mo)
Seymour14 (FOCUS)	III	IRI+5-FU	146	NR	9.2	15.2
Pluzanska 15 (LIFE Study)	III	IRI	148	NR	NR	15.9
Tournigand24	III	FOLFIRI	69	4	2.5	20.6
Grothey25	III	CAPIRI	34	21	5.1	16.5

Abbreviations: RR = response rate; PFS = progression-free survival; OS = overall survival; IRI = irinotecan; 5-FU = 5-fluorouracil; FOLFIRI = biweekly infusional plus bolus 5-FU/leucovorin plus irinotecan; CAPIRI = irinotecan plus oral capecitabine; NR = not reported.

Overall, the chances of responding to second-line chemotherapy after failing oxaliplatin-based first-line treatment are in the range of 4%–21% with either FOLFIRI or irinotecan alone. In general, FOLFIRI is used in Europe more frequently, while single-agent irinotecan is favored in the United States.

Second-Line Treatment After a Triplet Regimen

A recent Italian randomized trial demonstrated significant extension of OS and PFS and a higher response rate with the FOLFOXIRI regimen (irinotecan/oxaliplatin/5-FU/LV) compared with FOLFIRI.²⁹ It should be noted that up-front use of all available active chemotherapeutic agents makes the choice of second-line treatment problematic. Masi et al reported on a small series of patients receiving second-line chemotherapy in this setting.³⁰ Among 74 patients who received FOLFOXIRI as firstline treatment, 54 underwent second-line treatment. An irinotecan-based regimen was given to 23 patients with persistent peripheral neurotoxicity; oxaliplatin-based treatment was used in 3 patients with impaired liver function; 17 patients who had complete or partial response or had progression more than 3 months after the end of treatment were re-treated with the triplet; and 11 patients received singleagent 5-FU or irinotecan. A 47% response rate was observed with FOLFOXIRI retreatment, with a PFS of 8.3 months. This finding supports the concept of oxaliplatin rechallenge (or triplet rechallenge) first reported for FOLFOX by de Gramont.³¹ Patients receiving doublets following FOLFOXIRI had a 27% response rate and PFS of 6.3 months. These data should be interpreted with great caution, however, given the limited number of patients evaluated in the study.

Second-Line Treatment With Combinations Including Biologics

The only phase III data available at the time of this writing on the use of second-line combinations involving biologic agents are from the trials assessing bevacizumab²² and vatalanib²³ (summarized in Table 2). The ECOG E3200 trial provided the basis for FDA approval of bevacizumab in secondline treatment.²² A total of 822 patients failing irinotecan-based regimens were randomized to FOLFOX, bevacizumab, or FOLFOX plus bevacizumab. Significantly greater response rate, PFS, and OS were observed in the FOLFOX/bevacizumab arm. The bevacizumab-alone arm was closed early due to evidence of decreased survival.

The CONFIRM 2 trial investigated the efficacy of vatalanib added to FOLFOX compared with FOLFOX alone in patients failing first-line FOLFIRI.²³ Vatalanib is considered a very promising agent because it may inhibit several vascular endothelial growth factor (VEGF) receptors, but it failed to show added efficacy compared to the control arm. Of note, however, was the substantial benefit observed with the addition of vatalanib in those patients (approximately 30%) with high lactate dehydrogenase levels. The relationship between this condition and treatment response is now being focused on in the clinical development of vatalanib. AZD 2171, another potent inhibitor of VEGF receptor tyrosine kinases that may influence the effects of a key angiogenic factor (VEGF-A), is also in phase II/III evaluation (in combination with FOLFOX vs. FOLFOX alone) as second-line treatment for patients failing irinotecanbased treatment.

The epidermal growth factor receptor (EGFR) inhibitor cetuximab was approved for treatment of patients resistant to irinotecan on the basis of the BOND study, a randomized comparative phase II trial that allowed several lines of chemotherapy to be given before cetuximab or irinotecan/cetuximab.²⁶ Accordingly, the study yielded a mix of results obtained in second-, third-, and fourth-line treatment (thus, PFS and OS are not reported in Table 2, which summarizes second-line outcomes only). Subgroup analysis among patients receiving cetuximab monotherapy (27 patients) and cetuximab/irinotecan (41 patients) as true second-line treatment indicated response rates of 19% and 17%, respectively. Results of a phase III study comparing irinotecan with irinotecan/cetuximab as second-line treatment after FOLFOX failure (the EPIC trial) are expected in early 2007. A similar phase III study is also ongoing with the fully human anti-EGFR antibody panitumumab.

The entire field of second-line treatment may undergo additional changes when the results from studies of first-line chemotherapy combined with bevacizumab, cetuximab, or panitumumab (CRYSTAL, CALGB 80405, COIN, CAIRO II, and PACCE studies) become available.

CONCLUSIONS

A number of conclusions regarding secondline therapy for advanced CRC can be made based on available data. Second-line chemotherapy with active agents is superior to best supportive care alone. Following 5-FU failure, active regimens include irinotecan, FOLFIRI, FOLFOX, and IROX, with IROX showing superiority to FOLFIRI. Following irinotecan-based first-line treatment, FOLFOX is, in general, the best choice, and the combination of FOLFOX plus bevacizumab appears to be superior to FOLFOX alone. Following FOLFOX failure, irinotecan and FOLFIRI are currently the most appropriate options. Irinotecan plus cetuximab should emerge as an effective regimen.