Presently, three clinically distinct groups of metastatic colorectal cancer patients can be differentiated: 1) patients with resectable disease; 2a) patients who require conversion chemotherapy or (2b) patients with highly symptomatic or rapidly progressive disease; 3) patients with disseminated metastasis and mostly asymptomatic disease. While only few studies have engaged to investigate these subgroups separately, most clinical trials were performed in the whole population of mCRC patients.
It appears that group 3 represents the largest subgroup. For this reason, several clinical studies have produced results acceptable for this subgroup. Patients in this group do not clearly benefit from rapid remission induction and may rather profit from prolonged disease control.
A previous report by Grothey et al. analysed two trials where patients either received chemotherapy alone (N9741-trial: FOLFOX vs. IFL) or chemotherapy plus bevacizumab (AVF2107g: IFL + bevacizumab versus Bevacizumab alone) . In both trials, the superior regimen improved PFS and OS regardless of tumour response. This led to the conclusion that tumour response in mCRC is not a necessary determinant of therapeutic benefit. This conclusion can be extended to the point that most probably a large fraction of patients simply benefits from non-progressive disease. If disease control becomes an important endpoint, then trials have to be designed where strategies including reduced treatment intensity may reach this goal in a defined population of group-3 patients.
The present trial investigates an escalation strategy where patients start with a well tolerated single-agent chemotherapy (capecitabine) combined with bevacizumab. Patients treated on this study arm are allowed to escalate to CAPIRI plus bevacizumab once disease progression occurs. This escalation strategy is compared to a control arm where patients receive first-line chemotherapy with CAPIRI plus bevacizumab. After 6 months of treatment or if intolerable toxicity is faced, treatment can be de-escalated to capecitabine plus bevacizumab. In case of progression, it can later be re-escalated to CAPIRI plus bevacizumab. This innovative approach therefore offers two strategies which allow a flexible response to the individual requirements of the patient. The study seeks to demonstrate that patients attributed to group 3 have an adequate treatment benefit with a well tolerated low-intensity first-line regimen provided that further treatment is appropriately applied if necessary. As described in the above the CAPIRI regimen with a dose of capecitabine 800 mg/m2, irinotecan 200 mg/m2 previously investigated in two AIO randomized phase II trials is expected to have an acceptable toxicity profile [7, 11].
In the pre-antibody era, an escalation strategy has already been investigated in the CAIRO (CApecitabine, IRinotecan, and Oxaliplatin in advanced colorectal cancer) study. Koopmann et al. reported a trial arm where treatment was escalated in sequential steps from capecitabine (1st-line) to irinotecan (2nd-line) to CAPOX (3rd-line) . Comparison with the sequential application of CAPIRI (1st-line) and CAPOX (2nd-line) led to the interpretation that "combination treatment does not significantly improve overall survival compared with the sequential use of cytotoxic drugs" .
Another study of the pre-antibody era was the MRC-FOCUS trial. This study compared arm A) 5-FU/FA followed by irinotecan at progression to arm B) 5-FU/FA followed by combination chemotherapy and to arm C) combination chemotherapy from the onset. Also this large study challenged the assumption that in a non-curative setting maximum tolerable treatment must necessarily be used first-line .
Comparable results were reported by Bouché et al. who compared the sequence LV5FU2 followed by FOLFOX (2nd-line) followed by FOLFIRI (3rd-line) to the more conventional sequence of FOLFOX followed by FOLFIRI (2nd-line) . Again, this trial confirmed the notion that initially intensive regimens do not induce a superior outcome compared to well tolerated single-agent first-line strategies. However, it is important to state that this conclusion is true most likely for patients treated within group 3.
Since the CAIRO- and the MRC-FOCUS trial, perspectives on curability of disease have changed specifically with regard to the interpretation of surgical respectability . Also the introduction of antivascular and anti-EGFR treatment strategies had a marked impact on treatment outcome. The present trial picks up the theme of escalation chemotherapy, investigates this strategy specifically in group-3 patients and analyses the effect of chemotherapy in the setting of concomitant antiangiogenic therapy with bevacizumab.
In the control arm of our trial, patients may de-escalate treatment from initial combination therapy with CAPIRI plus bevacizumab to a maintenance therapy with capecitabine plus bevacizumab. De-escalation is planned after 6 months of treatment. However it can take place earlier in case of treatment-associated toxicity. A 6-month duration of first-line combination chemotherapy has been described in several other trials . Clearly, the optimal time point of de-escalation and the optimal maintenance regimen still have to be defined. The present trial will contribute to this important topic. Re-escalation from single-agent maintenance to combination chemotherapy is another important aspect of this strategy. Results previously obtained from the OPTIMOX1 trial indicate that reintroduction of chemotherapy had an independent and significant impact on overall survival .
The present trial does not simply investigate two treatment regimens, but rather compares two strategies of treatment . For this reason, time to failure of strategy (TFS) was chosen as a primary endpoint of the study. TFS is defined as the time between randomization and final failure of CAPIRI-Bev treatment in both treatment arms. This end point allows a wide range of options to manage mCRC patients while remaining on study thus providing a more complete assessment of a strategy's benefit .
In case of comparable TFS times, toxicity will be evaluated according to NCI CTC in a predefined score system using symptomatic grade 2-4 toxicities per cycle. This innovative design allows to analyze both, the therapeutic strategies and the associated toxicity.
When this study is performed, it is necessary to also define those patients who are clearly not candidates for this trial. Due to the selected control arm, all patients considered for the trial must be able to tolerate first-line combination chemotherapy with CAPIRI. The present study has not been specifically designed for elderly or frail patients and clearly excludes those with an ECOG performance status ≥ 2. It also excludes patients who present with symptomatic tumour disease. These patients need a fast treatment response which can best be achieved by intensive combination therapy.
While the present trial requires that parameters of the EGFR-pathway are determined, KRAS mutation of the tumour does not affect treatment within the trial. Since treatment efficacy of bevacizumab so far is not considered to depend on KRAS mutational status and since patients receive bevacizumab in both treatment arms, also patients with KRAS mutation can take part in the study.