Adding oxaliplatin or cetuximab to FOLFIRI resulted in higher RR, longer OS and higher rates of liver resection [6, 8]. To combine both approaches might be an effective way to intensify treatment in selected patients with good performance status and k-ras wild type tumours. However, both combinations were associated with increased toxicity compared to FOLFIRI alone. To prepare further trials, the aim of this phase I trial was to determine the MTD of the chemotherapy combination of cetuximab, irinotecan, oxaliplatin and 5-FU/FA for future studies in patients with non-resectable liver metastases.
In our study, we have shown that it is feasible to administer cetuximab in combination with oxaliplatin, 5-FU/FA and irinotecan. However, diarrhoea and neutropenia limited the maximum dose of irinotecan which is recommended for further trials to 125 mg/m2, lower than in the FOLFOXIRI schedule (165 mg/m2) without cetuximab .
Our findings of increased toxicity are in line with the POCHER trial investigating the same drugs in a chronomodulated schedule requiring a dose reduction during the trial . A French trial investigating cetuximab and FOLFIRINOX did not reduce the chemotherapy (180 mg/m2 irinotecan, 85 mg/m2 oxaliplatin and 5-FU/FA) but reported a diarrhoea grade 3/4 rate of 52%, and a neutropenia grade 3/4 rate of 38% .
The confirmed response rate of 75% (irrespective of the k-ras status) is promising and - in the historical comparison - numerically higher than FOLFOXIRI (response rate of 66% ) or cetuximab/FOLFIRI (confirmed response rates 57.3% in k-ras wild type and 39.7% in k-ras mutant patients ). However, the low patient number and the subsequently large confidence interval in our study do not allow any final conclusion, especially for subgroup analysis, i.e. regarding the unexpected finding that the response was slightly higher in k-ras exon 2 mutant patients or patients or the known phenomenon of numerically higher response rates in patients with metastases in liver or lung, only.
In this trial that did not select patients with potentially resectable metastases but included a relatively high proportion of patients with more than one metastatic site (55%), three patients were resected for metastases and one patient stopped chemotherapy prematurely because of an early complete response. The overall survival belongs to the longest reported in clinical trials for metastatic colorectal cancer that were not conducted in the setting of neoadjuvant therapy. Patient selection, the high number of patients with second line treatment (90%) and the concept of an intensive induction treatment may have contributed to the favourable outcome.
The short time until best response of 3.1 months might be important for neoadjuvant chemotherapy in potentially resectable metastases, as a higher number of chemotherapy cycles is associated with higher perioperative morbidity . Even though the distribution of PS is comparable to other trials investigating intensive chemotherapy combinations , we do not recommend this schedule for patients with an impaired PS, considering, that 75% of the patients in our trial had a WHO PS 0. Our regimen might be a treatment option for patients with potentially resectable liver metastases, in whom the chance of a curative resection outweighs the risk of the high toxicity of such an intensive regimen, and also in other patients with a high need for intensive treatment . The ongoing CELIM2 study is investigating this chemotherapy schedule in patients with k-ras wild-type liver metastases (NCT01802645).