TO THE EDITOR: We read with special attention the letter from Hochster, Grothey, and Childs about preliminary results of the Combined Oxaliplatin Neuropathy Prevention Trial. The aims of this study were to explore two approaches in first-line chemotherapy with infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX) bevacizumab for metastatic colorectal cancer— the Stop-and-Go strategy from the French Oncology Research Group (GERCOR) group developed in the Stop-and-Go strategy study and the prevention of neurotoxicity with calcium and magnesium (Ca Mg ) infusion. According to these authors, preliminary results have led to suspect a lesser response rate in the two arms of patients treated with FOLFOX-bevacizumab plus Ca -Mg , and the authors decided to stop the administration of Ca Mg . We would like to step back to the concept of Ca -Mg infusions for preventing oxaliplatin-induced neurotoxicity. Obviously, oxaliplatin-induced neurotoxicity is the drug-limiting toxic adverse effect. It alters patients’ quality of life and can lead to postponed or even interrupted treatments with oxaliplatin. Considering the unique profile of acute oxaliplpatin-induced neurotoxicity, we suspected that certain ionic channels could be involved as it has been described in channellopathies, or Na channels inhibitors poisoning, such as tetrodotoxin. We showed in electrophysiological studies with patch-clamp technique that oxaliplatin, and more precisely one of its two direct metabolites, the oxalate, which is a potent Ca chelator, interferes with certain ionic channels expressed on the peripheral neurones (Ca dependent voltage-gated Na channels). On the other hand, diamminocyclohexane platin, the active cytotoxic metabolite, had no effect on the Na channels. The hypothesis was that when oxaliplatin enters the neurones, it splits into diamminocyclohexane platin and oxalate, and the latter one chelates Ca and interferes with Na channel activity, leading to a neurone ehyperexcitability, and secondarily, to a chronic neuropathy. We tested Ca and Mg infusion as chelators of oxalate to inhibit its action on Na channels and could show in a retrospective study a dramatic decrease of acute and chronic neurotoxicity, with no impact on oxaliplatin efficacy. In the French multicentric “Neuroxa” study, patients were randomly assigned in a double blind randomized study designed to prove the interest of Ca and Mg infusions for preventing oxaliplatin neurotoxicity. One hundred forty-four patients have been treated with the FOLFOX 4 regimen, with 85 mg/m oxaliplatin every 2 weeks. They either received or did not receive calcium gluconate 1g and magnesium sulfate 1.5 g, administered in 250-mL glucose serum just before and just after oxaliplatin infusion. In the first 52 patients treated for metastatic colorectal cancer in first-line therapy, we reviewed the data in term of patients’ characteristics, efficacy, and neurotoxicity. These 52 patients are split into two groups, with and without Ca and Mg infusion, respectively. Since all the data are not collected and treated yet, we did not take off the blind. These preliminary results were presented at the 2007 International Society of Gastrointestinal Oncology meeting in Philadelphia, PA, and will be published in its newsletter. The two groups are well balanced in term of age, Eastern Cooperative Oncology Group (ECOG) performance status, primary tumor, number of metastases, and number of metastatic sites. There is neither difference in term of objective response rate (50% v 53%, P .45), nor in progression-free (12 0.6 SE v 12 0.5 SE, P .79) and overall survivals (25.1 4 SE v 25.5 4.1 SE, P .45). On the other hand, we report a significantly lower frequency and grade of oxaliplatin neurotoxicity in one group (5% v 24% grade 3 National Cancer Institute Common Toxicity Criteria, P .001). According to these preliminary results, we confirmed that there is no impact of Ca and Mg infusions on oxaliplatin efficacy, and there is a significant effect on its neurotolerance. These results seem to be in contradiction with those presented by Hochster et al. Some major points must be highlighted; we reported results on FOLFOX combination, and we have no background about a potential interference between Ca and Mg infusions and bevacizumab. However, de Gramont reports a high objective response rate, 62.5%, in a phase II study with patients treated with FOLFOX bevacizumab with Ca -Mg (personal communication from A. de Gramont, December 2007). According to these results, coordinators of the Combined Oxaliplatin Neuropathy Prevention Trial should make sure that other parameters have been taken in account, such as being certain that the compounds have not been mixed altogether. We always emphasize the importance of not coadministering oxaliplatin, flourouracil, folinic acid, and Ca -Mg infusions, and we recommended to infuse Ca -Mg just before and after oxaliplatin administration. One hypothesis is that the wish to reduce the time of infusions for outpatients could have lead to the administration of some compounds at the same time and in the same tubulure. This could have led to a diminished drug stability and, thus, a reduced activity. According to our results, we continue to propose Ca -Mg infusions to reduce oxaliplatin neurotoxicity, in FOLFOX regimens, provided that the compounds are delivered sequentially and not in the same time.