Modern pharmacological products of molecular research are not always directed at single diseases—for instance, Yang and colleagues’ study was in renal cancer, and bevacizumab has been tried in non-malignant diseases too. However, the main current interest in this agent is in metastatic colorectal cancer, and the paper by Cunningham’s group is on this disease too, albeit with a different drug. A quick scan of the online Web of Science shows that review articles mentioning these two drugs vastly outnumber articles reporting randomized trials on them. The outlook for patients with advanced colorectal cancer did improve with the introduction of irinotecan and oxaliplatin and the considerable interest in the two recently approved monoclonal antibodies ("-mab") presumably reflects the hope that further improvement can be achieved. The New England Journal of Medicine editorials accompanying #14 and #19 are cautious, however. We are not talking about cures for an advanced malignancy that has already spread, and patients’ responses can be judged in different ways. For example, in the Hurwitz study (#7) the primary endpoint was survival, which was 20.3 months (median) in patients given a standard therapy plus bevacizumab compared with 15.6 months in those given standard treatment plus placebo. In the European trial (#19) of cetuximab plus irinotecan versus cetuximab alone, the chosen main endpoint was partial radiologic tumor response; there were no complete responses and partial response rates were 22.9% versus 10.8%, respectively, while survival times were not significantly different. The Hurwitz trial and two others on bevacizumab have lately been subjected to a meta-analysis (F.F. Kabbinavar, et al., J. Clin. Oncol., 23[16]: 3706-12, 2005). The combined data, for 241 controls and 249 patients receiving bevacizumab, show that median survival-times were 14.6 and 17.9 months, respectively, while progression-free survivals were 5.6. and 8.8 months—both significant in favor of bevacizumab without being dramatic.

Cancer patients needing chemotherapy of any sort will usually have had or will later require surgery, so a reasonable question to ask about the safety of drugs that inhibit growth factors is whether they also interfere with wound healing. Bevacizumab can do so but, as Frank A. Scappaticci and colleagues note from their study of combined trial data (J. Surg. Oncol., 91[3]: 173-80, 2005), few patients experience this problem—13 out of 305 patients given bevacizumab as compared with 2 of 223 controls (roughly 4% versus 1%), the difference being ascribable to the group operated on after the drug rather than before it.