Chemotherapy for advanced colorectal cancer : Please aim for accuracy rather than hard hitting headlines

Editor—Michael and Zalcberg's editorial on chemotherapy for advanced colorectal cancer and the Colorectal Cancer Collaborative Group's systematic review of palliative chemotherapy for advanced colorectal cancer agree superficially.1,2 This makes the discrepancy between their conclusions surprising and raises certain questions of accuracy and presentation. The editorial overplays the value of such treatment, which was more scientifically assessed by the paper. Firstly, the editorial's statement that there is a 35% reduction in the risk of death for treated patients is an oversimplification of the results. In the systematic review, from which this conclusion was derived, fig 2 unambiguously shows that the risk of death is only about 16% less in the treated group.2 This translates to a median survival advantage of only four months and applies only to a period between 5 and 15 months from the start of treatment. This summary is more accurate than the one derived from the more complicated and inferred statistics used in the editorial. Secondly, the editorial tries to make a case for a palliative value of the treatment where the paper fails to find such a benefit. It makes that case for two subgroups of patients in single studies (references 4, 5, 6 in the editorial corresponding to references 15, 22, 17 in the paper). The result of this “splitting” contradicts the overall result in the paper: “data on the effect of chemotherapy on quality of life are inadequate to draw firm conclusions about the palliative benefit of chemotherapy.” Thirdly, the editorial glosses over the toxicity of chemotherapy, which could not be meaningfully summarised in the systematic review.2 Additionally, the literature possibly underestimates the toxicity of chemotherapy in view of the relatively younger population of the reviewed studies. The editorial lacks the expected balance of opinion. The impact of an editorial with a gripping headline in a journal such as the BMJ should be more seriously considered. The now freely available access to such an unbalanced message can lead to conflict between a terminally ill and vulnerable group of patients and their medical carers, resulting in unnecessary suffering and a possible break of trust.