Implementing a randomized controlled trial with granulocyte colony‐stimulating factor prophylaxis to improve the outcome of high‐risk patients with colorectal cancer in the complex clinical setting of multiple surgical interventions

Background All large multicentre trials with immune modifiers in sepsis faded despite promising preclinical results. Several reasons for this have been discussed in the literature and include patient heterogeneity, different phases of pathophysiological state and trial errors. There are two additional and important points: the interaction of immune modifiers with the complex clinical management was not analysed preclinically and in multicentre trials the surgical management was not standardized. Methods Before implementing the clinical trial, the interaction of granulocyte colony‐stimulating factor (G‐CSF) with various surgical interventions was analysed in clinical modelling randomized trials (CMRTs) in rats, including different antibiotics, heparin, H 1 and H 2 antagonists, abdominal lavage, and management of preoperative and intraoperative bleeding. Good clinical practice rules were applied and the complexity of clinical reality was modelled in the CMRTs. In the clinical trial, high‐risk patients (American Society of Anesthesiologists grades III and IV) with colorectal cancer were identified and randomized to the placebo and G‐CSF groups ( n = 40 per group). G‐CSF prophylaxis (300 μg) was given subcutaneously 12 h before, and 12 and 36 h after operation. The three primary endpoints were health status measured with a modified McPeek index over 6 months, self‐reported quality of life determined with the European Organization for the Research and Treatment of Cancer QLQ38 questionnaire and the true endpoint defined 6 months after operation. Results Implementation of the trial demonstrated that a high standardization of surgical management is not possible in multicentre trials since the surgical management of patients with colorectal cancer is heterogeneous and management of the most important complication, anastomotic leakage after colorectal surgery, is performed according to personal experience and is not based on high levels of evidence from clinical trials low evidence based medicine (EBM). For this reason the surgical management of patients in the trial should be defined, a centre‐based clinical guideline for the management of anastomotic leakage applied and adherence to the guideline recorded. Conclusion Without CMRTs to analyse the interaction of immune modifiers in complex settings and defined clinical management of patients, no positive results will be obtained from such clinical trials. © 2000 British Journal of Surgery Society Ltd