Method to Our Madness or Madness in Our Methods? Pitfalls in Trial Methodology

In the article that accompanies this editorial, the results of three clinical trials in patients with non–small-cell lung cancer (NSCLC) are reported. An important aim of an editorial can be to evaluate whether the design of the trial and the methods used to analyze the results permit readers to draw definitive conclusions that inform clinical practice or guide the development of future clinical trials. Each of these articles raises important issues about trial methodology that the reader may not have had the time to think about and that could influence the interpretation of the results. Ipilimumab is an exciting new targeted agent which specifically blocks the binding of cytotoxic T-lymphocyte antigen-4 (CTLA-4) to its ligands. This blockade augments T-cell activation leading to tumor regression. Ipilimumab was recently shown to improve survival in patients with advanced melanoma. 1,2 Lynch et al 3 conducted a randomized phase II trial in patients with NSCLC to estimate the additional activity of ipilimumab over chemotherapy alone using immune-related progression-free survival (irPFS) as an outcome measure. The use of a placebo for ipilimumab and the double-blind design strengthened the comparison of carboplatin and paclitaxel plus the immune-targeted agent with chemotherapy alone. Another important objective of the trial was to address the scheduling of ipilimumab and chemotherapy. Two schedules were tested: (1) ipilimumab administered concurrently with chemotherapy (concurrent), and (2) two cycles of chemotherapy followed by four cycles of chemotherapy and ipilimumab together (phased). The authors reported that the phased regimen improved irPFS significantly compared with the control regimen (hazard ratio [HR] 0.72; 95% CI, 0.50 to1.06; P .05), but the concurrent regimen did not (HR 0.81; CI, 0.55 to 1.17,P .13). It is the choice of the optimal regimen that we want to comment on. Not only is the phased regimen active but so is the concurrent one. Note that both HRs are less than 1 and in the Kaplan-Meier plots, the curves for both the phased and concurrent regimens are above that of the chemotherapy control (Fig 2 in publication). A larger sample size could well have resulted in the concurrent approach being significant too. It did not appear that the rules for selecting the optimum schedule were specified a priori. Nonetheless, the conclusion to support further investigation of phased-ipilimumab plus paclitaxel/carboplatin over the concurrent-ipilimumab chemotherapy regimen seems sensible given that the “phased” arm is consistently the best for all the outcome measures, and clearly superior for the progression (PFS) outcomes.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.