Analyzing overall survival in randomized controlled trials with crossover and implications for economic evaluation.

BACKGROUND: Offering patients in oncology trials the opportunity to cross over to active treatment at disease progression is a common strategy to address ethical issues associated with placebo controls but may lead to statistical challenges in the analysis of overall survival and cost-effectiveness because crossover leads to information loss and dilution of comparative clinical efficacy. OBJECTIVES: We provide an overview of how to address crossover, implications for risk-effect estimates of survival (hazard ratios) and cost-effectiveness, and how this influences decisions of reimbursement agencies. Two case studies using data from two phase III sunitinib oncology trials are used as illustration. METHODS: We reviewed the literature on statistical methods for adjusting for crossover and recent health technology assessment decisions in oncology. RESULTS: We show that for a trial with a high proportion of crossover from the control arm to the investigational arm, the choice of the statistical method greatly affects treatment-effect estimates and cost-effectiveness because the range of relative mortality risk for active treatment versus control is broad. With relatively frequent crossover, one should consider either the inverse probability of censoring weighting or the rank-preserving structural failure time model to minimize potential bias, with choice dependent on crossover characteristics, trial size, and available data. A large proportion of crossover favors the rank-preserving structural failure time model, while large sample size and abundant information about confounding factors favors the inverse probability of censoring weighting model. When crossover is very infrequent, methods yield similar results. CONCLUSIONS: Failure to correct for crossover may lead to suboptimal decisions by pricing and reimbursement authorities, thereby limiting an effective drug's potential.

Improvement in long-term outcomes with successive Total Therapy trials for multiple myeloma: are patients now being cured?

The concept of applying all active therapeutic agents in Total Therapy (TT) clinical trials for newly diagnosed multiple myeloma was pursued with the intent of developing curative treatment. The results of TT1 (n=231), TT2 (n=668) without or with thalidomide and TT3 with added bortezomib (n=303) have been reported. An update with median follow-up times of 17.1, 8.7 and 5.5 years, respectively, is provided. Conditional overall survival (OS) analysis from a 4-year landmark was applied to account for earlier protocol failure owing to disease aggressiveness and toxicities. Cumulative relative survival was computed in the context of age- and gender-matched US population, and interval-specific relative survival ratios were estimated to determine times to normal survival expectation. Based on Cox model-adjusted statistics, OS, progression-free survival and complete-response duration all improved with the transitions from TT1 to TT2 to TT3; improvement was also evident from time-to-progression estimates, 4-year conditional survival data and cumulative relative survival. Interval-specific relative survival normalized progressively sooner, reaching near-normal levels with TT3 in patients who attained complete response. Thus, a strategy using all myeloma-effective agents up-front seems effective at preventing, in progressively larger patient cohorts over time, the outgrowth of resistant tumor cells that account for ongoing relapses.

Kidney cancer: overall survival is an unsuitable primary end point.

Final results of the CALGB 90206 and AVOREN trials have failed to demonstrate an overall survival benefit for bevacizumab plus interferon compared to interferon monotherapy in patients with metastatic renal cell carcinoma. Progression-free survival or objective response rate might be more suitable primary end points in phase III trials.

The risk of death among adult participants in trials of antipsychotic drugs in schizophrenia.

In this paper we investigate mortality in short-term placebo-controlled trials conducted to demonstrate efficacy and safety of atypical antipsychotic drugs for the treatment of schizophrenic patients in the acute phase of illness. Arguments are provided, why short-term placebo-controlled studies are required. This is an integrated analysis of results from randomized placebo-controlled trials conducted pre-licensing for risperidone, olanzapine, quetiapine IR, ziprasidone, risperidone consta, aripiprazole, paliperidone, and quetiapine XR. Information was retrieved from study publications, EPAR, and from the FDA SBA, all in the public domain. 7553 patients were randomized in the remaining 23 short-term acute phase clinical trials. 2/5738 patients died after having been randomized to an active treatment. 5/1815 died in the placebo group. The crude odds-ratio for an increased death risk with placebo treatment is 7.92 (95% confidence interval (1.45 to 40.87), two sided P=0.0134). Death narratives show: (i) both deaths in active treatment groups occurred during randomized treatment period, (ii) two deaths in the placebo group of a trial in the elderly were observed in patients with severe co-morbidities not usually included in a randomized trial, and (iii) two further deaths in the placebo groups occurred 9 and 23 days after the end of the randomized treatment period. Under these circumstances the crude odds-ratio for an increased risk of death for patient with placebo as compared to active treatment is 1.58 (95% confidence interval (0.14-17.45), two sided P=0.71). Confidence intervals are generally wide indicating a still limited knowledge about a potential increase in mortality with placebo treatment. Unless, however, society is willing to take the risk that ineffective drugs are licensed and cause undetected harm thereafter, or is willing to restrict licensing to drugs that are superior to current treatments, short-term placebo-controlled trials in the acute phase of schizophrenia are necessary. Measures are proposed to minimize risks.

Influence of enrollment sequence effect on observed outcomes in the ADDRESS and PROWESS studies of drotrecogin alfa (activated) in patients with severe sepsis.

INTRODUCTION: We performed a study to determine whether an enrollment sequence effect noted in the PROWESS (recombinant human activated Protein C Worldwide Evaluation in Severe Sepsis) trial exists in the ADDRESS (Administration of Drotrecogin Alfa [Activated] [DrotAA] in Early Stage Severe Sepsis) trial. METHODS: We evaluated prospectively defined subgroups from two large phase 3 clinical trials: ADDRESS, which included 516 sites in 34 countries, and PROWESS, which included 164 sites in 11 countries. ADDRESS consisted of patients with severe sepsis at low risk of death not indicated for treatment with DrotAA. PROWESS consisted of patients with severe sepsis with one or more organ dysfunctions. DrotAA (24 microg/kg per hour) or placebo was infused for 96 hours. RESULTS: In ADDRESS and PROWESS, there was a statistically significant interaction between the DrotAA treatment effect and the sequence in which patients were enrolled. In both trials, higher mortality was associated with DrotAA use in the subgroup of patients enrolled first at study sites. Compared with placebo, PROWESS mortality was lower with DrotAA treatment for the second and subsequent patients enrolled, whereas in ADDRESS, mortality remained higher for the second patient enrolled but thereafter was lower for DrotAA-treated patients. Comparison of patients enrolled first with subsequent patients enrolled indicated that the characteristics of patients changed. Subsequently enrolled patients were treated earlier, were less likely to suffer nonserious bleeds (ADDRESS), and experienced fewer protocol violations (PROWESS). CONCLUSIONS: Analyses suggest that an enrollment sequence effect was present in the ADDRESS and PROWESS trials. Analysis of this effect on outcomes suggests that it is most apparent in patients at lower risk of death. In PROWESS, this effect appeared to be associated with a reduction of the DrotAA treatment effect for the first patients enrolled at each site. In ADDRESS, this effect may have contributed to early termination of the study. The finding of an enrollment sequence effect in two separate trials suggests that trial designs, site selection and training, data collection and monitoring, and statistical analysis plans may need to be adjusted for these potentially confounding events. TRIAL REGISTRATION: ADDRESS trial registration number: NCT00568737. PROWESS was completed before trial registration was required.