Mojojo, I have the smoking gun here about ASA404 (see below). The upshot is that the phase II trial showed an increase in survival of 5.2 months with a hazard ratio of 0.73, but a p-value of 0.33. So it had a good hazard ratio, but a bad p-value. Yet it was moved into phase 3 anyway. Contrast that to the bavi phase 2 second-line NSCLC trial with an increase in survival of 6.5 months (using the pooled data) with a hazard ratio of 0.52 and a p-value of 0.015. Here is the phase 2 paper. British Journal of Cancer (2008) 99(12), 2006 – 2012 FREE: http://www.nature.com/bjc/journal/v99/n12/pdf/6604808a.pdf "Median survival was 14.0 months in the ASA404-CP group and 8.8 months in the CP group (Figure 2). The risk of death was reduced by 27% in the ASA404-CP group, with a hazard ratio of 0.73, 95% CI 0.39, 1.38, and P=0.33" -------------------------------------------------------------------------------------------------------- From JCO August 1, 2011 vol. 29 no. 22 2952-2955. Clinical Development of Vascular Disrupting Agents: What Lessons Can We Learn From ASA404? Patricia M. LoRusso, Scott A. Boerner, Sally Hunsberger [snip] In phase I/II clinical trials, ASA404 showed promising results in combination with paclitaxel/carboplatin for the treatment of NSCLC and resulted in improved tumor response and increased time to disease progression, partial response, and 5-month median survival advantage in both patients with squamous NSCLC and those with nonsquamous NSCLC. These promising early results prompted examination of the combination in two global, large-scale, randomized phase III clinical trials, named ATTRACT-1 (Antivascular Targeted Therapy: Researching ASA404) and ATTRACT-2. ATTRACT-1 was designed to evaluate ASA404 in combination with paclitaxel and carboplatin as first-line treatment for stage IIIB/IV NSCLC of squamous or nonsquamous histology, whereas ATTRACT-2 evaluated the combination as second-line treatment for patients with advanced NSCLC. Both studies were designed with the primary end point of determining whether the addition of ASA404 significantly extended overall survival (OS). In the article that accompanies this editorial, Lara et al describe the results of the ATTRACT-1 trial, which was conducted at more than 200 sites in 20 countries and involved 1,299 patients, 649 of whom received ASA404. At interim analysis, no difference in OS was observed between ASA404 and placebo arms, and the clinical trial was halted for futility. In March 2010, a statement was released that indicated that first-line ASA404 therapy had failed to meet the primary end point of significantly extending OS for patients with advanced NSCLC who were enrolled onto the ATTRACT-1 clinical trial. Similarly, in November 2010, it was announced that there was no observed improvement in OS with administration of ASA404 as second-line treatment of patients with advanced NSCLC in the ATTRACT-2 phase III trial. Should we be surprised that the phase III studies produced negative results? Unfortunately, probably not. The basis of the decision to move forward with a phase III study was likely heavily influenced by results from the randomized phase II study of ASA404 plus paclitaxel/ carboplatin, including observed increases in median OS (8.8 months v 14 months, with a corresponding hazard ratio of 0.73), response rate, and time to progression in the ASA404 arm (Table 1). However, interpreting these statistics alone, without accounting for the variability associated with them, can be misleading. The P value is essential to weighing the evidence in a phase II study. It is the probability of observing differences as extreme or more extreme than what were observed, if the treatments are exactly the same. In the calculation of the P value, the sample size and variability of the estimates are taken into account. An alternative way to understand the P value of .33 that was observed in the phase II study of ASA404 is to consider what would occur if 100 studies were run in the same population, with each study having two treatment arms of 35 patients and both arms receiving the chemotherapy regimens with different placebos. [snip] So what has this trial taught us about advancing novel agents into the phase III arena? Carefully scrutinizing phase II data and defining efficacy differentials may be pivotal to decrease the failure rate of future phase III trials. Excitement over differences in median survival must be tempered when P values indicate the differences have a high probability of being a result of chance. Minimizing the failure rate can save valuable resources, not only in terms of funding, but as it relates to patients as well. Even if there is no evidence of increased toxicity or alterations in quality of life between the novel combination and standard arms, randomly assigning a patient to a phase III trial still requires a valuable time commitment from a patient who is racing against the clock. Additionally, such patients could have been considered for alternative trials that, predicated on possible patient preselection, may have resulted in a significantly improved outcome for them. Whenever possible, defining appropriate patient subsets for protocol recruitment may be the tool that significantly influences failure rates, heightens enthusiasm for additional trial recruitment, and, most importantly, increases patient survival. So why are many positive phase II trials not resulting in positive phase III trials? Perhaps it is because, in those situations wherein positive subsequently begets negative, we may never have had a truly positive phase II trial to begin with.
News 
Market Data 
Discover 
AI
