We are downgrading Lilly to Neutral from Outperform.
We believe investor reaction to the TIMI-38 data [from a phase III trial that compares Lilly's and partner Daiichi Sankyo's prasugrel blood-thinning agent with Plavix in a trial to prevent recurring heart attacks and strokes in people with cardiovascular disease] will be negative and Lilly stock will decline.
We believe prasugrel will be associated with statistically more bleeding than Plavix [which is co-marketed by Sanofi-Aventis and Bristol-Myers Squibb], but also significantly better effectiveness. Such a profile should be approvable given that safety likely is not only judged on a [statistical] p-value, but on relative benefit/risk, which should favor prasugrel for some patients.
However, this implies low potential and our prasugrel sales and Lilly earnings-per-share forecasts have been reduced accordingly. Lilly now appears poised to deliver industry-average growth through 2012.
TIMI-38 enrolled 13,614 patients with moderate to high-risk acute coronary syndrome undergoing percutaneous coronary intervention. Patients were randomized to prasugrel 60/10 mg daily or Plavix 300/75 mg daily for up to 15 months. The primary end point is the time of the first event of a composite cardiovascular death, myocardial infarction, or stroke. The major safety end points include thrombolysis in myocardial infarction (TIMI) major and minor bleeding unrelated to coronary artery bypass graft surgery.
TIMI-38 was designed to continue until 875 subjects with UA/NSTEMI [unstable angina and non-ST elevation myocardial infarction] reach one of the components of the primary end point. This provided 90% power to establish superiority relative to the primary composite end point in the UA/NSTEMI group.
Lilly/Sankyo based this calculation on the following assumptions: (1) 10.5% of subjects in the Plavix group would develop the primary end point within one year; (2) an average hazard ratio of 0.80; and (3) the time-to-first-event analysis based on a two-sided log-rank test used at a two-sided significance level (alpha) of 0.05 to assess superiority.
The comparison of the treatment groups relative to primary and secondary efficacy end points are carried out using time-to-first-event analysis via the Gehan-Wilcoxon test. The time-to-event methodology accounts for patients lost to follow-up and incomplete treatment duration/follow-up at the interim analyses.
Using a biostatistical consultant, we analyzed TIMI-38's stopping rules, the primary efficacy end point, and the safety end points in order to hone in on potential outcomes. Given that we do not have access to patient level data we could not perform a time-to-event analysis and therefore our statistical analysis is based on simple events rates. There are limitations with this approach including: an inability to do a reasonable evaluation of the comparative efficacy for similar lost to follow-ups (LFTUs) in the two treatment groups, and a high relative risk associated with what is required for statistical significance when events occur at a very low incidence (less than 1%-2%).
Three planned interim analyses were conducted to determine whether the trial should be stopped. These occurred when 250, 450, and 650 patients reached the primary composite efficacy end point. The trial could have been stopped for safety, futility, and overwhelming efficacy.
Safety was reviewed at all three interim analyses; the trial could have been terminated because of excessive mortality and life-threatening bleeds in the prasugrel arm or if the predicted incidence of life-threatening bleeding exceeded historic levels.
The futility assessment was done at the second and third interim analysis and the assessment for overwhelming efficacy was only done at the third interim analysis.
Given that the trial was not stopped for either overwhelming efficacy or futility, we focused on the safety stopping rules to gain further insight regarding prasugrel's and Plavix's bleeding rates. However the interim safety analyses were done only evaluating statistical differences between life-threatening bleeds and mortality between the two arms, or a bleeding rate higher than historical levels.
TIMI-38 was unlikely to be stopped on statistical differences in TIMI major and minor bleeds. Therefore the hurdle for early termination was very high. The fact that it was not stopped tells us little about the rate of TIMI major and minor bleeds. <<
>> Daiichi Sankyo hikes H1 outlook, nudges up full-yr
Fri Oct 26, 2007 5:49am EDT By Edwina Gibbs
TOKYO, Oct 26 (Reuters) - Daiichi Sankyo Co Ltd said its first-half net profit would beat its initial estimate by 36 percent aided by strong drug sales overseas, but it only nudged its annual outlook higher as it will likely spend more to buy rights to new drug candidates.
The Japanese drug maker, which is gearing up for a major U.S. expansion, has been benefiting from robust U.S. sales of its high blood pressure medicine Benicar and favourable currency rates.
It said it now expects net profit for April-September to come in at 60 billion yen ($525 million), better than its July forecast of 44 billion yen, although that would still be 10 percent below last year's result. The expected first-half rise was also due to the company postponing booking some costs until the second half.
The drug maker, formed when Sankyo bought Daiichi Pharmaceutical in 2005, has surprised itself and analysts with cost-cutting results and sales strength. For the full year, Daiichi Sankyo said it expects to book a net profit of 100 billion yen, which is in line with analysts' estimates.
The figure is 8.7 percent higher than its previous forecast and 27 percent better than its year-earlier result.
This week its shares were hit by news that it and Eli Lilly and Co (LLY) would stop giving their most important experimental medicine, the blood-clot preventer prasugrel, to patients in two small phase II trials because the dosage may need to be changed for certain patients. Prasugrel is a serious potential rival to Bristol-Myers Squibb's (BMY) Plavix, one of the world's biggest-selling medicines.
Analysts gave various reactions to the news, but many in Tokyo said they were reserving judgment until they see data from phase III testing, the results of which are due to be released on Nov. 4.
UBS Securities analyst Hirohisa Shimamura said in a report that there had been no change to his expectations that prasugrel was likely to show superior efficacy but an inferior safety profile such as minor bleeding, compared with Plavix. The likelihood of the Food and Drug Administration requesting follow-up data had increased, which would slow time to market, he said.
Prior to the [first-half results] announcement, Daiichi Sankyo's shares ended 0.3 percent lower at 3,150 yen after tumbling 10.3 percent a day earlier on the prasugrel news. <<
Market Data 
Markets 
AI
