HD Lighthouse Contributing Editor's Comment: A revolutionary new statistical method could allow all study participants to take the therapeutic agents themselves - no one would have to take a placebo.

The words "exciting" and "statistics" are rarely heard in the same sentence. But new biostatistics research by Dr. Xiaoping Xiong and his team may fundamentally change the way clinical trials are done, especially for orphan diseases like Huntington's. The new approach could allow all study participants to benefit from a new treatment, with no one relegated to a concurrent control group. So the method can accelerate medical research in general, and its special characteristics could provide more conclusive results to boot.

Based on 15 years of work, Dr. Xiong and his colleagues at St. Jude Children's Research Hospital and the University of Maryland Greenebaum Cancer Center have developed a new statistical 'sequential procedure' for monitoring clinical trials (see abstract below). It allows enrolling all patients in a research trial in the experimental group, with no need to establish the usual concurrent control group for comparison. (A common form of this control group, used when little is known, as in HD, is the placebo group.) Using the new approach, the experimental group is compared instead at various intervals throughout the study to an historical control group, consisting of patients who received some 'standard treatment' in a previous study.

At each point of comparison, experimental group data are compared to all of the historical control group data. This iterative analysis allows an early decision on whether the trial is showing positive results. Researchers using the new technique can pinpoint earlier that the findings up to that point would almost surely be the same if the entire original planned trial were carried out. So trials can often be concluded faster than planned.

The new analytic technique could help support exciting new studies in Huntington's disease and other conditions with small patient populations. When patient populations are small, as in HD, it can be hard to enroll enough subjects to divide them into statistically meaningful experimental and standard control groups. Using the new technique, statistically rigorous conclusions can be drawn even when no concurrent but only historical control group data are available.

The value of the new approach hits home in considering all the promising agents for HD - creatine, fish oil, CoQ10, trehalose, selective serotonin reuptake inhibitors (SSRIs), and others that are already available - without even considering those still to come. These available agents might be tested much faster and more conclusively using statistical techniques like those of Xiong and his team. Up until now, the practical constraints on testing so many agents in such a small patient population have been overwhelming.

The new approach is especially useful when the experimental treatment is believed to be better than the traditional treatment, because all patients can benefit from the new therapy. In this regard, too, the new method is especially apt for Huntington's, since there are numerous promising, readily available, relatively nontoxic agents still to be tested for efficacy.

At the same time, the new monitoring procedure allows the trial to be cut off reasonably quickly if the new treatment is not panning out.

The Xiong team approach is based on a statistical technique known as the sequential conditional probability ratio test (SCPRT). What is novel about the approach is its application to the experimental paradigm described above, a feat made possible by employing a mathematical model of Brownian motion. (Brownian motion is the minute random movement of particles in or floating on the surface of a fluid. Mathematical models of this motion are sometimes used in statistics to describe the behavior of random variables.)

The approach could be especially useful in combination trials, which are strongly needed in HD research. For example, using the new technique, the historical control group could consist of patients who have taken creatine, while the active experimental group could consist of those taking both creatine and fish oil.

Also important, all the advantages of the new approach mean that research can be accelerated as a whole. Studies can be done more easily and cheaply without a concurrent control group. Research results can be obtained faster using the new forms of statistical comparison to an historical group, provided the latter meets some basic requirements. All this may propel research in a critical area like HD.

When clinical trials are well conducted, Dr. Xiong and his colleagues say, the new statistical technique should be more rigorous than any other form of clinical trial, except for the RCT [the randomized clinical trial, the gold standard]. The RCT is best for dealing with larger patient populations and therapeutics whose value is not as clear, when larger numbers are needed to obtain valid results. The new technique is enormously promising, therefore, for circumstances in which RCTs are difficult to perform, as for orphan diseases like HD.

For the full press release from St. Jude's, see:
http://www.stjude.org/media/0,2561,453_2816_21507,00.html

For the full original research report, see:
http://www3.interscience.wiley.com/cgi-bin/abstract/112732033/ABSTRACT?CRETRY=1&SRETRY=0

-- Ann Covalt, M.A. and Malcolm Casale, Ph.D.
Posted to the HDL: 15 Oct 2006

A sequential procedure for monitoring clinical trials

X Xiong, M Tan, and J Boyett

In this paper, we develop a sequential procedure to monitor clinical trials against historical controls. When there is a strong ethical concern about randomizing patients to existing treatment because biological and medical evidence suggests that the new treatment is potentially superior to the existing one, or when the enrollment is too limited for randomization of subjects into experimental and control groups, one can monitor the trial sequentially against historical controls if the historical data with required quality and sample size are available to form a valid reference for the trial. This design of trial is sometimes the only alternative to a randomized phase III trial design that is intended but not feasible in situations such as above. Monitoring this type of clinical trial leads to a statistical problem of comparing two population means in a situation in which data from one population are sequentially collected and compared with all data from the other population at each interim look. The proposed sequential procedures is based on the sequential conditional probability ratio test (SCPRT) by which the conclusion of the sequential test would be virtually the same as that arrived at by a non-sequential test based on all data at the planned end of the trial. We develop the sequential procedure by proposing a Brownian motion that emulates the test statistic, and then proposing an SCPRT that is adapted to the special properties of the trial. # # #

Source: Statistics in Medicine, Aug. 10, 2006

printer friendly version

 

	Read the HDAC/HDLighthouse Forum. Post your comments