Friday, May 26, 2006

Probing Pagoclone EXPRESS

I was told that a manuscript of the Pagoclone study will soon be published with more details, which will also allow a discussion on effect sizes. So we need to wait a bit. In the mean time...

I looked at the press release a bit more, here are my thoughts:


“Indevus has completed what we believe is the largest pharmaceutical trial ever conducted for stuttering and the results are very exciting,” stated Glenn L. Cooper, M.D., chairman, president and chief executive officer of Indevus. “Our results today show that stuttering, a condition with no approved pharmacological treatment, is potentially treatable with Pagoclone. This study was designed as an exploratory trial to follow up on a limited number of observations of the effect of Pagoclone on stuttering during previous anxiety trials. The design of the EXPRESS trial enabled us to evaluate the condition from several clinical perspectives and we believe this provides us with a strong foundation to develop a clinical plan for further development.”
He says a lot without saying anything... He could be French or a politician... :-)


Gerald A. Maguire, M.D., associate professor, department of psychiatry, University of California, Irvine School of Medicine, stated, “Being a person who stutters and a physician who researches and treats stuttering, I am very excited about the results of this trial. As an investigator on this study, I saw first hand the positive impact pagoclone can have on the lives of patients. Consistent with the results of the entire study sample, more than half of my pagoclone treated patients had a clinically meaningful decrease in the severity of their stuttering. Although there is no cure for stuttering, pagoclone holds significant promise as a well-tolerated, effective and viable treatment for the millions of Americans who stutter.”

Jerry Maguire's statement is more revealing. My interpretation of his statement (which MIGHT BE COMPLETELY WRONG!) is the following: He is mostly excited about the trials, because it is the first large scale study that has been done and the drug shows more effects than any other drug but he is not excited because Pagoclone is a cure. And, not all stutterers benefit from Pagoclone, and those who do reduce their stuttering noticeably but do not eliminate it.



8-week, placebo controlled, double-blind, multi-center trial with an open label extension. There were a total of 132 patients randomized in the trial.

They have followed standard procedure.


EXPRESS (the name of the study) used the following measures:

Frequency and Duration Subscale of the Stuttering Severity Instrument Version 3 (SSI-3)
the Stuttering Severity Scale (SEV)
the Subjective Screening of Stuttering (SSS) Severity Subscore
the Clinician Global Impression-Improvement (CGI-I)
the Liebowitz Social Anxiety Scale (LSAS)
the Speech Naturalness Scale (SNS)


They claim that the measures ( SSI-3, SEV, and SSS) are validated stuttering measures.
My question:
What do they mean by "validated"? Have they done a battery of statistical checks to see how consistent the measure is? My experience is that the stuttering measures taken at week 0 are NOT normally distributed, and there are some significant outliers, i.e. some stutterers are extremely disfluent or fluent. A massive improvement of 1-2 extreme stutterers give statistical significant results for the whole group.


For all measures except SEV, the distributions of measured values for the control group and the placebo group were statistically significantly different at p-values ranging from 0.007 to .08; (typically p=0.05 is regarded as "worth noting" as only 1:20 is the result of a statistical fluctuation, and p=0.01 is considered "reliable").
My question:
1) What happens if you give the medication to normal people? I would guess that their CGI-I and LSAS would improve, too?
2) I want to see the effect size, i.e. I want to know how big the effect is.
3) Is the effect only for some people or all people?


The SEV measure is rated by clinicians from "no stuttering" to "extremely severe". They write: "The on-treatment effect of patients receiving pagoclone demonstrated a numerically superior rating versus patients treated with placebo (p=.18)."
My question:
The p-value is quite high, p=.18, which means that there is a 20% probability that the effect does not exist and is due to statistical fluctuation (i.e. chance).


On SSS, they write "The SSS Severity Subscore, measured at week 2, week 4 and week 8, is a validated, patient-rated assessment of stuttering that takes into account specific speaking situations that have taken place over the prior week. Pagoclone produced a statistically significant reduction at week 2 (p=.004) and week 4 (p=.05) and a trend for significant improvement at week 8 (p=.08) as compared to placebo."
My question:
What is interesting is that the p-values go up with time, from p=0.004, to p=0.05 to p=0.08. This makes me a bit suspicious that they experience a high / placebo effect which decreases with time. Listening at the reports from patients, they pretty well all know whether they are on Pagoclone or not, even though it is blind. So they might experience a high in their subjective assessment of fluency. This possibility can be eliminated by 4-months data.

4 comments:

Holger Stenzel said...

Hey Tom,

Before I take Pagoclone: "The drug has been commercially available for 3-5 years without major negative news and gossip."

wo bleiben denn hier Forscherdrang und Opferbereitschaft für die Wissenschaft?
Da hat das British Empire aber schon Zeiten mit mehr Pioniergeist erlebt!

Apropos: The Junkie ist auf Entzug, d.h. ich habe Olanzapine seit 12 Wochen abgesetzt.

Holger

Tom Weidig said...

Dafuer haben wir The Junkie! :-)

I am a bit cautious here, but that is my nature.

This is just the advice I would give to others.

However, I might take more risks... under some circumstances... maybe

Tom Weidig said...

Yes, I might be a bit over-cautious.

This is just my advice to others.

Maybe I would take more risks. I might try it out for a few weeks. But long-term use... I dont know...

Holger Stenzel said...

Hallo Tom,
ich finde bei der Bewertung der Ergebnisse der Pagoclone Studie sollte bedacht werden, was denn hier eigentlich realistisch zu erwarten ist. Das Cyclopyrrolonderivat Pagoclone, ein GABA-A Modulator mit der mutmasslichen Formel (+)-2-(7-chloro-1,8-naphthyridin-2-yl)-2,3-dihydro-3-(5-methyl-2-oxohexyl)-1H-isoindol-1-one wurde ursprünglich nicht für die Indikation Stottern entwickelt. Das wäre ja zur Zeit auch noch gar nicht möglich, da die mit dem Stottern assoziierten Stoffwechselvorgänge nicht genau genug bekannt sind um den Pharmaforschern als Basis zur Synthetisierung von Wirkstoffen zu dienen. Also wurde bei den bisherigen Medikamentenstudien versucht durch klinische Versuche mit verfügbaren Pharmaka anderer Indikation, so auch jetzt bei Pagoclone, sich im Umkehrschluss der Neuropharmakologie des Stotterns zu nähern.
Die Chance dabei gleich einen Volltreffer zu erlangen ist wohl eher gering. Ich denke es sind vielmehr die kleinen, stetigen Schritte, die zum Ziel führen. Von besonderem Interesse sollte es daher sein ob aus den begleitenden fMRI Untersuchungen der Pagoclone Studie neue Erkenntnisse gewonnen werden können.

Die Begeisterung von Gerry Maguire kann ich aber nachvollziehen, da es ja gerade der Beharrlichkeit und dem Engagement seiner Arbeitsgruppe an der UC Irvine zu verdanken ist, dass die pharmakologische Forschung zum Stottern in den letzten Jahren einen akzeptierten Platz einnehmen konnte und zahlreiche Anregungen in Form wissenschaftlicher Modelle erbracht hat.

Holger