The perils with the misuse of predictive power.

In early drug development, especially when studying new mechanisms of action or in new disease areas, little is known about the targeted or anticipated treatment effect or variability estimates. Adaptive designs that allow for early stopping but also use interim data to adapt the sample size have been proposed as a practical way of dealing with these uncertainties. Predictive power and conditional power are two commonly mentioned techniques that allow predictions of what will happen at the end of the trial based on the interim data. Decisions about stopping or continuing the trial can then be based on these predictions. However, unless the user of these statistics has a deep understanding of their characteristics important pitfalls may be encountered, especially with the use of predictive power. The aim of this paper is to highlight these potential pitfalls. It is critical that statisticians understand the fundamental differences between predictive power and conditional power as they can have dramatic effects on decision making at the interim stage, especially if used to re-evaluate the sample size. The use of predictive power can lead to much larger sample sizes than either conditional power or standard sample size calculations. One crucial difference is that predictive power takes account of all uncertainty, parts of which are ignored by standard sample size calculations and conditional power. By comparing the characteristics of each of these statistics we highlight important characteristics of predictive power that experimenters need to be aware of when using this approach.

Effect of single and repeat doses of casopitant on the pharmacokinetics of CYP450 3A4 substrates midazolam and nifedipine.

AIM: To evaluate the impact of single and repeated doses casopitant on the pharmacokinetics of single dose midazolam and nifedipine (CYP3A substrates) in healthy subjects. The effect on debrisoquine metabolism (CYP2D6 substrate) was also assessed. METHODS: Three open-label studies were conducted in healthy subjects. In the first study subjects received single dose 50 or 100 mg oral casopitant, single dose 5 mg oral midazolam and single dose 10 mg oral debrisoquine. In the other two studies subjects received repeated doses of 10 mg (study 2), 30, or 120 mg oral casopitant and single doses of 5 mg oral midazolam (study 2) and single doses of 10 mg oral nifedipine (study 3). Plasma concentration-time data were analyzed using standard non-compartmental methods. The effect of casopitant on all probes was assessed using geometric means ratios and corresponding 90% confidence intervals (CIs). RESULTS: The AUC(0,∞) of midazolam was increased 1.44-fold (90% CI 1.35, 1.54) and 1.52-fold (90% CI 1.41, 1.65) after co-administration with a single dose of 50 or 100 mg casopitant, respectively. Debrisoquine metabolism was unchanged. After 3 days of casopitant administration, midazolam AUC(0,∞) was increased 1.45- (90% CI 1.32, 1.59), 2.02- (90% CI 1.75, 2.32), and 2.67-fold (90% CI 2.18, 3.27) after co-administration with 10, 30 or 120 mg casopitant, respectively. After 14 days of casopitant administration, midazolam AUC(0,∞) was increased 1.51- (90% CI 1.40, 1.63) to 3.49-fold (90% CI 2.98, 4.08). After 3 days of casopitant administration, nifedipine AUC(0,∞) was increased 1.56- (90% CI 1.37, 1.78) and 1.77-fold (90% CI 1.54, 2.04) after co-administration with 30 or 120 mg casopitant, respectively. Similar increases in nifedipine exposure were observed after 14 days of casopitant administration. CONCLUSIONS: Casopitant is a dose- and duration-dependent weak to moderate inhibitor of CYP3A.

Anxiolytic effects of vestipitant in a sub-group of healthy volunteers known to be sensitive to CO2 challenge.

The pharmacological properties of two NK1 antagonists were studied in comparison with a benzodiazepine during a 7% CO2 challenge in a population of healthy volunteers selected for a high sensitivity to the challenge. In total, 19 healthy subjects, pre-screened for their responsiveness to the 7% CO2 test, took part in the randomised, double-blind, cross-over, incomplete block design study. After receiving treatment or placebo, the volunteers were subjected to three 7% CO2 challenges each for a time of 20 min. The treatment consisted of the administration of the following three active drugs: a single dose of benzodiazepine alprazolam (0.75 mg) and a single dose of the NK1 antagonists vestipitant (GW597599) (15 mg) and vofopitant (GR205171) (25 mg). Anxiety during the challenge was evaluated with Visual Analogue Scale-Anxiety (VAS-A) and with Panic Symptom List (PSL III-R). Respiratory parameters, heart rate and skin conductance were also recorded. Compared with placebo, vestipitant showed a significant reduction (p<0.05) in anxiety assessed on the VAS-A scale (ΔVAS-A%) while alprazolam significantly (p<0.01) attenuated the PSL III-R total score. Vofopitant did not show any anxiolytic effect. In the comparison analysis between placebo and drugs, none of the respiratory and other physiological parameters showed a statistically significant difference.

Functional effects of chronic paroxetine versus placebo on the fear, stress and anxiety brain circuit in Social Anxiety Disorder: initial validation of an imaging protocol for drug discovery.

Recent studies suggest that pharmacologic effects of anxiolytic agents can be mapped as functional changes in the fear, stress and anxiety brain circuit. In this work we investigated the effects of a standard treatment, paroxetine (20mg/day), in subjects with Social Anxiety Disorder (SAD) versus placebo using different fMRI paradigms. The fMRI sessions, performed before and after the treatment, consisted of a public exposition of recorded performance task (PERPT), an emotional face processing task (EFPT) and a 6-min resting state followed by an off-scanner public speaking test. Paroxetine significantly improved the clinical conditions of SAD patients (n=17) vs. placebo (n=16) as measured with Clinical Global Inventory - Improvement (CGI-I) while no change was seen when using Liebowitz Social Anxiety Scale, as expected given the small size of the study population. Paroxetine reduced the activation of insula, thalamus and subgenual/anterior cingulate cortex (ACC) in PERPT. Resting-state fMRI assessment using Independent Component Analysis indicated that paroxetine reduced functional connectivity in insula, thalamus and ACC when compared with placebo. Both paradigms showed significant correlation with CGI-I in rostral prefrontal cortex. Conversely, paroxetine compared to placebo produced activation of right amygdala and bilateral insula and no effects in ACC when tested with EFPT. No treatment effects on distress scores were observed in the off-scanner Public Speaking Test. Overall this study supports the use of fMRI as sensitive approach to explore the neurobiological substrate of the effects of pharmacologic treatments and, in particular, of resting state fMRI given its simplicity and task independence.

A randomized, controlled study comparing the effects of vestipitant or vestipitant and paroxetine combination in subjects with tinnitus.

OBJECTIVE: Tinnitus is a common symptom that demonstrates a significant comorbidity with anxiety and depression. The novel neurokinin-1 receptor antagonist, vestipitant, has anxiolytic properties and a good safety profile. Vestipitant was investigated for potential effect against chronic tinnitus as a stand-alone treatment and in conjunction with a selective serotonin reuptake inhibitor, paroxetine. STUDY DESIGN: Randomized, double-blind, crossover study. SETTING: Tertiary neurotologic and audiologic center with additional referrals from a secondary university hospital center. PATIENTS: Twenty-four adult patients with tinnitus were randomized into the study. MAIN OUTCOME MEASURES: Visual analogue scale (VAS) measurements of tinnitus loudness (intensity), pitch and distress, VAS measurements of arousal/anxiety, Tinnitus Handicap Inventory, Quick Inventory of Depressive Symptomatology, and plasma concentrations of trial drugs. RESULTS: No statistically significant treatment benefit effect was detected for tinnitus (intensity, pitch, and distress) VAS scores, arousal-anxiety VAS scores, Tinnitus Handicap Inventory, or tinnitus aggravation scores assessed on Days 1 and 14. However, a statistically significant worsening of tinnitus intensity and distress scores was observed after vestipitant compared with placebo for the mean data collected over the treatment period. No relevant differences in vestipitant plasma concentrations were observed between the subjects given the combination with paroxetine and those receiving vestipitant alone. No specific relationships were observed between tinnitus intensity and vestipitant plasma concentrations. CONCLUSION: Although well-tolerated vestipitant, alone or in combination with paroxetine, was not effective in ameliorating tinnitus in this patient group.