Review article: Single subject trials as a research instrument in gastrointestinal pharmacology

A critical examination of best dose analysis for determining cognitive-enhancing potential of drugs: studies with rhesus monkeys and computer simulations

RATIONALE

Best dose analysis involves identifying the dose associated with the greatest improvement in performance for each subject and comparing performances associated with these individually determined best doses to control performances.

OBJECTIVES

The current experiments were conducted to examine whether significant best dose effects might result from the selective analysis of data rather than an actual drug effect.

METHODS

Experiment 1 examined the effects of nicotine and methylphenidate on delayed matching-to-sample (DMTS) and self-ordered spatial search (SOSS) performances in rhesus monkeys (DMTS: n = 7; SOSS: n = 6) to determine the validity and reliability of best dose effects. Experiment 2 used Monte Carlo computer simulations to estimate the likelihood of obtaining a significant outcome when the best dose method was applied to randomly generated data sets for which no difference existed.

RESULTS

Significant effects were obtained when the best dose analysis was applied to performances from nondrug sessions, and best dose performances were not significantly different from the best nondrug performances. The doses identified as best doses from two nicotine dose-response curve determinations were unrelated, and the improvement associated with the best dose observed during the first dose-response curve determination was not reliable when the dose was administered repeatedly. Finally, there was a high likelihood of obtaining a statistically significant difference when no real difference existed.

CONCLUSIONS

Best dose analysis for the identification of potential therapeutic agents should be replaced by single-subject designs.

Novel study designs to investigate the placebo response

Background

Investigating the size and mechanisms of the placebo response in clinical trials have relied on experimental procedures that simulate the double-blind randomized placebo-controlled design. However, as the conventional design is thought to elucidate drug rather than placebo actions, different methodological procedures are needed for the placebo response.

Methods

We reviewed the respective literature for trials designs that may be used to elucidate the size of the placebo response and the mechanisms associated with it.

Results

In general, this can be done by either manipulation the information provided to the subjects, or by manipulation the timing of the drug applied. Two examples of each strategy are discussed: the "balanced placebo design" (BDP) and the "balanced cross-over design" (BCD) and their variants are based on false information, while the "hidden treatment" (HT) and the ""delayed response test" (DRT) are based on manipulating the time of drug action. Since most such approaches include deception or incomplete information of the subjects they are suitable for patient only with authorized deception.

Conclusion

Both manipulating the information provided to subjects (BDP, DCD) or manipulating the timing of drug application (HT, DRT) allows overcoming some of the restrictions of conventional drug trials in the assessment of the placebo response, but they are feasible mostly in healthy subjects for ethical reasons.

The placebo response in functional dyspepsia--reanalysis of trial data

Clinical trial data rarely identify factors contributing to the placebo response. We reanalysed the data from the placebo arm (n = 157) of a functional dyspepsia (FD) trial to determine predictors of placebo response including total GIS score, change in GIS score during run-in, type of FD symptoms, duration of illness, age, gender, body mass index (BMI), family occurrence of FD-like symptoms, smoking and alcohol consumption. The same response criteria were applied to the drug arm (n = 158) of the study. Based on the initial 40% response criterion of the study, 35 (22.2%) were classified as placebo responders (PR), whilst 122 (78.8%) were placebo non-responders (PnR). Response rates for the drug arm were 41.1 and 56.9% respectively. PR had significantly lower GIS scores compared to PnR at visit 1 (10.6 +/- 0.6 and 12.3 +/- 0.4, respectively, P = 0.035), but not at visit 2 with study medication dispensing (10.9 +/- 0.5 and 11.3 +/- 0.4). Hence, PR symptoms increased during run-in by 4.2% whilst PnR symptoms decreased by 6.3% (P < 0.005). Gender, age and duration and type of FD symptoms were not different between PR and PnR. Smoking was less prevalent in PR (3%) compared to PnR (21%) (P < 0.025). Increasing the criteria for the placebo response resulted in higher BMI for PR than for PnR (P = 0.035). None of the predictors for placebo response were able to distinguish responders from non-responders to the drug. Variables predicting the PR point towards behavioural and biological mechanism of the PR, operating simultaneously and independently.

New insights into the placebo and nocebo responses

In modern medicine, the placebo response or placebo effect has often been regarded as a nuisance in basic research and particularly in clinical research. The latest scientific evidence has demonstrated, however, that the placebo effect and the nocebo effect, the negative effects of placebo, stem from highly active processes in the brain that are mediated by psychological mechanisms such as expectation and conditioning. These processes have been described in some detail for many diseases and treatments, and we now know that they can represent both strength and vulnerability in the course of a disease as well as in the response to a therapy. However, recent research and current knowledge raise several issues that we shall address in this review. We will discuss current neurobiological models like expectation-induced activation of the brain reward circuitry, Pavlovian conditioning, and anxiety mechanisms of the nocebo response. We will further explore the nature of the placebo responses in clinical trials and address major questions for future research such as the relationship between expectations and conditioning in placebo effects, the existence of a consistent brain network for all placebo effects, the role of gender in placebo effects, and the impact of getting drug-like effects without drugs.

Examining clinical trial results with single-subject analysis: an example involving behavioral and nutrition treatment for young children with cystic fibrosis

OBJECTIVE

To examine the process of change in a clinical trial of behavioral and nutrition treatment for children age 18-48 months with cystic fibrosis (CF) using single-subject analysis.

METHODS

The 5-week treatment included nutrition counseling and child behavioral management training for parents and was designed to increase energy intake measured by diet diaries 600-800 calories per day.

RESULTS

Energy intake changed at each meal, only when treatment was introduced (week 1: snacks, 420 to 691; week 2: breakfast, 325 to 443; week 4: lunch, 350 to 443; and week 5: dinner, 373 to 460 calories per day). Total daily intake increased in a systematic fashion that exceeded the criterion set each week during treatment.

CONCLUSIONS

Toddlers and preschoolers with CF meet energy intake recommendations as a result of behavioral intervention. Single-subject research designs are important methodologies for advancing clinical investigation in pediatric psychology.

Identifying response to acid suppressive therapy in functional dyspepsia using a random starting day trial--is gastro-oesophageal reflux important?

BACKGROUND

Single subject trials offer an alternative approach to identify and characterize responders to a specific treatment.

AIM

To test a new single subject trials design, called random starting day trial, to identify acid-related symptoms in dyspepsia.

METHODS

A total of 119 patients with functional dyspepsia entered a 12-day, double-blind random starting day trial. All patients started on placebo and switched to omeprazole 80 mg/day at a randomized and blinded day between day 5 and day 9, with active treatment continuing for the rest of the trial. Based on changes of a daily symptom score, response was defined as a sustained > or =50% reduction of symptoms within 3 days of active treatment.

RESULTS

Thirteen of 119 patients (11%) were classified as spontaneous responders because of complete symptom relief before switching to omeprazole. Of the remaining 106 patients, 15 (15.6%) were classified as responders. Five of six (83%) responders compared with 28 of 53 (53%) non-responders had pathological reflux. Multivariate testing identified symptoms suggestive of gastro-oesophageal reflux predictive of response.

CONCLUSIONS

The random starting day trial design could identify a subset of dyspeptic patients with a uniform symptomatic response to acid-suppressive therapy. Response seems to be associated with gastro-oesophageal reflux. The random starting day trial needs to be further validated to be considered as a reliable instrument in clinical research.

Reproducibility of a symptom response to omeprazole therapy in functional dyspepsia evaluated by a random-starting-day trial design

BACKGROUND

Satisfactory treatment options for functional dyspepsia are lacking. Single subject trial designs may identify subgroups of patients with a uniform response to therapy.

AIM

To test reproducibility of response in a new random-starting-day trial design developed to identify acid-related symptoms in functional dyspepsia.

METHODS

One hundred and nineteen patients with functional dyspepsia completed a 12-day, double-blind random-starting-day trial with an initial placebo run-in followed by switch to omeprazole on a randomized and blinded day (between days 5 and 9) with active treatment continuing for the rest of the trial. Response was defined as a sustained > or =50% reduction of a daily symptom-score within 3 days of active treatment. Fifty-nine patients repeated the random-starting-day trial at relapse of symptoms.

RESULTS

After exclusion of placebo responders, 14% (15 of 106) were classified as responders in the first and 20% (10 of 50) in the subsequent random-starting-day trial series. Sixty-eight per cent (40 of 59) of the patients reproduced their initial response with a chance-corrected agreement of 0.29. Comparing response patterns using different symptom rating-scales showed good correlation (kappa 0.60).

CONCLUSION

Reproducibility of response in a random-starting-day trial was imperfect, mainly because of the low response rates and strict response criteria. Lack of symptom stability impairs the value of the random-starting-day trial and only patients with frequent and stable symptoms should be evaluated in this design.