Power prior models for estimating response rates in a small n, sequential, multiple assignment, randomized trial

A small n, sequential, multiple assignment, randomized trial (snSMART) is a small sample, two-stage design where participants receive up to two treatments sequentially, but the second treatment depends on response to the first treatment. The parameters of interest in an snSMART are the first-stage response rates of the treatments, but outcomes from both stages can be used to obtain more information from a small sample. A novel way to incorporate the outcomes from both stages uses power prior models, in which first stage outcomes from an snSMART are regarded as the primary (internal) data and second stage outcomes are regarded as supplemental data (co-data). We apply existing power prior models to snSMART data, and we also develop new extensions of power prior models. All methods are compared to each other and to the Bayesian joint stage model (BJSM) via simulation studies. By comparing the biases and the efficiency of the response rate estimates among all proposed power prior methods, we suggest application of Fisher's Exact Test or the Bhattacharyya's overlap measure to an snSMART to estimate the response rates in an snSMART, which both have performance mostly as good or better than the BJSM. We describe the situations where each of these suggested approaches is preferred.

Bayesian methods to compare dose levels with placebo in a small n, sequential, multiple assignment, randomized trial

Clinical trials studying treatments for rare diseases are challenging to design and conduct due to the limited number of patients eligible for the trial. One design used to address this challenge is the small n, sequential, multiple assignment, randomized trial (snSMART). We propose a new snSMART design that investigates the response rates of a drug tested at a low and high dose compared with placebo. Patients are randomized to an initial treatment (stage 1). In stage 2, patients are rerandomized, depending on their initial treatment and their response to that treatment in stage 1, to either the same or a different dose of treatment. Data from both stages are used to determine the efficacy of the active treatment. We present a Bayesian approach where information is borrowed between stage 1 and stage 2. We compare our approach to standard methods using only stage 1 data and a log-linear Poisson model that uses data from both stages where parameters are estimated using generalized estimating equations. We observe that the Bayesian method has smaller root-mean-square-error and 95% credible interval widths than standard methods in the tested scenarios. We conclude that it is advantageous to utilize data from both stages for a primary efficacy analysis and that the specific snSMART design shown here can be used in the registration of a drug for the treatment of rare diseases.

Design and analysis considerations for utilizing a mapping function in a small sample, sequential, multiple assignment, randomized trials with continuous outcomes

Small sample, sequential, multiple assignment, randomized trials (snSMARTs) are multistage trials with the overall goal of determining the best treatment after a fixed amount of time. In snSMART trials, patients are first randomized to one of three treatments and a binary (e.g. response/nonresponse) outcome is measured at the end of the first stage. Responders to first stage treatment continue their treatment. Nonresponders to first stage treatment are rerandomized to one of the remaining treatments. The same binary outcome is measured at the end of the first and second stages, and data from both stages are pooled together to find the best first stage treatment. However, in many settings the primary endpoint may be continuous, and dichotomizing this continuous variable may reduce statistical efficiency. In this article, we extend the snSMART design and methods to allow for continuous outcomes. Instead of requiring a binary outcome at the first stage for rerandomization, the probability of staying on the same treatment or switching treatment is a function of the first stage outcome. Rerandomization based on a mapping function of a continuous outcome allows for snSMART designs without requiring a binary outcome. We perform simulation studies to compare the proposed design with continuous outcomes to standard snSMART designs with binary outcomes. The proposed design results in more efficient treatment effect estimates and similar outcomes for trial patients.

Genetic Subtype-Phenotype Analysis of Growth Hormone Treatment on Psychiatric Behavior in Prader-Willi Syndrome

Prader-Willi syndrome (PWS) is a complex multisystemic condition caused by a lack of paternal expression of imprinted genes from the 15q11.2-q13 region. Limited literature exists on the association between molecular classes, growth hormone use, and the prevalence of psychiatric phenotypes in PWS. In this study, we analyzed nine psychiatric phenotypes (depressed mood, anxiety, skin picking, nail picking, compulsive counting, compulsive ordering, plays with strings, visual hallucinations, and delusions) recognized in PWS and investigated associations with growth hormone treatment (GHT), deletions (DEL) and uniparental disomy (UPD) in a cohort of 172 individuals with PWS who met the criteria for analysis. Associations were explored using Pearson chi-square tests and univariable and multivariable logistic regression analyses to control for confounding exposures. This observational study of the largest dataset of patients with PWS to date suggested the following genetic subtype and phenotype correlations in psychiatric behaviors: (1) skin picking was more frequent in those with DEL vs. UPD; (2) anxiety was more common in those with UPD vs. DEL; and (3) an increased frequency of anxiety was noted in the UPD group treated with GHT compared to the DEL group. No other significant associations were found between the genetic subtype or GHT including for depressed mood, nail picking, compulsive counting, compulsive ordering, playing with strings, and visual hallucinations. Further studies will be required before any conclusions can be reached.

Protocol for a randomized multicenter study for isolated skin vasculitis (ARAMIS) comparing the efficacy of three drugs: azathioprine, colchicine, and dapsone

Background

Skin-limited forms of vasculitis, while lacking systemic manifestations, can persist or recur indefinitely, cause pain, itch, or ulceration, and be complicated by infection or scarring. High-quality evidence on how to treat these conditions is lacking. The aim of this comparative effectiveness study is to determine the optimal management of patients with chronic skin-limited vasculitis.

Methods

ARAMIS is a multicenter, sequential, multiple assignment randomized trial with an enrichment design (SMARTER) aimed at comparing the efficacy of three drugs—azathioprine, colchicine, and dapsone—commonly used to treat various forms of isolated skin vasculitis. ARAMIS will enroll patients with isolated cutaneous small or medium vessel vasculitis, including cutaneous small vessel vasculitis, immunoglobulin A (IgA) vasculitis (skin-limited Henoch-Schönlein purpura), and cutaneous polyarteritis nodosa. Patients not responding to the initial assigned therapy will be re-randomized to one of the remaining two study drugs (Stage 2). Those with intolerance or contraindication to a study drug can be randomized directly into Stage 2. Target enrollment is 90 participants, recruited from international centers affiliated with the Vasculitis Clinical Research Consortium. The number of patients enrolled directly into Stage 2 of the study will be capped at 10% of the total recruitment target. The primary study endpoint is the proportion of participants from the pooled study stages with a response to therapy at month 6, according to the study definition.

Discussion

ARAMIS will help identify effective agents for skin-limited forms of vasculitis, an understudied group of diseases. The SMARTER design may serve as an example for future trials in rare diseases.

Trial registration

ClinicalTrials.gov: NCT02939573. Registered on 18 October 2016.

Predicting progression to type 1 diabetes from ages 3 to 6 in islet autoantibody positive TEDDY children

OBJECTIVE

The capacity to precisely predict progression to type 1 diabetes (T1D) in young children over a short time span is an unmet need. We sought to develop a risk algorithm to predict progression in children with high-risk human leukocyte antigen (HLA) genes followed in The Environmental Determinants of Diabetes in the Young (TEDDY) study.

METHODS

Logistic regression and 4-fold cross-validation examined 38 candidate predictors of risk from clinical, immunologic, metabolic, and genetic data. TEDDY subjects with at least one persistent, confirmed autoantibody at age 3 were analyzed with progression to T1D by age 6 serving as the primary endpoint. The logistic regression prediction model was compared to two non-statistical predictors, multiple autoantibody status, and presence of insulinoma-associated-2 autoantibodies (IA-2A).

RESULTS

A total of 363 subjects had at least one autoantibody at age 3. Twenty-one percent of subjects developed T1D by age 6. Logistic regression modeling identified 5 significant predictors - IA-2A status, hemoglobin A1c, body mass index Z-score, single-nucleotide polymorphism rs12708716_G, and a combination marker of autoantibody number plus fasting insulin level. The logistic model yielded a receiver operating characteristic area under the curve (AUC) of 0.80, higher than the two other predictors; however, the differences in AUC, sensitivity, and specificity were small across models.

CONCLUSIONS

This study highlights the application of precision medicine techniques to predict progression to diabetes over a 3-year window in TEDDY subjects. This multifaceted model provides preliminary improvement in prediction over simpler prediction tools. Additional tools are needed to maximize the predictive value of these approaches.

Risk factors for chemotherapy-induced nausea in pediatric patients receiving highly emetogenic chemotherapy

BACKGROUND

Little is known regarding risk factors for chemotherapy-induced nausea (CIN) in pediatric patients.

PROCEDURE

A secondary analysis was conducted of a previously published multicenter, prospective, randomized, single-blind, sham-controlled trial assessing the efficacy of acupressure in preventing CIN in pediatric patients receiving highly emetogenic chemotherapy. The primary outcome was nausea severity, self-reported using the Pediatric Nausea Assessment Tool. The relationships between acute and delayed nausea severity and patient- (sex, race, age, and cancer diagnosis) and treatment-related (chemotherapy, antiemetic prophylaxis, CIN, and vomiting control) factors were analyzed by a proportional odds generalized estimating equation approach. The acute phase started with administration of the first and continued for 24 hours after the last chemotherapy dose. The delayed phase started at the end of the acute phase and continued until the next chemotherapy block (maximum seven days).

RESULTS

In the acute and delayed phases, 165 and 144 patients provided data for analysis, respectively. Nonwhite race was significantly associated with higher acute phase nausea severity (OR, 1.7; 95% CI, 1.1-2.6). Poor CIN control in the acute phase (OR, 16; 95% CI, 4.0-64.6), diagnosis of a cancer other than a central nervous system (CNS) tumor (OR, 2.5; 95% CI, 1.2-5.3), and cisplatin administration (OR, 3.7; 95% CI, 2.1-6.0) were significantly associated with higher delayed phase nausea severity.

CONCLUSION

Acute phase CIN was associated with nonwhite race. Delayed phase CIN was associated with poor acute phase CIN control, diagnosis of non-CNS cancer, and receipt of cisplatin. These findings will inform future antiemetic trial design.

Acupressure bands do not improve chemotherapy-induced nausea control in pediatric patients receiving highly emetogenic chemotherapy: A single-blinded, randomized controlled trial

BACKGROUND

Chemotherapy-induced nausea and vomiting remain common, distressing side effects of chemotherapy. It has been reported that acupressure prevents chemotherapy-induced nausea in adults, but it has not been well studied in children.

METHODS

In this multicenter, prospective, randomized, single-blind, sham-controlled trial, the authors compared acute-phase nausea severity in patients ages 4 to 18 years who were receiving highly emetic chemotherapy using standard antiemetic agents combined with acupressure wrist bands, the most common type of acupressure, versus sham bands. Patients wore acupressure or sham bands continuously on each day of chemotherapy and for up to 7 days afterward. Chemotherapy-induced nausea severity in the delayed phase and chemotherapy-induced vomiting control in the acute and delayed phases also were compared.

RESULTS

Of the 187 patients randomized, 165 contributed nausea severity assessments during the acute phase. Acupressure bands did not reduce the severity of chemotherapy-induced nausea in the acute phase (odds ratio [OR], 1.33; 95% confidence limits, 0.89-2.00, in which an OR <1.00 favored acupressure) or in the delayed phase (OR, 1.23; 95% CL, 0.75-2.01). Furthermore, acupressure bands did not improve daily vomiting control during the acute phase (OR, 1.57; 95% CL, 0.95-2.59) or the delayed phase (OR, 0.84; 95% CL, 0.45-1.58). No serious adverse events were reported.

CONCLUSIONS

Acupressure bands were safe but did not improve chemotherapy-induced nausea or vomiting in pediatric patients who were receiving highly emetic chemotherapy. Cancer 2018;124:1188-96. © 2017 American Cancer Society.

Vitamin D and probiotics supplement use in young children with genetic risk for type 1 diabetes

BACKGROUND/OBJECTIVES

Vitamin D and probiotics are nutrients of interest in the context of type 1 diabetes (T1D). We assessed the prevalence of and factors associated with vitamin D and probiotic supplementations among young children with genetic risk of T1D.

SUBJECTS/METHODS

Use of supplements during the first 2 years of life was collected prospectively from 8674 children in The Environmental Determinants of Diabetes in the Young (TEDDY) study.

RESULTS

Single and/or multivitamin/mineral (MVM) supplements were reported by 81% of the children. The majority of participants in Finland, Germany and Sweden (97-99%) and 50% in the United States received vitamin D supplements that were mostly MVMs. Probiotics use varied from 6% in the United States to 60% in Finland and was primarily from probiotics-only preparations. More than 80% of the vitamin D and probiotics supplementation was initiated during infancy, and more than half of the uses lasted longer than a year. Being the first child, longer duration of breastfeeding, born in a later year, older maternal age and higher maternal education level were associated with both vitamin D and probiotics use. Shorter gestational age and mother not smoking during pregnancy were associated with a higher likelihood of probiotics supplementation only.

CONCLUSIONS

Vitamin D and probiotics supplementations are popular in children 0-2 years old and are associated with common factors. Data documented here will allow evaluation of the relationship between early childhood dietary intake and the development of islet autoimmunity and progression to T1D.

Maternal use of dietary supplements during pregnancy is not associated with coeliac disease in the offspring: The Environmental Determinants of Diabetes in the Young (TEDDY) study

Perinatal exposure to nutrients and dietary components may affect the risk for coeliac disease (CD). We investigated the association between maternal use of vitamin D, n-3 fatty acids (FA) and Fe supplements during pregnancy and risk for CD autoimmunity (CDA) and CD in the offspring. Children at increased genetic risk were prospectively followed from birth in The Environmental Determinants of Diabetes in the Young (TEDDY) study. CDA was defined as having persistently positive tissue transglutaminase autoantibodies (tTGA). Diagnosis of CD was either biopsy-confirmed or considered likely if having persistently elevated levels of tTGA>100 AU. Of 6627 enrolled children, 1136 developed CDA at a median 3·1 years of age (range 0·9-10) and 409 developed CD at a median 3·9 years of age (range 1·2-11). Use of supplements containing vitamin D, n-3 FA and Fe was recalled by 66, 17 and 94 % of mothers, respectively, at 3-4 months postpartum. The mean cumulative intake over the entire pregnancy was 2014 μg vitamin D (sd 2045 μg), 111 g n-3 FA (sd 303 g) and 8806 mg Fe (sd 7017 mg). After adjusting for country, child's human leucocyte antigen genotype, sex, family history of CD, any breast-feeding duration and household crowding, Cox's proportional hazard ratios did not suggest a statistically significant association between the intake of vitamin D, n-3 FA or Fe, and risk for CDA or CD. Dietary supplementation during pregnancy may help boost nutrient intake, but it is not likely to modify the risk for the disease in the offspring.

An examination of the efficiency of the sequential parallel design in psychiatric clinical trials

Background Psychiatric clinical trials have a high failure rate, even among agents which are known to be effective. Because of this high failure rate, a novel clinical trial design has been proposed which incorporates a second phase in which non-responders to placebo are randomly reassigned to drug or placebo.

Purpose The purpose of this research is to examine the efficiency of this new design compared to the conventional two arm clinical trial. We consider both binary and continuous endpoints.

Methods The limiting distribution of a class of weighted average test statistics is obtained for the binary case which allows analytic calculation of the power for a given set of parameters. For the continuous case, we examine the efficiency of seemingly unrelated regression and a weighted average statistic via simulation.

Results The novel design reduces the sample size 20—25% compared to the standard design under a wide range of parameters.

Limitations There are no actual trials with the novel design therefore assumptions of the effect size across two periods for actual psychiatric agents is unknown.

Conclusions The new design reduces sample size which in turn should reduce the cost of clinical trials. Further refinements of the design are possible including alternative test statistics and incorporation of additional data from placebo responders. Clinical Trials 2007; 4: 309—317. http://ctj.sagepub.com

A small n sequential multiple assignment randomized trial design for use in rare disease research

BACKGROUND

Clinical trials in rare diseases are difficult to conduct due to the limited number of patients available with each disorder. We developed a Phase 2 trial which is a small n sequential multiple assignment randomized trial (snSMART) design to test several treatments for a rare disease for which no standard therapy exists.

PURPOSE

This paper illustrates the design, sample size estimation and operating characteristics of an snSMART.

METHODS

We investigate the performance of a class of weighted Z statistics via computer simulations.

RESULTS

We demonstrate the increase in power over traditional single stage designs, and indicate how the power changes as a function of the weight given to each stage.

CONCLUSION

The snSMART design is promising in a rare disease setting where several alternative treatments are under consideration and small sample sizes are necessary.

Glutamic acid not beneficial for the prevention of vincristine neurotoxicity in children with cancer

BACKGROUND

Vincristine causes known side effects of peripheral sensory, motor, autonomic and cranial neuropathies. No preventive interventions are known.

PROCEDURE

We performed a randomized, placebo-controlled, double-blind trial of oral glutamic acid as a preventive agent in pediatric patients with cancer who would be receiving vincristine therapy for at least 9 consecutive weeks (Stratum 1 = Wilms tumor and rhabdomyosarcoma) or 4 consecutive weeks in conjunction with steroids (Stratum 2 = Acute lymphoblastic leukemia and non-Hodgkin lymphoma). At designated time points, a scored neurologic exam using the Modified Balis Pediatric Scale of Peripheral Neuropathies was performed to document neurologic toxicity.

RESULTS

Between 2007 and 2012, 250 patients were enrolled (Stratum 1 = 50, Stratum 2 = 200). The glutamic acid treated group did not have a significantly lower percentage of neurotoxicity compared to placebo treated group either overall or within stratum or age subgroups. The only subgroup which was suggestive of treatment effect was for age. Patients 13 years or older showed a larger benefit in favor of glutamic acid (P = 0.055) compared to patients less than 13 years (P = 1.00). Constipation was the most frequently reported (14%) Grade II or higher neurotoxicity.

CONCLUSION

Vincristine-associated neurotoxicity in pediatric oncology remains a frequent complication of chemotherapy for multiple diagnoses with an approximate 30% of patients affected. Glutamic acid is not effective for prevention in pre-adolescents. There is a suggestion of benefit in patients 13 years or older, but the study was not designed to provide adequate power to test the treatment effect within this age group alone.

A sequential enriched design for target patient population in psychiatric clinical trials

High placebo response is widely believed to be one major reason why many psychiatric clinical trials fail to demonstrate drug efficacy. In order to alleviate this problem, research has developed several enrichment designs, including the parallel design with a placebo lead-in phase, the sequential parallel design, and a recently proposed two-way enriched design. While these designs have been evaluated and discussed individually, their effectiveness against each other has not been rigorously compared. The current study examines the various enrichment designs simultaneously. Building on their strengths, we introduce a new improved design named' sequential enriched design' (SED) aimed at removing not only patients with high placebo response but also patients who do not respond to any treatment from the study. The SED begins with a double-blind placebo lead-in phase followed by a traditional parallel design in the first stage. Only patients who respond to the drug in the first stage are re-randomized to the drug or placebo at the second stage. We simulate data for a mixed population composed of four subgroups of patients who are predetermined as to whether they respond to drug or not as well as to placebo or not. By focusing on the target patients whose responses reflect the drug's efficacy,we evaluate the bias, mean squared error, and power for different designs. We demonstrate that the SED produces a less biased estimate for the target treatment effect and yields reasonably high power in general compared with the other designs.

An open-label safety and pharmacokinetics study of duloxetine in pediatric patients with major depression

OBJECTIVE

This preliminary, 32-week study assessed the safety, tolerability, and pharmacokinetics of duloxetine in pediatric patients (aged 7-17 years) with major depressive disorder.

METHODS

Patients received flexible duloxetine doses of 20-120 mg once daily, with dose changes made based on clinical improvement and tolerability. Pharmacokinetic samples were collected across all duloxetine doses, and data were analyzed using population modeling. Primary outcome measures included treatment-emergent adverse events (TEAEs), vital signs, and Columbia-Suicide Severity Rating Scale (C-SSRS).

RESULTS

Of the 72 enrolled patients, 48 (66.7%) completed acute treatment (18 weeks) and 42 (58.3%) completed extended treatment. Most patients (55/72; 76%) required doses ≥ 60 mg once daily to optimize efficacy based on investigator judgment and Clinical Global Impressions-Severity score. Body weight and age did not significantly affect duloxetine pharmacokinetic parameters. Typical duloxetine clearance in pediatric patients was ≈ 42%-60% higher than that in adults. Four patients (5.6%) discontinued due to TEAEs. Many (36/72, 50%) patients experienced potentially clinically significant (PCS) elevations in blood pressure, with most cases (21/36, 58%) being transient. As assessed via C-SSRS, one nonfatal suicidal attempt occurred, two patients (2.8%) experienced worsening of suicidal ideation, and among the 19 patients reporting suicidal ideation at baseline, 17 (90%) reported improvement in suicidal ideation.

CONCLUSION

Results suggested that pediatric patients generally tolerated duloxetine doses of 30 to 120 mg once daily, although transient PCS elevations in blood pressure were observed in many patients. Pharmacokinetic results suggested that adjustment of total daily dose based on body weight or age is not warranted for pediatric patients and different total daily doses may not be warranted for pediatric patients relative to adults.

Estimation of treatment effect for the sequential parallel design

The sequential parallel clinical trial is a novel clinical trial design being used in psychiatric diseases that are known to have potentially high placebo response rates. The design consists of an initial parallel trial of placebo versus drug augmented by a second parallel trial of placebo versus drug in the placebo non-responders from the initial trial. Statistical research on the design has focused on hypothesis tests. However, an equally important output from any clinical trial is the estimate of treatment effect and variability around that estimate. In the sequential parallel trial, the most important treatment effect is the effect in the overall population. This effect can be estimated by considering only the first phase of the trial, but this ignores useful information from the second phase of the trial. We develop estimates of treatment effect that incorporate data from both phases of the trial. Our simulations and a real data example suggest that there can be substantial gains in precision by incorporating data from both phases. The potential gains appear to be greatest in moderate-sized trials, which would typically be the case in phase II trials.

A two-way enriched clinical trial design: combining advantages of placebo lead-in and randomized withdrawal

A new clinical trial design, designated the two-way enriched design (TED), is introduced, which augments the standard randomized placebo-controlled trial with second-stage enrichment designs in placebo non-responders and drug responders. The trial is run in two stages. In the first stage, patients are randomized between drug and placebo. In the second stage, placebo non-responders are re-randomized between drug and placebo and drug responders are re-randomized between drug and placebo. All first-stage data, and second-stage data from first-stage placebo non-responders and first-stage drug responders, are utilized in the efficacy analysis. The authors developed one, two and three degrees of freedom score tests for treatment effect in the TED and give formulae for asymptotic power and for sample size computations. The authors compute the optimal allocation ratio between drug and placebo in the first stage for the TED and compare the operating characteristics of the design to the standard parallel clinical trial, placebo lead-in and randomized withdrawal designs. Two motivating examples from different disease areas are presented to illustrate the possible design considerations.

Recent developments in improving signal detection and reducing placebo response in psychiatric clinical trials

Recent (2007-2010) empirical and theoretical literature on associations of trial design features with signal detection and placebo response were investigated, along with data and analytic considerations. Trials with greater percentages of patients randomized to placebo had larger average drug-placebo differences in two comprehensive meta-analyses (MDD and Schizophrenia). Excluding patients with large responses during double-blind placebo lead-ins resulted in small increases in drug-placebo differences. Core factor subscales of the HAMD yielded larger drug-placebo differences than the HAMD total score. Direct likelihood-based (MMRM) and similar analyses provided better control of false positive and false negative results than LOCF and BOCF. Theoretical considerations suggested that the number of sites and number of countries can influence power, depending on the correlation structure in the data and on how sites and countries are chosen. Use of centralized ratings reduced placebo response and improved drug-placebo differences. However, the number of comparisons was too small to draw conclusions. Use of patient ratings and reducing the number of study visits reduced placebo response, but their effects on signal detection were unclear. Practical experience with novel designs such as the sequential parallel approach hold promise for improvements in signal detection. Given the complexities of signal detection and placebo response, no single strategy is likely to fully solve the problem and combinations of approaches may be most useful. Utilizing appropriate analytic techniques and randomizing an adequate fraction of patients to placebo are perhaps the most broadly applicable approaches.

Treatment-associated suicidal ideation and adverse effects in an open, multicenter trial of fluoxetine for major depressive episodes

BACKGROUND

Some reports suggest that a subset of depressed patients may experience suicidality - that is increase or emergence of suicidal ideation (SI) or behavior--after initiation of an antidepressant. The time course and clinical correlates of this phenomenon have not been characterized in detail.

METHOD

We conducted a secondary analysis of a multicenter, prospective, open, 12-week trial of fluoxetine 20 mg in outpatients with nonpsychotic major depressive episodes. Adverse effects and other clinical features associated with the emergence of suicidality, defined using item 3 of the Hamilton Depression Rating Scale, were examined using Cox regression models.

RESULTS

Among 414 subjects without SI at baseline, 59 (14.3%) reported SI on at least 1 postbaseline visit. In a Cox regression, emergence of activation and worsening of depression severity were independently associated with emergence of SI, along with female gender, younger age and having thoughts that life was not worth living prior to treatment. Treatment response and remission were significantly less likely among subjects who developed SI.

CONCLUSIONS

New SI was relatively common in this trial of fluoxetine and associated with the emergence of activation and overall symptomatic worsening. Whether prophylaxis against or aggressive treatment of adverse events can decrease emergence of SI merits further study.

Score tests for dose effect in the presence of non-responders

When only a certain proportion of subjects respond to treatment ('responders') or may never experience an event of interest (thus 'cured'), mixture models often lead to increased understanding of the treatment or disease process. This paper focuses on hypothesis testing in a dose-response framework and shows that increased power is possible by using a mixture model where both the logit of the response rate and the response mean are linear functions of the dose level. Three score tests are developed for testing an overall effect and permutation methods are used to control the type I error. Extensive simulations establish the power properties of the tests and show that our proposed score test has the best performance. The approach is illustrated by a multi-country clinical trial of rapid acting Intramuscular Olanzapine.

Olanzapine treatment of psychotic and behavioral symptoms in patients with Alzheimer disease in nursing care facilities: a double-blind, randomized, placebo-controlled trial. The HGEU Study Group

BACKGROUND

Patients with Alzheimer disease (AD) commonly exhibit psychosis and behavioral disturbances that impair patient functioning, create caregiver distress, and lead to institutionalization. This study was conducted to assess the efficacy and safety of olanzapine in treating psychosis and/or agitation/aggression in patients with AD.

METHODS

A multicenter, double-blind, placebo-controlled, 6-week study was conducted in 206 elderly US nursing home residents with AD who exhibited psychotic and/or behavioral symptoms. Patients were randomly assigned to placebo or a fixed dose of 5, 10, or 15 mg/d of olanzapine. The primary efficacy measure was the sum of the Agitation/Aggression, Hallucinations, and Delusions items (Core Total) of the Neuropsychiatric Inventory-Nursing Home version.

RESULTS

Low-dose olanzapine (5 and 10 mg/d) produced significant improvement compared with placebo on the Core Total (-7.6 vs -3.7 [P<.001] and -6.1 vs -3. 7 [P =.006], respectively). Core Total improvement with olanzapine, 15 mg/d, was not significantly greater than placebo. The Occupational Disruptiveness score, reflecting the impact of patients' psychosis and behavioral disturbances on the caregiver, was significantly reduced in the 5-mg/d olanzapine group compared with placebo (-2.7 vs -1.5; P =.008). Somnolence was significantly more common among patients receiving olanzapine (25.0%-35.8%), and gait disturbance occurred in those receiving 5 or 15 mg/d (19.6% and 17.0%, respectively). No significant cognitive impairment, increase in extrapyramidal symptoms, or central anticholinergic effects were found at any olanzapine dose relative to placebo.

CONCLUSION

Low-dose olanzapine (5 and 10 mg/d) was significantly superior to placebo and well tolerated in treating agitation/aggression and psychosis in this population of patients with AD.

Comparing time to onset of response in antidepressant clinical trials using the cure model and the Cramer-von Mises test

Currently available antidepressants have a delayed onset of action and there is considerable interest in developing new products which exhibit a shorter time to response. Previous authors have suggested using a cure model to characterize the time to response for the combination of the patients who respond and the patients who do not respond. In this paper, we use the cure model and propose a Cramer-von Mises statistic to compare the time to response distributions for patients who respond to therapy. The asymptotic null distribution of the statistic is derived. A bootstap procedure is also proposed for sample sizes typical in antidepressant clinical trials. Results of the simulation study suggest that for common designs and sample sizes used in such trials, the statistic has reasonable size and power properties as long as censoring is moderate.

Changes in adverse events reported by patients during 6 months of fluoxetine therapy

BACKGROUND

Although a period of 6 to 12 months of antidepressant therapy is recommended for most patients with depression, systematic examinations of the course of adverse events over time, the resolution of early-onset events, and the possible emergence of later-onset events are limited. We examined the safety of fluoxetine, 20 mg/day, in a large, prospective, long-term treatment trial, and we report a comparison of early- and late-onset adverse events and the course of adverse events over 26 weeks of treatment.

METHOD

Adverse events were recorded at each visit in a uniform format by open-ended questioning, regardless of perceived causality. New or worsened events reported in either the first 4 weeks of treatment (early-reporting interval) or weeks 22 through 26 of treatment (late-reporting interval) were compared.

RESULTS

Patients (N = 299) whose depression (DSM-III-R) remitted with 12 weeks of fluoxetine treatment entered continuation therapy, and 174 completed 26 weeks of therapy. All events that occurred in > or =5% of patients early in treatment decreased in frequency over time (p<.05), and no events occurred significantly more frequently during continuation therapy. No previously uncommon adverse events became common during long-term treatment.

CONCLUSION

Common adverse events associated with initiating fluoxetine treatment in depressed patients, including nausea, insomnia, nervousness, and somnolence, resolve in the majority of patients and become significantly less frequent with continued treatment over a 6-month period. No adverse events present initially become more frequent late in treatment. Therapy with fluoxetine, 20 mg/day, is well tolerated over 6 months, and most adverse events observed early in treatment resolve.

Randomised double-blind comparison of the incidence of tardive dyskinesia in patients with schizophrenia during long-term treatment with olanzapine or haloperidol

BACKGROUND

Tardive dyskinesia is important in the side-effect profile of antipsychotic medication.

AIMS

The development of tardive dyskinesia was evaluated in patients treated with double-blind, randomly assigned olanzapine or haloperidol for up to 2.6 years.

METHODS

Tardive dyskinesia was assessed by the Abnormal Involuntary Movement Scale (AIMS) and Research Diagnostic Criteria for Tardive Dyskinesia (RD-TD), it was defined as meeting RD-TD criteria at two consecutive assessments. The risk of tardive dyskinesia, the relative risk, incidence rate, and incidence rate ratio were estimated.

RESULTS

The relative risk of tardive dyskinesia for the overall follow up period for haloperidol (n = 522) v. olanzapine (n = 1192) was 2.66 (95% CI = 1.50-4.70). Based on data following the initial six weeks of observation (during which patients underwent medication change and AIMS assessments as frequently as every three days), the one-year risk was 0.52% with olanzapine (n = 513) and 7.45% with haloperidol (n = 114). The relative risk throughout this follow-up period was 11.37 (95% CI = 2.21-58.60).

CONCLUSION

Our results indicated a significantly lower risk of tardive dyskinesia with olanzapine than with haloperidol.

Outcome assessment and clinical improvement in panic disorder: evidence from a randomized controlled trial of fluoxetine and placebo. The Fluoxetine Panic Disorder Study Group

OBJECTIVE

Although panic attacks account for only a portion of the morbidity of panic disorder and panic attack frequency assessments are unreliable, studies of drug efficacy in panic disorder have generally used reduction in panic attack frequency as the primary measure of improvement. The authors studied the efficacy of fluoxetine treatment in panic disorder and measured the relative contributions of changes in symptoms to overall improvement.

METHOD

Patients with a diagnosis of panic disorder (N = 243) were randomly assigned to treatment with 10 or 20 mg/day of fluoxetine or placebo. Primary outcome measures were change in panic attack frequency and clinician-rated Clinical Global Impression improvement scores. Other assessments included a panic attack inventory, clinician-rated and patient-rated versions of the Panic and Phobic Disorder Change Scale, a phobia rating scale, the Hamilton Anxiety Rating Scale, the 21-item Hamilton Depression Rating Scale, and the Sheehan Disability Scale. Correlations were determined between outcomes in individual symptom domains and overall clinical outcome.

RESULTS

Fluoxetine, particularly the 20-mg/day dose, was associated with more improvement than was placebo in patients with panic disorder across multiple symptom measures, including global improvement, total panic attack frequency, phobic symptoms, and functional impairment. Global improvement was most highly correlated with reductions in overall anxiety and phobic symptoms and least correlated with reduction in panic attacks. Fluoxetine treatment for panic disorder was well tolerated, with a safety profile consistent with that observed for fluoxetine in other disorders.

CONCLUSIONS

These data provide support for the efficacy and safety of fluoxetine treatment in reducing panic attacks, phobic symptoms, anxiety, and depressive symptoms in patients with panic disorder. Reductions in panic attack frequency in subjects given either fluoxetine or placebo were less closely related to overall clinical improvement than reductions in phobic avoidance, anxiety, depressive symptoms, and functional impairment, suggesting that outcome measures in this disorder should be more broadly based.

Analysis of homogeneity of treatment effect in adaptive multicenter clinical trials

In this paper, an exact test for analyzing the homogeneity of treatment effect in adaptive multicenter clinical trials is proposed. Extensive simulation studies are performed to investigate the large sample behavior of a commonly used test statistic for testing homogeneity of treatment effect. When the sample size in each center is large relative to the number of centers, the asymptotic null distribution of the test statistic is reasonable. On the other hand, when the data are relatively sparse, the proposed exact test should be used to incorporate the adaptive nature of the design.

Blind, controlled, long-term study of the comparative incidence of treatment-emergent tardive dyskinesia with olanzapine or haloperidol

OBJECTIVE

Tardive dyskinesia is a serious and common complication of neuroleptic treatment. Olanzapine is a novel antipsychotic agent exhibiting regional mesolimbic dopaminergic selectivity and a broad-based pharmacology encompassing serotonin, dopamine, muscarinic, and adrenergic receptor binding affinities. The authors' goal was to compare the incidence of tardive dyskinesia among patients receiving olanzapine and those receiving the conventional dopamine 2 antagonist haloperidol.

METHOD

Data were analyzed from three actively controlled and blind long-term responder studies of subjects meeting DSM-III-R criteria for schizophrenia, schizophreniform disorder, or schizoaffective disorder treated with olanzapine (N = 707, up to 20 mg/day, 237 median days of exposure) or haloperidol (N = 197, up to 20 mg/day, 203 median days of exposure) who did not have evidence of tardive dyskinesia at baseline. All of the subjects had a chronic disease course (mean greater than 10 years), and there were no significant between-treatment group differences in demographic or disease characteristics. The Abnormal Involuntary Movement Scale and research diagnostic criteria for tardive dyskinesia were used to define the comparative incidence rates of long-term treatment-emergent tardive dyskinesia.

RESULTS

The incidence of newly emergent tardive dyskinesia at any visit after baseline, at the final visit, and at the final two clinical assessments was statistically significantly lower among olanzapine-treated patients than among haloperidol-treated patients.

CONCLUSIONS

These findings support an atypical extrapyramidal symptom profile and the potential of a significantly lower risk of tardive dyskinesia with olanzapine than with haloperidol among patients requiring maintenance antipsychotic treatment.

Coating of titanium alloy with soluble laminin-5 promotes cell attachment and hemidesmosome assembly in gingival epithelial cells: potential application to dental implants

The formation of a biological seal around the transmucosal portion of dental implants may be crucial for the long-term success of these therapies. Data to date suggest that the gingival epithelium attaches to dental implants through the formation of hemidesmosomes. Biochemical and genetic data indicate that the laminin isoform, laminin-5, a component of basement membranes, plays a crucial role in the assembly and maintenance of hemidesmosomes. We report the use of soluble laminin-5 as a biological coating of titanium-alloy to promote cell attachment of the gingival epithelial cell line, IHGK. Monoclonal antibodies reactive with laminin-5 depleted the coating solution of all cell attachment activity and blocked cell attachment to laminin-5-coated disks. Immunodepletion with antibodies to fibronectin had no effect. Finally, we demonstrate that IHGK cells assembled hemidesmosomes within 24 h of attachment to laminin-5-coated titanium alloy but not to the titanium alloy alone. These results suggest that soluble laminin-5 may have clinical applications as a dental implant coating to promote the formation of a biological seal.

Olanzapine versus haloperidol in the treatment of schizophrenia and schizoaffective and schizophreniform disorders: results of an international collaborative trial

OBJECTIVE

This international, multicenter double-blind trial was designed to compare the therapeutic profile of an atypical antipsychotic, olanzapine, with that of a conventional dopamine D2 antagonist, haloperidol.

METHOD

A total of 1,996 patients at 174 sites in Europe and North America were randomly assigned to treatment with olanzapine (N = 1,336) or haloperidol (N = 660) over 6 weeks. The primary efficacy analysis involved the mean change from baseline to endpoint in total scores on the Brief Psychiatric Rating Scale (BPRS). Secondary analyses included comparisons of the mean change in positive and negative symptoms, comorbid depression, extrapyramidal symptoms, and overall drug safety.

RESULTS

Olanzapine demonstrated clinical results superior to those of haloperidol on overall improvement according to the BPRS and on every secondary measure, including depression. Olanzapine was also associated with significantly fewer discontinuations of treatment due to lack of drug efficacy or adverse events. Substantially more olanzapine-treated patients (66.5%) than haloperidol-treated patients (46.8%) completed 6 weeks of therapy. Statistically significant advantages of olanzapine treatment were related to 1) change in negative symptoms, 2) extrapyramidal symptom profile, 3) effect on prolactin levels, and 4) response rate.

CONCLUSIONS

Olanzapine shows a superior and broader spectrum of efficacy in the treatment of schizophrenic psychopathology, with a substantially more favorable safety profile, than haloperidol. It meets several of the criteria for a novel atypical antipsychotic agent.

Using the proportional odds model to assess the relationship between a multi-item and a global item efficacy scale in a psychiatric clinical trial

The proportional odds model is illustrated in the analysis of two efficacy scales used in a phase II clinical trial involving 81 schizophrenic patients. The proportional odds model preserves the discrete, ordinal nature of one of the scales. The analysis of this data suggested that the relationship between the two scales is not captured by a linear proportional odds model. A linear model and a piecewise linear model for the explanatory variable were therefore compared using likelihood-based analyses. Residuals from both models were compared. Predicted probabilities for the ordinal categories were constructed from the estimated model. Extensions and limitations of the model for interpretation of other trials and for the planning of future trials are discussed.

Rapid spreading and mature hemidesmosome formation in HaCaT keratinocytes induced by incubation with soluble laminin-5r

HaCaT cells, an immortalized keratinocyte line, incubated in plastic wells in the presence of conditioned medium from 804G cells adhered and spread rapidly in less than 30 min. In contrast, cells plated in fibroblast or keratinocyte conditioned medium adhered poorly and remained rounded at 30 min. Immunodepletion of 804G conditioned medium with polyclonal antisera to laminin-5r, but not control antisera, abolished rapid cell spreading. Electron microscopy of HaCaT cells spread by incubation in 804G conditioned medium, but not control medium, revealed mature hemidesmosomes after 24 h. Rapid spreading was also observed in wells precoated with 804G conditioned medium or 804G cell-deposited matrix, but not with fibronectin, vitronectin, or laminin-1. Immunoblotting of 804G conditioned medium with anti-laminin-5r antibodies unveiled polypeptides of 150, 140, 135, and 100 kDa, identical by electrophoretic mobility to immunoreactive polypeptides in 804G deposited matrix. Our results suggest that addition of laminin-5r in a soluble form is sufficient to promote rapid spreading and hemidesmosome assembly in keratinocytes. The mechanism of soluble laminin-5r action may include efficient surface "priming" for cell adhesion. Soluble laminin-5r may have a physiologic role in morphogenesis and repair of the epidermis and may be of use for therapeutic applications.

Quantitative measurement of alpha 6 beta 1 and alpha 6 beta 4 integrin internalization under cross-linking conditions: a possible role for alpha 6 cytoplasmic domains

In epithelial cells integrins are segregated on discrete domains of the plasma membrane. Redistribution may also occur during migration or differentiation. However, little is known about the mechanisms that control such redistribution. Receptor internalization may be a part of one such mechanism. We developed a quantitative assay and measured internalization of two epithelial integrin heterodimers, alpha 6 beta 1 and alpha 6 beta 4, induced by cross-linking with specific antibodies. alpha 6 beta 1 is a receptor for EHS laminin, while alpha 6 beta 4 is a receptor for a component of the basement membrane. alpha 6 beta 4 plays an important role in the establishment of hemidesmosomes, and becomes redistributed on the epithelial cell surface when cells are in a migratory phase. We report that alpha 6 beta 4 is efficiently internalized in human keratinocytes. More than 25% of cell surface alpha 6 beta 4 was internalized at 30 minutes, after cross-linking with A9, an anti-beta 4 monoclonal antibody. alpha 6 beta 1 is also internalized, in melanoma and teratocarcinoma cells, with maximum values of 20% of total receptors expressed at the cell surface. No significant difference was observed between the alpha 6 isoforms A and B in these assays. To determine whether alpha 6 cytoplasmic domains could influence integrin endocytosis, we prepared chimeric constructs with the extracellular domain of a reporter protein (CD8), and the cytoplasmic domains of either alpha 6 A or alpha 6 B. Both alpha 6 cytoplasmic domains but not a control cytoplasmic domain promoted internalization of the chimeric proteins, after cross-linking with antibody. Internalization of alpha 6 integrins may have a role in redistributing these receptors at the cell surface. Furthermore, the cytoplasmic domains of alpha 6 may be involved in regulating integrin internalization.

Constrained randomization in a therapeutic efficacy trial

The use of valid constrained randomization is presented for a therapeutic efficacy trial. The construction and evaluation of randomization schemes are studied for two treatment replicates per block and for hypothesized linear gradients. Two different isomorphism classes of designs are compared using the variance of the mean squared error criterion. The results indicate that valid constrained randomization schemes can be constructed that are superior to ordinary randomization. A mastitis efficacy trial provided the specific problem to be solved.

The use of repeated measures analyses in developmental toxicology studies

In many toxicology studies, it is common to take the same measurements on an individual animal at several time points (e.g., body weight across days or weeks, activity levels either within a single test session or across days). Such repeated measures for both maternal and offspring endpoints routinely are incorporated into developmental toxicology studies to allow characterization of the profile of treatment-related effects over time. To permit a valid statistical analysis, repeated measures studies are characterized by: random assignment of treatment to experimental units, measurement of the variable(s) of interest at the same time points for all animals, and avoidance of systematic overlaying of treatment and potential response gradients. Assumptions for and advantages of the use of a repeated measures analysis of variance versus single-factor analyses at each time point are illustrated in two data sets from a rat perinatal and postnatal study. The variance-covariance structure of repeated measures designs dictates that adjustments are necessary to provide protection against inflated Type 1 error rates. Communication between statisticians and toxicologists which allow the implementation of such analyses can improve the interpretation of data resulting from repeated measures study designs.

Cell type-specific integrin variants with alternative alpha chain cytoplasmic domains

The integrin heterodimers composed of the alpha 6 subunit with the beta 1 or beta 4 subunit (alpha 6 beta 1 and alpha 6 beta 4) are receptors for laminin and basement membrane components, respectively. The alpha 3 beta 1 integrin recognizes laminin, collagen, fibronectin, or epiligrin. We report the identification of structural variants (A and B) of the alpha 6 and alpha 3 subunits, containing distinct cytoplasmic domains. The expression of one cytoplasmic domain or the other, based probably on alternative exon usage, is cell-type dependent. Most transformed cell lines express both alpha 6A and alpha 6B isoforms, as determined by mRNA amplification or antibody immunoprecipitation. In contrast, embryonic fibroblasts express exclusively alpha 6A, and embryonic stem cells express exclusively alpha 6B. In most normal tissues, both alpha 6 isoforms are detectable, but several tissues express either alpha 6A or alpha 6B. The alpha 3B mRNA was amplified from heart and brain, while all other tissues and cell lines tested contained only alpha 3A mRNA. Alternative cytoplasmic domains may provide a means for varying the cellular responses to the ligands of alpha 6 and alpha 3 integrins according to the cell type.

The major laminin receptor of mouse embryonic stem cells is a novel isoform of the alpha 6 beta 1 integrin

Laminin is the first extracellular matrix protein expressed in the developing mouse embryo. It is known to influence morphogenesis and affect cell migration and polarization. Several laminin receptors are included in the integrin family of extracellular matrix receptors. Ligand binding by integrin heterodimers results in signal transduction events controlling cell motility. We report that the major laminin receptor on murine embryonic stem (ES) cells is the integrin heterodimer alpha 6 beta 1, an important receptor for laminin in neurons, lymphocytes, macrophages, fibroblasts, platelets and other cell types. However, the cytoplasmic domain of the ES cell alpha 6 (alpha 6 B) differs totally from the reported cytoplasmic domain amino acid sequence of alpha 6 (alpha 6 A). Comparisons of alpha 6 cDNAs from ES cells and other cells suggest that the alpha 6 A and alpha 6 B cytoplasmic domains derive from alternative mRNA splicing. Anti-peptide antibodies to alpha 6 A are unreactive with ES cells, but react with mouse melanoma cells and embryonic fibroblasts. When ES cells are cultured under conditions that permit their differentiation, they become positive for alpha 6 A, concurrent with the morphologic appearance of differentiated cell types. Thus, expression of the alpha 6 B beta 1 laminin receptor may be favored in undifferentiated, totipotent cells, while the expression of alpha 6 A beta 1 receptor occurs in committed lineages. While the functions of integrin alpha chain cytoplasmic domains are not understood, it is possible that they contribute to transferring signals to the cell interior, e.g., by delivering cytoskeleton organizing signals in response to integrin engagement with extracellular matrix ligands. It is therefore reasonable to propose that the cellular responses to laminin may vary, according to what alpha subunit isoform (alpha 6 A or alpha 6 B) is expressed as part of the alpha 6 beta 1 laminin receptor. The switch from alpha 6 B to alpha 6 A, if confirmed in early embryos, could then be of striking potential relevance to the developmental role of laminin.

Target organ toxicity and leukopenia in Fischer 344 rats given intravenous doses of vinblastine and desacetyl vinblastine

The toxicity and leukopenia produced by vinblastine or desacetyl vinblastine were established in 1-month studies in rats. Groups of male Fischer 344 rats were given weekly intravenous doses of vinblastine or desacetyl vinblastine at doses of 0.08, 0.16, 0.32, 0.64, or 1.28 mg/kg. The target organ toxicity was similar for both compounds. Decreased cell production in the thymus, testes, and bone marrow was produced in the animals of the two highest dose groups for both compounds with the degree of severity greater in the highest dose group. All high dose rats (1.28 mg/kg) given desacetyl vinblastine and three rats given vinblastine died prior to study termination. Body weight loss was more pronounced in high dose rats given desacetyl vinblastine, but at lower doses there were no significant differences in body weight reduction for rats receiving either compound. Leukopenia occurred at all dose levels of 0.32 mg/kg and higher. During the first 2 weeks of the study, rats given 1.28 mg/kg of desacetyl vinblastine had a greater leukopenic response than rats given 1.28 mg/kg of vinblastine. It is concluded that both compounds produced similar target organ toxicity and leukopenia without deaths at doses of 0.32 and 0.64 mg/kg given once a week for 4 weeks. At the high dose (1.28 mg/kg) desacetyl vinblastine was more toxic than vinblastine resulting in greater mortality and body weight reduction.