Compositional ideas in the bayesian analysis of categorical data with application to dose finding clinical trials

Compositional random vectors are fundamental tools in the Bayesian analysis of categorical data.Many of the issues that are discussed with reference to the statistical analysis of compositionaldata have a natural counterpart in the construction of a Bayesian statistical model for categoricaldata.This note builds on the idea of cross-fertilization of the two areas recommended by Aitchison (1986)in his seminal book on compositional data. Particular emphasis is put on the problem of whatparameterization to use

Compositional ideas in the bayesian analysis of categorical data with application to dose finding clinical trials

Compositional random vectors are fundamental tools in the Bayesian analysis of categorical data. Many of the issues that are discussed with reference to the statistical analysis of compositional data have a natural counterpart in the construction of a Bayesian statistical model for categorical data. This note builds on the idea of cross-fertilization of the two areas recommended by Aitchison (1986) in his seminal book on compositional data. Particular emphasis is put on the problem of what parameterization to use

BAYESIAN PHASE I DOSE FINDING IN CANCER TRIALS

This dissertation explores phase I dose-finding designs in cancer trials from three perspectives: the alternative Bayesian dose-escalation rules, a design based on a time-to-dose-limiting toxicity (DLT) model, and a design based on a discrete-time multi-state (DTMS) model. We list alternative Bayesian dose-escalation rules and perform a simulation study for the intra-rule and inter-rule comparisons based on two statistical models to identify the most appropriate rule under certain scenarios. We provide evidence that all the Bayesian rules outperform the traditional ``3+3'' design in the allocation of patients and selection of the maximum tolerated dose. The design based on a time-to-DLT model uses patients' DLT information over multiple treatment cycles in estimating the probability of DLT at the end of treatment cycle 1. Dose-escalation decisions are made whenever a cycle-1 DLT occurs, or two months after the previous check point. Compared to the design based on a logistic regression model, the new design shows more safety benefits for trials in which more late-onset toxicities are expected. As a trade-off, the new design requires more patients on average. The design based on a discrete-time multi-state (DTMS) model has three important attributes: (1) Toxicities are categorized over a distribution of severity levels, (2) Early toxicity may inform dose escalation, and (3) No suspension is required between accrual cohorts. The proposed model accounts for the difference in the importance of the toxicity severity levels and for transitions between toxicity levels. We compare the operating characteristics of the proposed design with those from a similar design based on a fully-evaluated model that directly models the maximum observed toxicity level within the patients' entire assessment window. We describe settings in which, under comparable power, the proposed design shortens the trial. The proposed design offers more benefit compared to the alternative design as patient accrual becomes slower.

Intravenous immunoglobulin for treatment of mild-to-moderate Alzheimer's disease: a phase 2, randomised, double-blind, placebo-controlled, dose-finding trial.

BACKGROUND: Three small trials suggest that intravenous immunoglobulin can affect biomarkers and symptoms of mild-to-moderate Alzheimer's disease. We tested the safety, effective dose, and infusion interval of intravenous immunoglobulin in such patients. METHODS: We did a multicentre, placebo-controlled phase 2 trial at seven sites in the USA and five in Germany. Participants with probable Alzheimer's disease aged 50-85 years were randomly assigned (by a computer-generated randomisation sequence, with block sizes of eight) to infusions every 4 weeks (0·2, 0·5, or 0·8 g intravenous immunoglobulin per kg bodyweight, or placebo) or infusions every 2 weeks (0·1, 0·25, or 0·4 g/kg, or placebo). Patients, caregivers, investigators assessing outcomes, and staff at imaging facilities and the clinical research organisation were masked to treatment allocation, but dispensing pharmacists, the statistician, and the person responsible for final PET analyses were not. Treatment was masked with opaque pouches and infusion lines. The primary endpoint was median area under the curve (AUC) of plasma amyloid β (Aβ)(1-40) between the last infusion and the final visit (2 weeks or 4 weeks depending on infusion interval) in the intention-to-treat population. The trial is registered at ClinicalTrials.gov (NCT00812565) and controlled-trials.com (ISRCTN64846759). FINDINGS: 89 patients were assessed for eligibility, of whom 58 were enrolled and 55 included in the primary analysis. Median AUC of plasma Aβ(1-40) was not significantly different for intravenous immunoglobulin compared with placebo for five of the six intervention groups (-18·0 [range -1347·0 to 1068·5] for 0·2 g/kg, -364·3 [-5834·5 to 1953·5] for 0·5 g/kg, and -351·8 [-1084·0 to 936·5] for 0·8 g/kg every 4 weeks vs -116·3 [-1379·0 to 5266·0] for placebo; and -13·8 [-1729·0 to 307·0] for 0·1 g/kg, and -32·5 [-1102·5 to 451·5] for 0·25 g/kg every 2 weeks vs 159·5 [51·5 to 303·0] for placebo; p>0·05 for all). The difference in median AUC of plasma Aβ(1-40) between the 0·4 g/kg every 2 weeks group (47·0 [range -341·0 to 72·5]) and the placebo group was significant (p=0·0216). 25 of 42 (60%) patients in the intervention group versus nine of 14 (64%) receiving placebo had an adverse event. Four of 42 (10%) patients in the intravenous immunoglobulin group versus four of 14 (29%) receiving placebo had a serious adverse event, including one stroke in the intervention group. INTERPRETATION: Intravenous immunoglobulin may have an acceptable safety profile. Our results did not accord with those from previous studies. Longer trials with greater power are needed to assess the cognitive and functional effects of intravenous immunoglobulin in patients with Alzheimer's disease.

Safety and immunogenicity of a chimpanzee adenovirus-vectored Ebola vaccine in healthy adults: a randomised, double-blind, placebo-controlled, dose-finding, phase 1/2a study.

BACKGROUND: The ongoing Ebola outbreak led to accelerated efforts to test vaccine candidates. On the basis of a request by WHO, we aimed to assess the safety and immunogenicity of the monovalent, recombinant, chimpanzee adenovirus type-3 vector-based Ebola Zaire vaccine (ChAd3-EBO-Z). METHODS: We did this randomised, double-blind, placebo-controlled, dose-finding, phase 1/2a trial at the Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland. Participants (aged 18-65 years) were randomly assigned (2:2:1), via two computer-generated randomisation lists for individuals potentially deployed in endemic areas and those not deployed, to receive a single intramuscular dose of high-dose vaccine (5 × 10(10) viral particles), low-dose vaccine (2·5 × 10(10) viral particles), or placebo. Deployed participants were allocated to only the vaccine groups. Group allocation was concealed from non-deployed participants, investigators, and outcome assessors. The safety evaluation was not masked for potentially deployed participants, who were therefore not included in the safety analysis for comparison between the vaccine doses and placebo, but were pooled with the non-deployed group to compare immunogenicity. The main objectives were safety and immunogenicity of ChAd3-EBO-Z. We did analysis by intention to treat. This trial is registered with ClinicalTrials.gov, number NCT02289027. FINDINGS: Between Oct 24, 2014, and June 22, 2015, we randomly assigned 120 participants, of whom 18 (15%) were potentially deployed and 102 (85%) were non-deployed, to receive high-dose vaccine (n=49), low-dose vaccine (n=51), or placebo (n=20). Participants were followed up for 6 months. No vaccine-related serious adverse events were reported. We recorded local adverse events in 30 (75%) of 40 participants in the high-dose group, 33 (79%) of 42 participants in the low-dose group, and five (25%) of 20 participants in the placebo group. Fatigue or malaise was the most common systemic adverse event, reported in 25 (62%) participants in the high-dose group, 25 (60%) participants in the low-dose group, and five (25%) participants in the placebo group, followed by headache, reported in 23 (57%), 25 (60%), and three (15%) participants, respectively. Fever occurred 24 h after injection in 12 (30%) participants in the high-dose group and 11 (26%) participants in the low-dose group versus one (5%) participant in the placebo group. Geometric mean concentrations of IgG antibodies against Ebola glycoprotein peaked on day 28 at 51 μg/mL (95% CI 41·1-63·3) in the high-dose group, 44·9 μg/mL (25·8-56·3) in the low-dose group, and 5·2 μg/mL (3·5-7·6) in the placebo group, with respective response rates of 96% (95% CI 85·7-99·5), 96% (86·5-99·5), and 5% (0·1-24·9). Geometric mean concentrations decreased by day 180 to 25·5 μg/mL (95% CI 20·6-31·5) in the high-dose group, 22·1 μg/mL (19·3-28·6) in the low-dose group, and 3·2 μg/mL (2·4-4·9) in the placebo group. 28 (57%) participants given high-dose vaccine and 31 (61%) participants given low-dose vaccine developed glycoprotein-specific CD4 cell responses, and 33 (67%) and 35 (69%), respectively, developed CD8 responses. INTERPRETATION: ChAd3-EBO-Z was safe and well tolerated, although mild to moderate systemic adverse events were common. A single dose was immunogenic in almost all vaccine recipients. Antibody responses were still significantly present at 6 months. There was no significant difference between doses for safety and immunogenicity outcomes. This acceptable safety profile provides a reliable basis to proceed with phase 2 and phase 3 efficacy trials in Africa. FUNDING: Swiss State Secretariat for Education, Research and Innovation (SERI), through the EU Horizon 2020 Research and Innovation Programme.

Understanding the Continual Reassessment Method for Dose Finding Studies: An Overview for Non-Statisticians

The Continual Reassessment Method (CRM) has gained popularity since its proposal by O’Quigley et al. [1]. Many variations have been published and discussed in the statistical literature, but there has been little attention to making the design considerations accessible to non-statisticians. As a result, some clinicians or reviewers of clinical trials tend to be wary of the CRM due to safety concerns. This paper presents the CRM in a non-technical way, describing the original CRM with some of its modified versions. It also describes the specifications that define a CRM design, along with two simulated examples of CRMs for illustration.

Optimal design for dose finding studies on safety and efficacy

Magdeburg, Univ., Fak. für Mathematik, Diss., 2009

How Low Can We Go in Contrast-Enhanced CT Imaging of the Chest?: A Dose-Finding Cadaver Study Using the Model-based Iterative Image Reconstruction Approach.

RATIONALE AND OBJECTIVES: Dose reduction may compromise patients because of a decrease of image quality. Therefore, the amount of dose savings in new dose-reduction techniques needs to be thoroughly assessed. To avoid repeated studies in one patient, chest computed tomography (CT) scans with different dose levels were performed in corpses comparing model-based iterative reconstruction (MBIR) as a tool to enhance image quality with current standard full-dose imaging. MATERIALS AND METHODS: Twenty-five human cadavers were scanned (CT HD750) after contrast medium injection at different, decreasing dose levels D0-D5 and respectively reconstructed with MBIR. The data at full-dose level, D0, have been additionally reconstructed with standard adaptive statistical iterative reconstruction (ASIR), which represented the full-dose baseline reference (FDBR). Two radiologists independently compared image quality (IQ) in 3-mm multiplanar reformations for soft-tissue evaluation of D0-D5 to FDBR (-2, diagnostically inferior; -1, inferior; 0, equal; +1, superior; and +2, diagnostically superior). For statistical analysis, the intraclass correlation coefficient (ICC) and the Wilcoxon test were used. RESULTS: Mean CT dose index values (mGy) were as follows: D0/FDBR = 10.1 ± 1.7, D1 = 6.2 ± 2.8, D2 = 5.7 ± 2.7, D3 = 3.5 ± 1.9, D4 = 1.8 ± 1.0, and D5 = 0.9 ± 0.5. Mean IQ ratings were as follows: D0 = +1.8 ± 0.2, D1 = +1.5 ± 0.3, D2 = +1.1 ± 0.3, D3 = +0.7 ± 0.5, D4 = +0.1 ± 0.5, and D5 = -1.2 ± 0.5. All values demonstrated a significant difference to baseline (P < .05), except mean IQ for D4 (P = .61). ICC was 0.91. CONCLUSIONS: Compared to ASIR, MBIR allowed for a significant dose reduction of 82% without impairment of IQ. This resulted in a calculated mean effective dose below 1 mSv.

Phase I dose-finding study of cabazitaxel administered weekly in patients with advanced solid tumours.

BACKGROUND Cabazitaxel is approved in patients with metastatic hormone-refractory prostate cancer previously treated with a docetaxel-containing regimen. This study evaluated a weekly cabazitaxel dosing regimen. Primary objectives were to report dose-limiting toxicities (DLTs) and to determine the maximum tolerated dose (MTD). Efficacy, safety and pharmacokinetics were secondary objectives. METHODS Cabazitaxel was administered weekly (1-hour intravenous infusion at 1.5-12 mg/m2 doses) for the first 4 weeks of a 5-week cycle in patients with solid tumours. Monitoring of DLTs was used to determine the MTD and the recommended weekly dose. RESULTS Thirty-one patients were enrolled. Two of six patients experienced DLTs at 12 mg/m2, which was declared the MTD. Gastrointestinal disorders were the most common adverse event. Eight patients developed neutropenia (three ≥ Grade 3); one occurrence of febrile neutropenia was reported. There were two partial responses (in breast cancer) and 13 patients had stable disease (median duration of 3.3 months). Increases in Cmax and AUC0-t were dose proportional for the 6-12 mg/m2 doses. CONCLUSION The MTD of weekly cabazitaxel was 12 mg/m2 and the recommended weekly dose was 10 mg/m2. The observed safety profile and antitumour activity of cabazitaxel were consistent with those observed with other taxanes in similar dosing regimens. TRIAL REGISTRATION The study was registered with ClinicalTrials.gov as NCT01755390.

Efficacy and safety of a phospholipid emulsion (GR270773) in Gram-negative severe sepsis: results of a phase II multicenter, randomized, placebo-controlled, dose-finding clinical trial.

OBJECTIVE: To assess the survival benefit and safety profile of low-dose (850 mg/kg) and high-dose (1350 mg/kg) phospholipid emulsion vs. placebo administered as a continuous 3-day infusion in patients with confirmed or suspected Gram-negative severe sepsis. Preclinical and ex vivo studies show that lipoproteins bind and neutralize endotoxin, and experimental animal studies demonstrate protection from septic death when lipoproteins are administered. Endotoxin neutralization correlates with the amount of phospholipid in the lipoprotein particles. DESIGN: A three-arm, randomized, blinded, placebo-controlled trial. SETTING: Conducted at 235 centers worldwide between September 2004 and April 2006. PATIENTS: A total of 1379 patients participated in the study, 598 patients received low-dose phospholipid emulsion, and 599 patients received placebo. The high-dose phospholipid emulsion arm was stopped, on the recommendation of the Independent Data Monitoring Committee, due to an increase in life-threatening serious adverse events at the fourth interim analysis and included 182 patients. MEASUREMENTS AND MAIN RESULTS: A 28-day all-cause mortality and new-onset organ failure. There was no significant treatment benefit for low- or high-dose phospholipid emulsion vs. placebo for 28-day all-cause mortality, with rates of 25.8% (p = .329), 31.3% (p = .879), and 26.9%, respectively. The rate of new-onset organ failure was not statistically different among groups at 26.3%, 31.3%, 20.4% with low- and high-dose phospholipid emulsion, and placebo, respectively (one-sided p = .992, low vs. placebo; p = .999, high vs. placebo). Of the subjects treated, 45% had microbiologically confirmed Gram-negative infections. Maximal changes in mean hemoglobin levels were reached on day 10 (-1.04 g/dL) and day 5 (-1.36 g/dL) with low- and high-dose phospholipid emulsion, respectively, and on day 14 (-0.82 g/dL) with placebo. CONCLUSIONS: Treatment with phospholipid emulsion did not reduce 28-day all-cause mortality, or reduce the onset of new organ failure in patients with suspected or confirmed Gram-negative severe sepsis.

Adaptive designs for Phase I dose-finding studies

Improving methodology for Phase I dose-finding studies is currently of great interest in pharmaceutical and medical research. This article discusses the current atmosphere and attitude towards adaptive designs and focuses on the influence of Bayesian approaches.

Positioning and number of nutritional levels in dose-response trials to estimate the optimal-level and the adjustment of the models

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)