Clinical drug trials are costly and time-consuming. Bayesian methods alleviate the inefficiencies in the testing process while providing user-friendly probabilistic inference and predictions from the sampled posterior distributions, saving resources, time, and money. We propose a dynamic linear model to estimate the mean response at each dose level, borrowing strength across dose levels. Our model permits nonmonotonicity of the dose-response relationship, facilitating precise modeling of a wider array of dose-response relationships (including the possibility of toxicity). In addition, we incorporate an adaptive approach to the design of the clinical trial, which allows for interim decisions and assignment to doses based on dose-response uncertainty and dose efficacy. The interim decisions we consider are stopping early for success and stopping early for futility, allowing for patient and time savings in the drug development process. These methods complement current clinical trial design research.

The clinical trial is the most import study for the development of successful novel drugs. The aim of this dissertation is to develop innovative statistical methods to overcome the three main obstacles in clinical trials: (1) lengthy trial duration and inaccurate maximum tolerated dose (MTD) in phase I trials; (2) heterogeneity in drug effect when patients are given the same prescription and same dose; and (3) high failure rates of expensive phase III confirmatory trials due to the discrepancy in the endpoints adopted in phase II and III trials. Towards overcoming the first obstacle, we originally develop a hybrid design for the time-to-event dose escalation method with overdose control using a normalized equivalent toxicity score (NETS) system. This hybrid design can substantially reduce sample size, shorten study length, and estimate accurate MTD by employing a parametric model and adaptive Bayesian approach. Toward overcoming the second obstacle, we propose a new approach to incorporate patients’ characteristic using our proposed design in phase I clinical trials which considers the personalized information for patients who participant in the trials. To conquer the third obstacle, we propose a novel two-stage screening design for phase II trials whereby the endpoint of percent change in of tumor size is used in an initial screening to select potentially effective agents within a short time interval followed by a second screening stage where progression free survival is estimated to confirm the efficacy of agents. These research projects will substantially benefit both cancer patients and researchers by improving clinical trial efficiency and reducing cost and trial duration. Moreover, they are of great practical meaning since cancer medicine development is of paramount importance to human health care.

I performed two randomised controlled trials (RCTs) to determine the best treatments for enteric fever in Kathmandu, Nepal, an area with a high proportion of nalidixic acid resistant S. Typhi and S. Paratyphi A isolates. I recruited 844 patients with suspected enteric fever to compare chloramphenicol versus gatifloxacin. 352 patients were culture confirmed. 14/175 patients treated with chloramphenicol and 12/177 patients treated with gatifloxacin experienced treatment failure (HR=0.86 (95% CI 0.40 to 1.86), p=0.70). The median times to fever clearance were 3.95 and 3.90 days, respectively (HR=1.06 [CI 0.86 to 1.32], p=0.59). The second RCT compared ofloxacin versus gatifloxacin and recruited 627 patients. Of the 170 patients infected with nalidixic acid resistant strains, the number of patients with treatment failure was 6/83 in the ofloxacin group and 5/87 in the gatifloxacin group (Hazard Ratio, HR=0.81, 95% CI 0.25 to 2.65; p=0.73); the median times to fever clearance were 4.7 and 3.3 days respectively (HR=1.59 [CI 1.16 to 2.18], p=0.004). I compared conventional blood culture against an electricity free culture approach. 66 of 304 patients with suspected enteric fever were positive for S. Typhi or S. Paratyphi A, 55 (85%) isolates were identified by the conventional blood culture and 60 (92%) isolates were identified by the experimental method. The percentages of positive and negative agreement for diagnosis of enteric fever were 90.9% and 96.0%, respectively. This electricity free blood culture system may have utility in resource-limited settings or potentially in disaster relief and refugee camps. I performed a literature review of RCTs of enteric fever which showed that trial design varied greatly. I was interested in the perspective of patients and what they regarded as cure. 1,481 patients were interviewed at the start of treatment, 860 (58%) reported that the resolution of fever would mean cure to them. At the completion of treatment, 877/1,448 (60.6%) reported that they felt cured when fever was completely gone. We suggest that fever clearance time is the best surrogate for clinical cure in patients with enteric fever and should be used as the primary outcome in future RCTs for the treatment of enteric fever.

Des données complexes sont fréquentes dans les essais cliniques récents et nécessitent des méthodes statistiques adaptées. La recherche vaccinale du VIH est un exemple d’un domaine avec des données complexes et une absence de critères de jugement validés dans les essais précoces. Cette thèse d’Université concerne des recherches méthodologiques sur la conception et les aspects statistiques des essais cliniques vaccinaux du VIH, en particulier sur les critères de jugement d’immunogénicité et les schémas d’essai de phase I-II. A l’aide des données cytokiniques multiplex, nous illustrons les aspects méthodologiques spécifiques à une technique de mesure. Nous proposons ensuite des définitions de critères de jugement et des méthodes statistiques adéquates pour l'analyse des données d'immunogénicité multidimensionnelles. En particulier, nous montrons l’intérêt des scores multivariés non-paramétriques, permettant de résumer l’information à travers différents marqueurs d’immunogénicité et de faire des comparaisons inter- et intra-groupe. Dans l’objectif de contribuer à la conception méthodologique des nouveaux essais vaccinaux, nous présentons la construction d’un schéma d’essai optimisé pour le développement clinique précoce. En imbriquant les phases I et II d’évaluation clinique, ce schéma permet d’accélerer le développement de plusieurs stratégies vaccinales en parallèle. L’intégration d’une règle d’arrêt est proposée dans des perspectives fréquentistes et Bayesiennes. Les méthodes mises en avant dans cette thèse sont transposables à d’autres domaines d’application avec des données complexes, telle que les données d’imagerie ou les essais d’autres immunothérapies. / Complex data are frequently recored in recent clinical trials and require the use of appropriate statistical methods. HIV vaccine research is an example of a domaine with complex data and a lack of validated endpoints for early-stage clinical trials. This thesis concerns methodological research with regards to the design and analysis aspects of HIV vaccine trials, in particular the definition of immunogenicity endpoints and phase I-II trial designs. Using cytokine multiplex data, we illustrate the methodological aspects specific to a given assay technique. We then propose endpoint definitions and statistical methods appropriate for the analysis of multidimensional immunogenicity data. We show in particular the value of non-parametric multivariate scores, which allow for summarizing information across different immunogenicity markers and for making statistical comparisons between and within groups. In the aim of contributing to the design of new vaccine trials, we present the construction of an optimized early-stage HIV vaccine design. Combining phase I and II assessments, the proposed design allows for accelerating the clinical development of several vaccine strategies in parallel. The integration of a stopping rule is proposed from both a frequentist and a Bayesian perspective. The methods advocated in this thesis are transposable to other research domains with complex data, such as imaging data or trials of other immune therapies.