Revisiting Partnerships 2010: Designing Safe and Efficient Phase I Studies

Partnerships in Clinical Trials is taking place this March 30-April 1, 201a in Phoenix, Arizona. Every Friday leading up to the 20th anniversary of the event, we'll be recapping one session from Partnerships in Clinical Trials.

Designing Safe and Efficient Phase I Studies – Partnerships 2010

Workshop Presenters:
Mike Tanguay, BPharm, PhD, VP, Scientific Affairs, Anapharm, Inc
Eric Masson, PharmD, Director, Clinical Pharmacology, Discovery Medicine and Clinical Pharmacology, Bristol-Myers Squibb


The first presentation (Dr Tanguay) covers first in human trials for novel chemical entities. Dr Tanguay explores the trends driving more complex Phase I studies, the application of creativity in Phase I study design to enhance the scope of data obtained, and specifically the increased interest in evaluating QT potential within First in Human studies.

Attention to Study Drug Formulation: A key issue can be formulation and particularly with the frequency we see poorly soluble compounds. The speaker explores appropriate formulation optimization in the face of the demand for expediting initial study timelines. Noteworthy is that about half the time the formulation is simply a powder in bottle at the time of study initiation.

Keeping the Primary Objectives in Mind: Objectives of Phase I trials are reviewed including the specific examination of dose and dose range suitability for future studies and covering such topics as how to determine the maximum recommended starting dose in light of the EMEA Guidance for FIH trials published in 2007. NOAEL safety factor of 10 was applied in approximately 40% of trials reviewed.

Subject Population and Cohort Size: Patient population selection including entry criteria for aged and gender specific populations are discussed including when to introduce cohorts of the appropriate disease target population. Factors for inclusion of females of childbearing potential are reviewed and covered areas such as drug half-life. It is reported that only 11% of studies are male only and about 2/3 of trials currently include women of childbearing potential.

Number of subjects per cohort for typical MAD studies is optimized as 6 on active drug and 2 placebos. Six subjects should be sufficient to detect an adverse event with a frequency of 25% using an 80% probability level.

Dose Escalation and Timing: Dose escalation in these types of studies is managed on a case-by-case approach and guided by drug class. Classical approaches include doubling the dose at each cohort, a Fibonacci approach, etc. Ninety percent of studies will follow a doubling protocol for the low dose ranges while higher dose ranges will observe a more conservative model.

Interim blinded PK data analysis is recommended here despite the potential for study timeline effect the case being made for allowing more precise determination of stopping criteria in escalating dose studies.

It is pointed out that dosing strategy is coming under more scrutiny by regulatory bodies and should be well defined and utilize conservative staggering models within the protocol.

Study Design: The advantages of parallel versus cross-over designs are reviewed. While parallel designs may be generally considered safer as an approach to escalating dose studies, cross-over studies limit the number of subjects that need to be recruited. Since PK is not generally considered a first priority within a FIH study the author reviews various hybrid approaches. Examples are given within the slide deck accessible on the Partners 2010 website. Currently, the author states that about 20% of trials employ some type of hybrid approach.

Integrating Studies to Obtain Efficiencies: Efficiency in a Phase I program is driven by the need for obtaining proof of concept data as well as sponsor requirements for more data per trial. Integrated Phase I protocols are becoming more common. Forty percent of the authors are studies are now integrated SAD/MAD studies or studies including an additional are to elucidate food effect or, more rarely, effects on CYP2D6 poor metabolizers.

The author presents several case studies of ‘integrated’ SAD/MAD studies stating that this type of integrated trial has increased 4-fold in his clinic over the past year. It is stressed that an SAD cohort should have been completed at a higher dose than any pending MAD dose cohort in order to maintain a conservative safety margin within the higher risk study design. Drawbacks to this type of integrated study include inability to fully evaluate the SAD study prior to initiating the MAD study.

TQT studies are discussed in the context of obtaining intensive cardiac data within the escalating dose or FIH study. Advantages would include evaluation of electrocardiograms at substantially high doses prior to a full TQT study. Since positive controls are not employed this will not substitute for a complete TQT study and may or may not support a waiver for future TQT analysis.

Conclusions: In concluding the author reviewed questions frequently asked by Institutional Review Boards and regulatory agencies including Health Canada in considering First in Human Studies as well as guidance documents available to study design teams.

The second (Dr Masson’s) presentation comprises a case study on the conduct of Phase I Oncology studies and explores the use of healthy and patients.

Cytotoxics: Cytotoxicity has generally dictated the use of patients, however, recently, the development of targeted therapeutics has opened the way for using healthy volunteers. About 6 products of this class are presented were enumerated. The type of studies and compounds suitable for use of normal healthy volunteers (NHV) were explored as well as those precluding their use.

The use of NHV is attractive from a costs and resource perspective as well as timing. The cases reviewed compared use of NHV versus patients across several types of studies and concluded that a 2 to 3 fold savings can be achieved if NHV can safely be employed.

Biologics: The use of biologics and specific considerations including cytokine release was reviewed and appropriate study design (starting dose, etc) in light of half-life was discussed.

SAD/MAD studies are employed to define maximum tolerated dose and minimum biologic effect. Within this context the author discusses the use of combination drugs and the danger of overlapping toxicity along with the utility of SAD/MAD studies to categorize these risks.

QT/QTc studies in oncology were discussed along with potential confounding factors including excluding patients at extreme risk, con-meds, and electrolyte imbalance precipitated by patient nausea and vomiting.

Case Study: The author presents an ‘in-depth’ case study of the studies employed in the development and subsequent approval including the initial SAD study, the MAD study in patients with solid tumors, Food effect, ADME, and hepatic impairment. Drug-Drug interaction studies were completed in NHV subjects at drug levels comparable to those achieved in patients.



Author: Ronald Norton