Speakers


Elizabeth EISENHAUER
Queen's University,
Canada

Obtained her MD from Queen's University Kingston, Canada in 1976 and subsequently completed Royal College of Physicians and Surgeons (Canada) training in both Internal Medicine and Hematology. In 1982 she became Director of the Investigational New Drug Program (IND) of the National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) where her major responsibilities lie in identifying and bringing into clinical trial novel cancer agents. She is also a Professor in the Department of Oncology at Queen’s University.

Her major research interest has been the evaluation of new anti-cancer agents. Through her work in the IND Program at NCIC CTG coordinating over 170 phase I, II and III trials carried out in institutions in Canada, the US and Europe. These have included some of the first trials of paclitaxel and docetaxel, studies of topotecan, gemcitabine, various targeted antisense agents, angiogenesis inhibitors and small molecule signaling inhibitors.

Dr. Eisenhauer has also has served on committees of many national and international bodies including the National Cancer Institute of Canada, the Canadian Cancer Society, the Royal College of Physicians and Surgeons of Canada, the European Organization for Research and Treatment of Cancer, the American Society of Clinical Oncology and the US National Cancer Institute.



In 1998 she was the Michel Clavel Award lecturer at the NCI-EORTC Symposium on New Drugs in Cancer Treatment held in Amsterdam. In 2002 she was awarded the O. Harold Warwick Prize by the National Cancer Institute of Canada that recognizes a scientist whose research has had a major impact in Cancer Control in Canada. In 2010 The Society of Gynecologic Oncology of Canada (GOC) presented Dr. Eisenhauer with the GOC Presidential Medal Award in recognition of her outstanding contributions to gynecologic oncology in Canada and Abroad.

In 2002 she was appointed to the Board of Directors of the National Cancer Institute of Canada and from June 2006-February 2009 she was President of NCIC. As of September 2008, in addition to continuing her role as Director of the NCIC CTG IND program, she assumed the position of Chair of the Research Advisory Group of the Canadian Partnership Against Cancer, the federally funded agency charged with implementing the Canadian Strategy for Cancer Control. In this role she is co-chairing the Canadian Cancer Research Alliance, the alliance of Canada’s major cancer research funders.



Abstract Title: Clinical experience with targeted therapeutics and study design challenges

The "era" of clinical evaluation targeted therapeutics began 10-15 years ago with the first entry into the clinic of EGFR family receptor inhibitors (gefitinib and trastuzumab). Since then, the list of agents entering the clinic has grown at an exponential pace. Targeted inhibitors of receptor tyrosine kinases (e.g. EGFR, HER2, KIT, ABL, IGF1R, MET), pathway signalling molecules (e.g. mTOR, RAF, RAS, PI3k, AKT), extracellular matrix and angiogenesis (e.g. MMP, VEGF, VEGFR) have all been studied clinically. Often multiple agents affecting the same target from competing companies have been in clinical development at the same time.

There has been considerable debate amongst experts about how best to undertake the clinical development of targeted agents. Many have advocated the need for changes to clinical trial design, including development of new endpoints, strategies to identify predictive biomarkers and enrich trial populations, and changes to the usually discrete phase I to II to III transition to more seamless movement between trial phases, and adapting the design to emerging results. Despite all these ideas, little evidence was available at the beginning of this era to know: a) what endpoints are predictive of "success" (improved survival), b) when along the path of preclinical-clinical development is a predictive biomarker best selected? How should biomarkers be validated and diagnostics developed during clinical evaluation of new agents? Now that there is more than a decade of experience behind us, some observations about what has been learned can inform future work.

Phase I
Most phase I trials of targeted agents have utilized toxicity endpoints, rather than measures of target inhibition (pharmacodynamic measures), to define dose. For example in one review, of the 60 completed phase I studies, 36 used toxicity and eight used pharmacokinetic data as endpoints for selection of the recommended phase II dose. Non-traditional endpoints, such as measures of molecular drug effects in tumor or surrogate tissue or functional imaging studies, were not routinely incorporated into the study design and rarely formed the primary basis for dose selection [1].

In part this has been because toxicity must always be measured in any case and may be mechanistically related to drug effects. In addition, by knowing the highest tolerable dose, one is unlikely to select a dose that, in retrospect, was too low for efficacy. One major reason for use of the toxicity endpoint has been the challenge associated with measuring target inhibition in fresh tumour tissue. The frequency, timing of biopsies, availability of adequately validated assays and other considerations have tended to limit the use of PD measures as the primary endpoint. However, increasingly phase I trials are designed with a plan to confirm target inhibition via tissue biopsy or imaging studies at the very least at the expanded recommended phase II dose level [2]. Indeed the increased focus on ensuring appropriate phase I evaluation of pharmacokinetic, pharmacodynamic and early predictive biomarker measures have been labelled the “Pharmacologic Audit Trail” [3]. This approach is particularly important when drugs being studied are specifically designed to target mutant proteins, when knowledge of proof of principle early in development will be critical.

Phase II
The main concerns in phase II have centred on the belief that so-called “cytostatic” targeted agents may not cause tumour regression, thus rendering the use of the traditional objective response endpoint meaningless. Other intermediate endpoints such as stable disease or non-progression rates, progression free survival, serum marker changes, or metabolic changes by imaging (FDG PET) have been proposed instead. Many of these endpoints, in particular those assessing stable disease or time to progression, require randomized controlled designs to enable appropriate interpretation of the results [4]. In the setting of single agent investigation, this is problematic since often these agents are tested in the situation where no standard treatment is available. A randomized discontinuation design has been proposed as one that may be useful [5]. Despite the theoretical concerns about an objective response endpoint being useful, in a review of 89 phase II trials of single targeted agents, El-Maraghi found that most (61) used objective response as primary, or co-primary, endpoint [6]. Indeed, in this review, there was a strong relationship between the overall rate of response and ultimate success in phase III, suggesting that objective response, particularly for agents targeted critical mutated proteins or “oncogenic drivers” remains an important screening measure of clinical activity

Two other areas of debate in phase II have been identified:
i) Whether and how to enrich populations on the basis of molecular markers to those expected to benefit
ii) Whether randomized phase II -III design is most efficient to obtain speedy results.

With respect to the question of enrichment, it is clear that an overriding goal of all targeted agent development is to understand which patient subsets benefit most. Whether the start of phase II trials is the right place to begin to enrich for putative predictive markers depends on the weight of evidence around the utility of the enrichment factor and the ease with which it can be reliably measured. For example, when testing a drug designed to target mutant bRAF, it makes sense to ensure all, or a majority, of phase II (or even phase I) patients carry the relevant mutation. On the other hand, when targeting general processes such as angiogenesis or proliferation, the exact molecular subset of patients who benefit is often elusive until substantial clinical data is amassed. Many drugs fall between these extremes so another key role of the phase II study is to begin to amass data about the relationship between genetic characteristics of patients and clinical outcome. To do this, tissue collection must be mandatory and the appropriate translational research facilities readily available.

Phase III
There is little debate that the main goal of phase III trials remains to identify if new agents improve the length or quality of patient survival. Thus patient reported outcomes and overall survival are key endpoints. The phase III trial may be designed with early stopping for futility to limit patient recruitment to inactive regimens. Alternatively, a separate phase II randomized study may be undertaken prior to phase III with only those regimens/agents showing improvement in appropriate phase II endpoint(s) (survival or progression) moving into phase III. This approach is becoming widespread and the field would benefit from a systematic analysis to determine how reliable randomized phase II data are in predicting phase III success.

The other extremely important change in phase III design is the increasingly routine incorporation of biomarker studies. The phase III trial can be the place that putative biomarkers from phase II are validated. If so, the design of the study should ideally be such that the relative impact of the new treatment in biomarker positive and negative groups can be adequately determined. More often, biomarkers are identified retrospectively using tissue collected in the course of the trial to investigate the relative effects of treatment in various molecular subgroups (examples include EGFR amplification and mutation in lung cancer, KRAS mutation in colorectal cancer). A weakness with this approach is that only a subset of the randomized population may have tissue available, which can yield a biased sample.

In summary, phase I, II, and III trial design are evolving to enable not only evaluation of new agents but to include measures demonstrating the new agent acts as expected on its molecular target and identification of molecular markers associated with efficacy. Tumour tissue collection is becoming a mandatory component of such trials. Much is still to be learned about the optimal timing of biomarker studies and the best approaches to discovery and validation. Clinical endpoints may also continue to evolve but, at the present time, standard endpoints continue to provide useful information.

References
  1. Parulekar WR, Eisenhauer EA. Phase I trial design for solid tumor studies of targeted, non-cytotoxic agents: theory and practice. J Natl Cancer Inst 2004; 96: 990-7. Erratum in: J Natl Cancer Inst 2004;96:1640.
  2. Booth CM , Calvert AH , Giaccone G , et al. Endpoints and other considerations in phase I studies of targeted anticancer therapy: recommendations from the Task Force on Methodology for the Development of Innovative Cancer Therapies (MDICT). Eur J Cancer 2008; 44: 19-24 .
  3. Yap TA, Sandhu SK, Workman P, de Bono JS: Envisioning the future of early anticancer drug development. Nat Rev Cancer 2010; 10:514-23.
  4. Rubinstein LV, Korn EL, Freidlin B, Hunsberger S, Ivy SP, Smith MA . Design issues of randomized phase II trials and a proposal for phase II screening trials. J Clin Oncol 2005;23:7199-206.
  5. Fu P, Dowlati A, Schluchter M. Comparison of power between randomized discontinuation design and upfront randomization design on progression-free survival. J Clin Oncol 2009 ;27:4135-41.
  6. El-Maraghi RH, Eisenhauer EA. Review of phase II trial designs used in studies of molecular targeted agents: outcomes and predictors of success in phase III. J Clin Oncol 2008;26:1346-54.

Back to Speakers