When designing trials for molecular therapeutics, there are a number of key differences compared to the traditional design described above.
For phase I studies, the paradigm described above is challenged in a number of ways. In phase I studies of cytotoxics, toxicities against rapidly dividing normal tissues serve as biological surrogates and often define the MTD (Eckhardt et al., 2003). For molecularly targeted therapeutics, these clinical effects on normal tissues cannot be reliably used as surrogates, thus making the assessment of toxicological endpoints more complex. In addition, the standard design of increasing drug dose to toxicity may be unnecessary for optimal drug effect, and the use of MTD as a surrogate of effective dose may be inappropriate in the phase I setting (Parulekar and Eisenhauer, 2004). This has led to use of the term "optimum biological dose" to define a dose, often significantly below the MTD, which can be assessed by measurement of PD markers of biological activity of drug in tumor and surrogate healthy tissues. Another key difference between trials with molecularly targeted agents versus cytotoxics relates to optimal scheduling of drug. As indicated earlier, molecularly targeted agents are often cytostatic rather than cytotoxic in nature. Hence they are preferably given as a continuous (probably oral) administration rather than as intermittent pulsed cycles. This also takes into consideration the fact that there is likely to be less need for recovery of normal proliferating tissues (especially bone marrow) with targeted therapies compared to cytotoxics.
However, although the determination of optimum biological dose provides a logical framework within which to conduct phase I clinical studies of targeted agents, the experience to date of conducting these trials has been chastening. In a survey by Parulekar and Eisenhauer (2004), 57 phase I publications concerning 31 novel targeted agents were reviewed with regard to patient population, starting dose, methods of dose escalation, determination of recommended phase II dose (RPIID), and inclusion of correlative studies. Of the 57 studies, 36 used toxicity and 7 used PK data to halt dose escalation. Nontraditional endpoints, such as molecular effects on surrogate tissues, were rarely incorporated into trial design, and in only two trials was a targeted endpoint or surrogate tissue biomarker used as to determine RPIID. This study and others demonstrate the challenges in using nontraditional endpoints in the design of phase I trial (Gelmon et al., 1999; Korn, 2004). One of the most critical issues is the lack of appropriately validated PD biomarkers. This may relate to difficulties defining the desired target effect, and to practical issues in measuring these effects once they have been defined, for example, because of a lack of reliable assays or the problems in obtaining the required tumor specimens (Korn et al., 2001). Quite often there is simply a failure to plan ahead and implement validation of biomarkers so that they are available for use in the clinical development phase.
In addition, the fact remains that the new generation of targeted therapies can still cause significant toxicity. In certain circumstances, this is only apparent outside the time window of phase I DLT definitions. An example is the hypersensitivity pneumonitis seen with gefitinib (Konishi et al., 2005). For this reason, and the difficulty of obtaining validated PD biomarkers due to lack of appropriate preclinical data, some investigators still feel that it is not prudent to base definition of MTD on a biomarker (Adjei and Hidalgo, 2005). This does not mean, however, that obtaining biomarker data is not of scientific or practical value.
So what is the optimal current way to incorporate PD endpoints into phase I studies? Critical to the process is the development of hypothesis-based trials (Adjei and Hidalgo, 2005). One approach, which we favor, is to conduct a two-stage approach of "dose estimation" based on effects on normal tissues, followed by "dose confirmation" based on effects on tumor tissues (Hidalgo, 2004). PD assays that have been validated preclinically and are readily feasible should be used in conjunction with relevant PK analyses to construct meaningful PD-PK relationships. In practice, therefore, once surrogate tissue data demonstrate inhibition of target, it is important to determine whether the surrogate tissue results correlate with findings in tumor tissue. This is probably best conducted at or near the RPIID based on traditional toxicity endpoints, often in an expanded cohort of patients (e.g., 10-12 patients are often used). Functional imaging can be incorporated if indicated. In this way, multiple endpoints can be included and an RPIID can be incorporated based on both MTD and optimum biological dose. However, tolerability should probably still be used as the primary determinant (Eckhardt et al., 2003).
In response to the shift toward earlier use of correlative laboratory studies, and the need to reduce the time and resources expended during early drug development on candidates that are unlikely to succeed, regulatory authorities have streamlined requirements for exploratory trials (Collins, 2005). Recent draft guidance from the US Food and Drug Administration
(FDA) has considered exploratory Investigational New Drug (IND) studies as an appropriate tool to distinguish promising drug candidates from those less likely to succeed (http://www.fda.gov/cder/guidance/6384dft. htm#_Toc100638018). These are "early phase I exploratory approaches that are consistent with regulatory requirements, but that will enable sponsors to move ahead more efficiently with the development of promising candidate products while maintaining needed human subject protections." For example, exploratory IND studies can help sponsors gain an understanding of the relationship between a specific mechanism of action and the treatment of a disease, or select the most promising lead product from a group of candidates designed to interact with a particular therapeutic target in humans, or explore a product's biodistribution characteristics using various imaging technologies.
For phase II studies, one of the key issues is the question of tumor response. We have already discussed the fact that novel molecular cancer therapeutics are more likely to act in a cytostatic manner as a result of mechanism-based cell-cycle arrest or the induction of generally modest increases in apoptosis. They will not produce several logs of cell kill as was seen with alkylating agents or radiation in responsive cancers. As a result, these agents may be active on prolonged administration without causing significant tumor shrinkage, creating difficulty in assessment of these agents by the traditional phase II endpoint of radiological response. The challenge is therefore to produce innovative designs and endpoints for phase II studies, as argued by Ratain and Eckhardt (2004). One possible solution for such agents is to use progression-free survival as an alternative endpoint and to carry out the trials with crossover or randomized discontinuation designs and involving a range of doses including placebo. The randomized discontinuation design is increasingly being used, and was pivotal in the development of the multitargeted kinase inhibitor sorafenib. Using this design, patients with renal cell carcinoma showed significant disease-stabilizing activity compared to placebo and led to the subsequent approval of this agent (Ratain et al., 2006). However, despite the innovative trial design, PD biomarker data were not reported, which has led to uncertainly as to which kinase or kinases are the key targets for the drug. When biomarker studies are included in randomized phase II studies, the presence of a control group which is also sampled for the biomarkers could be helpful in further defining whether effects observed in the treated patients are truly related to the drug, and also in determining whether the presence of a biological change is related to a beneficial effect in the patient (Eckhardt et al., 2003).
The use of appropriate PD endpoints is one of the main ways that negative phase III studies can be avoided. These expensive, usually multinational studies have blighted the oncology literature in recent years, with a range of targeted therapies producing a series of high-profile failures, including matrix metalloproteinase inhibitors, epidermal growth factor receptor (EGFR) inhibitors, and farnesyltransferase inhibitors. The rapid movement from phase I to phase III trials without any knowledge of whether the drug was hitting the target, and in some cases (e.g., gefitinib) uncertainty as to optimal drug dosage, has proved a chastening experience. Therefore, to minimize the risk of negative phase III trials of novel agents, the use of properly conducted phase II studies with informative biomarkers is crucial.
As mentioned earlier in this chapter, with an agent that is expected to be clinically effective without causing tumor regression, an expanded phase II trial of 100-300 patients can be very valuable. In addition, to showing clinical benefit with improved statistical power, it may be possible to use the data from the larger number of patients to identify a subgroup of patients who are especially sensitive to the agent based on the molecular characteristics of the tumors, for example patients with mutations in the EGFR gene predicting response to gefitinib (Lynch et al., 2004).
Was this article helpful?
Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.