Intersection Between Stratified Medicine and Targeted Therapies

Genetic variability may affect the PK and PD of protein therapeutics and may thus have the potential to affect patient responses. To achieve broader utilization of Pgx tools, it is critical that the scientific community understand what such information will really bring to clinical practice. Individualized medicine has received great publicity in recent years, with articles forecasting the use of genomic tests to tailor new drugs and provide patients with targeted treatment. Although this type of statement is appealing, it must be tempered by the fact that Pgx data are expected to contribute to clinical decision-making but should not be expected to supplant other forms of clinical information.

Consider a Pgx test with regard to its predictive capacity at the patient level. In general, evidence-based medicine requires a population analysis of predictive value for any diagnostic test. Individual treatment decisions may be based on the test results and an understanding of the given test, plus a constellation of additional information that is unique to the patient. While a nominal hazard ratio may be associated with a positive test result, this prediction can never be applied to a single individual, since the information is probabilistic. Weiss et al. provides a review of Pgx tests in the context of clinical decision-making (135).

Looking forward, the use of Pgx tests seems more likely in cases where several treatment options exist. Given recent advances in drug development, molecular therapies are expected to dramatically increase the number of potential treatments for many disease indications. However, in some indications, restrictions in insurance coverage may limit the number of protein drugs a patient may try. On the basis of the price of protein therapeutics and the obvious benefit to patients when optimal therapy is selected early in treatment, new therapeutics may be subject to increased scrutiny by regulatory agencies.

Academic researchers have enthusiastically embraced Pgx as a method to aid in treatment decisions, with positive examples of genetic associations in both target- and exposure-mediated response factors. Regulatory agencies have added the results of such scientific research to drug labels (e.g., CYP2C9 and VKORC alleles in warfarin dose finding) and in some cases requested inclusion of Pgx information prior to licensure (e.g., KRAS mutation status in pan-itumumab response).

The question of how best to select patients for clinical development and rapid market entry is still unsolved, as stratification by candidate Pgx markers may require larger treatment populations in early clinical trials. In some cases, a putative predictive marker may be associated with a patient's prognosis and not related to the specific therapeutic (136). The speed of development may be compromised to more fully understand the role of the Pgx marker in the proposed treatment population. On the other hand, for some drugs developed in specific patient subsets (with label restrictions), post-marketing Pgx studies have shown that some patients testing negative for the marker may also benefit from treatment. This has been noted in the cases of trastuzumab and cetuximab (97,137). Refinements to the diagnostic testing paradigm may alter the indication as new information emerges from expanded clinical use, as in the case of tras-tuzumab (13,14). The rapid pace of drug development may preclude validation of Pgx-based diagnostics until after product launch if clear associations are not established in early trials. In this case, postlaunch strategies might use Pgx markers to refine the indication.

0 0

Post a comment