Companion Diagnostics and Pharmacogenetics – key aspects of personalised medicine

By Lab21 Limited

Over the past few years it has become increasingly clear that the classical pathways for pharmaceutical development are radically changing. New technologies in molecular diagnostics provide tools which support the advancement of personalised medicine and the ability to identify patients who are more likely to respond to particular therapies or who require dose modifications. The key role that these new diagnostic tests play in both the development cycle and the clinical use of the drugs provides significant new opportunities for diagnostic companies to partner much closer with pharmaceutical companies than they have in the past.

Personalised medicine has largely been driven by two factors. Firstly, by the fact that many licensed drugs can cause unacceptable levels of adverse drug related (ADRs) events in some patients but not others. Secondly, in the antiviral therapeutic area, particularly in HIV therapy, it became clear several years ago that individuals treated with anti-retroviral drugs often become rapidly resistant to some of their therapies. Rather than prolonging therapies with these expensive medications which were having no therapeutic benefit it became necessary to carefully monitor individual patients for the appearance of specific drug resistance patterns so that alternative therapies could be administered. This was really the first large-scale evidence of personalised medicine appearing in the clinic and is still a key factor in the antiviral area and becoming more so in other areas.

The statistics on the incidence of ADRs in the clinic (Figure 1) are sufficiently concerning that regulatory bodies such as the FDA are now beginning to seriously introduce new requirements into clinical trial guidelines. In the UK, the average patient with blood pressure problems is prescribed 4 different medications before the optimal one is found and it is estimated that, at any one time, 20-40% of patients being treated with prescription drugs is actually on the wrong drug.

Figure 1.

One of the major reasons for the presence of ADRs in the clinic is that clinical trials have not, until recently, addressed the fact that every individual patient is different and that a drug that works in one patient may not necessarily be effective in another. Moreover, any given drug could be effective in one person and cause serious side effects in another. The major reason for the differences in response of individual patients is the core concept of pharmacogenetics and this, inherently, is the cornerstone of personalised medicine.

The basic principles of pharmacogenetics come from understanding how drugs are normally metabolised in the body. One of the key families of enzymes involved in drug metabolism are the Cytochrome P450 enzymes particularly subtypes 2C9, 2C19 and 2D6. Each of these enzymes has now been extensively studied at the molecular level and numerous polymorphisms have been identified for each protein, each of which can significantly affect the way that an individual responds to a particular drug.

The clearest example of this is in the way that patients respond to warfarin, a drug whose metabolism is regulated by the activity of the 2C9 enzyme together with another protein called VKORC1. Warfarin is a notoriously difficult drug to administer because the dose required to maintain active anti-clotting activity varies a great deal between individual patients. Fortunately, sufficient clinical data has now been generated which is associated with the presence of specific variants of the 2C9 and VKORC1 proteins that it is now possible to govern the correct dosage according to the enzyme forms present in each patient. Moreover, technology is now available that can rapidly generate the genotypic information from simple cheek swabs so that the analysis can easily be introduced into mainstream medicine. The impact of pre-screening for genetic variants is so great that the FDA now recommends that patient genotypes should be tested before warfarin is administered.

This personalised medicine is likely to extend widely through clinical development and patient management as more information is generated about the ways that individual drugs are metabolised and specific genetic variants that may influence this. From a patients perspective this is a huge step forward because drugs will be used which are most likely to be efficacious and to have less likelihood of toxicity in a specific patient. From a pharmaceutical company’s perspective it also makes sense because clinical trials can now be stratified and only those patients who are likely to respond will be incorporated into the trials hence boosting the levels of efficacy in the studies and eliminating the appearance of unwanted toxicity. Given the numbers of drugs that have failed in clinical trials due to toxicity effects a case can be made to retrospectively re-analyse the data in the context of updated pharmacogenetic information with a view to resurrecting drugs which may actually be valuable when used in the correct and targeted patient group.

Apart from the CYP 450 genes, several other genes closely implicated in pharmacogenetic response have already been identified e.g. UGT1A and TPMT and the list will continue to grow so that ultimately pharmacogenetic information will be a required routine component of the majority of clinical trial protocols.

Pharmacogenetic parameters are not the only new development in the pharmaceutical trial business however. As well as markers that can identify patients who might respond poorly or not at all to specific medicines there is an emerging interest in biomarkers which can clearly delineate the efficacy of the new therapeutic and can give further detail into whether a particular drug will work in a specific patient. A clear demonstration of this in practice is the identification of the biomarker Her2/neu whose presence on the surface of a breast tumour indicates the probability that the patient will respond to drugs such as Tamoxifen. Interestingly, Tamoxifen is a drug that is metabolised by the CYP450 pathway so, technically, patients being considered for Tamoxifen treatment should be tested both to see if the tumour is likely to respond and to determine whether the metabolism of the drug warrants any dose adjustments.

More recently, again in the oncology area, Amgen had developed an antibody therapy for metastatic colorectal cancer which targets the Epidermal Growth Factor Receptor (EGFR) pathway for tumour inhibition. Following the initial submission of the clinical data for this drug the licence was rejected on efficacy grounds (May 2007). However following detailed retrospective analysis of the status of one of the genes in the EGFR pathway (KRAS) Amgen were able to show that the drug doubled median progression-free survival in patients with non-mutated (wild type) KRAS compared with patients receiving best supportive care alone. Following this re-examination of the data the drug was approved but only for use in patients whose tumours do not have a genetic mutation in the KRAS gene. Therefore all patients with metastatic bowel cancer should now be tested for the presence of the wild type KRAS gene before they are prescribed the drug. This is the first time that the European Commission has licensed a bowel cancer product with the stipulation that a predictive test should be carried out and it highlights the fact that more and more drugs will be licensed with the requirement to have these associated or “companion” diagnostics available prior to the prescription of the drug.

So, clearly, these new developments in the advent of pharmacogenetics and companion diagnostics offer major opportunities to create an interface between the diagnostic and pharmaceutical industry. Firstly, a pharmaceutical company may identify a particular biomarker during early phase drug discovery that they feel is required to be measured during the clinical development. Obviously this biomarker must be robust and reproducible in order that the assay can be performed in the reference laboratory where the trial samples are analysed. Secondly, if the drug progresses successfully through the trial, it may achieve a licence where the diagnostic is required as part of the licence. In this scenario then, in order to sell the drug, a fully validated commercial test must be made available in all the target markets for the drug and with full regulatory approval in each market. So there are great opportunities for the diagnostic players who can develop the prototype assays (e.g. DxS Genotyping for KRAS, Monogram Biosciences for Trofile in HIV) and even greater opportunities for those companies such as Lab21 who can not only develop, commercialise and manufacture the tests (both protein and DNA-based) but also have the clinical reference laboratory arms that can actually run the clinical trials.

These opportunities of synergy between the pharmaceutical companies and the diagnostic houses are clearly going to be a continuing and growing factor in the development and fulfilment of personalise medicine in the near future.