Companion Diagnostics for Personalized Medicine

The rapid advance of biomarker sciences has led to a new drug development continuum. Now, pharmaceutical companies rely on biomarkers, particularly genomic biomarkers, to elucidate disease pathways, stratify patient populations, and monitor efficacy. Supported by evidence-based laboratory analysis and clinical evaluation, biomarker science is making personalized medicine a reality, not only in drug development, but in patient care for oncology, clinical immunology, metabolic diseases, and other areas (1). Increasingly, companion diagnostics (CDx) are essential to a drug's regulatory approval and clinical use.

Highlights of Current Approved CDx
• HerceptTest and HER2 FISH for Herceptin
HER2 is used as a clinical diagnostic and predictive biomarker for metastatic breast cancer. The overexpression of the HER2 gene can be detected by immunohistochemistry (IHC) for the protein or by fluorescence in situ hybridization (FISH) of the HER2 gene. Patients that score within the criteria of the HerceptTest are eligible for Herceptin treatment, or subject to an additional highly concordant FISH test before treatment.

The HerceptTest represents an early success of a CDx for personalized oncology treatment. Co-developed by Genentech and Dako, the test received Food and Drug Administration (FDA) approval as an IHC test. The follow-on chromogenic in situ hybridization (CISH) test by Genentech and Invitrogen for the HER2 gene was approved in 2008, and followed by another version of CISH approved in 2011 for Genentech and Dako. Currently, clinicians use the HerceptTest as the first-line screen for patients with the over-expressing HER2 proteins (3, 4).

• KRAS Mutation Test for Cetuximab
Another example of a successful oncology CDx is the Therascreen KRAS RGQ PCR kit, developed by ImClone and Qiagen. FDA cleared this test as a predictive CDx for Cetuximab for the treatment of EGFR expressing colorectal cancer patients bearing the wild type KRAS gene (4, 5). Cetuximab was approved in 2004 for treatment of metastatic colorectal cancer, and in 2006 for head and neck cancer, before KRAS mutation status was demonstrated as an important predictor for the response to Cetuximab treatment in the colorectal patients (6, 7). Compared to patients with the KRAS mutation, patients of EGFR-expressing metastatic colorectal cancer with wild type KRAS respond well to Cetuximab, in terms both of overall survival and of progression-free survival. FDA cleared the CDx test in 2012 and changed the Cetuximab label to include the Therascreen KRAS RGQ PCR test as a requirement for treatment of metastatic colorectal cancer patients.

• A CDx for Metastatic Melanoma Treatment
In 2011 FDA approved an assay for the BRAF V600E mutation as a predictive CDx for Vemurafenib in the treatment of metastatic melanoma, an example of a successful partnership between drug and diagnostics development. The co-approval of the drug and the CDx in this case underscores the important paradigm shift for drug developers, diagnostics partners, and regulators.

Researchers discovered that the BRAF V600E mutation was found in about half of late-stage melanoma patients who were responsive to first-line treatment by Vemurafenib in a study comparing the drug to Dacarbazin (8). In parallel to the approval of Vemurafenib, FDA approved Roche's cobas 4800 BRAF V600 Mutation Test for the detection of the V600E mutation for patient selection (9). The BRAF V600E CDx was co-developed and evaluated in the clinical trial of Vemurafenib, making it one of the few successes in co-development of a drug with its CDx. This drug-diagnostics co-development strategy also sped fast-track approval for the drug after the CDx allowed a shorter clinical trial and produced more robust efficacy and safety data.

Companion Diagnostics

• The Vysis ALK Break Apart FISH Test for Crizotinib Another recent success of drug and CDx co-development is Pfizer's Crizotinib, a drug for non-small cell lung cancer (NSCLC). Abbott developed the Vysis ALK FISHprobe for ALK gene rearrangement for treating a small subset of NSCLC patients who have positive ALK rearrangement but with negative KRAS and EGFR.

Pfizer first developed Crizotinib as an inhibitor for c-Met and ALK in other clinical indications (10, 11). Then, researchers discovered the ALK gene arrangement in a small subset of NSCLC patients (12). Early study of a selected few patients who responded positively to Crizotinib led to selection of the subset of NSCLC patients with positive ALK rearrangement and double negative EGFR and KRAS mutation in the phase II trials (13).

To support the trial for the selective subset of the NSCLC patients, Abbott modified the early ALK probe for lung tissue to meet the analytical criteria of in vitro diagnostics (IVD). The superior response in both progression-free survival and overall response rate from the phase II studies led to accelerated approval from FDA of Crizotinib and Abbott's CDx.

An Evolving Regulatory Landscape
Both Crizotinib and Vemurafenib represent the successful co-development, clinical evaluation, and regulatory approval of a new class of personalized cancer therapies under the new CDx paradigm. Molecular diagnostic tests have also been developed and used in CLIA-certified labs as laboratory-developed tests (LDT) without the FDA approval process. Notably, this includes the IHC test for the KRAS mutation in support of Cetuximab and Panitumumab for colorectal cancer treatment (14).

FDA released guidance in 2014 (15) explaining that "an IVD companion diagnostic device is an in vitro diagnostic device that provides information that is essential for the safe and effective use of a corresponding therapeutic product." FDA views the analytical results of a CDx as critical to patient safety and effective use of the drug, and believes that it should regulate CDx through the premarket approval (PMA) or 510(k) review and approval process. This sets forth a significant shift from the LDT to the IVD model historically. FDA does encourage drug-diagnostics co-development, but stresses the importance of co-approving the CDx with its therapeutic drug if the safety and effectiveness of the drug depends on the use of the CDx.

The guidance also sets forth a provision of two exceptions with regard to the co-approval of a CDx and its companion therapeutic drug. The first is when the new therapeutic is intended for treatment of serious or life-threatening conditions, and the benefit from the therapeutic drug outweighs the risk of delaying approval of a CDx. The second is for revising the labeling of an existing drug to include the use of an unapproved CDx to deal with a serious safety issue.

Prior to FDA issuing this current guidance, there have been a number of publications reflecting FDA's perspective in more detail (16, 17). A CDx and its associated drug are under the oversight of three FDA centers: the drug is reviewed by either the Center for Drug Evaluation and Research (CDER) or the Center for Biologics Evaluation and Research (CBER), while the CDx is under the jurisdiction of the Center for Devices and Radiological Health (CDRH). Co-approval of a drug with its CDx presents challenges to the existing regulatory framework and these centers historically had their own policies.

While the fundamental guiding scientific principles of the regulatory framework for CDx are similar between the U.S. and European Union, significant differences remain. One of the key differences is that European Medicines Agency (EMA) requires co-development and approval of a CDx at the same time as the drug. However, a harmonization effort is underway to align the key differences between FDA and EMA guidance on development of CDx. One example is the proposed classification of CDx as high individual risk or moderate public health risk (category C), which would require EMA approval (18, 19), while FDA may determine if a CDx is subject to PMA or 510(k) review and approval on a risk-based analysis.

Considerations for Developing CDx
Because FDA considers CDx to be high-risk IVD, these tests must demonstrate analytical and clinical evidence for their use specifically with the intended drug. The development of CDx may start early in a pre-clinical study in which the biomarker may develop from an understanding of the drug's mechanism of action or the disease biology.

The preferred path for developing CDx is a prospective biomarker strategy in conjunction with the therapeutic product under development. This way, a biomarker may demonstrate its utility in patient stratification and undergo a proof-of-concept study, early assay development, and limited validation in the preclinical phase. A critical decision point in any biomarker development program is considering how a biomarker's utility will translate from the pre-clinical to clinical phase.

Factors such as pre-analytical variables must be carefully considered as part of the development plan. Establishing the criteria of pre-analytical variables avoids the serious pitfall of complicating analytical and clinical validation.

Despite various formats of CDx assays, the core analytical performance must include precision/reproducibility, sensitivity, analytical specificity, selectivity, stability, and instrumentation performance (16, 20). For a quantitative CDx test, accuracy must be established. As an IVD, the assay must demonstrate acceptable day-to-day, site-to-site, and operation-to-operation variation.

Clinical validation is the most decisive step in developing a CDx. As a result, it is imperative that adequate time be built into the planning and analytical validation that must be completed prior to the start of the pivotal clinical phase. This timeline may vary greatly from a short few months for an already-developed assay to many months for a de novo assay. Critical factors such as inclusion/exclusion criteria of test samples, sample collection and storage, and method transfer and validation in a CLIA-certified lab all must be considered and planned accordingly prior to commencing the clinical validation phase.

Unlike assay development and validation during the discovery and preclinical phase, in which a greater degree of fit-for-purpose flexibility may exist, a CDx is expected to be fully validated for all sample types and tests, along with a complete operation manual for the clinical site and the central clinical lab. Validation of an IVD generally follows the Clinical and Laboratory Standard Institute analytical guidance, and may differ significantly from early biomarker development.

The Way Forward
With the completion of the human genome project and the advance in genomic technologies, there are high expectations for personalized medicine, particularly in the oncology area. For these discoveries to bear fruit and for drugs to become more targeted, CDx testing is essential. So far, there have been a few CDx successes, including those discussed in this article, and FDA has responded to this evolving discipline with the release of a new guidance and white papers to lay the foundation for drug-companion diagnostics co-approval. Meanwhile, drug and diagnostics companies have embraced the opportunity and continue to navigate through the challenges in this evolving landscape.

Significant challenges remain ahead that involve many key stakeholders. These areas include understanding how biomarkers relate to disease biology, developing a translatable preclinical model to bridge preclinical and clinical studies, designing appropriate analytical criteria for early discovery, preclinical and clinical study, choosing technology platforms and critical reagents, the LDT versus IVD regulatory pathways, and more (2, 21–23).

Challenges also exist in establishing partnerships early on between drug and diagnostics development teams, either internally or externally. The majority of pharmaceutical companies lack extensive expertise in diagnostics device development, and they tend to design their drug development programs without sufficient support from clinical diagnostics experts.

Conversely, diagnostics companies still heavily focus on clinical diagnostics over theranostics, and view CDx as high-risk investments because the test is linked to the successful regulatory approval of the drug. Moreover, the market for a CDx might be significantly smaller than general clinical diagnostics. To this end, pharmaceutical companies are often the driving force in initiating a CDx program. Thus, early and strong partnership among the diagnostics and therapeutics sides are essential to the success of this joint endeavor.

References

Modur V, Hailman E, Barrett JC. Evidence-based laboratory medicine in oncology drug
Barrett JC, Frigault MM, Hollingsworth S, et al. Are CDx Useful? Clin Chem
Jorgensen J, Moller S, Rasmussen BB, et al. High concordance between two companion
Landais P, Meresse V, Ghislain JC, et al. Evaluation and validation of diagnostics tests forguiding therapeutic decisions. Therapie 2009;64:195–201.
Van Custsen E, Lan I, D'haens V, et al. KRAS status and efficacy in the first-line treatment of patients with metastatic colorectal cancer (mCRC) treated with FOLFIRI with or without Cetuximab: The CRYSTAL experience. J Clin Oncol, 2008 ASCO Annual Meeting Proceedings 2008;26(15s):2.
Lievre A, Bachet JB, Le Corre D, et al. KRAS mutation status is predictive of response to
Karapetis CS, Khambata-Ford S, Jonker DJ, et al. K-ras mutations and benefit from
McArthur G, Hauschild A, Robert C, et al. Vemurafenib improves overall survival compared to dacarbazine in advanced BRAF V600E-mutated melanoma: Updated survival results from a Phase III randomized, open-label, multicenter trial. European Multidisciplinary Cancer Congress 2011.
Anderson S, Bloom K, Chilling R, et al. Molecular testing of BRAF V600 mutations in clinical trials of the BRAF inhibitor Vemurafenib (RG7204/PLX4032) in metastatic melanoma – a comparison with Sanger sequencing. Eur J Cancer 2011;47:abstract 1403.
Kwak EL, Camidge DR, Clark J, et al. Clinical activity observed in a phase I dose escalation trial of an oral c-Met and Alk inhibitor, PF-02341066. J Clin Oncol 009;27:abstract 3509.
Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small cell lung cancer. N Eng J Med 2010;363:1693–703.
Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4-ALK fusion
Crino L, Kim DW, Riely GL, et al. Initial phase II results with Crizotinib in advanced ALK-positive non-small cell lung cancer. J Clin Oncol 2011;29:abstract 7514.
van Krieken JH, Jung A, Kirchner T, et al. KRAS mutation testing for predicting response to anti-EGFR therapy for colorectal carcinoma: Proposal for an European quality assurance program. Virchows Arch 2008;453:417–31.
Food and Drug Administration. In vitro companion diagnostic devices; Guidance for industry and Food and Drug Administration staff. http://www.fda.gov/downloads/MedicalDevices/Device RegulationandGuidance/GuidanceDocuments/UCM262327.pdf (Accessed August 2014).
Philip R, Carrington L, Chan M. US FDA perspective on challenges in co-developing in vitro companion diagnostics and targeted cancer therapeutics. Bioanalysis 2011;3:383–9.
Pignatti F, Ehmann F, Hemmings R, et al. Cancer drug development and the evolving
Senderowicz AM, Pfaff O. Similarities and differences in the oncology drug approval
Marton MJ, Weiner R. Practical guidance for implementing predictive biomarkers into
Schmidt C. Challenge ahead for companion diagnostics. JNCI News 2012;104:14–5.
Halim AB. The biggest challenges currently facing companion diagnostic advancement.
Parkinson DR, Johnson BE, Sledge GW. Making personalized cancer medicine a

Arron S.L. Xu, PhD, is the associate scientific director of Intertek Pharmaceutical Services. He is an analytical biochemist primarily focused on biomarker and biologics analytical science for drug development.

+Email: [email protected]