For my occasional series of blogs examining various facets of the companion diagnostics market, I asked some IVD regulatory experts to share their thoughts and opinions on the current regulations for companion diagnostics. The following is presented in a panel discussion format. The panel participants are Glen P. Freiberg, president at RCQ Consulting; Fred D. Lasky, PhD, principal consultant at Lasky Consulting; Leif Olsen, director of regulatory sciences at Hogan Lovells; and Bradley M. Thompson, JD, an attorney at Epstein Becker & Green.
How would you characterize the current regulatory environment for companion diagnostics?
Freiberg: The policy is clear for the testing and submission process. When new, targeted therapies require a diagnostic to qualify the patient for the therapy, it makes sense that the companion diagnostic meets the typical design control and submission requirements to ensure safety and effectiveness.
Lasky: It’s mixed. FDA has been very willing to discuss strategies with companies that know that a companion diagnostic is needed. Based on several conversations and my experience, pharmaceutical companies have little idea on how to approach IVD validation, and the Office of In Vitro Diagnostics and Radiological Health (OIR) is not prepared as of yet to provide more help. Sometimes the sponsor might not be pleased with the outcome of the meeting, but it is better to understand the challenges that must be met early in the development cycle, rather than later. That being said, it is always a challenge to know when to approach FDA. The more feasibility/preliminary data available for discussion, the better the agency can provide input on the subsequent plans for analytical and clinical validation. In my view, less mature companies, meaning with less IVD experience, are anxious to talk too early.
Olson: Based on my understanding from the industry and FDA’s recent comments during the annual meeting of the Association of Medical Diagnostics Manufacturers (AMDM), the current regulatory environment for companion diagnostics is improving, but it is in flux. There seems to be a cooperative spirit between FDA and the IVD industry in having open communication in resolving issues early in the pre-submission process and as they come up during the submission/review process.
Thompson: It’s opaque and ad hoc. This is an important time for companion diagnostics. Developers are desperately looking for guidance on the regulatory framework for companion diagnostic approval and an opportunity to participate in fine-tuning that framework so it reflects the reality of test development. Manufacturers often struggle with a lack of clarity and consistency in FDA standards, and recent success stories are more about succeeding in spite of the current environment rather than success of a regulatory framework. For a few years, FDA has been working on a draft guidance document to provide clarity to those co-developing novel companion diagnostics with therapeutics, and when published, it would give stakeholders an opportunity to comment and help refine the process. However, the document has gotten completely hung up on the issue of lab-developed tests (LDT). My understanding is that the Office of Management and Budget is holding up the release of the draft guidance document because of its concerns over the backlash from labs and Congress. Many in the industry have been trying desperately to shake the guidance document free, arguing that it really does not need to address the issue of LDTs. They’ve even proposed simple fixes to un-tether it from the LDT issue, such as a disclaimer that the guidance does not affect LDT policy. It really should be that simple. But alas, so far, no luck.
What are the regulatory challenges that IVD manufacturers encounter when trying to get approval for their companion diagnostics?
Lasky: An ongoing challenge is one of the chicken versus the egg. Is the diagnosis needed to preselect the appropriate patient population, or is a presumptive diagnosis better to increase the demographic parameters so the assay can better assign patients from a more general population? My experience with companies that have a new, innovative technology that is capable of increased analytical sensitivity, for example, believe the technology can solve a host of diagnostic issues. What often occurs is that there is a lack of clarity in the intended use of the device and its enhanced technology. This seems to be a more common occurrence in the field of oncology. As a first principle of any regulatory submission, the intended use for stated indications is the place to start.
Olson: The regulatory challenges that IVD manufacturers seem to be encountering are lack of communication from the drug companies they’re working with. There seems to be a certain amount of frustration on the part of the IVD companies as to receiving information regarding feedback from their studies and how certain results could impact the IVD. Other regulatory challenges involve potential changes to the IVD required during the study, which could delay FDA’s review and require additional studies.
Thompson: In addition to a lack of clarity, one of the biggest challenges is that in the real world, companion diagnostics and drugs are often not developed strictly in parallel. For example, a drug company might develop a drug, and then decide that a companion diagnostic would be useful. So the diagnostic is developed at a later stage. The problems become very logistical because FDA wants to try to shoehorn all development into a strictly parallel process that doesn’t always work.
Another challenge is that manufacturers that are developing a companion diagnostic and a therapeutic are going to be dealing with at least two centers at FDA. For the diagnostic, the Center for Devices and Radiological Health (CDRH), and for the therapeutic, the Center for Drug Evaluation and Research (CDER), or the Center for Biologics Evaluation and Research (CBER), or both. The centers do not see eye to eye on some issues, such as the standard for clinical utility of a companion diagnostic, which can lead to inconsistent requests and arguments that delay bringing innovative products to market. This is all the more reason for a guidance that will develop not only a center policy but also an FDA policy that all the centers will abide by.
Freiberg: In some ways, the process is simpler in that the therapy protocol will determine the diagnostic sample types and number of samples. If the demographics satisfy CBER or CDER, then it appears that OIR will accept that sample base for the companion diagnostic.
How do IVD manufacturers overcome these regulatory challenges in companion diagnostics?
Olson: IVD companies appear to be addressing these challenges by working with the drug companies and FDA to improve communication of information as early in the process as possible and throughout the submission process for resolution of the issues rather than delaying the communication to later in the process. In addition, through the use of FDA’s guidance documents and the pre-submission process, the IVD industry is overcoming some of the past issues it experienced. These improvements are resulting in improved organization of the submissions, defensible intended uses, improved explanation of the technology, adequate description of the studies, and more complete validation studies.
Thompson: Right now, these issues get resolved through a lot of negotiation and patience on the part of IVD companies. Unfortunately, the development of these products goes really slowly, in fits and spurts. The additional costs in terms of manpower or dollars to address these issues often come directly out of the companies’ research and development budgets, which means fewer resources for innovation, and fewer new products in the pipeline. That doesn’t advance the public health.
Freiberg: The major challenges for IVD manufacturers are timing and coordination. Such projects for them are routine. For CLIA labs attempting to enter the process with an FDA regulated companion diagnostic, the challenge is greater because they are not accustomed to the process.
Lasky: Experience! Having a strong team that understands the clinical need is essential. Understanding the parameters that FDA is looking for and how to design appropriate studies is key. Included on the team should be a biostatistician (statistical testing modes is a frequent point of discussion with FDA) who is versed in pulling apart issues such as pre-selection bias that involve discussions with clinicians. Development and operations people need to understand the principles of IVD verification and validation, and be versed in design controls and risk assessment.
Is the FDA aware of such challenges to IVD manufacturers in complying with companion diagnostics policies? If so, is the agency making efforts to amend them?
Thompson: As I’ve already indicated, FDA is aware that manufacturers need more guidance and is in fact trying to develop the guidance. But they’ve hit a brick wall, and they can't seem to break through the bureaucratic obstacles preventing them from releasing the guidance. So in the absence of guidance, pursuit of approval requires a lot of communication back and forth between the agency and the development team. Having a uniform guidance would be a more efficient approach, but in the absence of that guidance, early and frequent dialogues are necessary.
Lasky: I think FDA is aware, but it seems to choose to handle these challenges on individual bases. In strong measure, I can understand it because each program/project has its own needs, challenges, and nuances. More forthright suggestions from OIR as to how to validate an assay, what might be presented at pre-submission meetings to maximize outputs, including how far in advance of expected approval the meetings should be held, are some examples.
Olson: Based on the information I’m hearing from FDA, it’s my understanding that FDA is aware of such challenges and is encouraging industry representatives to come in to talk early in the process, such as during the pre-submission meeting. The IVD industry has recognized that the agency has provided faster and timelier responses, a willingness to explore new statistical strategies, and an open dialogue on assay performance and interactive labeling.