Director of Worldwide
Regulatory Affairs Liaison for Translational
Name: Ronald Salerno
Position: Director of Worldwide Regulatory Affairs Liaison for Translational Medicine Research, Wyeth Pharmaceuticals, 2001 — present
Background: Various Research and Manufacturing positions, Merck and Co., 1971 — 2001
Education: MS and PhD in Biological Sciences, University of Maryland
At the April 16-21 Drug Information Association meeting in Bethesda, Md., Ronald Salerno gave the keynote lecture. He has long been involved in encouraging the use of pharmacogenomics technologies by pharma, diagnostics, and regulators, and has been instrumental in the process that produced the pharmacogenomic data submission guidance and the recent concept paper on drug-diagnostic co-development.
Pharmacogenomics Reporter spoke with Salerno late last week to get his take on the recent DIA meeting and drug development in the co-development era.
You helped to organize the Drug Information Association meeting [last] week. What did you think of it?
The meeting exceeded our expectations. We had a great attendance, a 441-person attendee list, if I recall. This time there was quite an international presence — [attendees from] about 10 other countries were there, 10 percent were international. And as I recall, what stood out is that there were 19 representatives from Japan. So this is definitely on the global radar.
I remarked to others at the end of the meeting, I said, 'I've noticed a significant change from when we introduced the concepts that we're talking about today; three years ago, at our 2002 workshop, everybody was apprehensive and safe-harbor leaning,' and so forth.
We followed up with another meeting in 2003 in response to that [one, in which] the voluntary genomic data submission package guidance was issued as a draft, and even at discussions at that conference, [we] were still trying to clarify [issues], and there were questions about the process, the procedures, the implications.
Well, at [last week's] meeting, clearly, we were talking about resolving these issues — putting solutions to the barriers that could be a stumbling block in using the principles of pharmacogenomics for eventual personalized medicine down the road. And not only giving examples of voluntary genomic data submission packages that were submitted — and Andrew Dorner from Wyeth was the co-chair of that session, along with the associate director from [the FDA Center for Drug Evaluation and Research], Felix Frueh, who was hired specifically to lead the Interdisciplinary Pharmacogenomic Review Group at CDER. Frueh then put together a case scenario that was modeled after one of Wyeth's experiences, and it really flushed out a lot of the uncertainties — to my surprise, there were still uncertainties out there.
The second track addressed the [drug-diagnostic co-development guidance concept paper] that the FDA put together listening to industry, from our comments back in July, and really tried to be ready and to encourage industry to think about co-development of a device with a drug; [encouraging industry] to think about it now and start getting these processes in place, so at the end of the day you can come out and really target a therapy with a useful prognostic or diagnostic device. And a driver for getting that out in time was this workshop.
The third track was very impressive, too, because it really raised the concept of retrospective pharmacogenomic data analysis, and the utility of that. Just a couple of years ago, people just shrugged their shoulders and said, 'No way.' But today, we're thinking about, 'OK, how can we use it?' You just can't do a new clinical trial to do a genetic study prospectively on a large scale. It just doesn't pay off with the intention of targeting and shortening the process of drug development — it would actually do the reverse. So, if you use retrospective data in a stepwise fashion, in certain scenarios, the prospective study, which most people felt was in need, could be limited to a specific arm, for example.
[Another] workshop took it a step further towards the end of the process, and that is, 'What is our private experience and what are the issues about co-developing a device with a drug, and what could be the implications — all the way down to the label — and implications for the companies?'
The sixth one, Iman Barilaro [Associate Director, European Regulatory Affairs, Global Regulatory Affairs and Quality Assurance] from [Johnson & Johnson] put a great session together, it was on the second day, and it brought not only Larry Lesko and the FDA to the table, but it also brought Marisa Papaluca [Deputy Head of Sector, Safety, and Efficacy of Medicines of the European Medicines Agency], and I think that they are really focused on bringing the state of knowledge in pharmacogenomics in the European Union equal to — it gets a little competitive, I guess, at these agencies. But they're on board, and they're putting all the processes in place and guidances and as a matter of fact, there were some sidebar meetings between the FDA and the EMEA officials to clarify the joint meaning of sharing information on a voluntary genomic basis. A sponsor can request both agencies in the same meeting.
So, those details will be coming forward — they shared their concepts. And for the first time, we had the Japanese regulatory agency — represented by Dr. [Yoshiaki] Uyama [principal reviewer for the Japan Pharmaceuticals and Medical Devices Agency] — present an update on their guidance. And their guidance is not as clear and as well-oiled as the guidance that the FDA has been leading to. But they clarified that it's voluntary — people were reacting to their document six months ago, 'Do we have to do this? Do we have to report every clinical trial?' No. Everything is based on a voluntary [submission model]. And why do they want this information? So they can go back to the sponsors and say, 'Hey, we would like to know more about this, could you voluntarily discuss this with us to help us write a guidance?' So those are the principles behind this.
All of this is eventually going to lead to an [International Conference on Harmonization]. The ICH concept has been accelerated, we recognize that this is something very global that is going to impact drug development worldwide. And so it's early in the process that you would pull out the ICH, but there seem to be things that we can talk about — standardization, terminology — so that eventually we can harmonize. So I'm excited about that process.
Can you lay out what you talked about?
I think [the DIA meeting] met the original objectives that I laid out in my talk. And I pointed out this whole initiative in 2001 and 2002, and with Janet Woodcock and Larry Lesko publishing a paper in the Pharmacogenomics journal that came out in the January issue of 2002. It turns out that that was the incentive to get things rolling. From then it has just mushroomed, and pharmacogenomics is part of the [FDA's drug-development streamlining] Critical Path [initiative]. And interestingly, Steven Wong he mentioned he did a survey and one of the questions — he's going to send me the details of that information, that report — was, 'What do you think was the single-largest factor that is promoting the use of pharmacogenomics?' And the answer was, 'The FDA's encouragement and involvement and initiatives.' So I made a point of that at the end of the conference — we're all scientists here and we love data — and I have to applaud the FDA for really taking this on as a public-health responsibility, and recognizing that the emergence of this technology will advance public health.
Janet Woodcock gave such an inspirational talk. She's got it. She gets it and she's really down to Earth. You see the sincerity there. Somehow or other — how does she do it all? She put her thumb right on the issue — the next issue is biomarkers — how do we clinically validate and clinically use pharmacogenomic biomarkers? To me, that is the next hurdle in this process, and after the meeting, Larry Lesko, Felix [Frueh] and I were talking about possibly meeting in October to start getting some early feedback. But it's too early to begin Pharmacogenomics IV, because these major conferences — I, II, and III — are 10 to 12 monthly meetings, [were] well thought out, and involve a lot of people — 40 to 50 people. [It's] divide and conquer, in a sense. We really focus and have sub-teams.
What was hard is that they were working on one conference, and I'm going to call my secretary this morning, and we're going to set up a teleconference the first week in May. First we will make a teleconference with the organizing committee. What I'm getting back to is that we're all committed to a goal, and that is publications. Because this information is too valuable to let hang in cyberspace — so we're going to have six different tracks and six different publications, so that the authors get due recognition, so it doesn't have to go through the bureaucracy of 43 authors. And that information will be preceded by an introductory publication that Larry [Lesko] and I will put together. So my goal is to get that up and running in May, and then get it out somewhere — we have to decide somewhere in June.
I just got an email this morning from the IR conferences. They're having a workshop from July 11th to the 14th, and it's called Pharmacogenomics and R&D. The first sentence says, 'Dear Dr. Salerno, it was a watershed year in pharmacogenomics,' and it talks about the guidelines and all that stuff. I like words like that. We did what we said we would do — we got the final concept out, we got the concept paper out on co-development. And now, underneath all this, I think companies can go back to their senior management and with signals from Marisa Papaluca's presentation, when she mentioned the electronic [common technical document] — they were thinking about where do they insert pharmacogenetics. I've been saying this from the get-go, 'If you're going to develop a drug, and you obviously do PK, you do PD, and there's going to be a section called PG.' A drug developer certainly examines the metabolic pathway of a drug, and that may involve pharmacogenetics, or it may involve a cellular pathway, where receptors are involved. It's not required now, a lot of it is voluntary, but down the road, as drug development proceeds, information will be expected, and not just from the FDA, but also from the [EMEA] and from the Japanese [regulatory agency].
In the opening session, I reviewed what has transpired in the last few years, in terms of the guidance, and the impact that the guidance has had on the industry, highlighted the publications, and talked about the 2004 workshop that we had on the co-development as a lead-in, and pulled out some significant words that Janet Woodcock said at the July 2004 workshop. And she said, 'A PG test should be co-developed with a therapeutic drug to diagnose a patient disorder, identify responders, optimal dose regimen, risk probability, monitor response to drug treatment, monitor development of adverse reactions.' So she's pretty consistent in her message, but the key outcome of that July workshop was this concept paper. And since then there have been training workshops that haven't been too advertised, because they're basically pharma-FDA genomic training workshops, where we try to train the FDA staffers about functional genomics, biostatistics, drug metabolizing enzymes — we're actually planning one for transcriptional profiling in July, and that will be at the University of Maryland Hopkins Center there in Shady Grove. So the last thing I want to mention is that you can go to the Genomics at FDA website and it talks about the organization of the genomics group, and how voluntary genomic data submissions will be submitted and reviewed, and so forth. And then I simply gave an overview of the plan for the workshop — the objectives and the tracks.
How do you see the co-development concept paper and guidance to bear fruit in speeding drug development?
I think the concept paper — we're already planning to prepare a formal response. It's a rather detailed package. It covers analytical validation, as well as clinical validation and clinical utility. There's a lot of information in there that probably is in excess of what would be in the guidance. But it's going to give scientists an opportunity to see what's expected, and it's going to raise the awareness of what is expected in taking a test out of the laboratory and actually having it ready for market, and what does the device agency [the Center for Devices and Radiological Health] think about the parameters for analytical validation and clinical validation, and whether or not it has clinical utility. So this is going to be a topic of discussion internally within companies, and companies are going to share that information. And I think just the process of dialogue is going to filter up — it's going to filter up to the projects within the company to say, 'OK, you guys are using biomarkers to make clinical decisions in early Phase I, even pre-development, going from an animal to Phase I, and you're using biomarkers to have some proof of principle or concept or some decision point that lets you proceed and invest in a Phase II trial.' People in pharmacogenomics and biomarkers now would say, 'We've got to now start thinking about talking to a diagnostic partner to say that if we have evidence of some kind of activity — biological response at the end of Phase IIa — we really need to plan a Phase IIb to validate the principles of a diagnostic to decide on how we're going to use that diagnostic in a Phase III trial.'
So you can use that Phase III trial with the same patients, entering the same samples, not only to prove the value of the drug, but also to prove the value of the diagnostic. Right now the paradigm is [that] it's a rare exception that you'll have co-licensure of a diagnostic and a drug. It's simply the way that we do business now. Everybody wants assurance that the drug will succeed and be out there, and you'll find subgroups of patients that the drug is really active for one particular group, and then, 'Oh, there's a problem here — we don't have a diagnostic!' And then there's a catch-up quality-deficient process to get a diagnostic just to not delay the drug any more than you have to. So, we're trying to avoid that scenario. The willingness of the CDRH to work with CDER and CBER to facilitate this — this could even lead discussion with the Europeans — devices really aren't controlled from the European point of view. They're going to go back and think about how they can provide guidance to developers that would similarly urge — where there's targeted therapy, and where therapy depends on the presence or absences of a characteristic — that a drug and diagnostic come out at the same time.
There has been a lot of thinking this year, and I suspect that a draft guidance will be out by sometime in the first quarter of 2006. That's my personal viewpoint, but the way things go, it could be the first half of 2006.
What do you see as pharma's major concerns with the co-development guidance?
I don't think they're negative concerns. The concerns [probably] are: the definition of validation — what is clinical use? — and what is clinical utility? That is the challenge of approving a diagnostic. So I think what we're doing is that we're retrospecting — if we think back over a couple years about pharmacogenetics and pharmacogenomics, we're into terminology and defining the requirements. And I think there are steps going forward that will answer the concern of 'What is a validated product?'
Biomarkers in particular?
Yeah. How much data and how much certainty do you need to get that thing on the market? So that's the challenge, and at least the FDA has given us a starting point for discussion.