Leaders from nine regional centers affiliated with the Pharmacogenetics for Every Nation Initiative will gather at the University of North Carolina-Chapel Hill this week to discuss the development of a world pharmacogenomics map of common SNPs in various parts of the world.
The PGx global map project is one of the immediate goals of the PGENI, an undertaking of UNC's Institute for Pharmacogenomics and Individualized Therapy. Currently the initiative is operating in nine regional centers located in Jordan, Mexico, India, China, Brazil, Ghana, and South Africa. Ultimately, UNC hopes to have PGENI regional centers in more than 100 countries.
Each regional center has a lead investigator responsible for inking collaborations within the country to initiate genotyping studies involving local populations based on the DMET biomarker panel from Affymetrix, which includes 1,936 markers for 225 genes associated with drug metabolism. Once SNP associations for a region are established, PGENI plans to work with local health regulators to initiate educational programs and prioritize drugs in national formularies with pharmacogenetics data.
Once PGENI creates the PGx map, it will use genomically informed profiles of a given region to help health regulators determine the safest and most effective drugs for the broader populace within that region for various disease states. This will be a pharmacogenomically guided approach to prescribing drugs that will not require individual genetic testing for patients, which is too expensive in many of these countries,
"For developing countries there is not usually the lab medicine infrastructure to allow for testing," Howard McLeod, Fred Eshelman Distinguished Professor of Pharmacogenomics and Individualized Therapy at UNC's School of Pharmacy and head of PGENI, told Pharmacogenomics Reporter last week. "In those countries we have focused for now on genetic information at the public health level, for example, as part of the decision on priorities for the national formulary. The prescribing is then done in the current way, [without] genetics, but with the drugs deemed to have the best safety/effect profile for the country."
McLeod discussed the current efforts underway at PGENI and the initiative's future goals with PGx Reporter. Below is an edited transcript of the interview.
Can you give some background on PGENI?
One thing we've realized is that to be really effective in the developing world is to have a non-colonialist approach. Typically what has happened in the past is a bunch of white people fly in some place, do a few things for a while and then leave. The place is no better off than it has been in the past.
What we've tried to do is make sure we're building something that people want. We're doing it for the [people in the region] at their request. We're also making sure that this isn't all about the US. So we've set up nine regional centers: in Mexico City, Mexico; Rio de Janeiro, Brazil; Accra, Ghana; Cape Town, South Africa; Amman, Jordan; Calcutta, India; and Changsha, China. We hope to have a center in Russia and in Eastern Europe. Those regional centers are really the hub for those individual countries. The samples go to the regional centers, not to the US. The education happens in the regional centers, not in Chapel Hill. A lot of this is practical, so we don't have to sort out everyone's visas. Some of it is cultural. If I accidentally lift my foot up the wrong way, I offend half the room in the Middle East. Well, the local people in the regional centers know that and they won't lift their foot, whereas I might. There is also the issue of trust. When you're working with your neighbor, they really can't screw you. They are going to be your neighbor for a long time, and it's not just some foreigner flying in and flying out.
So, the [heads of the] regional centers are coming in next week to Chapel Hill for some additional training, for some face-to-face updating. Folks from India, China, Ghana, and Mexico will be in [this week]. And then some of the other groups will be on video conference.
One of the big initiatives we're starting is we're creating a world map of pharmacogenetic markers and the data will be gleaned using the Affymetrix DMET chips within the populations of the different regional centers. What we'll get is a much clearer view of what the genetic variation is around the world for SNPs we know have some relevance in pharmacogenetics. There is still a lot to be done in terms of discovering new SNPs and doing associations with disease and drug response. But right now we don't have a clue what the frequency of a variant of interest is in many parts of the world. We find something really interesting in the US and we start to apply it. Then we wonder why the drug doesn't work very well in another part of the world. This sort of a map will allow drug developers, academics, policy makers, and others, to at least start thinking using real data, what are some of the predictive differences, in terms of planning studies, but then also looking backwards and saying, 'We see these differences in our pharmacovigilence. There are known genetic differences. Let's go in and see if we can solve this.'
[ pagebreak ]
So, it's not pharmacogenomics in the usual sense that we might be doing it at UNC-Chapel Hill, where individual patients all get genotyped. But it's more about looking at populations and saying, 'That population is enriched for this risk allele, or is in enriched for a drug resistance genotype.' That sort of thing.
The research that you do could potentially be very useful for [multinational] drug companies. Do you plan to share with industry the data that you glean from genotyping the populations in these various regions?
The answer is, yes, maybe. Because we're working with ministries of health to help them design their national formularies, using genomic information from their populations to guide how they prioritize one medicine versus another for a given disease indication, we have purposely stayed away from a lot of industry funding. That has helped us with the World Health Organization, who otherwise would not work with us. And when we go to administrative health, we can say, 'We're not funded by Merck.'
So … it's very clear that many companies ... will benefit from this sort of data. They'll benefit in terms of understanding the US market, which comes from all over the world, but also in terms of opening up new markets. So, they may be trying to decide whether to go into South America or Asia. The data may guide them to South America, even though Asia might be a bigger market.
What we will not be doing, at least initially, is doing work that is directly funded by industry. The types of data we'll generate, some of it will go into the public domain. This will be summary data, the kind of data in a manuscript. There will not be a public database of raw information with which any [researcher] can jump in and play with. This is because of the way the ethics committees in the various countries want to protect the information.
What are the ethical, legal, and social issues associated with operating in so many different countries? Is it on a country-by-country basis or do you have an overarching set of policies for PGENI regional centers?
We have an overarching approach that we start with. But each individual country does need a separate approach. And since we're trying to do individualized therapy, we shouldn't be surprised [that] we have to do individualized ethics. In some countries we have to literally work with tribal leaders. There are other countries where the Ministry of Health makes internal decisions without a lot of public input. There are other places where community consultation is the preferred method. We have to be flexible on that account. But we have a pre-set general consent form, and we have an approach that we have taken that is uniform across the countries once we have gotten ethical approval. But [as for] the process of getting the OK within the country, I think every country has something unique about it.
Where do you find the ethical standards most diverge from those in the US?
One of the surprises in the ELSI area has been that globalization has been more thorough than I realized. What I mean by that is there might be differences in the quantity of ethical oversight, but there is really not a big difference in the quality of ethical oversight. I don't mean that every place is pristine. Certainly within the US there is a lot of variability. There is not a lot of diversity in the ethical oversight as I expected. They are all trying to achieve similar goals. I think the reason why is that the global health community has gone before us for a long time and brought with them certain standards of ethics that have influenced the local government.
Not every place has a Western ethical approach, but if they were going to get money from [the] Gates [Foundation] or from the World Bank or the WHO, they have to have a certain level of ethical oversight. So, they have already gone over those issues in broader terms. When we come in with a very specific question, it's really very straightforward. The hard part is making sure they really want us to help. We don't want to just go in and do something, leave, and then nothing ever happens. We try to make sure they are really engaged. If they are not, we go somewhere else.
Once people realize this is not a bunch of people flying in for an extended holiday, and that we really want to work with them, and by working with them, help them, not help ourselves, then it's been smooth sailing.
The people coming into these centers, who agree to be genotyped, how much say do they have in the way their data is used?
[ pagebreak ]
So far the way we've approached it has been the traditional academic model in that there are clear statements about the way the information could be used. We do put it in there that the data can be shared in a summary fashion with public and private entities. We inform them that PGENI may be paid for this data and that none of that money will pass through to the individual person ... People are either in or not in our [studies]. So, we try to be where clear with them about how the data can be used. This isn't a for-profit exercise. We're a non-profit organization. But that doesn't mean that you can't do things that bring in money. And we try to make it clear that might happen.
What are the various programs PGENI is working on?
Our main [goal] is to use [genotypic] information to help prioritize medicines. It's the public health version of individualized therapy ... We work with administrators of health who decide which drugs come into the country. Regardless of whether we're in a private health system or a public health system, upstream of that is the Ministry of Health staff that controls the door.
Now, some of our data might say that Drug B is the best drug for asthma. A health system may still choose to go with Drug A. That's their business, but they will be doing it knowing that there is data saying that's the wrong thing to do. But we're not the police. We're trying to allow a country to make decisions based on data from within the country. A country has to be able to afford the medicine, they have to make sure it's not a counterfeit, and sometimes they don't even have clean water to swallow medicines. There are problems [that go] way beyond genetics. But if you are going to commit to providing a formulary for your country, do you pick the drugs approved in Geneva by the WHO, or do you pick the drugs from that list that are more likely to be of benefit or avoid harm in your population? So, we're just trying to make the data more personal.
In these countries will you be working with people in regulatory agencies, equivalent to the Food and Drug Administration in the US, that evaluate drugs for efficacy and safety prior to market entry?
Some of the people we'll be working with have jobs very similar to the FDA. They are not doing approval in the same way because most of the world looks to the European Medicines Agency or the FDA approval as their first step ... We work with the equivalent of the pharmacy and therapeutics committee that an individual institution might have, where they take all the FDA-approved drugs and decide which should be prioritized ... If you have a situation where there is only one drug for an indication, then there is nothing to individualize. If you have a situation where there are two equivalent drugs, then you have a real need for an objective tool for picking between them.
In sub-Saharan Africa most of the population — over 70 percent — have some genetic variants that would cause them to not get the full benefit from the beta-agonist asthma drugs we would typically use. So, the first-line therapy for most of the world is not as effective. The pulmonologists in those countries know that because they have had to learn the hard way. But we can now tell that ahead of time. That data allows a country to pick the second recommended drug, rather than the first recommended drug, because they know with their local data that that drug is similar in efficacy, similar in cost, but more likely to be beneficial to their population.
What are some of these disease areas you're focusing on gathering SNPs in?
The WHO's Essential Medicines Program has approved drugs for more than 250 clinical indications ... Within that there are recommended drugs for various diseases. Within that we put prioritization information based on genetics on to that data. Pretty much every known disease is part of the algorithm. What we do bring in is the burden of disease in a given country. In a given country we look at what are the top 20 causes of mortality, what are the most expensive diseases in a country to manage. That sort of information goes into the final decision. But for every country we work with, we have an algorithm and we have output that is there for all the more than 200 disease on WHO's list.
How are you being funded?
To date, we've been funded principally from philanthropy, or individual donors, and from a small amount of money from the National Institutes of Health. We've been in discussions with the Gates Foundation, who have been very supportive of what we're trying to do. We're looking at some initiatives with them. There are other global health entities who are interested.
[ pagebreak ]
This basic approach is not foreign to global health, once we have pushed the issue a bit. Initially when we talk to global health entities, they say, 'What, are you crazy? We can't do this!' But then you remind them that it is done all the time for HIV virus, malaria, and tuberculosis, where the WHO goes and samples the bug around the world and makes recommendations about which drugs should be prioritized. We're just doing the human version of that. Then it's like, 'Oh yeah! We do that all the time.' People just haven't thought about it.
Is there something else you wanted to highlight about the project?
The bar for selecting among two equal therapies is very different than the bar for excluding someone from therapy. Too often our mindset in pharmacogenomics has been, 'Should we test or should we not, because we will either deny someone therapy or let them have therapy.' What we're talking about here is if we have two things of equal weight, just a feather will cause the scale to shift. The level of evidence is much different there. The field of pharmacogenetics has really lost sight of that.
We often try to have these heroic 50,000-patient studies out there. Really, a little bit of biological plausibility is all that is needed. We are learning a lot about how pharmacogenetics should be practiced in the US based on the developing world. As a field we can learn a lot from the developing world, because things are very practical there, and it causes us to be much more practical. We need more of that in the US.