Name: Alfredo Hidalgo-Miranda
Title: Head, Cancer Genomics Laboratory, INMEGEN
If anyone has had an insider's view on the development of the market for genomics in Latin America, it is Alfredo Hidalgo. When Hidalgo was helping to set up the core platforms at Mexico's National Institute of Genomic Medicine, or INMEGEN, in Mexico City years ago, one company even refused to sell equipment to the fledgling institute. These days, though, Hidalgo says that representatives from the same firm are visiting him to see if he is interested in making a purchase.
While the regional market is clearly of growing importance for vendors like Affymetrix and llumina, both of which established Latin American sales offices in recent years, such development is also significant for researchers like Hidalgo, who wishes to use genomic research tools to solve his country's greatest healthcare issues, including aggressive types of breast cancer tumors that are claiming the lives of many younger women.
In a presentation last month at the joint Human Genome Meeting and International Congress of Genetics in Singapore, Hidalgo detailed his laboratory's use of SNP microarrays to uncover a relationship between Amerindian ancestry and breast cancer subtype. BioArray News interviewed him afterward and below is an edited transcript of that conversation.
What health issues specific to Mexico are you trying to address through your research?
The National Institute for Genomic Medicine has several lines of research, mostly focused on … the most pressing health problems in Mexico, such as diabetes, obesity, immune diseases, and my area of research, cancer. Breast cancer is the first cause of death among the female Mexican population. We need to have a better idea of what is going on at the genome level in these types of tumors because the epidemiology is telling us that there is something different in our patients. We are facing very young patients with very aggressive tumors.
There is data that suggests that some breast cancer patients are being diagnosed a decade earlier in life than those diagnosed in the US, for example. We had some samples that are from women in their early thirties with an invasive, aggressive tumor. The sad thing is that almost 60 percent of women dying of breast cancer in Mexico are in the 30- to 50-year-old age range. This is also very important because we are losing a lot of women who are very active economically, and this is also a social problem. We haven't really found out why this is going on, but this is why we set up a program to analyze the genomic basis of breast cancer cases in Mexico.
When did you start that work?
When we set up INMEGEN in 2004, it was completely from scratch. For the first five years, I was involved in setting up the core units for the microarrays and sequencing. Also, as the first flagship project of the institute, I was involved in the Mexican Genome Diversity Project, which was an analysis of how different the genetic makeup of Mexicans was throughout the country. We selected samples from six different geographically distant regions and we analyzed SNP markers to see how different they were. And we found out that the main source of genomic diversity is the ancestral component between the Amerindian and European. You don't have to be a geneticist to see that there are differences in phenotypes in Mexico, but nobody really had measured that before at a genomic level. After that project, I wanted to return to research in cancer, and that's when we started the breast cancer project, back in 2008.
Were you initially using SNP arrays in that research?
Yes, we used SNP arrays made by both Illumina and Affy in the Mexican Genome Diversity Project. And then once we set up the labs to use the arrays, we began working with the cancer samples. I did my PhD in cervical cancer using comparative genomic hybridization on microarrays, so the next logical step was to use the SNP arrays for copy number analysis and it is very good, it's a nice tool, because from the genotypes you will infer not only copy number but the ancestry, you can use it for calculating tumor purity and ploidy, so it has been a very useful tool in our research.
In your presentation, you showed data correlating ancestry and tumor subtype.
It's part of the genetic characterization of the tumors. We didn't really know what the prevalence was of different subtypes defined by gene expression in the Mexican population, or any other Latino population. We really needed to know about the distribution of different subtypes. I didn't expect to see an enrichment of a certain subtype in our population. It came to our attention when we looked at samples from other cases, mainly the Cancer Genome Atlas of breast cases, which were mostly Caucasian, and saw an enrichment of the luminal B subtype of breast cancer in those samples. We are not really sure what is going on there. The HER2 and basal-like tumors, which are the most aggressive, were found at a similar amount to other populations. The main distinction between Mexican samples and other samples in terms of ancestry is most obviously the Amerindian component. That is why we decided to check if there was a correlation between the Amerindian component and the prevalence of luminal B. We did the analysis and, yes, the luminal B tumors that we are finding are in women whose genomes are between 50 and 70 percent Amerindian. So we are now trying to increase our sample number to see if this finding holds, but we also have other information. We are working on a genome-wide association study not only with Mexican samples but with breast cancer samples from other Latin American populations. And the most important GWAS hit is in a gene that is very important to establish the luminal B phenotype. So there might be something there related to the ancestry and the Amerindian background.
Do you think that approach of correlating Mexican population genetics with cancer research is transferrable to other countries with different populations?
I think so. Maybe it is not that evident, for example, in US Caucasians, where you will find a very homogeneous population. At the same time, reports have shown that ancestry has something to do with breast cancer subtypes — for example, in the African-American population, the number of triple-negative, basal-like tumors [that] are very aggressive in young women, and that is something they are looking at. So, obviously, there is something in the ancestry of these persons that will define what kinds of tumors they will get. Even though it has not been, in my point of view, properly assessed in the genomic analysis of breast cancer tumors in general, it is an approach that I think will get more attention.
You have also introduced exome sequencing as a tool in these studies.
Yes, we are trying to use genetic information from different levels, trying to find out what is going on with the DNA at a point-mutation level, and the translocations, and the copy number, and correlate that with subtypes defined by gene expression, and also trying to link clinical information with the ancestry and genetic make-ups of the tumors — so, combining different kinds of information to get a better picture.
How do you translate your findings back to making a difference in the treatment decisions in Mexico?
We are trying to push, in particular, tumor subtyping. Right now in Mexico, very few people can afford to send a sample for testing by OncotypeDx, for example, [in the US]. It is quite expensive and we are trying to push these kinds of methods using commercial arrays and established algorithms to do the subtyping. It is a tool for clinicians to work with. But it is a way to go. Unfortunately, clinicians are not really in touch with genomics and they don't think it is something that they will need for their patients but we are trying to push that also.
It seems that vendors are paying more attention to Latin America. Is there more research activity in the region?
Yes, the Latin American market is growing very fast. The number of potential patients that we get is huge. There is not only the biomed market, it is also the agricultural market. So I think they are finally taking a look at the opportunities in those markets for genomics. Almost all of the major companies have representatives in Mexico or their offices there. It's interesting, when we started the institute, we were among the first in Latin America using, for example, Illumina arrays. In that sense, we have been trying to push it also, to convince the vendors that there is a market there, and that there is interest in the scientific community to work with such technologies.
The other side is that, unfortunately, in the research arena, we are still not used to genomics. The genetics community in Mexico is more interested in Mendelian genomics, and that's what they work in. The idea of a microarray is interesting for them, but, for example, we don't have a lot of bioinformaticians to do the analysis. It has taken a while, but we are trying to change that, to find the better way to use genomics and hopefully get somewhere.
What are the challenges to doing this kind of research in Mexico?
There is resistance from both sides. When we founded the institute, there were a lot of discussions about why the country should invest in something as expensive as genomics when we still have children dying from diarrhea. You have to balance things. We still have problems from the third world, but we are facing health problems from the first world. From the political side, it is difficult because genomics is very expensive. They said we should face the most pressing issues and then we can think about genomics. But we argued that if we do not get into genomics now, we will miss the genomics train. We need to develop these kinds of technologies in Mexico in order to tackle those problems, not only cancer, not only the big stuff, but also problems related to our development.
On the other hand, I remember that we approached a certain company years ago, and when we asked if they would sell us the equipment, they said no, Mexico is not a market, so we are not interested. Their representatives even flew down and visited us and said, "No, I don't think the market is prepared for this." We ended up buying other stuff, that works very good, and they recently visited my lab to ask if we are interested in acquiring their equipment.
Government funding for research is an issue in a lot of countries right now. How are things in Mexico?
I think it's a worldwide problem. Funding from the state is not as good as it should be, particularly in countries like Mexico where investment in science and research is very limited. And there is a lot of competition. I think our institutions are fine. We get money from the government. But, for example, the funding to support our exome sequencing of tumor samples came through a program generated by Carlos Slim Helú, who is … the richest man in the world. He's decided to give some funds to scientific research, and that's why we were able to sequence these tumors. This project has analyzed breast cancer, head and neck tumors, lymphoma, cervical cancer, and soft-tissue sarcomas. The money he provided really helped us get into sequencing at a larger scale, in collaboration with the Broad Institute and the Carlos Slim Institute of Health.
What is next for your research?
We are going to move forward with our breast cancer project. I am really interested in demonstrating that there is a relationship between ancestry and a particular subtype. There is evidence that this is the case, but we really need to prove it. We are also increasing the numbers of breast cancer tumor samples that we are sequencing to get an idea if there is a set of mutations particular to Mexicans compared with Caucasians. I am trying to go from research to clinical applications. We are interested in setting up sequencing protocols to provide better diagnosis and selection of treatment. These are the most interesting projects I am working with.
And, just on the tech side, are you content with the microarray and sequencing platforms you are working with?
Not really. Our experience in Mexico is that it was really hard to set up the equipment because of the cost. Because of the tax system, we often wind up paying double what US customers pay. Getting the equipment is a challenge. And sometimes the vendors are not able to provide the service that you need. Sometimes the field application scientist has been trained once in their headquarters, so when you face a real problem that needs an experienced user to solve it, they are not up to it. This mainly concerns sequencing though. I think that Affy has a very good support team, and with Illumina, we are dealing directly with the [US], so it's not a problem.