World AIDS Day: An interdisciplinary approach to HIV/AIDS research

By Ciara Parsons, BA'15

Art Poon, PhD, assistant professor, Pathology & Laboratory Medicine, is a bioinformaticist using an interdisciplinary research approach to the problem of tracking HIV-1 transmission rates and detecting outbreaks.

“When people talk about interdisciplinary research, this is it. In some ways, you need three or four degrees to be able to contribute to this area of work,” said Poon, laughing.  

Combining mathematics, epidemiology, evolutionary biology and computer science, Poon developed an automated system of computer programs to analyze ‘genetic clusters’, which are based on comparing the virus genomes from different anonymized infections in a hospital database. The system is also able to generate outbreak management reports and data visualizations in ‘near real-time’ and within days of a new diagnosis.

Poon’s system was the first of its kind in the world, and he has consulted with clinics and hospitals at multiple sites in the United States and other countries around the globe to help them build similar programs.

Such ‘near real-time’ tracking is feasible because, like many other viruses, HIV-1 has an extremely high rate of evolution, rapidly picking up genetic changes within weeks of a new infection. Poon notes this is significant because these mutations can be used to detect changes in HIV transmission rates in different populations.

“When you see a group of HIV infections that are almost genetically identical, this indicates that something unusual is occurring,” he said. “If there has been a rapid sequence of transmissions between individuals, this means there has not been enough time for their respective viruses to accumulate genetic differences. These genetic clusters can then be used to detect an outbreak of transmission.”

The automated system that Poon developed was used to monitor an HIV population treatment database in British Columbia. Through genetic sequencing, which has the ability to pre-screen for elements like drug-resistance, a catalogue of HIV strains and mutations can be archived. With new sequences being deposited into the database each day, the system is consistently reanalyzing anonymized data, tracing trends and creating constellations of genetic clusters. The system recently detected an outbreak of transmitted HIV drug resistance, which led to a subsequent targeted public health response.

“In British Columbia, the system was able to impact and change the course of this particular outbreak. Cases did eventually begin to accumulate again, but they had much less transmitted drug resistance this time,” he said. This system was recently published as an implementation case study in The Lancet HIV journal.

Poon says genetic clustering is an intuitive and rapid means of analysis, but that there is more to be learned about the efficacy of this method. For instance, he discovered that, contrary to previous assumptions, a cluster of genetically similar HIV infections does not always represent an outbreak.

If new cases of infection in the cluster were diagnosed sooner than other populations, then the difference in sampling rates can also create genetic clusters in the database. Differences in access to health care services and engagement can cause some populations to appear to have higher rates of HIV transmission, when in reality the transmission rates are the same as everywhere else.

“This can represent a big problem when detecting outbreaks, because you may be misdirecting your public health responses. In these situations, you want to direct your efforts to where the transmission rates are the fastest and where the next cases are going to present,” said Poon.

To refine the genetic clustering method, he is developing new approaches to identify variation in HIV transmission rates. Poon believes that if genetic clustering is ‘done right’, it can improve public health responses, streamline costs and mobilize resources to high-risk areas in a timely manner. A new model-based clustering method from his lab was published last month in the journal PLOS (Public Library of Science) Computational Biology.

Ethical considerations also play a role in why he is scrutinizing genetic clustering and similar methods.

Genetic clustering has been used to prosecute individuals in court, and since HIV transmission with non-disclosure remains severely criminalized in Canada, he says great caution must be exercised with the utilization of such methods for public health. Poon recently delivered a talk on the ethical and legal implications and misinterpretation of genetic clusters at a symposium convened by the Canadian HIV/AIDS Legal Network in Toronto.

“In some research communities, the question of ethics doesn’t come up as often as it should,” he said. “I think we, as scientists in this field, need to provide more ‘open access’ resources for the communities who would benefit from our work so that we can communicate with them to talk about risks and benefits.

It’s important to use clear language so that the science is accessible.”

Poon says awareness is growing within HIV-positive communities about current applications of genetic clustering and that researchers like himself need to reach out to community members and advocates to discuss the scope and limitations of these methods.

Having the ability to make an immediate impact on an individual’s health is one of the most rewarding things about his work and Poon aspires to continue contributing to ongoing global efforts to end the HIV/AIDS pandemic through public health and cure research.

“In a world bursting with data, we are only limited by our imaginations and the questions we can ask,” said Poon.