A look at the statistics of clinical trials and lab assays with Yunda Huang

Vaccine and Infectious Disease Division

A look at the statistics of clinical trials and lab assays with Yunda Huang

Dr. Yunda Huang

VIDI staff scientist Dr. Yunda Huang grew up in China and has been living in the U.S. for 11 years, including five years at the Hutchinson Center. Huang feels the best measure of accomplishment for a biostatistician is how much one can help the progress of the scientific field. Her favorite quote is a saying by Albert Einstein: “Strive not to be a success, but rather to be of value.”

VIDI staff scientist Dr. Yunda Huang is both a people person and a numbers person.  So it makes perfect sense that she has found a career in studying the numbers of clinical trials, she said.  As a statistician at the Statistical Center for HIV/AIDS Research and Prevention (SCHARP), Huang’s work centers on the design and analysis of clinical trials and lab assays in the HIV Vaccine Trials Network (HVTN) and the Microbicide Trials Network (MTN). 

Much of Huang’s statistical work on clinical trial design boils down to the balancing act between enrolling enough volunteers to be able to determine the trial’s outcome with a level of certainty while not wasting resources to answer less important questions, she said.  Randomization, blinding, and ways to handle missing data are always at the center of trial design work in order to maximize the reliability and validity of the trial outcomes.

“If the vaccine is effective, you want to know that you have a well-powered study to see that,” Huang said.  “But if it’s not, you don’t want to waste resources to accidently see the opposite.”

Huang, along with other SCHARP statisticians including VIDI assistant member Dr. Holly Janes, also helped analyze the role the herpes virus HSV-2 plays in HIV-1 acquisition and progression.  By looking at participants in the Step study, an HIV vaccine trial conducted by the HVTN and Merck that was halted in 2007 as the vaccine did not protect against HIV acquisition, Huang and colleagues found that preexisting HSV-2 infection was one of the strongest risk factors for HIV acquisition among both placebo and vaccine recipients.  Their analysis also showed that HSV-2 did not alter the vaccine’s effects on HIV acquisition, viral load or progression.  Their findings suggest that design and analysis of future HIV clinical trials should take into account prior HSV-2 infection. 

In addition to leading the statistical analyses for follow up studies of Step participants, Huang plans to collaborate with VIDI member Dr. Peter Gilbert and VIDI co-director Dr. Steve Self to look more deeply into the modestly successful HIV vaccine trial recently concluded in Thailand, also known as the Thai Trial, which evaluated the combination of two vaccines previous found to be unsuccessful when used alone in a large, mainly heterosexual group of Thai volunteers.  Huang and the other scientists want to look for immune correlates of protection, or signals from the immune systems of vaccinated participants that researchers can detect in the lab, to identify both how the vaccine changes the immune system of vaccine recipients who were protected against HIV infection and to look for “markers” that could be used as early identifiers of an effective vaccine in future trials.

Besides her collaborations with trial and laboratory groups, Huang also works on her own area of research, which focuses on the statistics involved in  equivalency testing – studies that ask whether two compared interventions or laboratory methods  are equivalent in a certain measure, such as effectiveness of a vaccine or quantitative readouts of biological samples.  While on the surface this kind of study might sound similar to one asking, for example, whether one treatment is better than another treatment, the statistical testing procedures are quite different, Huang said.  Asking the question, is treatment A as good as treatment B comes with a different set of potential statistical errors than asking, is treatment A better than treatment B?

“It’s not as simple as just flipping the hypothesis,” Huang said.

Huang joined the Center five years ago when she and her husband moved to Seattle from San Diego, where she’d been working at the Salk Institute.  Huang was trained in China as a statistician, and then earned a doctorate in biostatistics from UCLA.  Her PhD research focused on statistics of cancer biology, so the transition to working on infectious diseases was a happy accident, Huang said.  “But once I came here I fell in love with the field,” she said. 

For more about Huang’s work, see her recent publication:

Simultaneous Evaluation of the Magnitude and Breadth of a Left and Right Censored Multivariate Response, with Application to HIV Vaccine Development.  Huang Y, Gilbert PB, Montefiori DC, Self SG. Stat Biopharm Res. 2009 Feb 1;1(1):81-91.