Alumnus interview: Herbert Struemper

A career in the pharmaceutical industry might not be what you’d expect from someone whose dissertation involved theoretical work in motion control. But that’s just what has happened with ISR alumnus Herbert Struemper (1997 ECE Ph.D.), who makes his living building models that help pharmaceutical companies address specific questions related to the development of new drugs.

Struemper works for the consulting firm Pharsight, modeling clinical data and developing simulations that support decision-making for drug manufacturers.

At ISR, Struemper worked with his advisor Professor P.S. Krishnaprasad in non-linear geometric control for unmanned aerial vehicles, satellites, and automobiles. After graduation, Struemper briefly was a lecturer in geometric control at the California Institute of Technology.

He developed an interest in systems technology and biology, building virtual models on drug interaction in the human body, and worked at the consulting firm Entelos from 1999 to 2007. At this firm Struemper built models of complex diseases and their underlying biology.

Struemper moved to Pharsight in 2007, building models that could help both shorten the length of time and reduce the expense involved in drug development.

A successful drug may cost some $800 million and take 14 years to come to market. In large part, this is because most pharmaceutical research and development is based on trial and error. Thousands of compounds are examined on the early end of drug discovery. Those that appear most promising are put into a pretrial phase, then tested on animals and finally on humans in clinical trials. Of the initial thousands, perhaps only 10 compounds make it as far as clinical trials.

Another difficulty is that in clinical trials, sometimes drugs that work in animals do not work in humans. The reverse also can be the case, but drugs that do not work in animals may be dropped from consideration before they reach clinical trials.

Mechanistic disease modeling (MDM) holds promise for drug companies in shortening time, reducing expense, and improving the efficiency of the discovery process. It can specifically address questions that arise during the clinical phase of development, such as what concerns the Food and Drug Administration might raise, or whether the company should continue with the research.

MDM uses clinical data and an understanding of underlying biology to build a model that mimics the progression of a disease.

The core model is based on public, mostly preclinical in vitro/in vivo data. One of Struemper’s tasks is to sift through clinical data contained in medical studies and decide which are the most important to be included in the model. “This is a difficult part of the process,” Struemper acknowledges. “The model’s decisions should be mostly data driven, so choosing what to include is very important.”

The biological markers are costly to build and maintain; Struemper took more than a year to build a model of a joint with rheumatoid arthritis. But once the model is created, its biology components can be used in modeling additional diseases.

“Right now it takes a lot of resources and time to start a new model,” Struemper says. “That means modeling is most useful for developing drugs for major diseases that many people suffer from, such as osteoporosis, rheumatoid arthritis, diabetes/obesity, asthma, and Alzheimer’s Disease.”

Struemper notes that drug company use of modeling is still in its infancy and needs to gain the trust of the industry as researchers become more experienced with its benefits. Long-term, Struemper expects modeling to play a significant role in drug discovery.

Struemper enjoys his applied engineering work because he has an impact not just on drug development but on people’s lives down the road. He envisions a time where, “if there are good, cost-efficient modeling methods, pharmaceutical companies will find it worthwhile to develop drugs for smaller-scale diseases as well as major diseases.”

Considering about the trajectory of his career, Struemper finds it interesting that he did not take biology courses in college, but got most of his education in this area post-graduation at the application level. He believes he succeeded in a new field because his degree was “not about the actual thesis, but the background of my education.”

“Matrix algebra, statistics and estimation theory all continue to be important. I do a lot of things like model building, cellular assays, forming hypotheses about what is important in the disease and forming theories about what to do,” he says.

“Because biology is so complex, I still consider the work I do to be a systems engineering problem.”

Published January 26, 2009