Success from Failure

On a sunny May afternoon, in a small conference room at AMAR's offices at George Mason University Enterprise Center, a small business incubator facility in Fairfax, VA, four people gather to discuss the myriad mistakes that companies can make as they shepherd a drug through the development process: Prakash, company founder and CEO Savita Prakash, an expert in marketing statistics (and his wife), along with Farouk Karoum, a director at AMAR, and Christopher Royce, who has managed clinical trials for antibioterrorist agents.

Many times, failure could have been predicted, Prakash says, and AMAR aims to predict it. The company employs a proprietary database of market research, based in part on detailed analyses of past successes and past failures in drug development. To predict the parameters that are most important for a product to succeed, they assess a project's funding, the new drug's potential competitors, dosing and animal models, and the patients being tested.

This last factor is particularly important, he says. Worldwide demographics change, and by the time a company's drug is ready for approval 10 to 15 years after the compound first began testing, the makeup of its target audience may have changed dramatically. "We haven't come to grasp that our population is changing," Prakash says. "We are running trials on the patients of yesterday."

The Short of It

Prakash, a trim man with warm eyes and a salt-and-pepper moustache, stands 5'4" tall - a fact that ironically launched his career. He grew up in Allahabad, a city in northwest India, and attended Allahabad University, one of the oldest schools in the country. Prakash graduated first in his class in biochemistry, with marks so high that they set a record for the school. With his academic record, he says, "I was guaranteed to get any [position] I wanted." So he took an exam for a coveted government position that would have set him up comfortably for life. He passed, but just before the start date, the government withdrew the offer, giving an explanation that sounds absurd to Western ears trained in political correctness: Prakash was told he was ever so slightly too short for the job.

"It was absolutely the best thing that happened," Prakash says. With the blessing of his father, a chemistry professor at a college near Allahabad, and little more than $20 to his name, Prakash headed off to the United States to study microbiology. He earned his doctorate in the lab of Richard Meyer, who studied antibiotic resistance at the University of Illinois. As Prakash's finishing date was drawing near, Meyer asked him to name his three top choices for a postdoc advisor. Not one to let circumstances decide his fate, Prakash slipped a piece of paper into Meyer's mail box on which, as requested, he'd written three names: Sydney Finegold, Sydney M. Finegold, and Dr. Finegold.

Finegold, one of the founders of the Infectious Diseases Society of America and a noted expert in developing antimicrobials for anaerobic bacteria, took on Prakash as his postdoc. After spending his time in Meyer's lab examining antimicrobial compounds in animal models, Prakash's fellowship in Finegold's lab "taught me how these things were working in human systems," he says. He also began to take courses at UCLA's School of Management at night. "It was very important for me to understand the business world," he says. "I never believed in [just] publishing papers. Even as a young man, I wanted to create something which would go on a pharmacy shelf."

He took his first industry position at Johnson & Johnson and then moved to Pfizer. There, he worked his way up to a position as assistant director for scientific affairs for international antifungal and anticancer programs, responsible for overseeing the development of drugs including Diflucan (fluconazole), a multi-billion dollar blockbuster and the biggest-selling antifungal drug in the world, which launched in 1990. After nine years at Pfizer, Prakash struck out for more entrepreneurial waters, serving as CEO of two small biotechs. Then, in 2003, the government came calling, and he spent time in management at the National Institutes of Health. After that, he launched and directed a new $1.6 billion program for the Department of Defense (DoD) to develop medical countermeasures for bioterror pathogens.

Two years ago, he returned to the entrepreneurial world to join his wife, Savita, at AMAR. The duo designs predictive models using more than 200 scientific and business parameters for each stage of development, from preclinical to postmarketing. They include data on similar drug development attempts, regulatory issues, and studies on the compound's mechanism of action, as well as feedback they gather from physicians about the drug's potential.

AMAR, which has 11 part-time and full-time employees and a network of about 60 special consultants, works with small companies, government research groups, and a few Big Pharma companies (including Pfizer). Convincing Big Pharma to take on their approach has been slow, says Savita Prakash; hardly anyone in middle management at large pharmaceutical companies stays through the entire development process, and most often, they don't want to deviate from established practice. "They need smaller companies to take the initiative," she says.

"Very few people have this rare blend of science talent and business knowledge," says John J. McGowan, deputy director of the National Institute of Allergy and Infectious Diseases, who worked with Prakash during his time at the NIH and the DoD. That combination helps him to see potential problems with a drug development project early on, says David Moskowitz, who became a client of AMAR when he served as managing director of healthcare at the investment firm Friedman, Billings, Ramsey. "For example, we'd think, let's talk to physicians," to ask whether a particular product might be useful. "But [AMAR] may say, 'Well, you may want to talk to the statisticians; ... the trial design may have particular aspects we need to explore before we go into the physician community."

Saving Drugs with Diversity

One of the biggest mistakes Prakash sees is that clinical trials are often simply conducted on the wrong patients. That conviction goes back to his days at Pfizer, where Prakash was approach by Kao Li, who at the time was a regional hospital coordinator on the sales team and was spearheading a new diversity initiative. Pfizer wanted to bring more people of color into its sales force, but Li quickly convinced Prakash that the company needed to widen its scope. The two asked Savita Prakash to crunch the numbers; her analysis showed that almost a third of prescribing physicians were either people of color or foreign-born. Those physicians needed to be convinced of the value of Pfizer's products, the trio argued, so they could convince their patients to enroll in clinical studies and take the products.

A year or two later, after both Li and Prakash had left the company, Li passed through Washington, DC, and spent the night as a guest in the Prakashs' home. The two sat on the living room floor, drinking Chinese tea and discussing clinical trial design late into the night. Already, anecdotal evidence that different ethnic groups metabolize drugs differently was strong. "But this is not reflected in the Physician Desk Reference because the clinical trials were not done with enough representatives of that population," says Li. Drug developers need to work different subsets of the population into Phase II trials, to see the effect of ethnicity early on, he and Prakash reasoned. "Maybe for some drugs it doesn't matter, but maybe for others it does," says Li.

Indeed, says Alfonso Alanis, CEO of Anaclim, an Indianapolis contract research organization specializing in minority recruitment in clinical trials, there's been little formal investigation of the topic. Recent data from the FDA suggests that more than 80% of subjects in clinical trials of diabetes drugs have been white, yet only about a quarter of newly diagnosed patients are white. "If you're trying to do good science, what you would like to have is a population of patients that is truly representative of the epidemiology of the disease you're studying," says Alanis, who spent seven years as chief medical officer at Eli Lilly and Company.

For AMAR, race and ethnicity represent a crucial variable in its models, which include technical issues specific to different ethnic groups, down to the diets that different groups eat, Prakash says. (Certain kinds of herbs, for example, can affect drug metabolism.) "Sure, you can hit all the boxes, and you might even get your drug approved, but as soon as it hits the market, it gets exposed not just to the patient on which you did the trial," says Prakash, who this year was the keynote speaker at a life science conference hosted by a new organization called the Diversity Alliance for Science.

Prakash likes to say that in his career in product development, he's had one success - Pfizer's Diflucan - and about 15 failures. Drug development is a science, Prakash says, but it is also an art. The art, he says, lies in analyzing the failures, start to finish, and making tweaks to fix the mistakes that were made. One of the most striking aspects of Prakash's outlook, says Li, is his ability to respond to the winds of change. He recalls a conversation the two had in Prakash's office at Pfizer, when Prakash looked at articles sitting atop his desk, and said, "If I should get a pink slip, I'm ready. I would just take my personal belongings, some photographs, put it in a box - one box - and move on."