In April of 2007, a research study on the use of Gleevec to treat early stage gastrointestinal stromal tumors (GIST) -- a type of gastrointestinal cancer -- ground to a screeching halt. Mayo Clinic biostatisticians Daniel Sargent, Ph.D. and Karla Ballman, Ph.D., were responsible. Their exploration of the study's early data revealed that trial participants were enjoying a startlingly significant reduction in the risk of cancer recurrence--so startling, in fact, that the trial was stopped, in order to expedite Gleevec as a bona fide treatment for reducing the risk of recurrence in GIST.

"Clearly, the best possible outcome is when something works even better than we thought it would," says Dr. Sargent. He explains that he's seen this same situation with three studies over the last several years, situations where the trials could be stopped more quickly than originally planned, due to the recognition of a clearly positive result.

Dr. Sargent credits the exceptional collaboration between Mayo investigators and its Biostatistics Department: "Because we very closely monitored the study and were up-to-date with how patients were doing . . .we could stop the trial more quickly than we had anticipated, because of a very favorable result. The great thing about that is you're able to get that therapy then out to patients more quickly than you would have been able to otherwise."

So what exactly is the role of a biostatistician on a research study? When investigators (medical doctors or laboratory scientists) generate a question that they think needs answering, they call in a biostatistician to help decide the study's parameters, such as how many patients are needed to generate sufficient data and what kinds of measurements will be needed to explore the question. Once the data has been collected, the biostatistician analyzes it from all angles.

"For example, if we are testing a new sort of therapy, we want to know first, did that therapy work overall?" explains Dr. Sargent. "We're also going to be interested in whether it worked in men, in women, in younger patients or older patients or patients of different ethnic backgrounds." He says he "explores and interrogates" the data, in order to gain the greatest amount of scientific knowledge out of what has been collected.

In the case of Gleevec, the study was designed to explore whether the drug could benefit patients who had undergone successful surgery in early stages of GIST, but still had the usual risk of recurrence.

It was thought that the study would require approximately 850 patients and six years in order to assess Gleevec's effect on preventing recurrence in early stage GIST. Instead, 4 _ years into the study, careful early analyses of the data revealed an exciting effect of the therapy.

"Gleevec actually reduced the risk of the cancer coming back by 2/3s, which was quite remarkable, and far better than we had even hoped might occur," explains Dr. Sargent. In light of these results, the study was ended, thus allowing Gleevec to go more quickly from being an experimental to an approved treatment for reducing the risk of recurrence in early stage GIST.

Dr. Sargent likes to use a sports analogy to explain the concept behind ending such a research study early. "If you set up a seven-game playoff series, the winner has to win four out of the seven games. You set it up this way because you think that the teams are pretty evenly matched, and so you predict that one game will go to one team, and another game will go to the other team, so you plan on requiring the full seven games in order to figure out which team is the best. Sometimes, though, one team will just win four games in a row, so you say 'We don't need to play games 5, 6, & 7, because one team is already clearly better than the other.'"

Dr. Sargent has been at this numbers game at Mayo for eleven years. A resident of Rochester since the age of nine, he went to the University of Minnesota for Bachelor's degrees in mathematics and statistics, and a Master's and Ph.D. in biostatistics. His pursuit of biostatistics was a natural for someone fascinated by mathematics, but compelled to use his or her talents and interests to help people. "That's why I find what I do so rewarding, being able to use the analytic skills and analytic training that I have to really improve people's lives."

Dr. Sargent says a number of factors drew him to work at Mayo, but most important was the highly collaborative and interactive environment. "Incredible researchers and scientists all have the same goal--research in which the patient comes first." Additionally, he cites Mayo Clinic's very strong tradition in biostatistics, with a department founded in 1929; it's long been one of the premiere biostatistics groups in the country. He says his love of the Rochester community rounds out his job satisfaction.

His dedicated focus on cancer research is more than a professional interest. He shared that a number of family members have died of cancer. "There's never any lack of motivation to do what I do; it's clear that cancer is a devastating disease, and you can't find anyone that hasn't had it affect themselves or their close family or friends." Dr. Sargent is himself a 25-year survivor of Hodgkin's Lymphoma, diagnosed and treated at age 10 at Mayo Clinic. So he finds himself in the position of having benefited from the same kinds of Mayo research, data analysis, and clinical care to which he currently contributes.

Dr. Sargent's personal research passion involves work that "only a biostatistician can do." This involves looking through huge data sets to try to find an early marker that predicts a disease outcome (referred to as a "surrogate endpoint"). An example involves treatment for colon cancer. Originally, studies of potential therapies looked at whether patients were living five years after completing treatment. Dr. Sargent's data analysis led to the realization that looking after only three years could very accurately predict whether patients were going to survive out until five years. Due to Dr. Sargent's publications and presentations, the FDA has approved three years as the new endpoint for study of experimental therapies for early stage colon cancer, making beneficial treatments available two years earlier. And Dr. Sargent is looking at the possibility that two years of experimental study probably works just as well: "For a disease like colon cancer, which is a very prevalent disease, if we can get treatments out to patients more quickly, it's going to help a lot of people."

- Rosalyn Carson-DeWitt, M.D.