In This Section      

Promise and Progress - Cover Story Sidebar: The Mathematics of Curing Cancer

The Time is Now: 2010-2011

Cover Story Sidebar: The Mathematics of Curing Cancer

By: Valerie Matthews Mehl
Date: November 11, 2010

Gary Rosner
Gary Rosner

Research is driven by data.  There is no translation of research to patient care, no personalized cancer medicine, without being able to translate the data that help steer the research and clinical teams towards what is working and away from what is not. 

Technology has advanced faster than the ability to analyze it. The burgeoning field of quantitative sciences is taking center stage as scientists work to manage immense amounts of data being generated by these new technologies that help us see and understand the mechanics of the cancer cell and the environment in which it thrives. 

Leading the charge at the Kimmel Cancer Center is Gary Rosner, new director of its program that invokes principles of statistics and mathematics for biology-specific inquiries to help speed the application of discovery to patient care.  His team of ten statisticians is a unique group in cancer science.  They need to know their own craft—mathematics—but they also need to understand biology and technology, so that they can draw inferences from laboratory and preclinical studies and help accurately design clinical studies.

 “I want our statisticians involved from the start when the study is being designed,” says Rosner.  The automated sequencing technology used today to study cancer genes probes millions of DNA sequences.  A special difference on a chip could affect the reading it gets.  “If we do not incorporate random placement of the probe, the results we get could be related more to an aberration in the machine than to biology,” explains Rosner.  Their work helps ensure that results are related to human biology and not simply idiosyncrasies of the equipment.

Rather than working from the periphery of research programs, Rosner wants his team to be an integral part of them, attending lab meetings and working directly with investigators.  He believes the scientific community needs to change the paradigm for Phase I and Phase II clinical trials to get new treatments to patients more rapidly.  He is advocating for something termed adaptive study design.  “We need to develop methods that include information from laboratory research and preclinical studies as well as data from other similar studies,” says Rosner.  “If we borrow strength from all of these things, then we might speed the pace of clinical trials.  We can more quickly find the right dose for an agent.  If it’s a targeted drug, we can more quickly determine if it’s hitting its target but also see if it’s hitting other targets as well.” 

Phase I trials are currently proposed to determine the safest dose and best way to administer a new agent.  Phase II trials continue to explore safety but also begin to look to see if they are having the desired effect.  Amazing responses, however, are not always dose related, Rosner points out, and recommends combining Phase I and II trials. “We must look beyond the limited perimeters of Phase I trials and develop a 30,000-foot view of how all of the pieces fit together,” he says. 

Rosner also calls for midstream analysis of clinical trials to determine earlier on if these potential new treatments show promise.  “If we can find out that something isn’t working, let’s abandon it or go back to the drawing board,” says Rosner.  He also wants to make sure that patients are getting the very best option.  He recommends that several potential therapies be tested simultaneously among larger numbers of patients.  As some reveal themselves as better performers, the trial can be adapted to move patients from drugs that aren’t working to those that are.

We must also look at how pharmacology and genetics interact to potentially improve responses but conversely to identify when they could result in a toxic side effect. 

As we begin the age of personalized cancer medicine, Rosner says we must use analyses that pinpoint patients with cancer gene profiles that indicate cancers respond better to specific therapies.  For example, researchers have found that lung cancer patients who have never smoked more often have mutations of the EGFR gene.  As a result, they are more likely to benefit from drugs that block EGFR signaling.  By the same token, we now know that certain mutations can trigger toxic responses to specific compounds. 

Understanding these interactions and using the data in designing and adapting clinical trials helps ensure that the safest and most effective treatments are identified for each patient.

Articles in this Issue

Director's Letter


Find Physicians Specializing In...