Using Big Data for Better Cancer Treatment Decisions

Using Big Data for Better Cancer Treatment Decisions

CAMBRIDGE, Mass., May 3, 2016--Cancer is a leading cause of death worldwide with 13 million deaths expected annually by 2030.   And while the disease is common, deciding which drugs best treat the disease is difficult.

Advanced cancer is often treated with a combination of multiple chemotherapy drugs, yet it can be hard for scientists and medical professionals to know which combination of drugs will be most effective and the least toxic. Even knowing which combinations to test can be costly and difficult.

In a study released in the current issue of Management Science, scientists led by Dimitris Bertsimas of the MIT Sloan School of Management developed novel ways to model the efficacy of Stage II and Stage III clinical trials for gastric and esophageal cancers.

Using the most current analytical techniques, the authors developed a database of over 400 published papers showing results of clinical trials for gastric and gastro-esophageal cancers. Working with his team, Bertsimas then created a model to predict which Stage II or Stage III trials would provide the best outcomes.

“Today, the majority of clinical trials are done by intuition,” said Dimitris Bertsimas, Boeing professor of Operations Research and co-director of the Operations Research Center at MIT.  “Our goal is to use the most up-to-date data analytic and machine learning methods to insure the best outcomes for cancer patients.”

While this study focused on designing combination chemotherapy regimens for gastric cancer, he believes data-driven tools leveraging databases of clinical trial results could prove useful in other settings.  Combination therapy is used to treatment many other cancers as well as other diseases such as hypertension and diabetes.