Longitudinal data can determine earlier, more targeted cancer treatment

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
cancer - 139.23 Kb

BOSTON—At Moffitt Cancer Center in Tampa, Fla., the goal is to follow patients over time, said medical oncologist William S. Dalton, PhD, MD, speaking during the Medicine 2.0 Congress. Longitudinal data combined with recognition of the uniqueness of every patient, help clinicians match the right patient to the right treatment, he said.

The opt-in study began six years ago at Moffitt. The Total Cancer Care Protocol “must be a public healthcare issue,” Dalton said. “We can’t expect all patients to come to an National Cancer Institute-designated cancer care center.”

Six years later, the program has more than 90,000 patients enrolled and about one-third of those have donated their tumor. Some cannot donate their tumor because it’s too small. The data are now organized into four portals, one each for researchers, patients, administrators and clinicians. “It’s great to collect data but figuring out how to harness and use [them] is a challenge.” Dalton said they have already collected 1 million petabytes of data.

Dalton discussed the eventual possibility of giving patients their data in a usable format, perhaps one that includes targeted care recommendations, such as a screening colonoscopy or mammogram. He said he hopes that would allow patients to participate in their care. The protocol also serves as a vehicle to communicate longitudinally through a patient’s lifetime. “We need to integrate new science into the clinical pathway to determine its value. If we don’t, we will bankrupt the system.”

A data factory must be developed to harmonize the data. The power of a database, he said, offers the potential of superior decision-making and the “ability to create evidence-based medicine models to illustrate the relative value of treatments for payers and other stakeholders.”

Dalton also called for the development of a rapid learning environment system. With trials utilizing the operational database already underway, “we can design by frequency or prevalence of a given marker in certain diseases.” That is most effective through the power of a dynamic rapid learning environment, however. “We can’t just do it at National Cancer Institute centers—it must be a national resource.” He said a grassroots effort has begun to build a national health informatics research exchange. That could be one way that newly diagnosed cancer patients can learn about their options from other patients.