Cancer Researchers Look Ahead to Breakthroughs in Personalized Medicine

Artificial intelligence and nano technologies are top opportunities to change the care dynamic.

Technology is helping to change the way cancer is predicted and treated, with speed and specificity unimagined even a decade ago.

“Transformation is happening,” said Mauro Ferrari, PhD, president and CEO of the Houston Methodist Research Institute. “I think 10 years from now, cancer treatment will be very different.”

Ferrari and his team are taking a systematic approach to treating and predicting cancer.

“The future of medicine has a lot to do with personalization. No two people and no two diseases are alike,” he said. “Personalization of medical intervention is certainly going to be a big thing.”

For example, millions of Americans have had breast cancer, but not all cancers are the same and each reacts to medicines differently.

“The same disease can have different manifestations in different people,” Ferrari said. Trying to pinpoint risk and what will work for a specific person is the key.

That’s where fast computers can come into play. Researchers at Houston Methodist are using artificial intelligence with powerful algorithms to help predict breast cancer and reduce costs by eliminating the need for painful and expensive tests. The software interprets mammograms and other patient data from charts to attempt to reduce the number of false positives. Usually, a positive or suspicious mammogram leads to a biopsy, many of which come back negative. “It’s costly and not good for patient care,” said Stephen Wong, PhD, PE, chair of the department of systems medicine and bioengineering at Houston Methodist Hospital.

“Right now, a lot of effort goes into reading mammograms.”

AI can change that. It can interpret key words from free text, compare lab results, read mammogram images and more to help come up with a more accurate breast cancer risk prediction than the current BI-RAD score. BI-RAD is short for Breast Imaging Reporting and Data System, which is a classification system radiologists use to categorize breast screening results. Currently, a patient with a suspicious BI-RAD score often undergoes a biopsy, even though the risk of cancer varies widely from 3 to 95 percent.

“Biopsies are not an easy way of doing things,” Wong said, “This is definitely a waste.”

The data for the AI comes from the electronic medical records of about 4 million patients in seven hospital systems within Houston Methodist.

“In the past, it was difficult to get there, but now it is at our fingertips,” Wong said, within that database “you can cut and dice [the data] however you like.”

Manually reviewing 50 patient charts could take a couple of clinicians 50 to 70 hours. The AI software can do 500 charts in just a few hours, freeing up clinicians do to other things.

“If I was doing it myself, it would take years,” Wong said. “AI allows me to do things I could not do in the past, no one has time to read so many reports.”

New technologies such as AI that lead to a reduction in hours spent reviewing charts and in the number of biopsies could mean lower costs.

“I think it could be beneficial because it could reduce unnecessary and ineffective treatment,” Ferrari said.

If a cancer is found, determining the right course of treatment the first time also could reduce costs and improve outcomes for the patient. Nano drug delivery devices could help pinpoint treatment and minimize side effects.

“It is a very tiny device, which is the size of a grain of rice, which can be inserted into a tumor,” said Alessandro Grattoni, PhD, chair of the department of nanomedicine at Houston Methodist Research Institute. The device stays within the tumor and delivers the drug there, unlike traditional chemotherapy treatments in which drugs go all over the body.

Current chemo treatment protocols are expensive and often take weeks and mean repeat visits for the patient, leading to additional costs and horrible side effects.

“[In chemotherapy] you need to use a lot of the drug to achieve the concentration of the drug into the tumor,” Grattoni said. “A lot of the body that is actually healthy absorbs the drug and suffers from it. We want to treat only the part of the body that is diseased, not the rest.”

This new type of treatment could do just that and also go after rogue cancer cells that may have metastasized throughout the body.

MORE ON TECHNOLOGY: Read About Tools, Tactics and Strategies That Influence Physician Leadership

The surgical procedure to insert the nano delivery device is minor and the side effects are minimal because only diseased tissue gets the drug. The nano particles are tiny, about 20,000 times smaller than a human hair, so the delivery device contains enough drug to treat the cancer.

“For as long as you have the drug within the device, the drug will pass through the nanochannel membrane and get into the tumor at a constant rate,” Grattoni said. “It’s a very simple concept that utilizes very high-end technology.”

That technology comes at a cost, but one that needs some perspective, Ferrari said. Even though the new predictive methods, drugs and delivery systems might cost more than the current treatments, personalization can be cheaper in the end.

“We go through years coming up with specific treatments,” he said. “We need to consider costs in a broad and comprehensive way,” Ferrari said.

Also, preventing a recurrence of a disease means a cost savings down the road. “You have to look at the impact to society in a different way.”

Both the artificial intelligence to predict breast cancer risk and the nano drug delivery devices are far from being used in everyday care. The AI project is in trials and results have been published in peer-reviewed journals.

“It’s promising, it’s going very well,” Wong said.

Grattoni and his team are studying the implanted nano drug delivery device in mice.

“Overall, if we are capable of demonstrating the effectiveness, this could be a real game-changer,” he said.

On average in the United States, bringing something from discovery to clinic takes 17 years and costs $2.7 billion. Ferrari is looking to lower that cost and speed up the timeframe.

“We are very focused on bringing discoveries to the clinic in the safest and most efficient fashion,” Ferrari said. “We look at things with a very high impact.”

Tiffani Sherman is a freelance journalist based in Florida.