Devices Test Drugs in Patients' Tumors

Because people with the same cancer can respond differently to the same therapy, it's important to “identify the best therapy and kill the tumor effectively the first time around,” says biophysicist Oliver Jonas, PhD, a postdoctoral fellow working with Robert Langer, ScD, at Massachusetts Institute of Technology (MIT) in Cambridge.

The best way to gauge a drug's effectiveness is to study it in a tumor's natural environment—inside the patient, says Richard Klinghoffer, PhD, chief scientific officer at Presage Biosciences in Seattle, WA. Cell cultures and animal models don't reproduce key features of the tumor's microenvironment.

Jonas and Klinghoffer are the lead authors of two studies describing experimental devices designed to simultaneously test multiple cancer drugs directly in the patient. Their work was recently published in Science Translational Medicine.

The most advanced version of Jonas's tool—a cylinder about the size of a grain of rice—can be loaded with up to 48 drugs, or combinations of drugs. With a standard biopsy needle, researchers inject the device into the tumor, where the drugs are released. The drugs are spaced far enough apart on the device that compounds in adjacent reservoirs will not seep into the same region of the tumor tissue. After 24 hours, the implant is removed, along with a bit of the surrounding tissue, with a larger-core needle. By staining the tumor samples with antibodies to cell death or proliferation markers, researchers can determine how well each drug worked.

A future version of this tool could potentially test more than 48 drugs without major changes to its size or design. Practically speaking, though, “it's probably easier to put two devices into one tumor to test 96 drug combinations,” Jonas says.

The MIT team tested its technology in mouse models of human prostate, breast, and skin cancers (Sci Transl Med 2015;7:284ra57). In one set of animals, the researchers measured tumor cell apoptosis in response to drugs loaded into the implanted device. These local cellular readouts correlated with tumor cell responses in a separate cohort of animals treated systemically with the same drugs.

Plans are under way to test the device this summer in patients with breast cancer, to show that the local molecular readouts, such as expression of cell death or proliferation markers, correlate with clinically relevant markers such as tumor shrinkage and long-term survival, Jonas says.

The Presage team, working with scientists at Fred Hutchinson Cancer Research Center in Seattle, developed a hand-held injection device with eight needles (Sci Transl Med 2015;7:284ra58). Doctors can use this device, called CIVO, much as they would administer a flu shot, Klinghoffer explains.

Guided by ultrasound imaging, the physician positions the device over the length of the tumor and pushes a lever to deliver up to eight drugs. Part of the tumor is removed 1 to 3 days later for analysis by immunohistochemical assays and high-resolution imaging. These analyses could help doctors choose the best drug for the patient.

The scientists have used CIVO successfully in mice engrafted with human tumors and in dogs with naturally occurring cancer. The team also tested CIVO in people with lymphoma. The mouse experiments showed that localized tumor responses predicted responses to the same drugs given systemically, and the research in dogs and people found no serious side effects with the microinjection procedure.

In a related commentary, R. Charles Coombes, MD, PhD, professor of medical oncology at Imperial College London, UK, writes, “These techniques offer a possible alternative to the ‘hit and miss’ way of using anticancer drugs in patients that has unfortunately become accepted practice” (Sci Transl Med 2015;7:284ps10).

For more news on cancer research, visit Cancer Discovery online at http://CDnews.aacrjournals.org.