A better way to model cancer

Later this year, the first US-based clinical trial to test whether an organoid model of prostate cancer can predict drug response will begin recruiting patients. Researchers will grow the organoids—miniorgans coaxed to develop from stem cells—from each patient’s cancer tissues and expose the organoids to the patient’s planned course of therapy.

If the organoids mirror patients’ drug responses, the results would support the model’s use as a tool to help guide therapy. “We are aiming for a precision-medicine system to treat prostate cancer,” said Hatem Sabaawy, MD, PhD—a co–principal investigator of the trial.

Organoid models were first described more than 40 years ago, but cancer researchers began studying them as better models of cancer only in the last decade. An organoid closely resembles the organ or tissue from which it was derived. Cancer cells grown in a dish—the standard method used to study the effect of cancer therapies—“can’t represent the complexity of cancer,” said Jeremy Rich, MD, of the department of stem cell biology and regenerative medicine at Ohio’s Cleveland Clinic Lerner Research Institute. Those conventional models are two-dimensional and bear little physical, molecular, or physiological similarity to their tissue of origin. With the high failure rate of cancer drugs in clinical trials, existing models do a poor job of predicting how patients will respond to treatment, he added.

Several research groups have shown that organoids grown from a patient’s own tumor cells mirror those tumors in terms of genes and other biochemicals expressed. Those findings suggest that such models can help guide personalized treatment decisions, Sabaawy said. Organoids from cancer tissues can be grown within 2–3 weeks and frozen for later study.

The prostate cancer clinical trial, which will take place at the Rutgers Cancer Institute of New Jersey in New Brunswick, will exploit tissues removed from patients with advanced metastatic prostate cancer. Researchers will sequence the DNA of each cancer tissue to create its molecular profile. They will also extract stem cells to grow organoids, which they will treat with the same therapy devised for the patients. Researchers will see whether using organoids can predict how patients will respond to treatment. The results will become part of a biobank of clinical profiles that will serve as a resource for future patients.

A similar trial of colon cancer patients in the Netherlands is ongoing, said Hans Clevers, MD, PhD, director of the Hubrecht Institute in Utrecht, Holland, and founder of Hubrecht Organoid Technology. Though results from the first few patients look promising, Clevers added, the study is in its early stages. Two more such trials, one in colon cancer and another in breast cancer, will launch in Holland later this year.

But although organoids hold hope of taking some guesswork out of cancer treatment, they don’t completely model cancer behavior in the body. Because they have no immune system or blood vessels, organoids can’t be used to evaluate whether immunotherapies or angiogenesis blockers will work, nor can they hint at whether a patient’s immune system will help or hinder cancer. To address those concerns, some researchers are using tests on mice to complement organoid models. Sabaawy, for example, is working with colleagues to implant patient-derived colon cancer organoids into mouse colons. The mice have been engrafted with a human immune system by using bone marrow cells from the same patient. The mice could serve as predictive models of immune therapy for each patient, he said.

Jatin Roper, MD, director of the Center for Hereditary Gastrointestinal Cancer at Tufts Medical Center in Boston, is working on a mouse model of human colon cancer as a research tool. Growing human colon organoids in mouse colons, where they are exposed to other cells found in the colon, allows researchers to observe the cancer in a more natural environment, he said. In his model of immune-deficient mice implanted with human-derived organoids, Roper has observed colon cancer metastasizing to the liver, which often occurs in human colon cancer. “It’s a system where you can test aspects of human biology that you can’t test in a dish,” he added. Roper and others are also using the gene-editing technology CRISPR (clustered, regularly interspaced short palindromic repeats) to induce specific mutations into organoids and then implanting them in mice to better understand how they drive cancer.

Such research has already highlighted cancer’s complexity. When Clevers grows organoids from different cancer cells from the same patient, he often sees some clones that are resistant to all drugs, even in patients who have never been treated. “The level of genetic diversity in these tumors is frightening,” he said. “It will be extremely difficult to find a cure.”

Nevertheless, better models can help tailor cancer therapy to each patient, reducing unnecessary treatments and the anxiety of waiting to learn whether a therapy worked. Clevers sees organoids playing a role in making targeted drugs more precise. Most chemotherapy drugs used in the clinic are not specific to any kind of mutation, he said. For older therapies, Clevers said he hopes that organoids can be used to help separate responders from nonresponders. Having the “tumor cells literally in our hands” offers many advantages, he said.

A version of this article originally appeared in the Journal of the Nation Cancer Institute.

Featured image credit: Intestinal organoid grown from Lgr5+ stem cells. St Johnston D (2015) The Renaissance of Developmental Biology. PLoS Biol 13(5): e1002149. doi:10.1371/journal.pbio.1002149 by Meritxell Huch. CC BY 4.0 via Wikimedia Commons.