BOSTON-Researchers have developed a new way to visualize pancreatic cancer to determine the molecular details of the disease and evaluate therapeutic opportunities using organoids, or miniature organs in vitro.
Organoids provide an opportunity to create cellular models of human disease, which can be studied in the laboratory to better understand the causes of disease and identify possible treatments. "The vision is that a pancreatic cancer patient has a surgical resection or biopsy, and organoids are prepared that are characterized to look for mutations in the patient's DNA and, in parallel, in organoids subjected to drugs active against that organoid. With molecular and therapeutic information, we can go back to the treating physician and suggest which drugs to treat the cancer," said David Tuveson, MD, PhD, Professor and Deputy Director of the Cancer Center at Cold Spring Harbor Laboratory in New York, in an interview with OT.
At the moment, organoids are an investigational approach. Tuveson and other researchers are attempting to determine whether this is a tractable model for pancreatic cancer patients, and whether it can be generalized to other cancers as well. Results from prospective studies now being conducted to evaluate this approach in pancreatic cancer patients "will within two years let us know whether this is a research tool or can help us take care of patients," Tuveson said.
Improved Therapeutics
Pancreatic cancer is becoming more common and is projected to become the second leading cause of death in the US by 2020. "We need better modalities," Tuveson said.
The disease is detected late. Most patients are detected with metastatic pancreatic cancer, which has one of the poorest five-year relative survival rates of any cancer, 2 percent. The overall five-year relative survival rate is 6 percent. If resected, only 20-25 percent of patients are alive at five years, he said.
Pancreatic cancer derives from mutations in key genes, including KRAS, tumor suppressor genes, and in some inherited mutations. "Inflammation promotes pancreatic cancer and can be caused by obesity, the microbiome, tobacco smoking, or intermittent alcohol drinking. We may be able to prevent mutations or inflammation through diet, avoiding tobacco, and limiting alcohol," Tuveson said, noting many mutations are due to DNA oxidation.
"We are developing accurate model systems to enable basic observations to establish dependencies and discover new therapeutics and diagnostics. We are learning more about disease onset and progression and finding new cancer vulnerabilities," he said.
Researchers are looking for therapeutics to kill the cancer, including more effective new drug combinations and novel therapies with increased efficacy, as well as biomarkers to detect pancreatic cancer earlier.
Previously, in vitro and mouse models have defined the early steps in cancer initiation. The new model adapts a system of culturing primary tissue to develop pancreatic ductal organoids. Researchers derive and culture organoids by growing a tumor or the normal duct from a human or mouse rapidly in three dimensions. "We are modeling disease progression relative to the normal pancreatic tissue in 3D culture," Tuveson said.
"We are learning from organoids to develop better therapeutics and diagnostics for a personalized approach," he continued. Organoids are used for precision medicine for biomarker discovery, therapeutic testing, and to identify resistance mechanisms, characterize genetic lesions, study interactions between cell types, and predict treatment response.
After deep sequencing of tumors and organoids from tumors, "we have found more mutations in organoids than in the primary tumor," he said, adding that organoids may miss a few mutations and may be unstable. In pancreatic cancer, the stroma masks the ability to see mutations. Organoids may be a new way to unmask mutations.
Future Study
Tuveson is leading the Actionable Loci in Pancreatic Cancer Specimens (ALPS) trial, which is being assessed by regulatory bodies. The trial, which plans to open this spring, will follow 50 patients with pancreatic cancer who had surgery and 50 patients with newly diagnosed metastatic pancreatic cancer, taking 300 specimens from each group.
"The idea is to develop organoids to identify drugs and demonstrate we can generate organoids in a timely fashion. The goal is to define molecular abnormalities in pancreatic cancer and to correlate this with specific therapeutic sensitivities that exist for that version of the cancer," he said. "If we make organoids on day one, will we have an answer for a physician in six months?"
Tuveson likens this process to a microbiologist identifying the bacteria present in an infection, and then recommending the best drug to kill the bacteria.
He noted seven drugs are FDA-approved for the treatment of pancreatic cancer. "We always look at them first, and then all other FDA-approved cancer drugs, as well as investigational drugs that may be promising," he said.
A spin off of the ALPS trial could help find better drugs. "Once we show that organoids can be generated, the next trial is to use organoids to direct patient management. We would have multiple arms open, similar to trials for lung cancer patients who are separated into groups based on the molecular characteristics of their tumors. Not just what the pancreatic tumor looks like, but how the tumor responds in the lab, trying to move it to be more like microbiologists," Tuveson said.
Organoids also can be used to find diagnostic biomarkers, both universal and personal. "Universal biomarkers will be useful for early detection of cancer, and personal biomarkers for longitudinal studies. We can identify multiple mutations in a patient's genome and look for epitopes on the surface of organoids," Tuveson said.
In the future, researchers may assemble a large collection of organoids in all variants of pancreatic cancer. "This bank of organoids will be used for therapeutic investigations for optimal strategies that can be provided to new patients on demand. We will still need individual organoids for personal diagnostics and to follow patients in the future, when we expect we will have better drugs and patients will become resistant to drugs for different reasons," Tuveson said.
He foresees a registry of cross species of organoids to reveal biomarkers of early pancreatic ductal adenocarcinoma. "We could develop a blood test for pancreatic cancer using a protein-based assay," he said.
Organoids "may be a new way to get tissue from patients as we care for them. The tissue may be a crystal ball into their disease. We can peer into the molecular causes of the cancer, look for therapeutic methods to treatment it, and develop diagnostics for their specific cancer," Tuveson said.
"It's a method that makes sense," he continued. "It's something that we could scale up and speed up. The idea is to do it rapidly, efficiently, and economically. This may be a new way to look at cancer in patients. I anticipate it has a good shot at playing a role of how we take care of pancreatic cancer patients in the future."