Author: Michele Cohen Marill
Joel Silverman is facing down a nightmarish cancer prognosis. What he’d thought was a benign cyst in his jaw was actually a rare cancer that grew stealthily, supplanting the bone. And even after the tumor was excised, an undetectable remnant in his bloodstream seeded metastatic lesions in his lungs. His doctors can do little beyond removing the lesions as they appear. This cancer, myoepithelial carcinoma, doesn’t have a standard chemotherapy treatment.
Silverman, 59, an internal medicine physician in Boca Raton, Florida, is accustomed to delivering both good news and bad to his patients, so he is realistic about his predicament. But he is also aware that science constantly pushes the bounds of what is possible. His hopes now hinge on a new paradigm of personalized medicine that will use a half million fruit flies to design and test a drug regimen tailored to his specific cancer. Not his type of cancer. His individual tumors.
Drosophila melanogaster, the tiny creatures of high school genetics experiments, are actually sophisticated models of human biology. Some 60 percent of Drosophila protein-encoding genes (known as the exome) have a parallel in humans. Drosophila can become drunk or obese, can develop diabetes or Parkinson’s disease, and, with some tricks of genetic engineering, can be induced to develop tumors identical to those in humans.
The signaling pathways among cells—the mechanisms that control cellular repair, for example—are the same in humans and flies. “If you find a drug that is going to affect [a cancer-inducing] oncogene in flies, there’s a good chance that will have an effect in humans,” says Norbert Perrimon, a developmental biologist and geneticist at Harvard Medical School who developed several of the techniques used in genetic research on flies.
That is the premise of a London startup called Vivan Therapeutics (formerly My Personal Therapeutics), which is harnessing a century of fruit fly genetic research along with modern genomic sequencing to create the cancer-fighting “Personal Discovery Process” in which Silverman’s hopes lie. (Perrimon is not affiliated with the company.) Their process is essentially a fly-based clinical trial for an individual patient: By giving hundreds of thousands of fruit flies the same cancer mutations as in a human patient, they can run thousands of drug screens in parallel, testing to see which are the most effective—and in what combinations—for combatting that particular tumor. This truly personalized medicine incorporates both common cancer drugs and those not normally used in cancer treatment.
Vivan Therapeutics is focusing on gastrointestinal, head, and neck cancers and rare cancers for which there is no established treatment. Ultimately, the company plans to develop a database of gene mutations and previously tested drug combos, enabling patients to receive suggestions for a regimen more swiftly. “For colorectal cancer, we already know that there are about five drug combinations that work for about 75 percent of the population,” says CEO and founder Laura Towart. “When we have a colorectal cancer patient, we would first test those five drug combinations along with 150 other drugs”—ones that have shown some effect in other screens—“and see if they rescue the flies. If they don’t work, then we would broaden out the screen.”
This could be the beginning of a new horizon in cancer care, a move beyond the current targeted therapy based on a single biomarker or commonly found mutation (such as the BRAF gene in melanoma). The fly-based process seeks agents against the combined, interactive effects of as many as 20 mutations in a single tumor. But at the moment it is still more of an oncological Hail Mary, an option for patients who have exhausted all their alternatives.
“I’m about two months away from finding out if there are other drugs and combinations that could actually save my life,” says Silverman, who has been taking a targeted therapy based on the detection of a mutation in the PIK3CA gene. It’s not clear if that drug is diminishing the lung lesions. “If they could stop what’s going on in my lung, my life is saved—or at least prolonged,” he says of Vivan Therapeutics.
The basic science underlying Vivan Therapeutics dates back to 1918, when Mary Stark, a little-known scientist in biologist Thomas Hunt Morgan’s famous Fly Room at Columbia University, identified tumors in Drosophila larvae and experimented with transplanting pieces of them into healthy larvae. Over the decades, the lowly fruit fly became an exquisite model of human disease. (Morgan received a Nobel Prize for his Drosophila work in 1933.) The fruit fly reveals attributes and treatments for disorders ranging from amyotrophic lateral sclerosis to aging, from epilepsy to eye disease—the source of enough discoveries to fill a book titled First in Fly. (The author, Harvard geneticist Stephanie Mohr, also contributes to an ongoing blog called Drosophila Models of Human Disease.)
When the Drosophila melanogaster genome was sequenced in 2000 (three years before the human genome), new possibilities arose for probing the genetic origins of disease. Developmental biologist Ross Cagan was studying the mechanisms of cancer in fruit flies, but in 2010 he turned the question around: Could the flies reveal cancer-killing drugs, even if the science wasn’t fully worked out?
He created the drug testing process in his lab at Mount Sinai Medical Center in New York City that has since been licensed by Vivan Therapeutics. “We’re exploring which drugs work, attacking the cancer network with a therapeutic network,” says Cagan, who recently moved his work to the University of Glasgow in Scotland.
First, scientists analyze the patient’s tumor, comparing its exome with the whole exome sequencing of the patient’s blood to identify the tumor’s protein-coding alterations. They select the changes most likely to drive the growth or proliferation of the tumor, based on their function or location. (A single tumor can contain hundreds of genetic alterations, but typically only five to 15 of them drive its growth.)
“There are many, many tumors that are not caused by one mutation. Or one mutation is compounded with two or three others that allow the cancer to grow, proliferate, and stay alive,” says Marshall Posner, a Mount Sinai oncologist specializing in head and neck cancer who has conducted fly research with Cagan but is not affiliated with the company.
Scientists next inject strands of synthetic bacterial DNA into fruit fly larvae to integrate the mutations into the genome. The location is precise; a colorectal cancer will be expressed in the fly’s gut, for example. Then they calibrate larvae development by altering the temperature of their environment, so the tumor is timed to kill the larvae in seven days. (Larvae typically metamorphose into flies in 10 to 11 days.)
Then these fruit fly “avatars” must propagate. Vivan Therapeutics uses about a half million fly larvae to test about 2,000 drugs and drug combinations, encompassing a version of most FDA-approved drugs that are currently in use, says the company’s chief scientific officer, Nahuel Villegas. For example, an anti-inflammatory or anti-hypertensive drug might have unexpected cancer-fighting properties when used with a tumor suppressor.
The larvae live in tubes in groups of 35—half with the tumor, half without to serve as the control group—feeding on drug-laced food. The healthy larvae have been tweaked with genetic alterations that make them shorter and fatter, so they can be distinguished from those carrying tumors. After seven days, their survival rates are compared. Every drug is tested on at least 300 larvae, and promising drug combinations are retested. The top candidates are ranked based on the survival rates, but ultimately the selection takes into account the human patient’s clinical history and their oncologist’s judgment. For example, a patient with an underlying cardiac problem might steer clear of a drug associated with cardiac concerns, Villegas says.
The entire process takes about six months, from whole exome sequencing of the tumor to drug recommendations. “We are under pressure to get it right and to get it fast. We want to give them the best option possible,” says Villegas. While the patients and their oncologists will make the decision about following any recommendations, the process is designed to expand and individualize the cancer-fighting armamentarium beyond what is currently possible.
So far, only a limited number of people have followed these fly-tested regimens. Cagan launched a clinical trial testing drugs in fruit fly avatars at Mount Sinai in 2016, and it took a few years to enroll 39 cancer patients. Nine patients received the screening and treatment options, and other patients stayed with standard treatment or haven’t yet used the recommendations, for various reasons (including the interruptions caused by Covid-19).
In 2019, Cagan, Posner, and colleagues published a report in Science Advances about a patient with metastasized colorectal cancer who had stopped responding to his chemotherapy. The fly model identified a promising combination of the anti-cancer drug trametinib and the bone-repairing drug zoledronate. The patient followed a regimen with the drugs, and the lesions shrunk by 45 percent, according to the case report. But after 11 months, new lesions appeared. Although the patient died, the sustained response to the novel treatment hinted at the promise of the personalized fly model. “It’s a success in the sense that it worked for a while,” says Cagan. “Now we can say maybe we’re moving in the right direction.”
Another case study done by the same team, published in the open access journal iScience, focused on a patient with metastasized adenoid cystic carcinoma, a rare and difficult-to-treat cancer of the salivary gland, who received a three-drug cocktail that was identified in the fly screening. The combination of the arthritis drug tofacitinib, the beta blocker pindolol, and anti-cancer drug vorinostat caused the cancer to stabilize for 12 months. But after a year, new mutations evolved that evaded the drugs and the cancer rapidly progressed. The patient died the following year.
While they believe that intervening earlier with personalized treatment would likely produce better results, the Vivan Therapeutics team acknowledges that their work hasn’t yet evolved enough for fruit-fly-derived regimens to supplant standard care. “We are very open with the patients in terms of what we know and don’t know,” says Cagan, who is on the Vivan Therapeutics scientific advisory board.
Still, this would expand the current scope of targeted treatments by tailoring them to individual patients. “The promise of precision medicine and targeted therapy is very real, but it’s [been] very limited,” says Posner.
Vivan Therapeutics recently began recruiting patients for clinical studies in the United Kingdom (20 patients) and Saudi Arabia (five patients), and expects to launch a study with at least five patients in Switzerland. Nine private patients from various countries are at various stages in the process, but none have yet received a fly-derived treatment. The company charges $15,000 for the Personal Discovery Process, but Towart says this barely allows Vivan to break even. Its broader plan is to develop a database of cancer mutations and use machine learning algorithms to identify drug combinations to treat them. Vivan Therapeutics has received grants from the European Union and the United Kingdom and is using tumor sequencing data from Genomics England to build profiles.
Eventually, most patients will simply need their tumor sequenced and then could take a drug cocktail identified by the AI system, rather than new tests in fly proxies; the AI system would search the database for a match with previously screened tumors and the resulting drug recommendations. That would greatly shorten the time to treatment, Towart says. But as new mutations arise, some patients won’t match a tumor profile in the databank—or some will want their own personalized avatars—so Vivan Therapeutics officials expect to continue conducting some fly trials.
For now, the search for the optimal cancer treatment still requires a half million fruit flies. That cumbersome process concerns Silverman’s oncologist, Hilary Gomolin at the Lynn Cancer Institute in Boca Raton. When a patient’s cancer has already metastasized, it’s hard to wait months for a new treatment. Insurance companies might not cover cancer drugs identified through a method they view as experimental. Silverman’s tumor is particularly challenging—a rare salivary gland cancer with an even rarer emergence in the jaw—and the effectiveness of any treatment may be limited.
But Gomolin, who once worked with Posner at the Dana-Farber Cancer Institute in Boston, remains open-minded. “I’m an eternal optimist,” he says. “I hope they can find something that helps Joel.”
Ideally, oncologists want a more individualized way to treat each patient’s tumor. Researchers are pursuing other models to explore tumor vulnerabilities, including organoids or engineered tumors in zebrafish or mice. But ultimately, any approach will need to be validated in studies that confirm the effects in a sizable number of patients, says Sam Klempner, an oncologist who specializes in gastrointestinal cancers at Massachusetts General Hospital.
“Our standards of care are not good enough for many tumors. We definitely need new platforms,” says Klempner, whose research focuses on targeted therapies. “Ultimately, the power of many model systems is the generation of large data sets that we can mine for patterns and shared vulnerabilities in the tumors.”
So far, Silverman has been able to keep on seeing patients and carrying on with his life. He just wants more time—time to spend with family and see his youngest child, who is a freshman, graduate from high school. And perhaps his results will add important information to the fly bank that could shape future treatments. “Whether it works for me or not, I’m convinced it will help somebody down the road,” he says. “It’s worth all the effort to support them and help them move forward.”