Experts are increasingly turning to a research network called Matchmaker Exchange to find diagnoses for patients with rare genetic disease and chart new paths to discovery and even potential treatments.


Leah Eisenstadt

In late 2010, Nick and Nicole Coslov welcomed two baby boys into their family: fraternal twins, Cameron and Drew. Born second, Cameron spent two days in intensive care to treat low blood sugar, and other medical challenges soon followed. He had trouble feeding and struggled to lift his head. He was late to hit other milestones as an infant and, by 18 months, had notable delays in speech, fine motor skills, and overall development.

The pattern of his symptoms didn’t fit any known disorder. “We’ve got all these clues, and even living in New York City with some of the best healthcare in the world, we couldn’t find what was wrong,” said Nick Coslov, an executive at a self-storage developer based in Manhattan. Without knowing the cause of his disorder, Cameron’s medical team was unsure how to help him. Over the next few years, the family pursued every avenue they could: “You pick a specialist, we’ve been there,” said Coslov. However, all their visits to physical, speech, and occupational therapists and multiple neurologists brought limited improvement and little insight into the true nature of his unique condition.

The Coslovs went on to spend a decade on their diagnostic odyssey, during which they hoped to find other patients like Cameron and solve the mystery of his disorder. After a few wrong turns, their search eventually brought them to a team of geneticists at the Broad Institute of MIT and Harvard and to a rare disease research network known as Matchmaker Exchange, through which the Coslovs were at last able to confirm the genetic root of Cameron’s disorder. Through the network, they found a handful of patients with the same malfunctioning gene and similar symptoms, and a scientist across an ocean who is studying these patients and determined to find a possible new treatment for them.

“Without that connection, we’d never have stumbled upon the answer,” said Coslov. “Getting a diagnosis was a relief, because now we can focus our efforts on something real. And yet, it’s not all that different on a day-to-day basis, because there’s still so little known about his illness.”

Cameron is one of more than 400 million people worldwide who live with one of over 7,000 described “Mendelian diseases” — rare disorders that arise from genetic alterations, or “variants,” in a single gene. For one-third of those conditions, the underlying genetic cause is unknown.

Patients with suspected genetic disorders can have their DNA sequenced and analyzed. For some, this uncovers a DNA variant known to cause disease. However, more than half still remain undiagnosed after DNA sequencing, often because the variants found in their genomes are extremely rare and haven’t yet been studied or linked to a human illness. It’s a frustrating outcome for families who hoped the analysis would yield answers, not more questions.

Matchmaker Exchange is a diagnostic game-changer that has cracked the case for many of these patients. Launched in 2015, it is an online platform that researchers use to connect patients around the world who have a Mendelian disease and no diagnosis. The hope is that two or more individuals who have similar symptoms and variants in the same gene would be “matched” through the network, which can provide researchers with the evidence and the confidence they need to prove a gene’s role in disease, help define a new genetic disorder, and give families much needed diagnoses.

Matchmaker Exchange, which was spearheaded by multiple researchers around the world including Broad scientist Heidi Rehm, finds matches using genetic and clinical data from roughly 150,000 families added by more than 11,000 contributors in 88 countries. Each day, researchers and physicians add hundreds of undiagnosed patients like Cameron to the network. Some match with others already in the platform, bringing medical insights and turning their one-of-a-kind illness into one that’s simply very, very rare. Others remain in the network, waiting for a match that could one day put a name to their illness and connect them with other families facing the same challenges.

Beyond its diagnostic potential, Matchmaker Exchange has become an essential research tool used by most experts in the rare disease field today. Most, if not all, rare disease gene discoveries made in the last several years have involved Matchmaker Exchange. Since its launch, the network has been used to define dozens of candidate genes as disease-causing and generate knowledge that can help future patients arrive at a diagnosis more quickly after clinical sequencing.

“There’s not a group doing rare disease gene discovery today that’s not using Matchmaker Exchange,” said Rehm.

There’s not a group doing rare disease gene discovery today
that’s not using Matchmaker Exchange.
Heidi Rehm

A diagnosis, however, is not the end of a patient’s journey. Uncovering the cause of Cameron’s disorder is just the first step in a new quest to understand the meaning of his unique genetic makeup, and to explore whether the misspellings in his genome can possibly be offset by a drug. Yet while his diagnosis has given the Coslovs optimism, it also creates more uncertainty, as they wait to see if this new path will lead to more than just insight, but real improvements to his quality of life.

A diagnostic journey

Today, Cameron is a cheerful 11-year-old who, despite his difficulties with walking and communication, has a great sense of humor and positive attitude. He loves playing with his peers and his brother, Drew, whom he admires. “Cameron is always smiling and laughing,” Nick Coslov said. “I’m biased because he’s my son, but he really has the best disposition, and doesn’t bear the weight of any of this.”

As with many families living with undiagnosed diseases, the Coslovs struggled for years to get an accurate diagnosis. When Cameron was three years old, his neurologist felt certain that he had cerebral palsy, a group of disorders that impair movement, balance, and posture. However, the course of his disease soon threw that diagnosis into doubt. Cameron even underwent an experimental stem cell treatment for cerebral palsy, but it didn’t improve his symptoms. In addition, brain imaging showed none of the tissue damage that usually underlies the motor disorder.

With that diagnosis in question, Cameron’s orthopedist at the Hospital for Special Surgery in New York, David Scher, suspected a genetic disorder and suggested that the family focus on finding the true cause of his illness. A local clinical geneticist who examined Cameron agreed. Some unknown molecular misspelling in his genome, possibly a single letter of DNA among 3 billion, was likely causing a chain reaction of biological disruption that led to his unique pattern of symptoms and traits.

In 2016 the family arranged to have Cameron’s DNA tested using exome sequencing, a technique that reads out the roughly two percent of the human genome that encodes for proteins, the working parts of the cell. If the sequencing turned up a genetic variant already known to cause a human disease, the Coslovs would have gotten a relatively quick diagnosis. But instead, the analysis yielded a handful of suspicious variants that were rarely mentioned in the scientific or medical literature. This suggested that his genetic disorder was likely to be exceedingly rare — he could perhaps be the only person in the world with the condition.

“With an ultra-rare disease like this, you have to be your own advocate,” said Coslov, “especially when you are the only family you know going through this.”

Families facing rare diseases intuitively know the power of connection. When a diagnosis seemed far from reach, the Coslovs suspected that finding other patients like Cameron might bring them some clues. Patients with diseases that are rare — affecting fewer than 200,000 people — can turn to social media to learn more about their illness and join a community. Yet, for ultra-rare diseases that affect only a handful of people in the world, those resources may not exist, especially when the condition doesn’t yet have a name.

With an ultra-rare disease like this, you have to
be your own advocate.
Nick Coslov

So the Coslovs created a website to share notes about his medical history, the genetic variants he carries, and photos of him through the years. A video they shared features Cameron at age eight in a physical therapy session, smiling as he labors to walk across the room unassisted.

His parents had mixed feelings. They worried that finding others like him might reveal a more difficult future than they’d imagined, one with a more profound disability or a shortened life expectancy. Still, they hoped someone, somewhere, might have a child with the same challenges or variants in the same gene and reach out to connect. No one did.

Meanwhile, Cameron’s doctors homed in on one set of variants in his exome sequence, mutations in a gene called PI4KA. The protein it encodes, phosphatidylinositol 4-kinase alpha (PI4KA), is an enzyme that speeds up biochemical reactions that help nerve cells mature and support other key activities in the body. Defects in the enzyme might explain Cameron’s movement difficulties. But with no known PI4KA-related disorder and no other patients like him that they knew of, the finding offered little more than a hunch.

From hunch to hypothesis

Relentless advocates for their son, the Coslovs applied in July 2019 to the Rare Genomes Project at the Broad. Project members work with patients and advocacy groups to help diagnose and discover genetic causes for rare disorders through genetic sequencing and analysis, and then share this information with the research and medical communities. The project was launched in 2017 by former Broad researcher Daniel MacArthur and institute member Heidi Rehm, a human geneticist who helps lead rare disease efforts at the institute and at Massachusetts General Hospital. As Rehm explained, “The project allows us to engage with patients more deeply than in other efforts, and they become partners in our research.”

Rehm also co-leads the NIH-funded Center for Mendelian Genomics at the Broad along with associate member Anne O’Donnell-Luria, an assistant professor of pediatrics at Harvard Medical School (HMS) and a faculty member at Boston Children's Hospital, and institute member Michael Talkowski, an associate professor of neurology at HMS. Beyond helping individual patients, the center aims to improve the understanding of rare, unstudied disorders, while sharing results that can benefit other families.

Cameron’s case was a familiar one to the Rare Genomes Project team, which has worked with more than 300 families affected by rare disease, with patients ranging from 2 months of age up to nearly 90 years, although most are children.

“Some of our participants are very new to this and have hit barriers with access to testing from the get-go,” said Melanie O’Leary, who is principal clinical genomics specialist, operations lead, and one of several genetic counselors on the project. “Others have been looking for a diagnosis sometimes for decades, but they couldn't get their insurer to cover genetic testing, or maybe they still remain undiagnosed after initial testing. That's where we aim to help.”

In August 2019, the Coslovs were accepted into the program. Cameron and his family members were enrolled by clinical project coordinator Brian Mangilog, one of three team members focused on family engagement. Project scientists aim to deeply examine participants’ genetic material to look for any suspicious variants, so they sequenced the entire genome of Cameron, his brother, and his father and analyzed the results while combing through his medical records. Their goal was to whittle a list of thousands of candidate variants down to a few. Everyone has several million variants across their genome, so it can be a difficult task.

The project’s analysts act as medical detectives, gathering evidence and building the case for a candidate genetic variant until they’re fairly certain it is causing a patient’s disorder. With enough evidence, the goal is to help the patient reach a diagnosis, possibly understand what’s going wrong in the body to cause the disorder, and ideally, find a researcher studying the gene or disease, or a clinical trial they can join.

“We are driven by the ability to give some people answers that can provide solace or help them plan medical care,” said Lynn Pais, a senior clinical genomic variant analyst with the project.

To investigate a suspected gene variant, the analysts consult scientific literature to see what’s known about the gene. They use computational models to predict the impact of a variant on the function of the protein encoded by that gene and evaluate whether that might lead to the patient’s symptoms. And they query resources like the Genome Aggregation Database (gnomAD) to evaluate whether the variant is common in the population, meaning it is unlikely to cause disease.

“I try to find the one variant — the needle in the haystack — that actually explains the phenotype that’s reported in their medical records,” said Stephanie DiTroia, a senior clinical genomic variant analyst at the Broad.

In some cases, the analysts locate a variant that is clearly pathogenic in a gene that’s already been linked to a disorder. In a matter of minutes, the team can make a diagnosis, which can then be passed along to the patient’s medical team after validation in a clinical laboratory.

Other cases are more challenging to diagnose. “It can take years,” Rehm said.

In cases like Cameron’s, with a few suspicious variants but no clear answer, one approach is to rely on the power of numbers — to find more people like the patient, as the Coslovs had tried to do with their website. “We try to make patterns by finding multiple people with similar-ish classes of variants in the same gene to see if they have the same phenotype, like a statistical study across humans,” said O’Donnell-Luria. Even a single unrelated, yet similar, case can be enough to take a gene from a hunch to a strong candidate, with each additional individual who is identified helping to build the case.

In the past, researchers and clinicians would act as “genetic matchmakers,” reaching out to colleagues individually or presenting their unsolved cases at scientific meetings, in hopes of finding a match. “Asking around is an incredibly inefficient way to do this, and it’s biased towards who you know and how well-connected you are,” said O’Donnell-Luria.

To turn serendipity into something more systematic, over the past decade research groups have launched various platforms not only for analyzing genomes, but also for matching cases. Rare disease databases like GeneMatcher, DECIPHER, myGene2, and the Broad’s own seqr platform allow clinicians, researchers, diagnostic labs, and sometimes even patients themselves to enter their candidate gene and attempt to find other individuals with variants in the same gene or other researchers who study it.

As these databases were starting to help generate diagnoses and new disease-related gene discoveries, Rehm and her colleagues suspected that there could be many more matches if the databases connected with each other. “What if there’s a match out there, in somebody else’s database? How do we bring the different databases together?,” she wondered.

At a scientific meeting in 2013, Rehm and her colleagues gathered leaders from each of the platforms to explore the idea of a unified network that would allow them to work together more seamlessly. In 2015, they launched Matchmaker Exchange to link these once-siloed databases.

Today, Matchmaker Exchange connects eight matchmaking platform nodes, including the Broad’s seqr database, in addition to two knowledge bases that include model organisms and scientific literature. The project was one of the first initiatives using standards and policies developed by the Global Alliance for Genomics & Health, an organization aimed at responsible sharing of genomic data to benefit human health. Matchmaker Exchange allows users to simultaneously query data from more than 150,000 individuals stored across the databases, which manage and store their own sets of data on genetic variants, traits, and symptoms.

Matchmaker Exchange participants

Matchmaker Exchange connects eight matchmaking platform nodes and two knowledge bases.

Many of the cases analyzed by the Broad’s rare disease team end up in Matchmaker Exchange, including Cameron’s. The Broad’s Rare Genomes Project researchers, in poring over Cameron’s DNA sequence and clinical information, zeroed in on variants in the PI4KA gene as strong candidates, just as his other geneticists had done earlier in their search for a diagnosis.

Hoping for more cases and clues, the team in early 2020 entered Cameron’s case into Matchmaker Exchange, through Broad’s seqr database. “At the Broad, we try to get as close as we can to a genetic diagnosis or a candidate variant, and then we use Matchmaker Exchange to connect with other researchers who are accumulating cohorts of similar patients or doing functional work that will tip the scales of evidence for that candidate,” said O’Leary. “The families who participate hope that we can come back to them with information that’s useful. Matchmaker Exchange helps us build confidence that a variant underlies their condition, so that we can report it back to the family.”

Meet the Broadbent family, who waited a year to find a match.

Not all cases in Matchmaker Exchange get matched right away – some have to wait, perhaps for years. One of these patients is five-year-old Emma Broadbent, who in 2020 received a presumptive diagnosis of a new ultra-rare disorder through the Rare Genomes Project. After a year in Matchmaker Exchange, her case was matched with another patient overseas, only the second to be diagnosed with the new illness caused by a variant in the long non-coding RNA gene known as CHASERR. Matches like this increase the level of confidence in a patient's diagnosis, and can give a family momentum to advocate for research and funding aimed at new targeted treatments.

A match for Cameron

In April 2020, after the Coslov family’s decade-long search for a diagnosis, O’Leary called them with some good news. After their analysis, the Rare Genomes Project team suspected that the PI4KA gene was a likely cause of his illness. Even better, through Matchmaker Exchange, they’d found a researcher in Spain, Aurora Pujol, who is not only an expert in the PI4KA gene, but was also studying nine other patients who carry PI4KA variants and have similar kinds of symptoms as Cameron.

O’Leary explained, however, that the match was still tentative, and asked if they’d agree to share more information about Cameron with Pujol. His symptoms weren’t exactly the same as any of the other patients — this could be due to distinct misspellings of the gene carried by each individual, but it could also mean he had a different genetic disorder. A more detailed account of his symptoms would help the researchers determine if he indeed had a form of the same disease.

Fortunately, Cameron had also been accepted and enrolled into the Undiagnosed Diseases Network (UDN) at the National Institutes of Health (NIH), a research study known well to families affected by rare disease. In summer 2020, a few weeks after O’Leary’s call, the family spent a week at the NIH in Bethesda, Maryland, where Cameron underwent an array of medical tests and imaging scans by a team of specialists, who also pored over his medical records and early genetic results. The results of the tests at the NIH clarified the physiological impacts of his illness, which helped confirm Cameron’s match with the other nine families. Although Cameron’s first geneticists and the Rare Genomes Project team had each suspected PI4KA, the match and the NIH test results provided stronger evidence implicating the gene in his illness.

“This match was the first time we found someone who was actually working on the gene and could say it was more than a hunch,” said Nick Coslov. “Without that, we’d have never figured it out.”

The step forward brought both optimism and fear for Cameron’s parents. “It’s exciting to think there might finally be something we can do about his illness,” Coslov said. “But throughout our journey, the anticipation of a diagnosis has also been scary, not knowing what we’d find out about his illness on the other side.”

The researcher in Spain, Aurora Pujol, was perhaps the ideal match for the Coslovs. As a medical geneticist, she diagnoses and treats patients with ultra-rare diseases of the brain’s white matter, and she also studies the molecular roots of the disorders as a research professor at the Catalan government research agency ICREA and at the Bellvitge Biomedical Research Institute near Barcelona. Her lab is part of the Spanish network for rare diseases, CIBERER, and also a member of the NIH’s Undiagnosed Disease Network International (UDNI). “Ultra-rare diseases have no borders,” Pujol said.

Pujol’s lab aims to find tailored treatments that work for particular genetic disorders. She’s developed tools to study the functional effects of genetic variants in cells and animals in the lab, and bioinformatics approaches to glean insights from existing data on related genes. Her studies of a rare disease called adrenoleukodystrophy led to preclinical tests that resulted in new orphan-designated drugs to treat the illness.

The more patients there are, the better is the evidence.
Aurora Pujol

Pujol found the cohort of patients with PI4KA variants through a project she launched in 2016 with the UDNI, with support from the Spanish government, to recruit 400 adults and children from across Spain with suspected, yet undiagnosed, disorders of the brain’s white matter. Among these were two adults and two children with PI4KA variants and similar symptoms, though they varied in severity. Hoping for more evidence that PI4KA caused their disorder, Pujol entered the cases into Matchmaker Exchange through the GeneMatcher platform: “The more patients there are, the better is the evidence,” she said. The search uncovered six more similar cases in Europe and the United States, including Cameron.

Pujol’s team then requested samples of skin fibroblast cells from the patients that her group could examine in the lab. They discovered that Cameron’s cells are unable to properly produce lipids that participate in cell membrane functions, such as the myelin sheath that insulates cells of the nervous system, causing nerve signals particularly in the brain and spinal cord to slow down or even stop. The finding explained Cameron’s symptoms and confirmed his diagnosis: PI4KA-related leukodystrophy, a new, ultra-rare disorder.

Pujol and her colleagues recently described the new illness in a paper in the journal Brain, a huge step forward for the families. Once a new genetic disorder is described in a publication, doctors and clinical laboratories will know to look for it, and as more patients are diagnosed, the body of knowledge will increase. The scientific community’s formal identification of the disorder also helps build momentum to raise awareness around the disease and gives patients clout to raise funds to spur more research.

“But we still need a treatment,” Pujol said. She’s a fan of drug repurposing — using an already-approved medicine to overcome the effects of a genetic variant, instead of spending many years developing a new drug from scratch. Her group learned of a cancer drug known to block the PI4KA protein, but she needed one to do the opposite. The leukodystrophy is due to so-called “loss-of-function” mutations in PI4KA, so a tailored treatment would need to restore the protein’s function, not block it.

After connecting the Coslov family with Pujol, O’Leary at the Broad continued to help the family navigate this new world of scientific research. “I think a lot of people would say, ‘Here’s a researcher, good luck’,” said Nick Coslov. “But Melanie checks in every few weeks and gives us resources. She’s been awesome.”

O’Leary learned about Pujol’s difficulty finding a drug candidate for Cameron’s disorder, and suggested to her that Broad Institute scientists might be able to help. She connected Pujol with researchers in the Broad’s Center for the Development of Therapeutics, which runs the Drug Repurposing Hub, a library of compounds including 2,500 FDA-approved drugs.

The Broad scientists are now pursuing a large-scale study of these compounds to search for a treatment for Cameron’s disorder. If feasibility tests show promise, they plan to expose Cameron’s cells to the drugs and observe if any restore function of the PI4KA protein, using assays developed in Pujol’s lab and modified for high-throughput testing. If they find a promising compound, Pujol could initiate a small clinical trial of the potential treatment on the handful of patients diagnosed with the new disorder.

“It doesn’t matter if the child is the only one in the world or has a disease shared by a thousand people,” she said. “As a physician-scientist, my strongest desire is to help as many people as I can, but I don’t mind making a tailored treatment for even just one family.”

Hopes for a better future

The Coslovs are optimistic that with the right treatment, Cameron’s symptoms and delays can be reversed or lessened. Until then, they’ll continue his therapy sessions, aimed at improving and maintaining his movement and communication abilities. They plan to start a foundation to raise money for more research, and they want to build a supportive community with the other families facing PI4KA-related leukodystrophy who are scattered around the globe.

“My hopes are that someday we can come up with a therapy that will remove as many of his obstacles and disabilities as we can,” said Nick Coslov. “We’re just hoping for something that will improve his quality of life.”

Meanwhile, three-quarters of the nearly 150,000 cases in the Matchmaker Exchange network are still unsolved and waiting for a match. With new cases added to the network every day, the chances for a match increase. Planned improvements to Matchmaker Exchange could help those unsolved cases, too. Currently, the platform only searches for those genes that a research team has flagged as suspicious, but Rehm hopes they can improve its sensitivity soon with new ways of sharing data that can allow querying of an entire genome’s worth of data to find matches, even when a researcher has not flagged any candidates yet.

The diagnostic odyssey for a patient with a rare genetic disease takes on average ten years, in part because many disorders are so poorly understood. Researchers are working hard to discover their genetic roots, but there are still several thousand without a known cause, in addition to thousands of potential disease-causing variants still to characterize. By connecting patients and researchers and by sharing data, Matchmaker Exchange gives future patients hope that they won’t have to wait so long for answers.

Video production by Scott Sassone, Russell Murachver, and Kevin Middleton, and additional reporting by Tom Ulrich, Broad Communications.