Tag: Medication

Posted on

Glioblastoma: Testosterone Supplements Linked to 38% Lower Risk of Death

Researchers at Cleveland Clinic have discovered that androgen hormones such as testosterone can limit the growth of glioblastoma tumors in men. Results published today in Nature show that men receiving testosterone supplements for reasons unrelated to cancer showed a 38% lower risk of death compared to patients not taking these supplements. 

These findings are surprising because testosterone is known to contribute to the growth of other forms of cancer in men, such as prostate cancer, where hormone therapy is used routinely to decrease levels of androgen hormones and block cancer progression. However, these hormones were found to play a very different role in glioblastoma, an aggressive form of brain cancer that is more commonly diagnosed in men. 

“This outcome is a welcome surprise and may potentially offer a lead for new treatments for a kind of cancer that is deadlier in men,” said Anthony Letai, MD, PhD, director of NIH’s National Cancer Institute (NCI).  

In a mouse model of glioblastoma the researchers found that reducing levels of androgen hormones induced overdrive on the hypothalamus-pituitary-adrenal (HPA) axis, a brain circuit that controls reactions to stress and many physiological processes including hormone secretion. This caused a spike in stress hormones that led the brain to reinforce the protective function of the blood-brain barrier and create an immunosuppressive environment in the brain, reducing the ability of immune cells to fight against the tumor. 

“The brain has evolved to keep stuff out and that includes immune cells from elsewhere in the body. It’s a delicate tissue that often doesn’t want huge immune reactions,” said Justin D. Lathia, PhD, professor of cancer sciences and scientific director of the Brain Tumor Center at Cleveland Clinic.

Importantly, this effect was only observed in male mice. In females, changes in testosterone levels did not produce the same effects.

These findings were then confirmed in human samples obtained from 1,300 men with glioblastoma participating in the NIH database Surveillance, Epidemiology, and End Results (SEER). An analysis showed that men who received supplemental testosterone for reasons unrelated to their glioblastoma diagnosis had a 38% lower risk of death than other male patients. 

More research will be needed to better understand the complex pathway activated by testosterone and other androgen hormones. While the current study identified inflammation in the hypothalamus as a potential trigger of HPA axis activation, future work will look for the exact mechanism glioblastoma tumors employ to induce this reaction from an entirely different region of the brain.  

Lathia noted that, although these results do not establish a causal link between testosterone and patient outcomes for men diagnosed with glioblastoma, the study opens the door for future clinical trials that dive deeper into the link between androgen hormones and glioblastoma tumor growth. He added, “An obvious follow-up study would be to find out whether androgen deprivation, which is a common treatment for cancer, is actually detrimental for glioblastoma.” 

 

The post Glioblastoma: Testosterone Supplements Linked to 38% Lower Risk of Death appeared first on Inside Precision Medicine.

Posted on

Blood Test Reads Cell Spatial Environment to Predict Immunotherapy Response

A new study published in Nature has uncovered a universal map of the tumor microenvironment shared across 17 different cancer types—and, crucially, demonstrated that its features can be detected from a blood test to predict whether patients will respond to immunotherapy.

The work, led by Aaron Newman, PhD, at Stanford University and Aadel Chaudhuri, MD, PhD, of the Mayo Clinic describes nine distinct “spatial ecotypes”—stereotyped communities of immune and structural cells that organize themselves in consistent patterns around tumors, regardless of the cancer type. The researchers then developed an AI-based algorithm capable of inferring the proportions of those ecotypes from circulating tumor DNA in plasma, translating a tissue-level biological map into a minimally invasive liquid biopsy readout.

A universal architecture

The researchers studied more than 100 tumor specimens across 10 cancer types, using their tools to map gene expression patterns across nine cell types at varying locations throughout the tumor. They identified nine distinct spatial ecotypes—neighborhoods roughly the diameter of a human hair—and found that these patterns were conserved across all tumors studied. Some ecotypes clustered at the border between tumor and healthy tissue; others appeared deeper within the tumor mass. Several of the ecotypes correlated with immunotherapy response, pointing to potential clinical utility in guiding treatment decisions.

The central discovery is that tumors are not randomly organized. Across carcinomas of the breast and prostate, melanoma, and more than a dozen other cancer types, the same nine spatial ecosystems appear—each with a distinct cellular composition, gene expression signature, and physical location relative to the tumor mass.

“Cells exist not in isolation, but in little communities or neighborhoods or social networks,” said Newman. “These characteristics were highly stereotypical, highly conserved across different carcinomas and melanomas. Each spatial ecosystem has its own common features that differ from other spatial ecosystems—you can think of them as the parts list that most tumors are made of when we think about the soil that nurtures the cancer cells.”  And, each spatial ecotype has its own internal social network—the cellular behaviors, or genetic programs it is carrying out, are influenced by the cells around it.

Chaudhuri, a clinician scientist, offered a grounding analogy. “Whether you’re in Tokyo or Minneapolis or San Francisco, you have a downtown, and as you go farther out, you have Midtown, suburbs, and then rural farmland. That’s how it is for the tumor microenvironment as well—close to the tumor cells, you have spatial ecotypes seven, eight, and nine; go farther away and you have spatial ecotypes one, two, three, and four. That map is consistent across all these different cancer types.”

Predicting immunotherapy response

The clinical implications center on immune checkpoint inhibitors, a class of drugs that has transformed oncology but still fails a significant portion of patients. Analyzing 15 immunotherapy cohorts in tissue, the team identified two spatial ecotypes—SE7 and SE8—characterized by T cells positioned at and within the tumor, associated with a pro-inflammatory, anti-tumor immune response, and strongly linked to immunotherapy benefit. Conversely, a fourth ecotype, SE4, associated with wound-healing biology, was strongly predictive of resistance.

“Higher levels of SE7 and SE8 strongly forecast the patient’s response to immune checkpoint inhibitors,” said Newman. “And conversely, higher levels of SE4 portend a worse outcome—resistance to immune checkpoint inhibitors.”

Those tissue-based findings were then validated in blood. Using a deconvolution algorithm the Newman lab developed—the same class of computational approach used to identify cell-type proportions from bulk sequencing data—the team showed that plasma samples taken before treatment could recapitulate the spatial ecotype signals found in paired tumor biopsies, and that those liquid-biopsy-derived ecotype levels predicted patient outcomes with striking fidelity.

“Without looking at their tumor biopsy, just taking a blood sample, we could determine the levels of the liquid spatial ecotypes from their tumors directly from the blood,” Newman said. “Those levels very strongly predicted their outcomes once they received immunotherapy.”

Beyond the single biopsy

Perhaps the most consequential aspect of the platform is what it enables over time. Today, a patient’s tumor microenvironment can be assessed only at the moment of biopsy—a single, static snapshot that may be compromised by sampling bias and cannot be repeated without invasive procedures.

“There is no existing clinical assay to access these features of the cancer over time,” Newman noted. “For the first time, this gives us those insights.”

The team has already begun longitudinal work, with funded studies planned to track melanoma patients serially through immunotherapy—monitoring how spatial ecotype levels shift on treatment and whether those dynamics outperform existing tools like CT imaging, which can struggle to distinguish true progression from pseudo-progression.

Chaudhuri emphasized the broader paradigm shift. “For years and years, we have been completely tunnel-visioned on malignant tumor cells and their mutations, measuring their mutations as surrogates for MRD and things like that,” he said. “For the first time ever, we’ve opened our eyes and we can look at the tumor microenvironment, which is absolutely critical for precision oncology.”

From lab to clinic

The technology has been licensed to a startup company, Liquid Cell Dx, co-founded by Newman and Chaudhuri, with the goal of developing and validating a clinical-grade assay. The team has already presented blinded clinical validation data at AACR, applying cut points learned from the Nature paper cohort to an entirely independent patient population at a separate institution—without knowledge of outcomes until unblinding.

Active translational programs are now underway in non-small cell lung cancer, muscle-invasive bladder cancer, and mesothelioma, examining spatial ecotypes in the context of chemotherapy combinations, neoadjuvant settings, and radiotherapy.

Newman was recently awarded an AACR Trailblazer Award for the serial monitoring work in melanoma.

“The goal here is for this not to just end at the Nature paper, for it to benefit humanity,” said Chaudhuri. “We want to take the technology, get it into patients, develop it into a clinical assay, and really show that we can bring next-level precision into oncology.”

 

The post Blood Test Reads Cell Spatial Environment to Predict Immunotherapy Response appeared first on Inside Precision Medicine.

Posted on

Pancreatic Cancer Shares Genetic Drivers with Obesity and Diabetes

Researchers at the University of Birmingham have found that the same genes are active in pancreatic cancer, obesity, and diabetes. Their findings, published in Cancer Medicine, could finally provide an explanation to why metabolic disease is a major risk factor for pancreatic cancer. 

“We know that people with obesity or diabetes tend to have worse outcomes from pancreatic cancer, but the biological reasons have not been clear,” says Animesh Acharjee, PhD, associate professor of integrative analytics and AI at the University of Birmingham and senior author of the study. “Our study shows that the same genes and inflammatory pathways are active in both metabolic disease and pancreatic cancer, which helps explain this link and points to new opportunities for identifying high‑risk patients and developing more targeted treatments.”

Treatment options are currently limited for patients diagnosed with pancreatic cancer, a form of cancer that is often diagnosed at advanced stages. Only about 15% of patients are eligible for surgery, and about 80% of them relapse after treatment. 

Previously, the researchers had identified a series of genes that were consistently altered in metastatic pancreatic tumors. In the current study, they examined whether these same genes also play a role in metabolic disorders such as obesity and diabetes, which are increasingly recognized as risk factors for pancreatic cancer.  

First, the team analyzed genetic data from publicly available datasets to study how six key drivers of pancreatic cancer behave in healthy individuals compared to people with obesity. These included the ITGAM, PECAM1, CCL5, STAT1, STAT2, and CD44 genes, which are involved in inflammation, immune cell recruitment, and lipid metabolism processes. All six genes were found to be upregulated in individuals with obesity. 

Single-cell RNA sequencing of patient tumor samples revealed that a subset of immune cells, including macrophages and monocytes found within the tumor microenvironment, expressed these core six genes at higher levels than other cells. This discovery suggests this group of cells may be key drivers of tumor progression and recurrence and a potential therapeutic target for the development of targeted therapies.  

Taken together, these findings indicate that immune and inflammatory pathways that drive metabolic disease also play a major role in pancreatic cancer, where they could be involved in immune evasion and recurrence after surgery. Future work will investigate whether modulating the activity of these genes could reduce the chronic inflammation and immune dysregulation that drive the recurrence of pancreatic cancer to improve the success rate of this procedure. Targeting these pathways could offer new therapeutic strategies to manage pancreatic cancer, especially in patients with underlying metabolic conditions.

“This study highlights how chronic inflammation and metabolic dysfunction can intersect with cancer biology,” says Simon Jones, PhD, professor in musculoskeletal aging at the University of Birmingham and team lead for the NIHR Biomedical Research Centre. “Understanding these shared mechanisms is essential if we are to improve outcomes for patients who are living with multiple long‑term conditions alongside cancer.”

 

The post Pancreatic Cancer Shares Genetic Drivers with Obesity and Diabetes appeared first on Inside Precision Medicine.

Posted on

Comparing Usual Care With Coordinated Clinician and Patient Use of Mobile Technology in Primary Care for Patients With Major Depressive Disorder: Practice-Based Pilot Study

Background: Major depressive disorder (MDD) affects millions of Americans each year and is often diagnosed and treated in primary care. Evidence shows that self-management techniques, shared decision-making (SDM), and goal setting are effective strategies for managing MDD, but the required collaboration between patients and primary care clinicians can be difficult. Primary Care Path is a program for supporting depression management in primary care that includes a patient-facing mobile app and an accompanying care team–facing web interface. Leveraging programs that provide clinician-facing software with companion patient-facing mobile technology may help patients and physicians align depression treatment and management goals, support effective SDM, alleviate barriers, and improve both clinical care and patient outcomes. Objective: To pilot-test the use of Primary Care Path for MDD management in primary care and evaluate the impact of its use on depression treatment, symptoms, goal setting and attainment, and SDM. Methods: Four primary care clinical practices in the United States were assigned to program use (2 practices; intervention) versus usual care (2 practices; control). Intervention practices used the Primary Care Path program in their clinics and engaged patient participants in app use for 18 weeks. Clinical care teams engaged with the patient-informed program portal primarily during patient encounters (in-person, virtual or calls). Patient participants were smartphone users aged 18 years and older who were being treated for MDD. Patient participants received online surveys (medication changes, Patient Health Questionnaire-9 [PHQ-9], goal setting and attainment questions, and Shared Decision-Making Questionnaire-9 [SDM-Q-9]) at baseline, 6, 12, and 18 weeks. Results: A total of 76 patient participants (34 intervention; 42 control) were enrolled; the majority were female (27/34, 79%; 32/42, 76%), White (31/34, 91%; 40/42, 95%), non-Hispanic/Latino/a (29/34, 85%; 40/40, 100%), and employed (26/34, 77%; 34/42, 81%). Control patient participants’ conversations with their medical providers increased over the study period, while intervention patient conversations with their medical providers decreased over time. At week 18, intervention participants felt more successful than control in achieving their personalized treatment goals. More intervention patient participants initiated antidepressant medication by weeks 12 (=.03) and 18 (=.04) and switched medications by weeks 6 (=.009) and 12 (=.04) versus control. All patient participants demonstrated significant improvement in PHQ-9 scores throughout the study period (<.001), with no difference in change by group. Clinicians and patients indicated using the program to support SDM, but no significant differences were observed in SDM-Q-9 between intervention and control. Conclusions: Preliminarily, the use of this digital health program related to earlier medication optimization, earlier conversations between patients and medical providers, and patient attainment of goals that matter most to them, indicating that coordinated use of the program by both patients and clinical team members may enhance MDD management in primary care clinical settings.
Posted on

Advanced Cardiovascular-Kidney-Metabolic Syndrome Linked to Increased Cancer Risk

Researchers from Japan are calling for increased cross-disciplinary collaboration after showing that people with advanced cardiovascular-kidney-metabolic (CKM) syndrome are at increased risk for cancer.

They explain in Circulation: Population Health and Outcomes that CKM syndrome is a conceptual framework, proposed by the American Heart Association (AHA) in 2023, that captures the interconnected nature of cardiovascular, kidney, and metabolic diseases and reflects the shared risk factors and pathophysiological mechanisms of these diseases.

The current study findings suggest that this framework could also “serve as a valuable, non-invasive stratification tool in precision oncology and preventive medicine,” said Hidehiro Kaneko, MD, PhD, the study’s lead author and associate professor in the department of cardiovascular medicine at the University of Tokyo in Japan.

He told Inside Precision Medicine: “By identifying individuals with advanced CKM (particularly stages 3 and 4), clinicians could tailor and potentially intensify cancer screening protocols for these high-risk patients. This enables more personalized surveillance and early detection strategies that bridge the gap between cardiometabolic management and cancer prevention.”

According to AHA statistics, nearly nine out of 10 adults in the United States have at least one component of CKM syndrome, which includes high blood pressure, abnormal cholesterol, diabetes, obesity, and reduced kidney function.

Although these components of CKM syndrome have each been associated with an increased risk for certain cancers, the relationship between CKM stage and the risk for incident cancer is unclear, as current knowledge is largely derived from studies of individual components rather than the integrated syndrome.

To address this, Kaneko and team analyzed administrative claims data for more than 1.3 million people living in Japan. Of these, 12.5% had CKM stage 0, 9.8% had stage 1, 31.7% had stage 2, 36.3% had stage 3, and 9.8% had stage 4.

The researchers report that, over a median follow-up of 3.4 years, increasing baseline CKM stages were associated with significantly greater cancer incidence.

Specifically, the incidence of cancer was 81.2 cases per 10,000 person–years in people with CKM stage 0, increasing to 97.2, 105.1, 250.9, and 257.7 cases per 10,000 person–years in those with CKM stages 1, 2, 3, and 4, respectively.

After adjustment for age, sex, alcohol consumption, and physical inactivity, individuals with CKM stage 1 or 2 at baseline did not have a significantly increased risk for cancer relative to those with CKM stage 0. However, people with CKM stages 3 and 4 had significant 25% and 30% higher risks for cancer, respectively, than those with stage 0.

When the team analyzed the data by cancer type, they found that the incidence of colorectal, stomach, lung, renal pelvis and ureter, pancreatic, non-Hodgkin lymphoma, bladder, liver, kidney, thyroid, leukemia, and gallbladder cancers increased progressively with higher baseline CKM stages. There was a similar pattern for prostate cancer in men and for breast, cervical, and uterine cancers in women.

Conversely, there was no clear association between baseline CKM stage and the incidence of esophageal cancer, malignant melanoma, or Hodgkin lymphoma.

The associations between CKM stage 3 or 4 and cancer risk were stronger in men than in women and in people younger than 65 years of age relative to older individuals. However, people aged 65 years and older were also at increased risk for cancer even when they had CKM stage 1 or 2 relative to stage 0, whereas younger individuals were not.

Kaneko said that the study highlights a critical need for increased awareness of the link between CKM syndrome and cancer.

“While physicians and the public generally understand that interconnected metabolic and kidney conditions lead to heart disease and stroke, the integrated CKM syndrome is rarely viewed as a significant driver of cancer,” he remarked. “Our study demonstrates a clear, stage-dependent increase in incident cancer risk as CKM progresses, underscoring the need to recognize cancer as a major potential consequence of this multisystem syndrome.”

Kaneko suggested that “this can be addressed by shifting public health messaging to emphasize that proactive lifestyle modifications—such as weight management, a healthy diet, and regular exercise—provide dual protection against both cardiovascular events and cancer.”

He added: “Within the medical community, we should promote the CKM staging framework as a comprehensive health assessment tool, encouraging cross-disciplinary collaboration among cardiologists, nephrologists, endocrinologists, and oncologists to manage these overlapping risks holistically.”

The post Advanced Cardiovascular-Kidney-Metabolic Syndrome Linked to Increased Cancer Risk appeared first on Inside Precision Medicine.

Posted on

Researchers’ Spinout Focuses on Simplifying Viral Vector Purification

U.S.-based researchers have developed a portfolio of peptide ligands for purifying viral vectors for gene therapies and have launched a company that develops affinity technologies for biopharmaceutical manufacturers.

ChromaGenix, a spinout from North Carolina State University (NC State), commercializes peptide ligands as an alternative to the traditional protein ligands used in affinity chromatography.

According to Stefano Menegatti, PhD, CSO at ChromaGenix and a professor in the Department of Chemical and Biomolecular Engineering at NC State, the ligands are cheaper and less likely to trigger an immune response in patients than the protein ligands traditionally used.

“Protein ligands have been a fantastic enabler of advanced biological therapies over the past two decades, but they do have shortcomings,” he explains. “Proteins can denature or degrade, potentially releasing immunogenic fragments, which poses a certain level of risk.” As such, Menegatti says, protein ligands have a short lifetime and must be replaced frequently, adding further to production costs. They can also bind too strongly to the product, making it harder to recover it from the chromatographic step.

By contrast, peptide ligands, which are very small proteins, overcome these issues, he says. They can be produced synthetically, making them cheaper than proteins. As they don’t have a complex structure, they can be cleaned under harsh conditions without becoming inactive. They have a longer lifespan and a much lower immunogenicity risk, Menegatti says. Also, as they’re small, he explains,they can be cleared during final product filtration.

This “represents a new frontier of gene therapy manufacturing, says Menegatti, “as it boosts the efficiency of the viral vector manufacturing pipeline.”

Having developed peptide ligands for a wide variety of viral vectors, ChromaGenix is already selling to many companies, Menegatti says. The researchers and the company are now hoping to move beyond gene therapies.

“Our next chapter is going to be developing ligands for the purification of therapeutic cells, starting with CAR [chimeric antigen receptor] T cells,” he says.

The post Researchers’ Spinout Focuses on Simplifying Viral Vector Purification appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

Matthew Rabinowitz: Engineering a New Era of Diagnosis

Jonathan D. Grinstein, PhD, North American Editor of Inside Precision Medicine, hosts a new series called Behind the Breakthroughs that features the people shaping the future of medicine. With each episode, Jonathan gives listeners access to his guests’ motivational tales and visions for this emerging, game-changing field.

Matthew Rabinowitz switched from engineering and computational research to medicine after a breakthrough on the Human Genome Project. He realized that telecommunications, aerospace, and machine learning technologies could help him understand human biology. His shift in focus was influenced not only by scientific interest but also by personal loss, including the deaths of family members affected by genetic conditions. These experiences convinced him that current diagnostic methods were inadequate, especially for patients and families in critical situations.

After founding Natera, Rabinowitz and his team developed Panorama Prenatal Test, a noninvasive prenatal test. This technology uses DNA variant analysis, Bayesian statistical methods, and machine learning to detect genetic conditions in cell-free fetal DNA in maternal blood samples. It increased accuracy, accessibility, and reduced invasive procedures. Myome, his new project, uses whole genome sequencing to find rare diseases. Myome uses AI models to assess cancer and cardiovascular disease risks using genomic and clinical data to improve early detection and prevention.

In this episode, Rabinowitz discusses regulatory constraints, fragmented data systems, and difficulties translating complex genetic information into clinical decisions. In the long run, he wants to create blood test diagnostics that can predict health and allow proactive medical intervention.

Rabinowitz uses several technical engineering and computational concepts, including:

  • Packet Switching: a method of dividing data into smaller units that are transmitted independently and reassembled at their destination
  • Transformer Model: a type of artificial intelligence system that processes entire datasets simultaneously to identify relationships between elements, widely used in modern AI systems such as GPT
  • Gradient Descent: an iterative method used in machine learning to minimize error by adjusting model parameters

This interview has been edited for length and clarity.

 

IPM: What originally drew you into applying engineering and machine learning to genetics and clinical diagnostics?

Rabinowitz: There was all this incredible work happening in the early 2000s around the Human Genome Project, along with applications of signal processing and machine learning, which is what I focused on during my electrical engineering training.

There were really three catalysts for me.

One was in 2003, when my sister gave birth to a child with Down syndrome at one of the top hospitals in the country, and they didn’t know until he was born. I spent six days flying around trying to help. They went through one procedure after another and after six days, the baby died from complications. It was absolutely horrific.

Second, I couldn’t believe that we had all these advanced technologies in our phones, laptops, and spacecraft, but they hadn’t made their way into clinical diagnostics. At that point, I felt I had to apply my background in signal processing and early machine learning to these problems.

The third reason was about 15 years ago, I lost a child due to a genetic condition, an absolutely devastating experience. After going through that, I felt there was a path I needed to follow.

The engineer in me took over. It felt like a problem I had to solve. It was like being struck twice: unrelated events, but the same kind of tragedy.

That’s when we used a pregnancy sample to apply for NIH funding to improve prenatal testing. We got that grant, then several others, and ultimately built Panorama, which has transformed pregnancy care globally.

From there, one thing led to another. Now, through Myome and companies like Natera, we’re working on projects that could save the U.S. healthcare system around $200 billion per year. It’s been a very personal mission. 

 

IPM: What are you seeing today with whole genome analysis that feels fundamentally new or different?

Rabinowitz: We’re now diagnosing conditions with whole genome analysis that simply weren’t detectable before. Myome has largely led the charge.

When I look at these case studies today, I get the same feeling I had 20 years ago. How were we not able to see this before?

We’ve spent a lot of time extracting signal from noise so you don’t need multiple sequential tests. You can start with the whole genome and layer analyses. This includes SNPs, CNVs, difficult deletions, tandem repeats, mitochondrial DNA, and methylation.

One example: an eight-year-old with developmental delay, autism, and hypotonia had already undergone exome sequencing with no findings. We identified a subtle deletion about one kilobase involving a single exon too large for short-read breakpoints and too small for coverage changes. That finding completely changed the child’s life.

Another example: a man in his mid-20s with dystonia, convulsions, and vomiting had undergone standard neuromuscular panels. They missed a tandem repeat very difficult to detect with short-read or exome sequencing. We developed new statistical methods and identified the breakpoints, which changed his life.

More broadly, rare disease costs in the U.S. are about $1 trillion annually with ~47 physicians involved over a 4–7 year diagnostic journey and massive lost productivity. The fact that we can now catch these cases is remarkable.

On the pregnancy side the belief was that the issue was solvable, but the technologies were limited. People at the time used shotgun sequencing and looked at DNA quantity. We instead analyzed SNPs between individuals.

We built a massively multiplexed PCR system ~20,000 primers in a single reaction. The challenge is noise, cross reactions, and primer dimers. We developed a machine learning optimization so every primer is tuned relative to every other, standardizing thermodynamics across the system.

From there, we built a statistical framework integrating across trillions of hypotheses, crossover events, noise, and fetal fraction. When it converges, you see a clear maximum likelihood peak that tells you what’s happening. If not, you know something is wrong.

This allowed us to detect things others couldn’t: very high sensitivity for aneuploidy and structural variants like microdeletions.

We could detect triploidy and vanishing twins, determine zygosity, and even de novo mutations, which are more than five times as common as Down syndrome. It was a completely new approach combining passion, engineering, and statistics.

 

IPM: Why can’t we have one universal test that does everything?

Rabinowitz: The short answer is that there’s so much more we can extract from each sample, especially with AI.

These transformer models trained to predict the next word require learning enormous context. With gradient descent, backpropagation, and large datasets, the performance is extraordinary. We didn’t fully appreciate the significance early on but today the possibilities are enormous.

That said, you can’t have one universal test. First, sample context matters. In pregnancy, you’re analyzing fetal cell-free DNA very different from adult disease testing. Second, it’s not just blood. There are many analytes. Beyond DNA, you need methylation, RNA, proteins. Most diseases require a multi-analyte approach. Third, regulation. You need to validate each test rigorously. You can’t validate everything at once across the genome. We also have variants of unknown significance. If you look for everything, interpretation becomes a problem.

So you have to focus your inquiry and ensure results are validated and actionable. That said, from a single blood draw, we can already do an incredible amount.

 

IPM: How has cfDNA and noninvasive testing evolved with AI?

Rabinowitz: Around 2017–2018, Natera began applying deep learning to diagnostics. [It was] one of the first large-scale uses in genetic testing. We had used neural networks earlier (e.g., in HIV mutation analysis) but this was different.

We applied convolutional neural networks to detect microdeletions in low fractions of cell-free DNA. A key example is 22q11.2 deletion syndrome. This occurs in about one in every 1,500 to 2,000 pregnancies. It’s more common than many screened conditions. Early detection allows intervention at birth. We initially used classical statistics, but after generating millions of samples, we trained deep learning models. The AI learned noise patterns and edge cases better than we could model, like Kasparov versus Deep Blue.

In a study of ~20,000 patients, we saw 100% sensitivity for larger deletions and ~83% for smaller ones, with a specificity of 99.95%. That translated to a positive predictive value (PPV) of ~53%, compared to 3–5% clinicians are used to.

Despite this, adoption has been slow due to reimbursement and guidelines, which is frustrating, because many children still miss early diagnosis.

 

IPM: Looking forward, how is AI transforming broader healthcare and genomics?

Rabinowitz: Today, we’re combining whole genome sequencing with AI and clinical data to predict disease risk far more accurately than even five years ago.

Across ~30 major diseases we can now predict susceptibility at a transformative level. If applied broadly, for example to people over 45, we could save over $200 billion annually by catching diseases earlier. Many interventions are simple like diet and lifestyle. Even small improvements matter. Every 1% increase in sensitivity can mean ~$7 billion in savings.

We’re also predicting neoantigens for personalized cancer vaccines, training on real patient outcomes—something that wasn’t possible before. And we’re building foundational genomic models, like language models, that learn the structure of the genome itself.

So across diagnostics, treatment, and prevention, AI is fundamentally transforming the field.

 

IPM: You mentioned earlier that you underestimated neural networks. What changed your perspective?

Rabinowitz: Around 2005, we were applying machine learning to genetics. We weren’t wrong, but we underestimated neural networks. We worked on HIV drug resistance. predicting which mutations respond to which therapies.

We used lasso regression, support vector machines, [and] carefully constrained models. Neural networks didn’t perform as well, which is what we expected. Our mindset was to control complexity to avoid overfitting. What we didn’t anticipate was massive data, stochastic training, and compute power, which allowed neural networks to escape local minima and scale. 

In 2010, I had a patent on training neural networks with memory. It lapsed because Stanford didn’t maintain it. That was right before Google Brain scaled these approaches. The lesson is to stay open-minded. Technology can open possibilities you don’t see coming.

 

IPM: How do you see the future of diagnostics evolving from a single blood draw?

Rabinowitz: We’re moving toward a world where a single blood draw can tell us an enormous amount. Historically, progress was slow: blood cell counting in the 1800s, automation in the mid-1900s, cell-free DNA in the 1990s. Since then, progress has been explosive. From one sample, we can identify incidental findings (e.g., rare diseases, pharmacogenomics, and predictive risk) across many conditions. We can also detect cancer noninvasively through circulating DNA.

The capabilities are remarkable, but the genome is complex—three billion bases, with interactions that require enormous data to model.

 

IPM: What challenges remain in making these technologies widely usable?

Rabinowitz: Two main challenges. First, data, standardizing and aggregating clinical data across institutions, is historically very fragmented. AI is helping, but more coordination is needed.

Second, education. As we generate more information, explaining it to doctors and patients becomes harder. What’s known, what’s uncertain, and what action to take. At Myome and Natera, we invest heavily in genetic counseling. But across the field, there’s a tendency to simplify by withholding information. That won’t scale. We need better ways to communicate complexity responsibly.

 

IPM: How important is diversity and multi-ethnic data in building accurate models?

Rabinowitz: It’s absolutely critical and still underserved. Many models were trained on homogeneous populations, limiting accuracy. We’ve focused on building multi-ethnic models using diverse datasets and functional genomics, but we still need more data from underrepresented populations.

For example, in cardiovascular disease, we built a multi-ethnic model using large datasets.

We were able to reclassify ~50% of patients in the intermediate-risk category, identifying who is truly high risk (>20%) versus low risk (<5%). That improved decision-making significantly with over 10% improvement in classification. When followed over time, outcomes matched predictions closely.

This has huge implications for individuals and for healthcare systems. By identifying risk earlier and intervening, often with simple lifestyle changes, we can reduce costs and improve outcomes at scale. That’s why building diverse, high-quality datasets is so important. It’s one of the most powerful ways to improve healthcare globally.

 

The post Matthew Rabinowitz: Engineering a New Era of Diagnosis appeared first on Inside Precision Medicine.

Posted on
<![CDATA[Explore how neuromelanin-sensitive MRI noninvasively tracks long-term dopamine and noradrenaline system changes.]]>
Posted on

Autonomous pathology research using agentic AI shows potential in oncology

Nature Medicine, Published online: 05 May 2026; doi:10.1038/s41591-026-04403-9

The agentic artificial intelligence tool SPARK is able to reproduce pathology-based reasoning and produce biological hypotheses and relevant diagnostic, prognostic and predictive cellular parameters. The output of SPARK has the potential to advance the understanding of tumor biology and enable the development of diagnostic, prognostic and predictive tools for pathology and oncology.