What are the opening hours?

Monday 9am - 5pm Wednesday 9am - 5pm Friday 9am - 5pm Saturday 9am - 5pm

Anil Bajnath, MD, 7030 Hi Tech Drive Suite 101, Hanover, MD (2026)

08/04/2026

Over the past decade, one of the most exciting ideas in cancer research has been that the trillions of bacteria living in our gut help decide how well cancer treatments work. Study after study has reported that patients who respond to immunotherapy carry particular bacteria the non-responders lack. The obvious hope followed: find the beneficial microbe, give it to everyone, and improve outcomes. A new review in Nature Reviews Microbiology suggests the story is more subtle, and in an important way more useful.

The problem is that the studies could not agree on which bacterium mattered. One pointed to Akkermansia, another to Bifidobacterium, another to Ruminococcus. For a while this looked like a field that could not replicate its own results, usually a sign something is wrong. The review reframes it. These different bacteria, the authors argue, are doing the same job. They all end up boosting the same kind of immune cell, the cytotoxic T cells that attack tumors. If many unrelated microbes can produce the same effect, then no single one is the true cause. What matters is the function they share, not their names.

This idea, called functional redundancy, changes what a useful test would even look like. Instead of asking which species a patient has, we should ask what their microbial community can do. The review shows the same principle explains a series of otherwise baffling contradictions. A gut compound called butyrate can either strengthen or weaken the anti-tumor immune response depending on the situation. A bacterium named Fusobacterium can help tumors grow in one location while helping the immune system fight them in another. Even a "good" microbe like Akkermansia is beneficial at moderate levels but harmful at high ones. The effect depends on dose, location, and the individual patient, not on the microbe alone.

The authors also review the concrete ways the microbiome influences cancer care. Some gut bacteria produce DNA-damaging toxins that can seed cancer years earlier, one of which leaves a fingerprint now tied to colorectal cancer in younger adults. Others chemically alter chemotherapy drugs, sometimes weakening them, sometimes making them more toxic. And microbiome-based treatments, including transplanting stool from responders into non-responders, have shown early promise in melanoma.

They are careful about the limits. Most studies are small, many measure bacteria in ways that distort the picture, and cancer patients differ from healthy people for many reasons unrelated to their tumors. Most of the engineered therapies remain experimental.

The takeaway is a course correction. Rather than hunting for a single magic microbe, the field should measure what the microbial community does, follow patients over time against their own baseline, and treat the human body and its microbes as one connected system. It is a harder path, but likely the one that reaches the clinic.

The full review is worth reading.

08/03/2026

How do hormones shape the developing human brain? It is one of the harder questions in biology to answer honestly, because the obvious experiment, deliberately altering hormone levels in a developing fetus, is something no one can ethically do. So most of what we know comes from animals, which differ from us in important ways, and from observing children affected by hormonal disorders after the fact. A new study in Cell Press Blue offers a third path.

The researchers used neural organoids, small three-dimensional clusters of brain tissue grown from human stem cells that reproduce key features of early brain development. Working with organoids from one male and one female donor, they chronically exposed them to activators and blockers of seven hormonal signaling systems, including the s*x hormones, stress hormones, thyroid hormone, and several others relevant to how the brain wires itself. The exposure spanned a period equivalent to the first trimester of human development. They then examined the results from every angle they could: reading gene activity in bulk and in more than 150,000 individual cells, imaging roughly five million cell nuclei, and measuring steroid hormone chemistry directly.

One hormone pathway dominated. Retinoic acid, a derivative of vitamin A already known to guide brain patterning, produced the strongest effects at every level. It pushed cells to stop dividing and mature into neurons, shifted their regional identity toward the back of the developing nervous system, and visibly changed the architecture of the tissue and even the size of cell nuclei. That consistency, the same signal showing up in gene activity, cell behavior, and physical structure, is a good sign the model is capturing something real.

The more surprising result was convergence. Hormone systems that act through entirely different receptors nonetheless ended up affecting the same underlying sets of genes, particularly those governing fat metabolism, protein quality control, and the chemical tags that switch genes on and off. In other words, several distinct hormonal disturbances seem to drain into shared downstream machinery. Notably, this overlapped with genes recently linked to the resilience of neurons from people who lived past 100.

The authors are candid about the limits. This is a foundational resource rather than a finished mechanistic study. It rests on two cell lines, one dose, and one time point, and the organoids lack blood vessels, immune cells, and the protective blood-brain barrier of a real brain. They are careful to call their model neural rather than cortical, because certain markers of the brain's outer layer were low.

Even so, the value is the map itself. If very different endocrine disruptions converge on the same developmental pathways, that gives scientists a concrete place to look when asking how environmental chemicals and hormonal disorders raise the risk of conditions like autism and ADHD.

The full paper is worth reading.

07/27/2026

One of the quieter facts of human aging is that the thymus, a small organ behind the breastbone that trains the immune system, steadily shrinks from adolescence onward. A new study in Nature Communications suggests this shrinkage may do more harm than we realized, and that reversing one small consequence of it could help older cancer patients respond to treatments that currently fail them.

The researchers began with a well-known feature of aging: older bodies carry more inflammatory immune cells. Specifically, a class of cells called myeloid cells starts pumping out inflammatory signaling molecules, and these cells accumulate inside tumors, where they appear to help cancer grow faster. In aged mice with breast tumors, disease progressed more quickly and survival was worse than in young mice. The same inflammatory pattern showed up in blood and tumor samples from older human patients.

The clever part of the study was working out what causes this. The team used two techniques to separate an immune cell from its surroundings. In one, young and old mice were surgically joined to share a blood supply; in another, immune stem cells from one age were transplanted into a mouse of the other age. The result was consistent and surprising. Old immune cells placed in a young body calmed down, and young immune cells placed in an old body turned inflammatory. In other words, it was not the age of the cell that mattered most, but the age of the body it lived in. Something circulating in young blood was keeping inflammation in check.

That something turned out to be thymulin, a small peptide made by the thymus whose levels fall with age until they essentially disappear in old animals. When the researchers gave thymulin to aged mice and to human immune cells from older donors, inflammation dropped. It worked by blocking a master inflammatory switch inside cells called NF-kB.

Then came the most striking result. A widely used class of cancer immunotherapy drugs, checkpoint inhibitors, often works poorly in older patients. In aged mice, one such drug had no effect on its own. But when the researchers combined it with thymulin, the tumors suddenly responded, growing more slowly and improving survival. Importantly, thymulin did little in young mice, suggesting it works by restoring something aging has taken away rather than by boosting a healthy immune system.

The authors are careful about the limits. This work was done in mouse breast cancer models and human cells, not in patients, and they note that the findings may not extend to every cancer type or translate cleanly from mice to people.

Still, the idea is compelling. If part of why older patients respond poorly to immunotherapy is a lost thymic signal, then replacing that signal is a concrete, testable strategy for the group of patients who most need better options.

The full paper is worth reading.

07/23/2026

Twenty years ago, a group of systems biologists proposed a new model for medicine. They called it P4: predictive, preventive, personalized, and participatory. Medicine would forecast disease before symptoms appeared, act early, tailor treatment to the individual, and treat patients as genuine partners rather than passive recipients. Two decades later, a Perspective published in the Journal of Translational Medicine makes an uncomfortable observation. The vision has mostly remained a vision.

The reason, the paper argues, is not a shortage of technology. It is a shortage of precision in how we use words.

Consider "multi-omics," a term now attached to almost any test that measures more than one thing. The paper proposes a minimum standard: at least three distinct molecular layers, measured from the same sample or the same person, with enough analytical validation to be interpreted clinically. Studies that simply combine published datasets from unrelated groups of people would not qualify. It also sorts these layers by how often they need measuring. Your genome is stable and gets sequenced once. Epigenetic and protein markers shift over months to years. Metabolites, microbiome, and wearable data change by the hour and the day, which makes them the practical readout for tracking whether an intervention is actually working.

"Personalized" receives the same scrutiny, and the definition offered is genuinely demanding. A truly personalized system must estimate where you are biologically right now, project where you are heading if nothing changes, and simulate what would happen under different interventions. By that standard, a recommendation drawn from population averages is not personalized no matter how many biomarkers preceded it.

To illustrate, the paper presents metabolomic data from 2,072 adults aged 18 to 85, sampled by dried blood spot. Nine metabolites tracked in a roughly straight line with chronological age, all statistically significant. These are not arbitrary markers. They map onto identifiable biology: mitochondrial energy production, antioxidant capacity, protein turnover, and nitrogen handling. That matters because a number you can trace back to a mechanism is a number a clinician can act on.

The limitations are stated plainly. The data are cross-sectional, meaning they compare different people at different ages rather than following the same people over time, so they cannot yet establish how any individual ages. The panel has not been replicated on another platform or in an independent group, and the exclusion of people on certain common medications may have produced a healthier sample than the general population. The findings are described as illustrative rather than clinically validated, and the blockchain prototype holds only 50 records with independent auditing still pending.

That restraint is the point. The paper also proposes a four-stage ladder for validating these systems, ending in randomized trials against standard care. In a field crowded with confident health scores and proprietary algorithms, defining the burden of proof before making claims is the harder and more valuable contribution.

The full Perspective is worth reading.

07/21/2026

There is a stubborn assumption in how we think about vascular dementia, the second most common form of dementia after Alzheimer's. We tend to treat it as a blood flow problem, and we assume that if we could just restore circulation to the brain, the cognitive decline would follow. A study by Zhao and colleagues in the Journal of Neuroinflammation complicates that picture in a useful way.

The researchers modeled chronic cerebral hypoperfusion, the reduced brain blood flow that underlies most vascular dementia, by surgically narrowing blood supply in rats. They then treated one group with nicotinamide adenine dinucleotide, better known as NAD+, a coenzyme central to how cells produce energy and manage stress. What they observed is telling. Even after cerebral blood flow had drifted back toward normal by the eighth week, the untreated animals still showed memory deficits, neuronal loss, and brain inflammation. Restoring flow was not enough. The injury had taken on a life of its own.

NAD+ changed that. Treated rats performed markedly better on a standard spatial memory test, lost fewer neurons in the cortex and hippocampus, and showed calmer immune activity in the brain. The inflammatory signals that damage neurons, molecules such as IL-1β, IL-6, and TNF-α, fell substantially.

The more interesting part is where the researchers located the root of the damage. Using electron microscopy, they looked inside microglia, the brain's resident immune cells, and found their mitochondria swollen and structurally broken. These damaged mitochondria appear to be the source of the reactive oxygen species, the corrosive molecules, that keep the inflammation burning. NAD+ repaired that mitochondrial damage and quieted the oxidative stress.

To understand why, the team traced the effect to a specific molecular pathway governed by two proteins, Sirt1 and PGC-1α, which together help maintain healthy mitochondria. Hypoperfusion suppressed both. NAD+ restored them. The most convincing experiment came last: when the researchers genetically boosted Sirt1 alone in cultured microglia, they reproduced nearly all of NAD+'s benefits, restoring mitochondrial function and lowering the oxidative burden. That is the difference between noticing that two things happen together and showing that one actually drives the other.

The authors are careful about limits. This is work in rats and in cultured cells, not in patients. One inflammatory marker, COX-2, behaved differently in living animals than in the dish, which they acknowledge openly. And they describe the pathway as only part of the story rather than the whole explanation.

Still, the implication is provocative. If the enduring damage in vascular cognitive decline is being driven from inside the mitochondria of the brain's immune cells, then protecting those mitochondria may matter as much as restoring blood flow, and possibly independent of it. For a field that has struggled to translate perfusion fixes into cognitive gains, that is a meaningful shift in where to aim.

The full paper is worth reading in detail.

07/20/2026

There is a quiet revolution happening in how we understand what makes cancer so hard to cure, and a new review in Nature Reviews Drug Discovery lays it out clearly.

The subject is something called extrachromosomal DNA, or ecDNA. To picture it, start with what you already know. Most of your genes sit on chromosomes, long threads of DNA that get copied and split evenly every time a cell divides, one full set to each daughter cell. That even split is the rulebook of inheritance. ecDNA breaks the rulebook. It is a small loop of DNA that has broken free of the chromosomes and floats separately in the nucleus. These loops were first spotted in childhood cancers sixty years ago and dismissed as oddities. We now know they are anything but.

Here is the crucial detail. A chromosome has a centromere, a molecular handle that machinery grabs to pull it cleanly into each new cell. ecDNA has no such handle. So when a cancer cell divides, its loops scatter unevenly, and some daughter cells inherit many copies while others inherit few. Because these loops often carry the very genes that drive the cancer, this random scattering lets a tumour rapidly change how many copies of a cancer gene it has, cell by cell. When a drug arrives that attacks cells with high copies, the tumour simply favours the low-copy cells, waits out the treatment, and rebuilds. It is a moving target by design.

The numbers have forced a rethink. ecDNA was once believed to appear in barely 1% of cancers. Newer genome-sequencing methods find it in roughly 14 to 17% of tumours overall, and in about 30% of neuroblastoma, an aggressive childhood cancer. A study of nearly 15,000 patients in the United Kingdom put the figure at 17.1%. Tumours carrying ecDNA tend to be more aggressive and are linked to shorter survival, and the loops often carry the most notorious cancer genes, including MYC and EGFR.

The hopeful part is that this strange biology also creates weak points. The loops tend to cluster together into little factories of gene activity, and breaking up those clusters shuts the activity down. Even more striking, different loops travel together when a cell divides. In one experiment, a drug aimed at one cancer gene also collapsed the copies of a second, currently undruggable gene, because the two loops were being inherited as a pair. That opens a genuinely new strategy: hit the cancer gene you cannot drug by targeting the one riding alongside it.

The authors are honest about the limits. No drug specifically targeting ecDNA has been approved yet, and the first candidate is only now entering clinical trials. Detecting ecDNA reliably in everyday hospital samples is still difficult and expensive, and we do not yet have the evidence to use it in routine treatment decisions.

It is a genuinely fascinating piece of science with real stakes for patients.

07/17/2026

There is a paradox at the heart of cellular cancer immunotherapy, and a new review in Nature Reviews Immunology from the Surgery Branch of the National Cancer Institute states it with unusual clarity.

The therapy in question is adoptive cell transfer: surgeons remove a fragment of a patient's tumour, grow the immune cells living inside it — tumour-infiltrating lymphocytes, or TILs — to enormous numbers in the laboratory, and infuse them back. The cells they are hunting for recognise neoantigens: mutated proteins that exist only in cancer cells and nowhere else in the body, which makes them nearly ideal targets. The approach works. In February 2024, thirty-six years after TILs were first used in patients, the first TIL therapy for a solid cancer was approved for metastatic melanoma, producing objective responses in roughly 30% of patients who had already failed checkpoint blockade.

But the review, by Sri Krishna, Paul Robbins, Frank Lowery and Steven Rosenberg, is candid about why success has been so hard to extend. The first problem is that anti-tumour immunity is almost universally present and almost universally failing. Across TILs from more than 200 patients with metastatic cancer, over 80% already carried neoantigen-specific T cells. Their tumours grew anyway. The second problem is that these targets are essentially unrepeatable between people. Among 73 melanoma patients, 133 distinct neoantigens were identified — and not one was shared by any two patients. Among 75 gastrointestinal cancer patients, all but one of 124 neoantigen reactivities was unique. Only about 2–3% of a tumour's mutations produce a peptide the immune system can genuinely see, which is why the authors argue for stricter language: an untested mutated peptide is a "candidate neopeptide," not a neoantigen.

The deepest problem is the one that gives this review its intellectual weight. The surface markers that let researchers find tumour-reactive T cells — PD-1, CD39, TIGIT — are markers of exhaustion. The cells that recognise the cancer best are the cells worn out by having recognised it for years. When the authors examined melanoma infusion products, patients achieving complete responses had not received more PD-1-positive cells; they had received more CD39-negative, CD69-negative cells, a less differentiated "stem-like" population. One patient's stem-like anti-tumour clones were still detectable more than six years after infusion. Reactivity and durability, in other words, live in different cells.

The authors are equally careful about limits. Whether tumour mutational burden predicts response to cell therapy is unresolved: positively associated in one 27-patient study, only weakly associated in an independent 64-patient cohort, and irrelevant in low-mutation epithelial cancers where responders and non-responders look identical. Tumour escape is also real — roughly 24–50% of epithelial cancers already show loss of an HLA gene copy before any immunotherapy begins, and losing that molecule renders the target invisible.

The full review is a rigorous, honest map of what we know and what we do not. It is worth reading in its entirety.

07/15/2026

When food moves through your digestive tract, it feels automatic, almost mechanical. But a new review in Communications Biology makes the case that the muscle layer responsible for that movement, called the intestinal muscularis, is far more than a motor. It is a living communication network where nerves, immune cells, and muscle constantly talk to each other, and when that conversation breaks down, digestive disorders follow.

At the center of this network are specialized immune cells called muscularis macrophages, which live permanently within the gut muscle layer. For years, macrophages were understood mainly as cells that clean up debris and fight infection. This review shows they do something much more surprising here: they actively keep the gut's own nervous system alive. The gut has its own extensive network of neurons, sometimes called the "second brain," and these macrophages provide ongoing support signals that these neurons need to survive. When researchers experimentally removed these macrophages in animal studies, the gut's neurons began to die off, and gut movement slowed down significantly.

But the relationship has a dark side. The very same macrophages that support neurons under normal conditions can turn against them during illness or surgery. After abdominal surgery, for example, these macrophages become activated and inflamed, and this actually damages the surrounding neurons, temporarily paralyzing normal gut movement, a common and uncomfortable complication called postoperative ileus. Similarly, in Crohn's disease, these macrophages infiltrate the nerve-rich areas of the gut wall and contribute directly to nerve damage and disrupted movement.

Infections reveal this dual nature even more clearly. In one type of bacterial infection, gut macrophages were shown to protect nerve cells from being lost during a future infection, essentially building a kind of protective memory into the tissue itself. In a viral infection model, on the other hand, disrupting communication between macrophages and another support cell type worsened nerve damage, showing just how carefully balanced this system needs to be.

The review also highlights the surprising role of gut bacteria in all of this. Beneficial compounds these bacteria produce, called short-chain fatty acids, appear to directly influence how well gut muscle contracts, how gut nerves function, and whether local immune cells stay calm or become inflamed. This adds to a growing body of evidence that gut bacteria don't just affect digestion in a general sense, they actively help maintain the machinery that makes digestion possible.

The authors are honest that important pieces of this puzzle remain unclear, particularly the precise, direct communication between these immune cells and the surrounding muscle cells in humans, which has been difficult to study directly.

Taken together, this review suggests that conditions like irritable bowel syndrome, inflammatory bowel disease, and even complications after surgery may be best understood not as isolated nerve problems or muscle problems, but as breakdowns in an integrated communication network. That reframing could shape how future treatments are designed.

Read the full paper here: https://doi.org/10.1038/s42003-026-10594-1

07/14/2026

Every cell in the body carries a built-in self-destruct program called apoptosis. It is not a malfunction, it is a normal, tightly controlled process that removes damaged or unneeded cells throughout life, from early development to daily tissue maintenance. When this program breaks down, the consequences run in both directions: cells that should die instead survive, a hallmark of cancer, or cells that should survive die anyway, contributing to conditions like neurodegenerative disease. A 2019 review in the International Journal of Molecular Sciences pulls together decades of research on exactly how this cellular decision gets made, and gets un-made.

The review describes two main routes into apoptosis. One starts inside the cell, triggered by internal damage like a broken piece of DNA, and runs largely through the mitochondria, the same structures known for producing a cell's energy. The other starts from outside the cell, when specific "death receptors" on the cell surface get triggered by external signals. Both paths eventually converge on a family of enzymes called caspases, which carry out the actual work of dismantling the cell in an orderly way.

What makes this review compelling is how it shows that the decision to live or die often comes down to a single molecular switch rather than a fixed sequence of steps. One of the clearest examples involves a leukemia drug called venetoclax, which works by blocking a protein called BCL2 that normally protects cancer cells from dying. Researchers discovered that a single mutation in BCL2, a change of just one amino acid, reduced the drug's binding strength 180-fold, essentially disabling the treatment. This single mutation was enough to let leukemia cells resist a drug that had been working well.

The review also highlights small regulatory molecules called microRNAs, which can push cells toward or away from apoptosis by controlling the activity of key genes. Some microRNAs found in resistant pancreatic and breast cancer cells actively suppress the cell death program, helping tumors survive chemotherapy. Understanding which microRNA is doing what, in which cancer, is opening new possibilities for treatment.

The authors are honest that some corners of this system remain poorly understood. In particular, they note that very little research attention has gone into how mitochondrial DNA itself is chemically regulated during apoptosis, even though mitochondria are so central to the whole process, and even though early evidence links this kind of regulation to diseases like ALS.

The bigger picture the review paints is that as more cancer drugs are designed to specifically trigger this cell death program, understanding its many small on-off switches, not just the general pathway, becomes essential to figuring out why some treatments stop working and how to prevent that.

Read the full paper here: https://doi.org/10.3390/ijms20174133

07/13/2026

When doctors talk about detecting cancer from a simple blood draw, the phrase making the rounds is "liquid biopsy." The idea is elegant: tumors shed tiny fragments of DNA, RNA, and other molecules into the bloodstream, and if we can find those fragments, we can catch cancer, track how it responds to treatment, or catch it coming back, all without surgery. A new review in Nature Reviews Genetics lays out both how far this technology has come and the surprising new direction it is heading next.

The core challenge the authors describe is almost mathematical. In someone with early, small-volume cancer, the fraction of tumor DNA floating among all the other DNA in a tube of blood can be lower than 1 in 10,000. That is an extraordinarily faint signal, and one of the field's most ambitious efforts to find it, a large methylation-based blood test studied in the CCGA and PATHFINDER trials, achieved very high specificity, over 99 percent, meaning it rarely gave false alarms, but caught only about 39 percent of stage I cancers, and did not meet its main goal of catching more late-stage cancers earlier.

Rather than simply trying to sequence more deeply, researchers are now pursuing two more creative paths. The first is combining multiple types of biological signal at once, not just tumor DNA mutations but also patterns in how DNA fragments break apart, which cell types released it, and information carried in RNA, mitochondria, and other circulating particles. In one pilot study, analyzing DNA fragment patterns alone correctly identified 91 percent of early-stage lung cancers out of 750 total participants. The second path is stranger and more striking: rather than just listening for whatever signal happens to be there, what if you could make tumors release more of it? In mouse studies, injecting antibodies that protect tumor DNA from being cleared from the bloodstream improved detection more than tenfold, pushing sensitivity for cancer detection from under 10 percent to over 75 percent.

This matters clinically already. In colorectal cancer, detecting tumor DNA in the blood after surgery is one of the strongest known predictors of cancer coming back, raising recurrence risk by 10 to 20-fold, and a major clinical trial recently showed that using this blood test to guide chemotherapy decisions let some patients safely avoid unnecessary treatment.

The authors are candid about real limits. No blood test that screens broadly across many cancer types has yet been approved for general clinical use, costs currently range widely from about £100 to £5,000, and techniques that intentionally boost how much tumor material enters the bloodstream carry their own safety risks that need careful study before they reach patients.

Looking ahead, the authors describe a future where blood or urine samples taken at home could be used for frequent, low-burden monitoring, potentially catching cancer recurrence or treatment side effects earlier than a hospital visit ever could.

Read the full paper here: https://doi.org/10.1038/s41576-026-00974-y

Anil Bajnath, MD

08/04/2026

08/03/2026

07/27/2026

07/23/2026

07/21/2026

07/20/2026

07/17/2026

07/15/2026

07/14/2026

07/13/2026

Address

Opening Hours

Website

Alerts

Contact The Practice

Shortcuts

Share

Category

Monday	9am - 5pm
Wednesday	9am - 5pm
Friday	9am - 5pm
Saturday	9am - 5pm