The drug, dubbed STX209, is further along than any other treatment for an autism-related disorder. If it passes the next, more rigorous testing phase, it could be in clinics within a few years.

At a meeting on Saturday of the National Fragile X Foundation, Seaside Therapeutics, a small biotech in Cambridge, Massachusetts, announced that in 15 children who had both fragile X syndrome and severe social impairments, STX209 significantly improved scores on several tests of social behaviors.

The drug didn’t work on everyone in the trial, but for some, it was remarkable. According to the lead investigators, some participants improved so much that they stopped taking other medications, such as antidepressants and antipsychotics, which can have nasty side effects. The researchers are indefinitely continuing an ‘open-label’ extension study, in which the participants knowingly take the drug while the company collects data on safety and long-term effects.

In 2002, Mark Bear, one of Seaside’s founders, famously showed that mouse models of fragile X syndrome have excessive activity of the chemical messenger glutamate. Accordingly, STX209 stimulates gamma-amino butyric acid type B receptors, which effectively dampen glutamate signaling. Most of the other drugs in development for fragile X either also stimulate pathways that inhibit glutamate, or directly block glutamate receptors.

Although STX209 showed a trend of improving irritability and social problems in the wider group of 54 children and adults with the syndrome, the results were not statistically significant.

Seaside is also running a preliminary open-label study to test whether STX209 improves irritability in 30 children with autism. About one in three of individuals with fragile X syndrome also have autism.

It’s unlikely that drugs developed for fragile X will improve the full range of symptoms on the autism spectrum, but given the current dearth of treatments, even small gains are reason enough for good cheer.

No one has pinpointed genes or pathways that cause the condition and, partly because it can be triggered by ordinary stressors, it’s difficult to diagnose. Its biology now becomes even more baffling with the release of two new mouse models of compulsive behaviors, each implicating a different type of brain cell.

Three years ago, Guoping Feng‘s team created mice that compulsively groom themselves by deleting the SAPAP3 gene. SAPAP3 makes a protein expressed exclusively at neuron connections in the striatum, a deep region that’s important for planning movements.

Circuits in the striatum are also highlighted in one of the new studies, which appeared in May in Nature Medicine. By knocking out part of SLITRK5, which encodes a synaptic protein found in the striatum, researchers created mice whose intense self-grooming leads to severe facial lesions.

The second new report looked at mice carrying mutations in the HOXB8 gene. Scientists first noticed in 2002 that these animals feverishly groom themselves and their littermates, but didn’t know why. In the 28 May issue of Cell, they reported that HOXB8 is expressed only in microglia, immune cells that originate in the bone marrow and then migrate to many regions across the brain.

Although the two studies finger very different systems, they might begin to explain how and why OCD overlaps with other psychiatric illnesses, such as autism. For instance, some people with autism have movement problems, or abnormally big striata.

There are also many genetic and neurobiological links between the immune system and autism. Most relevant, a study presented at a meeting last year found that postmortem brain samples from individuals with autism have large numbers of microglia.

HIV has a remarkable ability to resist antiviral drugs and hide in the body, so the idea of eradicating the virus seemed impossible. Suggesting otherwise, researchers feared, could create false hope and complacency.

In the past few years, however, there are increasingly loud whispers about a cure for HIV. The year 2007 saw the clinical debut of integrase inhibitors, which prevent HIV from inserting into the host genome. The following year, a bone-marrow transplant eliminated the virus from the body of an infected German man. Last year, breakthroughs in cell-culture techniques allowed researchers to screen for drugs that can lure HIV from its hiding places.

“I was very pessimistic five years ago, but we had to try,” says Warner Greene, director of the Gladstone Institute of Virology and Immunology in San Francisco. “And as we’ve tried, I’ve become much more optimistic that we might be able to achieve a drug-free remission.”

His optimism is understandably tinged with caution, however, as scientists promising eradication were proven wrong once before.

In the summer of 1996, researchers attending the eleventh international AIDS meeting in Vancouver trumpeted data showing that certain combinations of antiretroviral drugs can suppress HIV to undetectable levels in the blood.

The buzz grew over the course of the following year. In May 1997, David Ho’s group at the Aaron Diamond AIDS Research Institute in New York reported in Nature that HIV levels in eight people dropped by two orders of magnitude within ten days of receiving a particular three-drug combination and were undetectable within eight weeks.

Using a mathematical model, the researchers estimated that, barring any complications, the drugs could eradicate the virus from an infected person in less than three years.

If that sounded too good to be true, it was. In the same issue of the journal, Bob Siliciano’s group from Johns Hopkins University inspected the small number of dormant immune cells that harbour HIV. Because these cells do not replicate, they are impervious to treatment. Once activated, however, his team found that these cells can start pumping out the virus into the blood and lymph nodes.

Two other groups described these so-called latent reservoirs in 1997, and two years later Siliciano’s team estimated that it would take about 60 years for drugs to flush out HIV from these stores. “It was a real blow, I think, to people who were interested in eradication,” Siliciano says.

The field instead shifted focus to preventing resistance by using combinations of three or more drugs — dubbed highly active antiretroviral therapy (HAART) — and to decreasing the side effects of treatment.

These efforts were hugely successful: there are 32 approved HIV drugs, and at least a dozen more in the pipeline, which, together, can suppress the virus for decades.

“Now that HAART works so well,” Siliciano says, “we’re turning to the next step: can we actually cure anybody?”

…read the rest of my latest at Nature

But some insurance companies have resisted paying for the newer tests, called chromosomal microarrays, because they are more expensive than older techniques. This delay in technological uptake could be keeping many children from receiving crucial early treatment for their conditions. Now, an expert group is calling on large medical associations to adopt microarrays as the preferred genetic tests for children with unexplained autism, developmental delays or other birth defects.

Traditional karyotyping techniques are still the best choice for conditions such as Down’s syndrome that are caused by gross chromosomal abnormalities and are easily recognized by clinicians. But most developmental disorders show a range of symptoms and can arise from more subtle genetic glitches, such as microscopic DNA deletions or duplications. That’s where the much more sensitive microarrays come in.

…read the rest of my latest in Nature Medicine

Topol has launched a company, the West Wireless Health Institute, that is working on real stethoscopes and related technologies — using iPhones and other wireless devices. Here’s what he told the San Diego Union Tribune:

This technology is already here, introduced in February 2010, by GE (a device called “Vscan”). I use this for all the patients I see, at no cost except for initially purchasing the pocket echo device, and have been able to markedly reduce the number of full echoes that are needed (which cost more than $1,500, take 40 to 45 minutes, and require another appointment to be set up to get the test). Each year in the United States, over 8 million heart echoes are done at a cost of well over $10 billion. If we can cut that at least 10 (percent) to 20 percent, it has enormous potential.

As far as taking one’s vital signs, this is right around the corner. Sensors on the wrist can be used to get blood pressure, heart rate, oxygen concentration in the blood, breathing rate, temperature, and this will ultimately be displayed on the cell phone. Very exciting for people who need this, and potentially worrisome for inducing “e-hypochondriacs” for those who don’t.

The sensors will undoubtedly play an enormous role in the years ahead, since they can measure virtually anything that makes us tick, anytime, anywhere, continuously — it is just a matter of using these appropriately, validating the improvements, and making sure they can reduce the costs of health care.

As researchers reported this week at a cancer conference in Chicago, a drug (PLX4032) that had worked wonders on reversing mutations in melanoma had no effect on the same mutation in colon cancer.

The results are frustrating, of course, but also fascinating, as Andrew Pollack reports nicely in the NYTimes:

Enthusiasts for the targeted drug have been saying for years that tumors will eventually be characterized by their molecular profiles — which mutated genes they have — rather than where in the body they occur. Names like breast cancer and lung cancer will be supplanted by terms like B-RAF-positive or EGFR-positive tumors. And drugs will be chosen based on that profile, the way antibiotics are generally selected based on the pathogen that is causing the infection, not on where in the body the infection occurs.

Massachusetts General Hospital, for instance, is running a clinical trial testing a drug from AstraZeneca on any type of cancer — providing it has a mutation in the gene B-RAF, the same gene that is the target of PLX4032.

But the test of PLX4032 in colon cancer suggests that the location of the tumor still does matter, that it will not be just a case of looking at the target. There are other examples as well. Erbitux and Vectibix do not work in colon cancer patients with a mutation in a gene called K-RAS. But the relationship between the mutation and the effectiveness of Erbitux does not seem to hold in lung cancer.

Experts say that in certain cancers a genetic mutation might be present but is not really driving the cancer, as it might be in other types of cancer.

“You’ve got to target a specific disease, not a biomarker,” said Dr. Mark J. Ratain, an oncologist at the University of Chicago.

Last summer, I wrote about scientists trying to figure out how to use data gleaned from implanted electrodes in order to predict the onset of epileptic seizures.

Unfortunately, long-term use of those metal electrodes has major drawbacks: irritation, inflammation and scarring. Now comes a new kind of bio-compatible electrode, made of silk! From Technology Review:

This week in the journal Nature Materials, the team reports using a silk electrode device to measure electrical activity from the surface of the brain in cats. Silk is mechanically strong–that means the films can be rolled up and inserted through a small hole in the skull–yet can dissolve into harmless biomolecules over time. When it’s placed on brain tissue and wetted with saline, a silk film will shrink-wrap around the surface of the brain, bringing electrodes with it into the wrinkles of the tissue. Conventional surface electrode arrays can’t reach these crevices, which make up a large amount of the brain’s surface area.

“A device like this would completely open up new avenues in all of neuroscience and clinical applications,” says Gerwin Schalk, a researcher at the Wadsworth Center in Albany, NY, who is not affiliated with the silk electrode group. “What I foresee is placing a silk-based device all around the brain and getting a continuous image of brain function for weeks, months, or years, at high spatial and temporal resolution.”

Apparently though, real doctors aren’t so enamored. A group recently published a study of “bioethicical and professionalism content” of House and Grey’s Anatomy. From Nature Medicine‘s blog:

Their detailed study concluded that the physicians in these fictional dramas followed guidelines for informed consent in 43% of medical cases. Moreover, when patients refused treatment, the doctors respected their decision in about half the cases. Here’s the real catch, though: The study found that interpersonal relationships, both among the doctors and with their patients, were ‘exemplary’ a mere 5% of the time.

I recommend reading the whole post — they’ve got some great clips of “egregious examples”. But I think the authors of the study should, well, take a chill pill. And lay off my TV!

David Weatherall, a professor at the, ahem, Weatherall Institute of Molecular Medicine at the University of Oxford, agrees with me. He points out that genetic science has (among other things): found the cause of many single-gene disorders, such as cystic fibrosis and Huntington’s; uncovered how cholesterol metabolism works, leading to new drug treatments; and has shown us how acquired mutations (from, say, cigarette smoke) can cause cancer.

On the other side is James Le Fanu, a general practitioner and book author from London. He agrees that now, 10 years since figuring out the sequence of the human genome, geneticists generate “billions of bytes of biological data each week” and have published “fascinating insights”. He just doesn’t think they matter much in the clinic: “And yet for all this cornucopia of new facts and knowledge, its influence on everyday medical practice remains scarcely detectable.”

He gives two possible explanations for this slow start: A) genetics may not be all that important to disease; and B) we don’t understand even the basic function of a gene, let alone how they work together to create disease.

I have two objections to Le Fanu’s argument. The first is that 10 years is not a long time! Yes, this is extremely complicated science, which is why I often marvel at just how much has been learned about our genome in only 10 years.

The second problem is that Le Fanu made a negative argument, rather than a positive one. He says we shouldn’t bother spending money on genetics to understand medicine. OK, so what should we study, then? He ends his commentary with this doozy: “The current dominance of medical genetics threatens to bury the true spirit of intellectual inquiry under an avalanche of undigested (and indigestible) facts.” On the contrary, I think what he’s advocating — giving up on a young and extremely productive field — is a much better example of “burying intellectual inquiry.” Well, since we don’t understand a lot of this stuff, let’s not even bother trying to.

Of course genetics is not everything, and of course, the more we find out, the more that remains a mystery. But isn’t that what science is all about?

Credit: Emily Singer, Technology Review

Most people have heard of the robots that allow surgeons to see inside the body in 3D, and make super-precise and super-small movements, all without cutting more than a 1-inch incision. (Check out the website of the da Vinci robotic system, for instance.)

But, as Emily Singer reports in an interesting Technology Review feature (with slide show, and video!), a lot of surgeons are disillusioned with the flashy, clunky and expensive technology:

The group had varying concerns–if and when the robot will outperform traditional laparoscopy; the learning curve associated with the technology; whether it allows less experienced surgeons to perform more complex surgeries. But everyone agreed on two points. The technology isn’t advancing fast enough or dropping in price quickly enough. “The system is very expensive because only one company makes it now,” says [Hiep] Nguyen. “We need more competition to drive down price.”

da Vinci, apparently, is the only game in town. Sounds like a good project for industrious biomedical engineers!