“The first report from a big public-private project to improve genetic testing reveals it is not as rock solid as many people believe.”
By Steven Salzberg on 5/31/2015 05:00:00 AM
Your genes are not your fate.
Nonetheless, genes can tell us a lot about our risk for disease, and sometimes they can tell us how to lead healthier lives.
The landscape of the human genome is vast and mostly unexplored. There is far, far more that we don’t know than we do know. This is why it’s so important that as we test people for genetic mutations, we share information as much as possible. Genes have different effects on different people, and a mutation that is harmful in one group might be harmless in another. If we’re ever to fill in all the gaps in our knowledge, we have to pool our resources by sharing data.
This is the goal of ClinGen, a new project described in this week’s New England Journal of Medicine. ClinGen relies on a public database called ClinVar, where researchers, clinical labs, clinicians, and patients can deposit information about genetic variants and their effect on disease.
What makes this effort so complex is that many disease-causing mutations are quite rare. If two patients share the same disease and the same genetic mutation, and if we’ve never seen that mutation before, can we claim that the mutation cause the disease? Not really: we might have to wait until we’ve seen many more patients before we can have any confidence in the cause-and-effect relationship. This is why sharing data is so important: more data means more power to discriminate between real effects and coincidences.
ClinGen has already shown us that our knowledge has flaws: perhaps 17% of the disease associations that clinicians have reported might be incorrect. More precisely, the NEJM paper reports that out of more than 118,000 mutations that have a clinical affect, almost 13,000 were reported by more than one lab. Of those, 17% were interpreted inconsistently, with some labs calling mutations "pathogenic" while other labs calling them benign.
I disagree with the headline of an AP report headlined “flaws in gene testing,” which reported that
This is a straw man argument: experts in the field don't think genetic testing is “rock solid”, and the 17% rate of disagreement in disease-mutation association is not that surprising. What's more, these conflicts don’t always represent flaws, but rather holes in our knowledge that we just haven’t filled yet.
Disagreements over the effects of a genetic mutation represent opportunities to clarify and improve our genetic knowledge base. And as the ClinGen authors emphasize, it’s only by sharing our findings that we can discover and correct these flaws. This is why clinicians, scientists, and labs need to adopt a culture of sharing results, as the ClinGen project has emphasized (see their “Wall of Fame” of labs and institutions that have shared data).
All of us have mutations in our DNA; geneticists estimate that each of us carries perhaps 30 new mutations that even our own parents don’t share. The vast majority of these, fortunately, are harmless. Even mutations that do affect our health usually only have a small influence, changing our risk for disease but not actually causing it.
I'd be remiss if I didn't point out that sometimes, your genes are your fate. If you have certain mutations in the CFTR gene, you will have cystic fibrosis, and some mutations in the HTT gene will inevitably cause Huntington’s disease.
As geneticist Luke Jostins cogently wrote, identical twins usually don’t die from the same thing. So even though genes may influence our fate, we still have control over it. Knowing your genetic risks can be valuable, but we have a lot more to learn about how our genes influence our health.
By Steven Salzberg on 5/18/2015 05:09:00 AM
As spring turns into summer, we spend more time outdoors, exercising, gardening, or just walking around. And for many people, more exercise means knee pain. Count me among the afflicted.
Several people, including my orthopedic specialist, have suggested that I try injections of hyaluronic acid to treat my knee pain. Many people swear by it, and even though I looked into this two years ago (and rejected it as ineffective), I thought I would look again. Perhaps the evidence had changed.
Superficially, these injections sound reasonable. Hyaluronic acid is already inside your knee and helps to lubricate and cushion the joint. Adding more lubricant seems like a good idea; after all, it works for cars, bicycles, door hinges, or any other creaky joint.
But not for knees. Just recently, Anne Rutjes, Peter Jüni and their colleagues published a very large review of the evidence on knee injections. They looked at 89 trials involving over 12,000 adults and found that in the trials that were properly controlled (blinded), hyaluronic acid injections had either no effect or a “clinically irrelevant effect”; that is, too small a difference to matter to the patient. (They also found five unpublished trials that showed no effect at all; this is a good example of the so-called “file drawer effect”, where studies that don’t have the desired outcome are stashed away in a drawer and never published.)
Rutjes and colleagues also pointed out that there’s a real risk of harm when you inject something deep into the knee joint. To quote their summary for patients (published along with the article),
“Viscosupplementation [injection of hyaluronic acid] may provide little if no pain relief or function improvement in patients with knee osteoarthritis. It also seems to increase the risk for adverse events."
If you don’t believe Rutjes, perhaps you’ll believe the American Academy of Orthopaedic Surgeons (AAOS), who reviewed a large body of evidence and wrote:
“We cannot recommend using hyaluronic acid for patients with symptomatic osteoarthritis of the knee. Strength of recommendation: Strong.”
After looking at numerous studies, the AAOS concluded that there was no clinically meaningful benefit from these injections. This document is their official recommendation for physicians.
I was disturbed, though, to find that the AAOS information for patients directly contradicts their own recommendations to clinicians, and suggests that knee injections might work. On their patient-oriented web page, they offer the wishy-washy comment that “hyaluronic acid does seem to have anti-inflammatory and pain-relieving properties … [that] may last for several months,” though “not all patients will have relief of pain.”
What the heck? Are they orthopedists trying to hide the truth from their patients? Why would they do that?
I checked to see if the AAOS patient information was merely out of date, but no, it was reviewed in 2014, and the clinical recommendations appeared in 2013. I have to assume that the authors of the patient information page are not the same doctors who established the official recommendations for clinicians–but it seems unlikely that they are ignorant of their own organizations official guidelines. So why don’t they warn patients that hyaluronic acid injections don’t work? Is it because hyaluronic acid treatments are a significant money-maker for some orthopedists?
I can hear the response from some doctors already: "In my experience," they say, "this works for some patients." Sorry, but one doctor's experience (or "clinical judgment", as some call it) doesn't trump science. That's why we do experiments, to determine whether or not our subjective impression is correct. In this case, the science is clear.
Incidentally, the Mayo Clinic also fails on this topic. Their site states that hyaluronic acid
“can be injected into your knee to improve mobility and ease pain. Relief may last as long as six months to a year.”
They cite no evidence to back this up. Patients looking at either the Mayo site or the AAOS site will be misled into thinking that these injections might be worth a try.
WebMD does a much better job, writing that
“one study published in Rheumatology found that hyaluronan was no better at reducing joint pain than a placebo injection of salt water. An analysis of seven different studies published in the Journal of Family Practice did not recommend the treatment, since the benefits–if it had any at all–were so slight.”
Note that unlike the Mayo Clinic page, the WebMD page mentions specific studies and names the journals.
So what does work for knee pain? Physical therapy and exercise to strengthen the muscles around the knee is still the best course, according to the AAOS's clinical guide. Weight loss also helps, by reducing the load on the knee. The only medical intervention that helps is an NSAID pain reliever such as ibuprofen. (See my 2013 article for other treatments that don’t work, including acupuncture (not even close) and glucosamine-condroitin supplements).
As for me, I bought a knee brace even though the AAOS says it doesn’t help. But it feels like it’s helping, and it was cheap. And my knee is definitely feeling better. Why not take advantage of a placebo effect once in a while?
By Steven Salzberg on 5/07/2015 03:56:00 AM
We’re still spending vast amounts of time and money trying to counter the ill effects of a discredited, retracted paper from 1998 that claimed to find a link between the MMR (measles, mumps, and rubella) vaccine and autism. Even after the The Lancet retracted the study, and even after the British Medical Council revoked the medical license of its lead author, Andrew Wakefield, many people continue to withhold vaccines from their children because of a fear that somehow, despite all the evidence to the contrary, vaccines might cause autism. Vaccines, I hasten to add, have saved millions of lives and are probably the greatest medical advance of the past two centuries.
Now another study has appeared to add more weight to the evidence about the safely of the MMR vaccine. The new study by Anjali Jain and colleagues, just published in the Journal of the American Medical Association, looked at a huge number of children–95,727–for evidence of any link between autism and the MMR vaccine.
The results were not surprising, to those who have been following the science. To quote the conclusions directly:
“Receipt of the MMR vaccine was not associated with increased risk of ASD [autism spectrum disorder], regardless of whether older siblings had ASD. These findings indicate no harmful association between MMR vaccine receipt and ASD even among children already at higher risk for ASD.”
That should settle it, right? But then, dozens of previous studies should have already settled this question. Unfortunately, due to the ongoing activism of anti-vaccine groups such as Age of Autism, (who already attacked this new study) and to conspiracy theorists such as Robert F. Kennedy Jr. (whom I wrote about last summer, and who was campaigning against vaccines in Vermont just last week), misguided claims that vaccines cause autism or neurological problems persist.
Here are the numbers from the new study. The authors compared vaccinated children to unvaccinated children, using a huge database of medical claims that included at least 5 years of followup. (This was an "observational" study, by necessity–it would be unethical to withhold vaccines from children on purpose.) The relative risk for autism in children who had 2 doses of the MMR vaccine (the recommended amount) compared to unvaccinated children was 0.74. In other words, a child was somewhat less likely to be diagnosed with autism if he or she were vaccinated.
Even more surprising was the relative risk among children who had an older sibling with autism: in this smaller group, children with 2 doses of MMR were just 44% as likely to be diagnosed with autism as unvaccinated children. This statistically significant finding indicates, unexpectedly, that vaccines might actually protect children from autism.
The authors were quick to note that there are other good reasons for this apparent protective effect of vaccines: in particular, if parents of autistic children withheld vaccines from their younger children, this could explain the effect. Why? Because we know that autism has a genetic component, and that if one child has autism, his younger sibling is more likely (because they share many genes) to have autism as well. Jain and colleagues explained that if these parents withheld vaccines–because of fears spread by the anti-vaccine movement–then their children could contribute to the apparently lower rate of autism in children who were vaccinated. The authors couldn’t rule out a protective effect of vaccines, but scientifically it seems unlikely, and they wisely offered an alternative explanation.
So: once again we have a large, carefully conducted study showing that the MMR vaccine does not cause autism, and even finding evidence that vaccinated children have lower rates of autism. Let's hope this study helps to end the anti-vax movement, so that we can soon stop spending time and money trying to refute their long-discredited hypotheses and instead focus on trying to understand the true cause.