Wednesday, February 20, 2013

Amazing - a measured story on the benefits of DNA sequencing

Here's a welcome, and all too rare example of a measured story about DNA sequencing and what it can offer.  Gina Kolata writes in the New York Times about sequencing increasingly being used in attempts to explain rare, particularly pediatric disorders.  Rather than overselling the promise of treatment and cure, she writes that it isn't always a panacea, and is proving to sometimes be successful and sometimes not.

The piece is about sequencing being done for clinical purposes, primarily at Baylor College of Medicine in Houston.  The sequencing center there is increasingly busy these days with exome and whole genome sequencing, a service that is only just beginning to be affordable -- and often covered by insurance.
Demand has soared — at Baylor, for example, scientists analyzed 5 to 10 DNA sequences a month when the program started in November 2011. Now they are doing more than 130 analyses a month. At the National Institutes of Health, which handles about 300 cases a year as part of its research program, demand is so great that the program is expected to ultimately take on 800 to 900 a year. 
But, here's the dilemma.  Most people who are looking to whole genome sequencing for diagnosis are doing so because they've got condition or disease that is so rare, or so poorly understood, that it has perplexed their physicians.  Often many physicians.  They may already have been genotyped for known conditions that their symptoms suggest, but with no success. These traits are often called 'orphan' diseases for these types of reasons.

The rarity of the condition means there's unlikely to be a cure, or even treatment for the disorder, even if sequencing does find the cause.  That is because developing enough knowledge, and then some drug or other approach, takes lots of time and many cases to study in a well-controlled way.  Indeed, Kolata writes that the success rate for finding a genetic "aberration" is only about 25 - 30%, and finding a causal mutation results in improved management for only 3% of patients, and treatment and a "major benefit" for only 1%.  So, what's the point?

As an example, some diseases have clearly familial occurrences -- that is, among several close relatives, and in some of those a gene has been guessed at and a likely causative variant found in the gene.  These are rather clear-cut cases, to the point that the trait can come to be defined in terms of the gene or process in which the gene participates -- even if most patients have neither variants in the named gene, nor affected relatives.  That's a stumbling block to understanding, if the studies of the disease are restricted to these clearly-caused cases, as often occurs.  So, for many people with such diseases, genotyping won't show anything that seems relevant.  Very discouraging.

Still, there's more to human life than what's in the flesh.  For many people, it's important just to know why they or their child has this rare disorder.  And there are practical reasons as well.  As one parent told Kolata after her child's mutation was identified, “It really became definitive for my husband and me. We would need to do lifelong planning for dependent care for the rest of his life.”  And, with a definitive diagnosis, insurers are more likely to cover medical costs without question and the  patient is treated with more understanding.

Even without a diagnosis of that clear kind, there may be satisfaction of a sort in knowing that one has at least looked, and also to have become part of a data base that may, someday, be large enough and well-studied enough to lead to other discoveries (such as genetic variants found in other genes that come under suspicion).

Unlike the fly-by-dusk outfits that sell genetic risk assessments to pay for their yachts, the studies we are referring to occur in professional, clinical settings and are done by geneticists and genetic counselors, not businessmen.  These are the well-established, fully licensed and professional, legitimate clinical contexts in which this type of work should be done.  And these investigators are not just collecting data, but typically committed as their main job, to figuring out what to do about the diseases.

Why is it so hard to find causal mutations?
There are several very simple reasons why even common diseases are hard to characterize at the gene level.  First, many different genetic disruptions or modifications can give similar effects.  By analogy, there are many ways to fiddle with a car so that it goes slower than it did when it was new.  Just because it doesn't perform up to specs, doesn't mean we know what the cause is (and cars are actually a whole lot simpler in this kind of respect than people are). 

Second, most of the effects are individually very small, so that a disease is the result of multiple contributions of sub-par genes or environmental experiences.  And these first two reasons -- many different contributing factors, and each of them individually minor -- imply that each affected person may be affected for a different reason.

Third, many factors, including major ones, are rare enough in the population that we simply can't get enough instances of their mutated state in the kinds of data that are usually being collected (for example, case-control studies) to generate statistically detected evidence.  Even a major effect can be buried among a mountain of minor ones if the major isn't common enough.

Different kinds of data can certainly reveal different kinds of causes, and studies of various types are being designed.  Often, this is rather superficially rationalized to obtain funding for very large, expensive, studies.   Our science establishment, like our culture generally, believes that SuperSize must be better.

Another reason for frustration is that very rare traits simply defy many types of statistical approaches.  Often, diagnosis is not consistent, cases go unreported or mis-diagnosed, different studies can't be compared.  Family studies, which are statistically very powerful in some situations, are often hampered by such factors, or by the fact that relatives who may be deceased were not diagnosed (maybe that was simply not possible during their lives).

So there are issues of all sorts surrounding the challenge. BUT, when enough is known, and the approach is a responsible, professional one, genetic counseling can be extremely effective.

No comments: