To the Editor: We read with interest the letter by Drs Strong et al.1 about our op-ed piece in the Los Angeles Times, but we stand by our statement that “for most people…whole genome sequencing is an absurd medical test.”2

We agree with Dr Strong and colleagues that some patients can benefit from whole-genome or -exome sequencing—notably, individuals with rare phenotypes of likely genetic origin for whom conventional testing has been noninformative. Recent studies confirm that exome testing increases the number of patients in this group who receive a genetic diagnosis. However, the finding needs to be put in perspective; such patients are very rare.

Perhaps more important, the value of whole-genome sequencing for this purpose is likely to be short-term because comprehensive sequencing is now serving a discovery purpose. As the genetic etiology of rare conditions is clarified, these patients will be better served by targeted gene panels based on what we are now learning, because it is always preferable to avoid generating large amounts of potentially distracting additional data when addressing a focused clinical question.3

Much the same can be said for the other promising use of whole-genome sequencing: analyzing somatic changes in cancer tissues to inform therapy. The value of comprehensive sequencing in oncology is uncertain, but now this approach also serves a discovery purpose. Ultimately, the number of genes that provide useful information for cancer therapy will be finite and, again, most patients will be better served by targeted panels.

Thus while there may be some patients for whom whole-genome sequencing continues to be valuable, the number is likely to shrink over time as our knowledge accumulates.

Yet the public discourse about whole-genome sequencing suggests something very different: that the information will have universal value as a guide to individualized health care. Our commentary in the Los Angeles Times was motivated by the misleading nature of this discourse and the dangers that flow from it.

A person’s genome is not only an ineffectual way to predict risk for most diseases but also a potential source of confusion and misdirection. First, as we noted, the noise-to-signal ratio is not merely high, it is astronomical. In short, for the general public it is a recipe for a lot of false alarms. Second, many quantitatively accurate genetic risks tend to be misleading because their effects are small relative to the contribution of a myriad of social and environmental factors.4 A recent large cohort study demonstrated this nicely in type 2 diabetes; the effect of a genetic risk profile could be measured but was trivial compared with the effect of body weight.5 In fact, the study suggested that a genetic risk profile would produce inaccurate information, underestimating diabetes risk in many overweight and obese people and overestimating it in many people with normal body weight.

Finally, we also agree with Dr Strong and colleagues that “Any medical test would be absurd for most (healthy) people at any given time.” Unfortunately, that is how this particular test is being promoted. Geneticists have a responsibility to present genomic technology to the public in a more balanced fashion. Exaggerating the benefit of whole-genome sequencing, particularly as a useful test for most people, amounts to making a promise we cannot keep. Worse, it will lead to harm if people believe it.

Disclosure

The authors declare no conflict of interest.