Personalised care and the genome

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e3174 (Published 03 May 2012) Cite this as: BMJ 2012;344:e3174
  1. Megan Doerr, genetic specialist,
  2. Charis Eng, Hardis chair
  1. 1Genomic Medicine Institute, Cleveland Clinic, Cleveland, OH 44195, USA
  1. engc{at}ccf.org

Using family history to target genomic investigation may be the way forward

One of the universal goals of medicine is to give each patient the most appropriate care. To this end we constantly seek new ways to tailor care. Nothing is more personal than a person’s genetic make-up. Recently, Roberts and colleagues sought to determine the predictive capacity of whole genome sequencing for 24 common diseases using mathematical modelling based on twin studies, and they concluded that this technique does not provide a useful risk assessment for the most common diseases.1 As many have pointed out, neither twin studies nor mathematical modelling are new approaches to assessing heritability, but this study’s findings struck a chord with both the popular and the scientific press, with real time comment from CNN Health (http://thechart.blogs.cnn.com/), Time (www.time.com/time/), Nature NewsBlog (http://blogs.nature.com/news/), and the BMJ.2

Whole genome sequencing—the complete reading of the roughly 3.2 billion base pairs of DNA that makes each individual unique—captures the imagination of scientists and the public alike. Today, little more than a decade since the human genome was first sequenced, the cost of whole genome sequencing is a fraction of what it was, and the richness of the data has increased 100-fold. With this decrease in cost comes increasing excitement that this technology will be the crystal ball we seek, identifying predisposition and improving prevention for common diseases. But the question remains, how is whole genome analysis useful in the personalisation of care?

For patients with symptoms, genomic approaches to assessment and management are already personalising care. Whole exome sequencing—the sequencing of all of the protein coding regions of the genome (about 1% of the total genome)—has been particularly valuable in the identification of rare genetic variants with high functional impact. Exome sequencing overcomes limitations associated with earlier gene hunting techniques, including the challenges of linkage analysis and locus heterogeneity. Discoveries that have come from exome sequencing include the identification of the genes responsible for Freeman Sheldon syndrome, Schinzel-Giedion syndrome, and Kabuki make-up syndrome.3 4 Currently, the characterisation of cancer genomes by whole exome sequencing is improving cancer care, identifying mutations, and elucidating new targets for both existing and new treatments. Collaborative approaches, including the development of large genomic data sharing groups, such as the International Cancer Genome Consortium, will further expedite the translation of genomic discoveries from bench to bedside to improve the personalisation of care for patients with cancer.5

When we physicians are lucky we get a chance to prevent disease in a well patient. It is not just clinicians who are interested in keeping their patients healthy; there is demand in the public domain for individually tailored advice on disease prevention.6 Various forms of genomic testing are now available for the healthy patient, including personal genomic screening and newly affordable whole genome sequencing. How effective is this application of genomic technology?

Personal genomic screening, also known as commercial genomic screening or direct to consumer genetic testing, is a form of genomic profiling offered by commercial companies directly to the public. Currently, most of these tests use single nucleotide polymorphism (SNP) analysis for risk assessment. SNP analysis compares the frequency of common genetic variants, often discovered by genome-wide association studies, between populations to measure susceptibility to common multifactorial disease. Unlike genomic querying techniques, which look for uncommon genetic change with exceedingly high predictive or diagnostic value (or both), SNPs are common variations and most have a very modest effect size, with a median odds ratio of 1.33.7

The clinical usefulness of personal genomic screening is low. Comparisons of the technique with family history, the current gold standard for clinical risk assessment, have found alarming discrepancies.8 When family histories were used to assess the risk of cancer, 22 individuals were found to be at hereditary risk. By contrast, personal genome scanning found only one of these 22 to be at high risk of cancer. Furthermore, case examples of egregious misinterpretation of results are common.9 For this reason, many professional societies have published position statements and policies on the use of personal genomic screening, and legislation has been passed in several European countries to regulate its use.10 11 A recent excellent review found insufficient evidence to conclude that genomic profiles are useful in measuring genetic risk for common diseases or in developing personalised diet and lifestyle recommendations for disease prevention.12

Although it is important to keep our eyes on the prize, premature implementation of new technology rarely succeeds and often confounds. Instead, a phased approach should be instituted. Firstly, healthcare providers and consumers should be encouraged to take advantage of existing clinical genetics applications. The current practice of genetics very ably uses family health history and single gene or oligogene testing for disease risk assessment to guide and change management decisions for many conditions.13 14 15

At the heart of the conclusions drawn from Roberts and colleagues’ findings—that whole genome sequencing is not useful for clinical prediction—may lie a desire to use a “one or the other” approach (existing genetic clinical standards or whole genome sequencing and mathematical modelling). However, using the clinical context and family health history to target the genomic regions to be analysed in the data provided by whole genome sequencing would be a powerful next step. For example, if there is a family history of colon cancer and diabetes, then all the germane genes and regions should be analysed. This approach combines clinical context with focused genomic information. Pre-test genetic counselling will include “contracting” with (informing) the patient of this focused strategy. With more research, more clinical information on variants of unknown importance will accumulate. Logistical and ethical questions include the duty to inform the patient of such advances and whose duty it is to inform.

For whole genome screening to deliver on its promise, not only must terabytes of data be processed and stored, but equally enormous amounts of phenotypic data, information on family history, and behavioural data must be collected to put genomic discoveries into the correct clinical context. Clearly, consumer participation in both research and cutting edge clinical genomics is important. The next step in harnessing the power of the genome for public health will be to develop collaborative projects that are clinically useful for the healthy patient.


Cite this as: BMJ 2012;344:e3174


  • Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.