Use of DNA biomarkers in personalised medicine

It is well recognised that the use of biomarkers is becoming an increasingly important component of clinical medicine in general, and there is no doubt that this will have important specific future applications in Sport and Exercise Medicine. The use of DNA biomarkers in personalised medicine, in general, has recently been reviewed.9 I would like to highlight a few important points from this review, and apply these to genetic biomarkers for ERI's, in particular Achilles tendinopathy as one example.

First, it is important to note that biomarkers that are associated with injuries could fall into one of a number of categories, including biomarkers that (1) relate to the diagnosis or severity of injury—diagnostic biomarker, (2) discriminate between non-injured individuals and those with early asymptomatic injury—screening biomarker, (3) predict the likely course of the injury in a defined clinical population under standard treatment conditions—prognostic biomarker or (4) forecast the likely response to treatment (efficacy and safety)—predictive biomarker. Therefore, before a clinician even considers ordering an investigation to test the patient for the presence of a genetic biomarker, there should be a strong evidence for its specific use in any of the aforementioned defined categories. Researchers in this area should also carefully consider and identify the exact category of the biomarker that is investigated and then pursue a rigorous, systematic, investigative approach through a series of carefully planned phases of studies.

The phasic process of investigating biomarkers, before they can be applied clinically in personalised medicine, was also briefly discussed in this review.9 For example, diagnostic and prognostic biomarkers need to be studied in a number of phases starting with discovery, followed by assay development and validation, retrospective validation and refinement, prospective investigation, randomised trials and finally health economics studies. Similarly, there are guidelines for study designs to investigate predictive biomarkers for clinical application. It is clear that, before any genetic biomarker can be used confidently in clinical medicine, rigorous scientific investigation is required, involving multiple centres with large (usually in excess of 1000) sample populations. The reviewers conclude that a priori planning of research strategies is vital for the identification and validation of biomarkers for use in personalised medicine.9 As clinicians, we should take careful note of these guidelines as they form the basis in evaluating the scientific evidence that is required before genetic biomarkers should be used in clinical practice for the prevention, diagnosis and/or management of ERI's.

Is there currently strong enough evidence that clinicians should assess genetic biomarkers in patients with an ERI such as Achilles tendinopathy?

Most studies investigating the relationship between an injury and possible genetic biomarkers have been genetic association studies. While it is not the purpose of this editorial to review the strengths and limitations of this study methodology, it should be pointed out that there are several reviews and guidelines to assess the quality and credibility of these genetic association studies.10–13 This is important, as the strength of evidence for any genetic biomarker for the evaluation of an ERI should be tested using these guidelines. As one example, the evidence for genetic biomarkers in lumbar disc degeneration has recently been reviewed.14 In this systematic review, over 1300 research publications were initially identified; after rigorous quality assessment and data synthesis, 52 genetic association studies were included in the analysis. The result showed that none of the genetic variations studied reached a strong level of evidence and only five candidate genes were identified with a moderate level of epidemiological evidence. The authors therefore concluded that, although multiple genetic polymorphisms appear to be associated with lumbar disc degeneration, the level of the association remains weak.14 This review stresses the significant limitations of the current status of genetic association studies in this injury.

The potential application of genomics in the prevention and management of Achilles tendinopathy has also recently been reviewed.15 In this review, the authors highlight several important limitations in the application of research data in this field and they indicate that further research is required to address these limitations. In contrast to the body of knowledge that, for example, exists for genetic association studies in lumbar disc degeneration, there are still very few studies (referenced above) that constitute the main body of knowledge on genetic biomarkers for Achilles tendinopathy. More importantly, in the case of Achilles tendinopathy, most of the studies are from a single laboratory that mainly used the same initial cohort of injured and control individuals. In all these studies, it is also acknowledged that the sample size of this cohort and control group is small, particularly when compared with the majority of other genetic association studies (eg, lumbar disc degeneration cohorts). In addition, there is also very weak evidence for replication of results by other researchers. This association was only investigated in one other cohort, with an even smaller sample size. The strength of these studies, and their methodology, should also be tested using established guidelines for reporting genetic association studies (STREGA statement).13 Furthermore, many more studies in different populations and significantly larger cohorts are needed so that cumulative evidence guidelines can be applied.11 ,16 There is also a need to provide strong ‘biological evidence’11—evidence as to the specific function of a variant or associated gene, which can make it a plausible candidate for association with an injury. For example, one of the most promising genetic associations with Achilles tendinopathy are the variants within the COL5A1 3′-UTR. However, until recently, the function of the COL5A1 3′-UTR was not known. In only a single study (five cases and five controls), sequence variants in the 3′-UTR of the COL5A1 gene have been shown to alter mRNA stability, and this may have implications for Achilles tendinopathy.17 Therefore, there is limited ‘biological evidence’ for genetic biomarkers in Achilles tendinopathy, and this also requires further investigation. Finally, as mentioned previously, a potential genetic biomarker should first be defined (diagnostic, screening, prognostic or predictive) and then studied in phases to accumulate sufficient evidence for its possible role in personalised medicine.9

Should a clinician currently be testing for genetic biomarkers related to injuries such as Achilles tendinopathy?

In my opinion, testing for genetic biomarkers in clinical care of ERI is premature. Although research for genetic biomarkers and the possible incorporation of these into multifactorial injury prediction and treatment models is very important, current evidence that genetic biomarkers are related to ERIs is still weak and requires further study. Once a sufficient body of knowledge (including large multiple prospective cohort studies, studies in independent laboratories, genome-wide association studies, use of rigorous methodology based on established guidelines, studies on ‘biological evidence’) has been accumulated, can data be subjected to a rigorous, independent, epidemiological, biological and clinical evaluation using established guidelines and criteria. A final step before regular use of any genetic biomarker would also be to consider the cost implications of the test (cost-efficacy and cost-benefit). Only then can clinicians become confident in applying tests of genetic biomarkers to improve injury prevention, diagnosis, treatment and prognosis.

Future perspectives for clinicians

Personalised medicine and the inclusion of genetic biomarkers in the diagnostic, therapeutic and prognostic armamentarium of the clinician is an emerging reality. With this, however, comes a responsibility to always act in the best interests of the patient by making sure that (1) the test has a clear indication; (2) the test is for a specific purpose for which there is sound scientific evidence; (3) the result of the test will significantly alter the course of clinical decision making, which is to the benefit of the patient and (4) the cost is justified. I would strongly caution Sport and Exercise Medicine clinicians against current testing for genetic biomarkers for ERI, unless this forms part of a well-planned scientific research programme. This does mean that there are significant opportunities for clinician researchers to further the understanding in this field by forming research collaborations with colleagues working in the field of genetics and molecular biology. I will certainly keep this as a focus of my clinical and research attention over the next few years so that I can continuously improve the clinical service to my patients.