LAUSR

BACKGROUND Current clinical phenotyping of musculoskeletal pain provides very limited evidence- based support to personalized medicine. This paper discusses the potential of somatosensory phenotyping to contribute to personalized medicine for prognosis and prediction of treatment effects. METHODS Highlight of definitions and regulatory requirements for phenotypes and biomarkers. Appraisal of the literature on somatosensory phenotyping in musculoskeletal pain. RESULTS Somatosensory phenotyping can identify clinical conditions and manifestations that may affect treatment decisions. However, studies have shown inconsistent associations of phenotyping measures with clinical outcomes, and the strength of association is mostly weak. Most somatosensory measures have been developed for research, are too demanding to find large acceptance in clinical settings, and have uncertain clinical usefulness. CONCLUSIONS Current somatosensory measures will unlikely be validated as strong prognostic or predictive biomarkers. However, they still have the potential to support personalized medicine. Including somatosensory measures in biomarker signatures, that is, a set of measures that are collectively associated with outcomes, is potentially more useful than aiming to the identification of single biomarkers. Furthermore, somatosensory phenotyping may be introduced as part of patient's evaluation to contribute to better-informed and personalized treatment decisions. To this purpose, a change in the way research currently approaches somatosensory phenotyping is warranted. A pathway is proposed that involves: (1) the identification of clinically applicable measures that are specific to clinical conditions; (2) the association of somatosensory phenotypes with outcomes; (3) multi-site replication; and (4) the determination of clinical benefits in randomized controlled trials. SIGNIFICANCE Somatosensory phenotyping has the potential to support personalized medicine. However, current measures do not seem to meet the criteria for being strong prognostic or predictive biomarkers, most of them are too demanding to find large acceptance in clinical settings, and their clinical usefulness has not been proven. The value of somatosensory phenotyping can be more realistically determined by re-orienting research to the development of simplified testing protocols, applicable to large-scale clinical practice, and tested for clinical usefulness in randomized controlled trials.