The European ME/CFS Biomarker Landscape project: an initiative of the European network EUROMENE

Myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS) is a common and severe disease with a considerable social and economic impact. So far, the etiology is not known, and neither a diagnostic marker nor licensed treatments are available yet. The EUROMENE network of European researchers and clinicians aims to promote cooperation and advance research on ME/CFS. To improve diagnosis and facilitate the analysis of clinical trials surrogate markers are urgently needed. As a first step for developing such biomarkers for clinical use a database of active biomarker research in Europe was established called the ME/CFS EUROMENE Biomarker Landscape project and the results are presented in this review. Further we suggest strategies to improve biomarker development and encourage researchers to take these into consideration for designing and reporting biomarker studies.


Biomarker in ME/CFS
Although the exact pathogenesis of myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS) is still unknown, the most plausible hypothesis is that it is a complex multifactorial syndrome in which immunological and environmental factors play a crucial role. In addition, the severe fatigue, post-exertional malaise, cognitive impairment, and autonomic dysfunction that delineate the disease point to the involvement of both the nervous system as well as metabolic disturbances [1]. Infection by various pathogens, including herpes viruses and enteroviruses, but also intracellular bacteria, are known as triggers of disease. The complex clinical picture and the disagreement on potential pathomechanisms make ME/CFS a controversial entity and compel the research for disease biomarkers that could aid in the diagnostic and clinical management. Biomarker per definition may include both markers with a certain sensitivity and specificity for diagnosing ME/CFS as well as those which may allow to classify subtypes of the disease, be of value as indicators of prognosis, and to be predictive for response to treatment [2].

The EUROMENE ME/CFS Biomarker Landscape project
EUROMENE is a network of researchers and clinicians from 17 European countries and one COST (Cooperation in Science and Technology) near neighbor country on ME/CFS supported by the European COST program within Horizon 2020 (http://www.cost.eu/ COST_Actions/ca/CA15111).
The aims of EUROMENE are to foster strategies for collaboration and harmonization of diagnosis and research, and to compile an inventory of clinical and scientific data in ME/CFS. The Biomarker working group will also try to develop guidelines for the usage of biomarkers and synchronization of biomarker research.
As a first step, a database for active biomarker research in Europe was established called the EUROMENE ME/CFS Biomarker Landscape project. To achieve this, EUROMENE members performed a search for publications on biomarkers within their countries. The search strategy used the medical subject headings (MeSH) term "chronic fatigue syndrome", which includes myalgic encephalomyelitis, and the respective country, and selected all publications from the last 5 years (2012-2016). The searches were reviewed by members of the Fig. 1 Biomarker studies were categorized as metabolic, immunological, neurological or infection-associated. The data was visualized as total numbers of studies (size of cake) per category (piece of cake) from each country, and the numbers of active biomarker research groups is indicated in the countries. EUROMENE countries are indicated by grey (dark grey countries with published studies, light grey those without studies) and non-EUROMENE by white Scheibenbogen et al. J Transl Med (2017) 15:162 biomarker working group. Studies not involving patients with ME/CFS, non-biomarker, and sole treatment studies were excluded, only one review article was included.
A total number of 39 studies were identified. Studies were categorized as being immunological, infectionrelated, metabolic or neurological. We summarize the findings in Fig. 1, which shows the number and type of studies identified in each country, represented by pie charts-their sizes being proportional to the number of identified studies, and their pieces representing the distinct categories of the studies. The number of research groups working on ME/CFS biomarkers in the EU countries is also illustrated in Fig. 1. Countries from which no publications on ME/CFS biomarker could be retrieved are shown in light green/grey, and European countries not participating in the EUROMENE are shown in white. The references listed per countries are shown in Table 1.

Discussion
So far there is no single biomarker available for diagnostic use in ME/CFS. Most studies identified here were exploratory in design and lack sex and age-matched control groups or validation cohorts thus having a low evidence level as summarized for the immune marker studies in Table 2 [42]. Some studies report inconsistent data, too. For example an expansion of transitional and naïve B cells and reduced plasmablast levels was reported in one study [14], but could not be confirmed in two other studies [4,12]. Immune cell phenotype and function analyses are, of course, hampered by variations in sampling and methodological differences between laboratories as most flow cytometric assays are not standardized. Further, immunological biomarkers reported mostly show alterations in subgroups only or with wide overlap to healthy control groups. Such heterogeneous results may be related to the fact that subgroups of ME/CFS patients exist with different immunological pathomechanisms. This concept is supported by the existence of clinical subgroups with heterogeneity in disease onset (infection-versus non-infection triggered), the variability of immune-associated symptoms, and the divergent response to B cell depletion therapy [43]. Research activity in infection markers on ME/ CFS across Europe is sparse; however, there is currently no evidence from the available literature that there is a specific serological signature aiding in diagnosis of ME/CFS. Similar to immunological markers, there is no single neurological or metabolic marker with sufficient specificity and sensitivity as a tool in ME/CFS diagnosis yet. However, recent studies analyzing multiple metabolites could show specific alterations in the majority of ME/CFS patients [37,[44][45][46] pointing to a probably common and specific metabolic profile. Further, metabolic studies consistently revealed different gender-related patterns [37,44,46]. Thus, instead of searching single markers fitting for diagnosing all patients, multiplexed determinations of biomarkers analyzing pathways together with patient stratification, may be necessary to develop diagnostic assays with sufficient sensitivity and specificity [47].

Conclusions
Heterogeneity of biomarker studies with different case definitions, low number of patients, lack of matched control groups, missing validation studies and potentially subgroup heterogeneity are possible reasons why no Immunologic [3] France Metabolic [28,29] Germany Metabolic [30] Immunologic [4][5][6][7] Neurologic [23] Ireland Immunologic [8] Italy Metabolic [31][32][33][34] Infection [18,19] Latvia Infection [20,21] Netherlands Metabolic [35,36] Norway Metabolic [37] Immunologic [9,10] Neurologic [24,25] Poland* Immunologic [11] Serbia Metabolic [38] Spain Metabolic [39] Immunologic [12] Infection [22] Sweden Immunologic [13] UK Metabolic [40,41] Immunologic [14][15][16][17] Neurologic [26]  diagnostic biomarkers are available yet. Further, as result of the low amount of funding in CFS/ME research few and often small studies were performed so far. Therefore, strategies to improve the quality and to facilitate the comparability of biomarker studies are needed (summarized in Table 3). This starts with well-defined patient cohorts Authors' contributions CS designed the study and research guidelines, reviewed data received from the different partner countries. CS and JB were major contributors in writing the manuscript. HF reviewed and analyzed the data, did research for the non-EUROMENE countries and prepared figures and tables. All other authors collected and reviewed data for their own country. MM is head of the EUROMENE cooperation group within COST network. All authors read and approved the final manuscript.