Current literature includes several definitions of “biological markers” or “biomarkers.” In 1998, the National Institutes of Health (NIH) Biomarkers Definitions Working Group was constituted and, in 2001, the following definition of biomarker was published: “a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention” (1). In the same year, the International Program on Chemical Safety proposed another definition of biomarker: “any substance, structure, or process that can be measured in the body or its products and influence or predict the incidence of outcome or disease.” Both, these two definitions were established starting from a previous statement of the World Health Organization, describing a valid biomarker as: “almost any measurement reflecting an interaction between a biological system and a potential hazard, which may be chemical, physical, or biological. The measured response may be functional and physiological, biochemical at the cellular level, or a molecular interaction” (http://www.inchem.org/documents/ehc/ehc/ehc155.htm).

Given the crucial role that biomarkers can play in allowing and improving drug development progression, the economic investments and academic efforts in last decades were highly focused on the identification of valuable biomarkers for the most health impacting medical conditions. Indeed, over $ 2.5 billion were recently granted by the NIH to research proposals aimed to the identification of novel biomarkers. However, this did not result in the expected translation into clinical practice, especially with respect to biological markers identified by proteomic studies (2).

Together with genetic studies, research in the field of psychosis has focused on the identification of inflammation-related biomarkers and mediators of the stress response in peripheral blood samples. Another recent and growing field of research on possible clinic and therapeutic use of specific biomarkers in this mental disorder has been focused on the imbalance between free radical production and the antioxidant system functioning, resulting in redox dysregulation and oxidative stress. In this review, we analyzed studies on humans, found on PubMed from January 1^st 2010 to January 31^st 2017 and focused on the clinical use of inflammation and stress response blood biomarkers, as well as redox dysregulation biomarkers, in detecting the “premorbid” psychotic state and early phases of the disease. Their pharmacological implications in predicting and monitoring the individual response to antipsychotic medication are discussed.

The literature source for the writing of this review was represented by papers present in PubMed from January 1^st 2010 to January 31^st 2017, found by using the following keyword combinations: biomarkers AND psychosis; biomarkers AND first psychotic episodes; biomarkers AND psychosis high risk state; biomarkers AND antipsychotics; biomarkers AND psychosis AND inflammation; biomarkers AND psychosis AND stress-response; biomarkers AND psychosis AND cortisol; biomarkers AND psychosis AND oxidative stress; biomarkers AND psychosis AND redox dysregulation; biomarkers AND first psychotic episode AND inflammation; biomarkers AND first psychotic episode AND stress-response; biomarkers AND first psychotic episode AND cortisol; biomarkers AND first psychotic episode AND oxidative stress; biomarkers AND first psychotic episode AND redox dysregulation; biomarkers AND psychosis high-risk state AND inflammation; biomarkers AND psychosis high risk state AND stress-response; biomarkers AND psychosis high risk state AND cortisol; biomarkers AND psychosis high risk state AND oxidative stress; biomarkers AND psychosis high risk state AND redox dysregulation; biomarkers AND antipsychotics AND inflammation; biomarkers AND antipsychotics AND stress-response; biomarkers AND antipsychotics AND cortisol; biomarkers AND antipsychotics AND oxidative stress; biomarkers AND antipsychotics AND redox dysregulation.

A total number of 1,341 records was obtained with 73 record duplications, which were removed from further screening steps. The following inclusion criteria were applied for the next screening (based on titles and abstracts) of the remaining records (1,268): (1) language of publication (only publication written in English were considered); (2) type of publication (only original research articles, reviews and meta-analysis were considered); (3) subjects of the study (only studies on humans were considered); (4) source of samples (only data obtained from blood samples and derivatives were considered). Based on these criteria, 1,070 records were excluded and 198 full-text articles were assessed for eligibility. We further screened them excluding works (total number 139) which: (1) did not include control group subjects; (2) were missing of a specific description of patients’ inclusion and exclusion criteria: (3) did not report sociodemographic informations for the included patients and controls; (4) did not clearly indicate the diagnostic criteria used for the identification of the considered psychotic stage; (5) did not include a clear indication of the medication type and related doses. Thus, the final number of studies included in the qualitative synthesis was 59 (a PRISMA diagram is provided in Figure 1). The total number of references reported in the bibliography of this review also includes additional publications cited in the Introduction section and in the introducing paragraphs of the other sections of this manuscript.

Recent data have indicated a pathogenetic link between blood alterations and the psychotic disorder spectrum (3). Blood is considered the most easily accessible biological sample during any stage of the disease (4). This has justified significant efforts in trying to determine blood molecular correlates of schizophrenia and psychosis for early therapeutic interventions (5). In this context, data derived from drug-naive patients, subjects at ultra high risk of psychosis or at first psychotic episode, have been considered extremely precious. Several findings in this field have been published, especially regarding blood biomarkers of inflammation and stress response.

Several publications have identified redox imbalance as a crucial player in the pathogenesis of psychosis. Most of the data on this subject have been obtained on both pharmacologic and non-pharmacologic rodent models of this mental disorder (64–71). Increasing research interest is going toward the possibility to translate the findings obtained on animal models toward humans. Indeed, a consistent number of works are focusing on the measurement of reactive oxygen species amount, oxidative damage markers, and antioxidant defense functioning in different types of samples derived from psychotic medicated and unmedicated patients (72).

Table 2 reports an overview of the most relevant findings about the use of redox status-related biomarkers both in different stages of psychosis development and as predictors of response to medication.

Identifying a biomarker for psychosis would certainly lead to several advantages, such as the possibility of an early diagnosis, of a stratification of patient population as well as of a progression toward personalized therapies. Indeed, with respect to this last point, several extrinsic and intrinsic factors (such as genetic aspects, environment and neuronal alterations at both cellular and structural levels) may account for the interindividual variability in the response to the pharmacological compounds commonly use both to treat a frank psychotic state or to slower the progression toward this mental disorder. Another crucial benefit will also consist in the possible combination of individual genetic data with specific biomarkers in order to more effectively treat single cases of psychosis. Furthermore, even if a proposed biomarker could not be considered reliable for a group of patients, it might be predictive, or even associate, with specific aspects of the disease such as its severity or particular symptoms (e.g., positive versus negative symptoms of psychosis) in a single subject, probably because his/her unique genetic asset.

However, more than in other medical fields, the efforts of the scientific community have not been yet translated in what expected. Indeed, so far, most of what is considered a biomarker candidate for psychosis is instead only a potential biomarker. That said, important questions arise: if, for other medical fields, research has succeeded in identifying clinical applicable biomarkers, why it should not happen for psychiatric disorders and, in particular, for psychosis? Thus, which major differences exist with respect to other psychiatric conditions and to non-psychiatric pathologies? A possible attempt to answer to these questions should consider the complex nature of mental disorders with respect to other non-psychiatric conditions, requiring a more global approach in which the different pathogenetic elements should be adequately taken into account. Furthermore, with respect to other psychiatric disorders, the possibility to follow-up psychotic patients in longitudinal studies, which will provide a tool to overcome some time-related limitations in the field of biomarker research, is challenging given the high rates of noncompliance (especially to medication) of psychotic patients. Another important aspect to consider in psychosis is the presence of several comorbidities that significantly decrease the possibility that an identified biomarker candidate can be routinely used for the clinical practice. Thus, so far, available literature clearly indicates the absence of biomarkers for psychosis, in particular, for the identification of the ultra high-risk state, first psychotic episodes, and pharmacological response. Furthermore, the absence of a deep and comprehensive understanding of all the different players involved in psychosis pathogenesis has been highlighted leading to several speculations about the molecular mechanisms underlying the onset and progression of psychotic symptoms and the increasing “bad habit” to directly translate data obtained from biomarker research in animal models of this mental disorder to humans. Therefore, future research in this field will certainly need a more global overlook toward psychosis pathogenesis but, mostly, a more cautious approach in considering and diffusing obtained results.

Writing manuscript, SS. Revising manuscript content, LT. Approving final version of the manuscript, SS and LT.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.