Personalised Medicine: implication and perspectives in the field of occupational health.

"Personalised medicine" relies on identifying and integrating individual variability in genomic, biological, and physiological parameters, as well as in environmental and lifestyle factors, to define "individually" targeted disease prevention and treatment. Although innovative "omic" technologies supported the application of personalised medicine in clinical, oncological, and pharmacological settings, its role in occupational health practice and research is still in a developing phase. Occupational personalised approaches have been currently applied in experimental settings and in conditions of unpredictable risks, e.g.. war missions and space flights, where it is essential to avoid disease manifestations and therapy failure. However, a debate is necessary as to whether personalized medicine may be even more important to support a redefinition of the risk assessment processes taking into consideration the complex interaction between occupational and individual factors. Indeed, "omic" techniques can be helpful to understand the hazardous properties of the xenobiotics, dose-response relationships through a deeper elucidation of the exposure-disease pathways and internal doses of exposure. Overall, this may guide the adoption/implementation of primary preventive measures protective for the vast majority of the population, including most susceptible subgroups. However, the application of personalised medicine into occupational health requires overcoming some practical, ethical, legal, economical, and socio-political issues, particularly concerning the protection of privacy, and the risk of discrimination that the workers may experience. In this scenario, the concerted action of academic, industry, governmental, and stakeholder representatives should be encouraged to improve research aimed to guide effective and sustainable implementation of personalised medicine in occupational health fields.

Introduction

New technologies that are emerging in biomedicine are allowing the collection of a huge amount of information, potentially useful to promote precise and targeted approaches of prevention, diagnosis, and therapy. Traditional medicine follows a one-size-fits-all approach, and as a consequence, not everyone responds to stimuli in the same way. On the contrary, personalised medicine is intended to convey the idea that although medical approaches are infrequently developed for single individuals, subgroups of patients can be better characterized and targeted in more specific ways (Figure 1) (48, 50, 60). Personalised medicine aims to integrate genomic with biological, physiological, and physical parameters together with environmental information, such as lifestyle, to depict a more comprehensive picture of each individual and improve dedicated preventive and therapeutic interventions (23, 24).

Figure 1

The two different approaches of traditional medicine and personalised medicine

The American National Research Council’s Toward Personalised Medicine defined personalised medicine as: “The tailoring of medical treatment to the individual characteristics of each patient...to classify individuals into subpopulations that differ in their susceptibility to a particular disease or their response to a specific treatment”. As the definition suggests, the novelty of personalised medicine lies in the possibility of guiding health care decisions toward the most effective treatment for a given patient, and thus, improving care quality (16). As a consequence, personalised medicine looks at a host of data across a population and aims to define a patient’s response to a specific disease, based on integrated complex information, also developing interventional strategies (23).

The American National Institute of Health (NIH) “All of Us” Research Program is inviting one million people across the U.S., which are expected to share the data generated over more than 10 years from sequencing, electronic medical records, personal reported information, and digital health technologies. These data will be the topic of analyses to improve the comprehension of disease mechanisms and pathogenesis and to drive the definition of a health care agenda. This scenario may contribute to a novel paradigm of healthcare expected to impact the medical approach throughout the lifespan and to enhance the consolidated applications of personalised health care such as reproductive counseling and prenatal testing at conception, promotion of healthy aging, and molecular autopsies (16).

The focus of personalised medicine relies on identifying emerging approaches that improve disease prevention and treatment, taking into account individual variability in genetic, environmental, and lifestyle factors (34). Although encouraging perspectives have been reported for the application of a personalised medicine approach into clinical settings, its current role in healthcare is still in its infancy, due to the complexity of an overall exposure assessment (i.e. exposome) and its integration into the interpretation of susceptibility factors. In this view, occupational medicine could offer a promising field for the development of personalised medicine approaches. However, a future, sustainable application of such innovative strategies in occupational health should overcome ethical and societal concerns related to the application of personalised medicine into a preventive and not a therapeutic approach, aimed to ensure a safe workplace for all. The present review aims to debate and illustrate the current applications of personalised/precision medicine and to propose its potential applications in occupational health.

Current applications of Personalised Medicine in diagnosis and therapy

Personalized medicine is based on the concept that individual variability in DNA is linked to disease causation. This is true for rare diseases, for which DNA sequencing has improved the clinical evaluation of many patients. One of the most striking examples of diagnostic and therapeutic improvements in the field is that of Cystic Fibrosis (CF). CF is a multisystemic genetic disease with an autosomal recessive inheritance that affects approximately 70,000 people worldwide and that is caused by variants in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genetic understanding of CF and its categorization into molecular subgroups based on the retained functionality of the mutated CFTR channel have paved the way to the design and approval of specific drugs able to (at least partially) recover the CFTR channel functionality depending on the molecular defect they target (2).

Besides, several genomic applications of personalized medicine also contribute to healthcare at several time points throughout the lifespan. For instance, preconception genetic screening can predict the risk of transmitting pathogenic variants to offspring (3). During pregnancy, genetic testing, including whole-genome sequencing, can be performed to assess chromosomal abnormalities of the fetus (45). At birth, genetic analysis can be used to diagnose several critical conditions (14, 49). Later on, during life, similar approaches can be applied to diagnose and treat many diseases, including cancers (as described below), metabolic, and chronic diseases (16, 39).

Nonetheless, interventions taking advantage of genetic information alone have been limited, mainly due to problematic side effects, such as the progressive decline of penetrance estimates of pathogenic variants as more unaffected subjects are screened. Indeed, reclassification of variants initially considered as pathogenic is a common problem and genetic databases are constantly updated (21, 70).

Pharmacogenomics, probably the earliest application of personalized medicine, is the application of genomic information to individual variation in drug response phenotypes. Different drug responses can range from inadequate therapeutic efficacy to serious, potentially life-threatening adverse drug reactions. There are several reasons for variability in drug response, such as drug interactions, poor compliance, the failure of selected drug therapy to target the causative disease mechanisms, and disease-related changes in drug concentrations or efficacy (17).

The genotyping of VKORC1 gene, which is involved in biochemical activation of the blood clotting factor vitamin K, and Cytochrome CYP2C9, which is a member of the Cytochrome p450 family led to some success in warfarin dosing optimization (1, 29, 63).

Another successful application of personalized medicine procedure is that of ‘Monogenic Diabetes’, which refers to types of diabetes which are due to a single gene mutation (28). Indeed, it has improved the efficacy of Sulfonylureas (SU) (28). Among the fourteen known forms of Maturity onset diabetes of the young (MODY), those caused by HNF1α and HNF4α mutations (MODY 3 and MODY 1, respectively) are the most common. Patients with MODY 1 and 3 are usually diagnosed before the age of 25 years and commonly do not show any features of insulin resistance. They are frequently misdiagnosed as Type 1 Diabetes and start life-long insulin therapy. On the contrary, a correct diagnosis enables transitioning of these patients to SUs therapy.

Neonatal Diabetes Mellitus (NDM) is diagnosed within the first six months of life and is mainly monogenic. Therapy of patients with NDM may be considered as a paradigmatic example of benefits due to pharmacogenomics application to personalized medicine. Indeed, children with NDM are typically misdiagnosed as Type 1 Diabetes and initiated on insulin therapy but the response is often unsatisfactory (40, 44). On the contrary, due to the genetic variations they carry, if correctly diagnosed, their therapy can be transitioned to SU with a great response maintained over 10 years (7, 47). The successful personalised approach used for the management of monogenic forms of diabetes could be applied also in those multifactorial forms (e.g. Type 2 Diabetes), thus making the control of diabetes more efficient toward prevention of its complications and improvement of the quality of life of the affected people (17, 52).

The application of personalised medicine approaches to the early diagnosis and cure of tumors is one of the greatest challenges of biomedical research. Personalised oncology uses tumor molecular profiles to define diagnostic, prognostic, and therapeutic paths for the specific tested cancer (41, 56). Tumor molecular profiles include DNA, RNA, and protein alterations, as well as epigenetic modifications (12).

The molecular characterization has been the standard approach for differential diagnosis and therapy of various tumors, including lung carcinoma (32). For instance, the occurrence of specific somatic mutations in the epidermal growth factor receptor (EGFR) gene and rearrangements in anaplastic lymphoma receptor tyrosine kinase (ALK) gene allow the personalization of therapy with targeted kinase inhibitors (5, 42, 43). Likewise, BRAF inhibition in BRAF-mutant melanoma tailors the best course of treatment for each patient including surgery, immunotherapy, targeted therapy, chemotherapy, and radiotherapy (46, 61).

The molecular characterization and consequent stratification into different subgroups have permitted the development of targeted treatment strategies also for breast cancer in combination with standard therapies (4, 20).

Historically, the first cancer personalised therapy was imatinib for chronic myeloid leukemia, which initiated the development of a series of therapies targeted at specific alterations. Targeted therapies have improved outcomes in cancer patients with targetable molecular/genetic alterations (22, 41, 55).

Cancer immunotherapy is a recent breakthrough in the field of oncology by prolonging the survival of patients with rapidly fatal cancers. Inevitably, the application of immunotherapy for both diagnosis and targeted treatment requires the personalised approach. The idea to employ the immune system to treat neoplastic disease is at the basis of the cancer immunotherapy approach, which has been recently growing. Indeed, immunotherapy has achieved outstanding therapeutic efficacy in several types of tumors, including breast cancer, colorectal cancer, non-small cell lung cancer, renal cell carcinoma, and melanoma (57, 66). Tumors present antigens such as fetal developmental proteins (i.e. oncofetal antigens), oncogenic viruses (i.e. oncoviral antigens), or neoantigens that are generated as a result of somatic mosaicism. Attempts to directly target these antigens have been disappointing; however, the research activity carried out permitted to unveil the critical mechanisms for the activation of T lymphocytes and modulation of their autoinhibitory pathways mediated by various checkpoint receptors, such as cytotoxic T lymphocyte-associated protein 4 (CTLA4), programmed cell death 1 (PD1; also known as PDCD1), and others (15, 19). Thus, various molecules, known as “checkpoint inhibitors” targeting these proteins were developed and subsequently approved for therapy (25, 53, 73).

Personalised medicine applications in occupational health

Efforts in personalised medicine may offer unique opportunities for occupational health, as it can support a suitable redefinition of the processes employed for risk assessment, able to take into consideration the complex interaction between occupational and individual factors. These procedures are intended to integrate a mechanistic understanding of occupational risk factors, information on specific conditions of workplace exposure, and workers’ susceptibility due to “omic” features in a “gene-environment” perspective (54). “Omic” techniques may support a better identification of the hazardous properties of the xenobiotics as the detected alterations, i.e. genomic, epigenomic, proteomic, metabolomic changes, can be the result of the exposure. On the other hand, these features may function as effect modifiers of the exposure-related outcomes as they can influence the metabolism of xenobiotics, impacting their toxicokinetic and dynamic behaviors. In this regard, some examples can illustrate the current application of omic technologies into occupational health fields (65). In the case of solvent exposure, the analysis of the toxicogenomic and epigenomic profiles has revealed useful to better understand the mechanisms by which benzene may cause leukemia (59, 72). Two genetic variants in benzene key metabolizing enzymes, myeloperoxidase and NAD(P)H:quinone oxidoreductase (36), but also polymorphisms in genes involved in DNA double-strand breaks repair and genomic maintenance, i.e. WRN, TP53, BRCA2, BLM and RAD51, were demonstrated to influence occupational susceptibility to benzene hematotoxicity in exposed workers (38, 58). Single nucleotide polymorphisms in interleukin (IL-1A, IL-4, IL-10, IL-12A) and molecular adhesion (vascular adhesion molecule,VCAM-1) genes were associated with a statistically significant decrease in total white blood cell counts (37). Epigenetic alterations in the methylation of blood DNA samples were detected in gas station attendants and traffic police officers exposed to different levels of benzene compared to unexposed controls (6). Airborne benzene was associated with a significant reduction in LINE-1 and Alu I methylation, as an early benzene-induced change in normal methylation patters. Increasing airborne benzene levels were associated with the hypermethylation in p15 and hypomethylation in MAGE-1, initial alterations potentially preceding greater methylation changes determined in several tumors, including leukemias. Proteomic analysis revealed that protein profiles were significantly different in benzene exposed workers compared to controls, particularly, concerning the up-regulation of T cell receptor chain, FK506-binding protein and matrix metalloproteinase-13 (27).

Concerning exposure to metals and metalloids, different genome expression patterns were evaluated in subjects with and without arsenical skin lesions using RNA from peripheral blood lymphocytes. Four hundred sixty-eight genes were identified to be differently expressed between the two groups. Such analysis may provide insights into the underlying processes of arsenic-induced disease that may represent potential targets for chemoprevention studies to reduce arsenic induced skin cancer in the exposed population. Additionally, up-regulated expression of genes involved in inflammatory pathways, possibly related to arsenic-associated atherosclerosis, including several cytokines and growth factors has been identified (69). A significant dose-dependent DNA hypermethylation of the promoter region of p53 gene was observed in DNA of arsenic-exposed people compared to controls (11). Proteomic profiling of sera in a group of smelter workers with a mixed exposure to arsenic and lead, identified five discriminatory protein peaks that could form a proteomic signature providing higher sensitivity and specificity in detecting metal mixture exposure than single protein markers (71).

Overall, this different personalised approach may lead to the hazard identification phase of risk assessment to include not only the intrinsic toxicological profile of substances, but also how these may interact with the organisms. Additionally, this may offer the opportunity to generate comprehensive toxicologically relevant information on molecular changes more quickly and more accurately than ever before, supporting the identification of new hazards through enhanced coverage of biological or biochemical pathways during toxicological analyses (10). This approach might be even more important to support suitable risk assessment in emerging occupational scenarios, characterized by low-doses of exposure, employment of innovative materials (like chemicals at the nanoscale), as well as in settings where complex mixtures are used.

“Omic” derived information can explain potential variabilities in internal doses of exposure, dose-response relationships, no observable adverse effect levels, as well as in outcomes, overall promoting a deeper understanding of increased susceptibility of certain subpopulations. Additionally, the opportunity offered by personalised medicine to elucidate the pathway from a molecular initiating event to a health effect may characterize a new paradigm for assigning risk assessment based on known molecular mechanisms of toxic injury (64), enabling the development of early, high-quality biomarkers in the possible exposure-disease continuum, changing how occupational exposures are monitored and managed (9, 67).

All the abovementioned information will serve to characterize risks according to an individual perspective, to guide preemptive interventional strategies that may, interestingly, exploit knowledge acquired at an individual level into collective strategies for health protection (Figure 2). To extrapolate individual data into a population-based risk assessment and management processes characterizes the innovative perspective of the application of a personalised medicine approach in a preventive framework. As an example, once individual alterations have been confirmed to be functionally relevant with both xenobiotic exposures and specific phenotypes, these should be considered in the attempt to establish primary preventive measures, that may be protective for the vast majority of the population, including most susceptible subgroups (13). However, according to a pathophysiology perspective, deeper knowledge should be acquired concerning the magnitude or frequency of susceptibility conditions in the population and the variations related to geography and race.

Figure 2

Research fields and future perspectives of personalised medicine application in the occupational risk assessment process. The figure shows the fields of research (on the left) and future perspectives (on the right) of personalised medicine application into risk assessment processes in occupational settings

Additionally, also the experience acquired in some specific occupational settings may provide a greater advantage to enlarge personalised solutions into occupational health practice. In some situations, workers, like military services employees involved in war missions (8, 9) and astronauts employed in human space flights (51, 62) face unpredictable and unexpected challenges coupled with isolation, remote operations, and extreme resource limitations that can take advantage from a personalized medicine approach. In such peculiar settings, where it was not possible to assess and avoid risks, or where failure or adverse events should be avoided at all costs, great contribution derived from personalised medicine. It was important to prevent disease development that resulted essentially for the military mission success in the short term, and for adopting a risk management strategy in the long-term (8). Furthermore, in the case of astronauts in the international space station, pharmacogenetic testing served as a pre-treatment diagnostic option to prevent therapy failure or side effects during a definite period of self-medication (62).

However, while in these contexts a personalised medicine approach seem more necessary to maintain the well-being of workers avoiding the manifestation of the adverse effect that could hinder the success of job outcomes, in routine conditions, personalised strategies should be aimed to significantly improve primary preventive measures including the control of the exposure and the adoption/implementation of collective and individual measures focused on specific group/subject features. The emerging and ongoing Covid-19 pandemic, in this contest, has pointed out the relevance to define conditions of vulnerability to an increased risk of infection and severe disease manifestation that may require personalised preventive actions.

Furthermore, personalised medicine approaches may help in accumulating evidence regarding multiple exposure memory systems to prevent/predict adaptation to environmental and occupational challenges over time (26). Following occupational insults, short-term and reversible responses can occur, but also long-term and sometimes permanent changes in organ structure and function can develop (67). Indeed, flexibility and adaptability to subsequent challenges can be reduced. In this perspective, personalised medicine may take advantage of computational methods to study metabolic resilience in case of repeated exposure to variable occupational risk factors, i.e. chemical, biological or physical risks to define populations that can benefit from specific preventive measures.

Overall, innovative technologies and big data, that are driving personalised medicine forward, may lead to advances in occupational health, not only limited to disease prevention but also involving the overall health promotion of occupational populations taking into account both workplace conditions and personal issues (30, 31). This may take advantage of the assessment of a variety of biological factors that may be utilized to assess the health state of individuals according to a system-wide approach.

Discussion

The personalised medicine model promises that the fundamentals of health care, including assessments, medical decisions, and treatments, will be customized to the individual (35). This overall supports the definition of preventive and treatment interventions tailored to higher-risk groups, to effectively improve population health. In the occupational health field, this seems extremely important considering that a personalised medicine approach may inform adequate risk assessment and management processes to better protect the health of exposed workers, also through the application of advanced “omic” technologies.

However, to achieve an effective application of personalised medicine in the workplace, some challenging issues, relative to clinical diagnostic testing, biomarker validation, and biobanking should be carefully faced. In fact, despite the promising role of “omic” technologies in occupational health monitoring and bio-surveillance, such platform tools can exert their usefulness only when they provide information on relevant analytes indicative of exposure, effect, or susceptibility to toxicants. This requires to define quality standards on the design and performance of “omics” based studies aimed to identify successful candidate biomarkers, with sufficient specificity, sensitivity, and ease to apply to clinical/occupational settings. “Omic” technologies, in this regard, offer great opportunities, particularly concerning the ability to identify molecular biomarkers related to key events in pathways that may allow the assessment of risks and the possible exposure-disease continuum. This interesting aspect appears essential to understand the mechanisms underlining the occurrence of various adverse outcomes, and possible conditions that may increase the susceptibility of individual workers or homogeneous groups of workers.

“Omic” technologies can generate a huge amount of data posing a central challenge in their wide application. This relates to the great number of substances, as well as of molecular and cellular endpoints that can be efficiently tested and simultaneously evaluated (68). Indeed, uncertainties emerge concerning the generation, storage, processing, analysis, interpretation of “omic” data and comparison of data sets generated on different “omic” platforms. As the field of systems biology moves steadily forward, ways to integrate and compare data from different “omics” need to be developed so that the study outcomes can be reliably verified and confidently integrated into regulatory hazard and risk assessment.

A particular challenge for innovative biomarkers will be to determine the reproducibility of the “omic” data, the concordance and repeatability within experiments and labs considering also the preliminary experience in the field. All these aspects, together with suitable timing for data analysis, the appropriateness of bio-specimen collection and storage aspects of biobanking are important for the analysis success. Biobanks, intended as collections, archives, and databases, are one of the pillars of personalized medicine (33). Their power relies on the amount of high-quality samples of and related information available for current and future research. Thus, carefull attention must be payed at all the steps of the so-called “sample-to-answer” workflow for the quality of “omic” biomarkers, i.e. from the sample collection and handling, through biobanking, and final analyses to achieve accurate storage of a great number of biological materials for future investigation. This will ensure scientific practice to assume an even more global nature with increased interdisciplinary, and dynamic collaborations between clinicians, researchers, lab staff, bioinformatics, statisticians and other data analysts (33). In this scenario, the development of software enabling a system-biology approach appears essential to understand exposure-associated metabolic changes also with the support of computational modeling.

The wider ability of doctors to use patients’ genetic and other molecular information as part of routine medical care is necessary. Healthcare professionals may increasingly find themselves needing to interpret the results of genetic tests, understand how that information is relevant for prevention approaches and treatments, and convey this knowledge to patients. This raises concerns regarding the use of new molecular biology techniques in human biomonitoring and the interpretation of the obtained data. This poses some questions concerning risk communication strategies in the “omic” technology era.

Furthermore, successful implementation of personalised medicine into occupational health settings requires some ethical, legal, social, and political considerations, particularly concerning the privacy and protection of personal health information. When using innovative techniques, in fact, some important issues, other than scientific and technical aspects, need to be addressed to ensure their sustainable and equitable development (12). These include equity or distributive justice, supervision and control, as well as resources. In a therapeutic scenario, equity relies on the decisions and consequences associated with personalised health interventions that may cause health care disparities in minority populations, socioeconomically depressed contexts, or in relations to the individual motivation in caring health. Conversely, in an occupational preventive scenario critical “equity” issues include the correct balance between the benefit to the worker in terms of preventive action, and the cost in terms of possible discrimination they may experience in the workplace. In principle, personalised medicine strategies applied in occupational medicine contexts should not result in discrimination or reduction of job opportunities for the workers involved (18). Indeed, supervision and control is necessary to face some relevant issues concerning i.e. whether a group of workers should take “omic” tests, which is the role of the determined alterations in leading preventive actions, what may happen whether someone does not agree to undergo testing, and how to assure workplace equity. Questions regarding informed consent for testing and how the large amount of data generated for each individual will be managed, distributed and reported, remain of paramount importance (12). Additionally, it should be noticed that, if “omic” tests will not be covered by adequate scientific knowledge and by personalised and stratified prevention measures, risk assessment and management will become a two-tier system. Cost is also an issue with personalised medicine. On the one hand, technologies and the following adopted preventive strategies are expensive to carry out. In this context, reimbursement of these procedures is also likely to become an issue. On the other hand, there is the possibility that constant measurement of multicomponent data and clinical samples can improve health while decreasing costs considering also that such analyses may, in turn, support the efficient use of resources toward those populations at higher risk of adverse effects.

Overall, to achieve a sustainable application of innovative technologies, ethical and practical challenges must be balanced against the enormous potential across all the disciplines of occupational health, from prevention to early diagnostics and management. Moreover, personalised medicine will not be realized by research cohorts alone and its successful transition into routine settings will need to embody a cultural shift, to include national, institutional, and even individual practices and perceptions. This should include both technical and sociopolitical efforts (8).

In this scenario, the concerted action of different expertise, offered by medical and biotechnological scientists, but also derived from academic, industry, governmental bodies and stakeholder representatives should be encouraged to improve research efforts aimed to define the opportunities and challenging issues related to the personalised medicine application to guide its effective and sustainable implementation in occupational health fields.

No potential conflict of interest relevant to this article was reported by the authors