Personalised medicine: not just in our genes

Explainable discovery of disease biomarkers: The case of ovarian cancer to illustrate the best practice in machine learning and Shapley analysis

OBJECTIVE

Ovarian cancer is a significant health issue with lasting impacts on the community. Despite recent advances in surgical, chemotherapeutic and radiotherapeutic interventions, they have had only marginal impacts due to an inability to identify biomarkers at an early stage. Biomarker discovery is challenging, yet essential for improving drug discovery and clinical care. Machine learning (ML) techniques are invaluable for recognising complex patterns in biomarkers compared to conventional methods, yet they can lack physical insights into diagnosis. eXplainable Artificial Intelligence (XAI) is capable of providing deeper insights into the decision-making of complex ML algorithms increasing their applicability. We aim to introduce best practice for combining ML and XAI techniques for biomarker validation tasks.

METHODS

We focused on classification tasks and a game theoretic approach based on Shapley values to build and evaluate models and visualise results. We described the workflow and apply the pipeline in a case study using the CDAS PLCO Ovarian Biomarkers dataset to demonstrate the potential for accuracy and utility.

RESULTS

The case study results demonstrate the efficacy of the ML pipeline, its consistency, and advantages compared to conventional statistical approaches.

CONCLUSION

The resulting guidelines provide a general framework for practical application of XAI in medical research that can inform clinicians and validate and explain cancer biomarkers.

Individualized Dosage of Tacrolimus for Renal Transplantation Patients Based on Pharmacometabonomics

The clinical pharmacodynamics of tacrolimus in renal transplant patients has significant interindividual variability. T lymphocytes were selected to study the pharmacodynamic response of tacrolimus, which was significantly correlated with renal function and the outcome of renal transplant patients. Ultra-performance liquid chromatography-quadrupole time-of-flight mass spectroscopy (UPLC/Q-TOF-MS) was performed to obtain the metabolic profiles of 109 renal transplant patients. A partial least squares (PLS) model was constructed to screen potential biomarkers that could predict the efficacy of tacrolimus. Multinomial logistic regression analysis established a bridge that could quantify the relationship between the efficacy of tacrolimus and biomarkers. The results showed a good correlation between endogenous molecules and the efficacy of tacrolimus. Metabolites such as serum creatinine, mesobilirubinogen, L-isoleucine, 5-methoxyindoleacetate, eicosapentaenoic acid, N₂-succinoylarginine, tryptophyl-arginine, and butyric acid were indicated as candidate biomarkers. In addition, the key biomarkers could correctly predict the efficacy of tacrolimus with an accuracy of 82.5%. Finally, we explored the mechanism of individual variation by pathway analysis, which showed that amino acid metabolism was significantly related to the efficacy of tacrolimus. Moreover, orthogonal partial least squares discriminant analysis (OPLS-DA) showed that there was no difference in key metabolites among different pharmacodynamic groups at 1 month and 3 months after dose adjustment, suggesting that pharmacometabonomics is a useful tool to predict individual differences in pharmacodynamics and thus to facilitate individualized drug therapy.

Long Short-Term Memory Network for Development and Simulation of Warfarin Dosing Model Based on Time Series Anticoagulant Data

Background

Warfarin is an effective treatment for thromboembolic disease but has a narrow therapeutic index, and dosage can differ tremendously among individuals. The study aimed to develop an individualized international normalized ratio (INR) model based on time series anticoagulant data and simulate individualized warfarin dosing.

Methods

We used a long short-term memory (LSTM) network to develop an individualized INR model based on data from 4,578 follow-up visits, including clinical and genetic factors from 624 patients whom we enrolled in our previous randomized controlled trial. The data of 158 patients who underwent valvular surgery and were included in a prospective registry study were used for external validation in the real world.

Results

The prediction accuracy of LSTM_INR was 70.0%, which was much higher than that of MAPB_INR (maximum posterior Bayesian, 53.9%). Temporal variables were significant for LSTM_INR performance (51.7 vs. 70.0%, P < 0.05). Genetic factors played an important role in predicting INR at the onset of therapy, while after 15 days of treatment, we found that it might unnecessary to detect genotypes for warfarin dosing. Using LSTM_INR, we successfully simulated individualized warfarin dosing and developed an application (AI-WAR) for individualized warfarin therapy.

Conclusion

The results indicate that temporal variables are necessary to be considered in warfarin therapy, except for clinical factors and genetic factors. LSTM network may have great potential for long-term drug individualized therapy.

Trial Registration

NCT02211326; www.chictr.org.cn:ChiCTR2100052089.

Serum Predose Metabolic Profiling for Prediction of Rosuvastatin Pharmacokinetic Parameters in Healthy Volunteers

Rosuvastatin is a well-known lipid-lowering agent generally used for hypercholesterolemia treatment and coronary artery disease prevention. There is a substantial inter-individual variability in the absorption of statins usually caused by genetic polymorphisms leading to a variation in the corresponding pharmacokinetic parameters, which may affect drug therapy safety and efficacy. Therefore, the investigation of metabolic markers associated with rosuvastatin inter-individual variability is exceedingly relevant for drug therapy optimization and minimizing side effects. This work describes the application of pharmacometabolomic strategies using liquid chromatography coupled to mass spectrometry to investigate endogenous plasma metabolites capable of predicting pharmacokinetic parameters in predose samples. First, a targeted method for the determination of plasma concentration levels of rosuvastatin was validated and applied to obtain the pharmacokinetic parameters from 40 enrolled individuals; then, predose samples were analyzed using a metabolomic approach to search for associations between endogenous metabolites and the corresponding pharmacokinetic parameters. Data processing using machine learning revealed some candidates including sterols and bile acids, carboxylated metabolites, and lipids, suggesting the approach herein described as promising for personalized drug therapy.

Review article: distinctions between ileal and colonic Crohn's disease: from physiology to pathology

BACKGROUND

Ileal and colonic Crohn's disease seem to be two separate entities.

AIMS

To describe the main physiological distinctions between the small and the large intestine and to analyse the differences between ileal and colonic Crohn's disease.

METHODS

The relevant literature was critically examined and synthesised.

RESULTS

The small and large intestine have fundamental distinctions (anatomy, cellular populations, immune defence, microbiota). The differences between ileal and colonic Crohn's disease are highlighted by a heterogeneous body of evidence including clinical features (natural history of the disease, efficacy of treatments, and monitoring), epidemiological data (smoking status, age, gender) and biological data (genetics, microbiota, immunity, mesenteric fat). However, the contribution of these factors to disease location remains poorly understood.

CONCLUSION

The classification of ileal and colonic Crohn's disease as distinct subphenotypes is well supported by the literature. Understanding of these differences could be exploited to develop more individualised patient care.

Predicting Range of Initial Warfarin Dose Based on Pharmacometabolomic and Genetic Inputs

Anticoagulation response to warfarin during the initial stage of therapy varies among individuals. In this study, we aimed to combine pharmacometabolomic and pharmacogenetic data to predict interindividual variation in warfarin response, and, on this basis, suggest an initial daily dose range. The baseline metabolic profiles, genotypes, and clinical information of 160 patients with heart valve disease served as the variables of the function of the last international normalized ratio measured before a patient's discharge (INR_day7 ) to screen for potential biomarkers. The partial least-squares model showed that two baseline metabolites (uridine and guanosine), one single-nucleotide variation (VKORC1), and four clinical parameters (weight, creatinine level, amiodarone usage, and initial daily dose) had good predictive power for INR_day7 (R²  = 0.753 for the training set, 0.643 for the test set). With these biomarkers, a machine learning algorithm (two-dimensional linear discriminant analysis-multinomial logit model) was used to predict the subgroups with extremely warfarin-sensitive or less warfarin-sensitive patients with a prediction accuracy of 91% for the training set and 90% for the test set, indicating that individual responses to warfarin could be effectively predicted. Based on this model, we have successfully designed an algorithm,"IniWarD," for predicting an effective dose range in the initial 7-day warfarin therapy. The results indicate that the daily dose range suggested by the IniWarD system is more appropriate than that of the conventional genotype-based method, and the risk of bleeding or thrombus due to warfarin could thus be avoided.

Categories, components, and techniques in a modular construction of basket trials for application and further research

Basket trials have become a virulent topic in medical and statistical research during the last decade. The core idea of them is to treat patients, who express the same genetic predisposition-either personally or their disease-with the same treatment irrespective of the location of the disease. The location of the disease defines each basket and the pathway of the treatment uses the common genetic predisposition among the baskets. This opens the opportunity to share information among baskets, which can consequently increase the information of the basket-wise response with respect to the investigated treatment. This further allows dynamic decisions regarding futility and efficacy of individual baskets during the ongoing trial. Several statistical designs have been proposed on how a basket trial can be conducted and this has left an unclear situation with many options. The different designs propose different mathematical and statistical techniques, different decision rules, and also different trial purposes. This paper presents a broad overview of existing designs, categorizes them, and elaborates their similarities and differences. A uniform and consistent notation facilitates the first contact, introduction, and understanding of the statistical methodologies and techniques used in basket trials. Finally, this paper presents a modular approach for the construction of basket trials in applied medical science and forms a base for further research of basket trial designs and their techniques.

A Unified Conceptual Framework for Metabolic Phenotyping in Diagnosis and Prognosis

Understanding metabotype (multicomponent metabolic characteristics) variation can help to generate new diagnostic and prognostic biomarkers, as well as models, with potential to impact on patient management. We present a suite of conceptual approaches for the generation, analysis, and understanding of metabotypes from body fluids and tissues. We describe and exemplify four fundamental approaches to the generation and utilization of metabotype data via multiparametric measurement of (i) metabolite levels, (ii) metabolic trajectories, (iii) metabolic entropies, and (iv) metabolic networks and correlations in space and time. This conceptual framework can underpin metabotyping in the scenario of personalized medicine, with the aim of improving clinical outcomes for patients, but the framework will have value and utility in areas of metabolic profiling well beyond this exemplar.

Precision Medicine for Nutritional Management in End-Stage Kidney Disease and Transition to Dialysis

Chronic kidney disease (CKD) is a global public health burden. Dialysis is not only costly but may not be readily available in developing countries. Even in highly developed nations, many patients may prefer to defer or avoid dialysis. Thus, alternative options to dialysis therapy or to complement dialysis are needed urgently and are important objectives in CKD management that could have huge clinical and economic implications globally. The role of nutritional therapy as a strategy to slow CKD progression and uremia was discussed as early as the late 19th and early 20th century, but was only seriously explored in the 1970s. There is a revival of interest recently owing to encouraging data as well as the increase of precision medicine with an emphasis on a personalized approach to CKD management. Although part of the explanation for the inconclusive data may relate to variations in study design and dietary prescription, diversity in genetic make-up, variations in the non-nutritional management of CKD, intra-individual variations in responses to dietary and nondietary treatment, psychosocial factors, and dietary compliance issues, these all may contribute to the heterogeneous data and responses. This brings in the evolving concept of precision medicine, in which disease management should be tailored and individualized according not only to clinical manifestations but also to the genetic make-up and biologic responses to therapy, which may vary depending on genetic composition. Precision nutrition management also should take into account patient demographics, social, psychological, education, and compliance factors, which all may influence the therapeutic needs and responses to the nutritional therapy prescribed. In this review, we provide a novel concept of precision medicine in nutritional management in end-stage kidney disease with a transition to dialysis and propose how this may be the way forward for nutritional therapy in the CKD population.

Evaluation of person-level heterogeneity of treatment effects in published multiperson N-of-1 studies: systematic review and reanalysis

OBJECTIVE

Individual patients with the same condition may respond differently to similar treatments. Our aim is to summarise the reporting of person-level heterogeneity of treatment effects (HTE) in multiperson N-of-1 studies and to examine the evidence for person-level HTE through reanalysis.

STUDY DESIGN

Systematic review and reanalysis of multiperson N-of-1 studies.

DATA SOURCES

Medline, Cochrane Controlled Trials, EMBASE, Web of Science and review of references through August 2017 for N-of-1 studies published in English.

STUDY SELECTION

N-of-1 studies of pharmacological interventions with at least two subjects.

DATA SYNTHESIS

Citation screening and data extractions were performed in duplicate. We performed statistical reanalysis testing for person-level HTE on all studies presenting person-level data.

RESULTS

We identified 62 multiperson N-of-1 studies with at least two subjects. Statistical tests examining HTE were described in only 13 (21%), of which only two (3%) tested person-level HTE. Only 25 studies (40%) provided person-level data sufficient to reanalyse person-level HTE. Reanalysis using a fixed effect linear model identified statistically significant person-level HTE in 8 of the 13 studies (62%) reporting person-level treatment effects and in 8 of the 14 studies (57%) reporting person-level outcomes.

CONCLUSIONS

Our analysis suggests that person-level HTE is common and often substantial. Reviewed studies had incomplete information on person-level treatment effects and their variation. Improved assessment and reporting of person-level treatment effects in multiperson N-of-1 studies are needed.

Translating Lung Microbiome Profiles into the Next-Generation Diagnostic Gold Standard for Pneumonia: a Clinical Investigator's Perspective

Severe bacterial pneumonia is a major global cause of morbidity and mortality, yet current diagnostic approaches rely on identification of causative pathogens by cultures, which require extended incubation periods and often fail to detect relevant pathogens. Consequently, patients are prescribed broad-spectrum antibiotics in a “one-size-fits-all” manner, which may be inappropriate for their individual needs and promote antibiotic resistance.

Severe bacterial pneumonia is a major global cause of morbidity and mortality, yet current diagnostic approaches rely on identification of causative pathogens by cultures, which require extended incubation periods and often fail to detect relevant pathogens. Consequently, patients are prescribed broad-spectrum antibiotics in a “one-size-fits-all” manner, which may be inappropriate for their individual needs and promote antibiotic resistance. My research focuses on leveraging next-generation sequencing of microbial DNA directly from patient samples for the development of new, culture-independent definitions of pneumonia. In this perspective article, I discuss the current state of the field and focus on the conceptual and research design challenges for clinical translation. With ongoing technological advancements and application of computational biology methods for assessing clinical validity and utility, I anticipate that sequencing-based diagnostics will soon be able to positively disrupt the way we think about, diagnose, and treat pulmonary infections.

The Influence of Genotype Information on Psychiatrists' Treatment Recommendations: More Experienced Clinicians Know Better What to Ignore

BACKGROUND

This study applies attribute nonattendance to medical decision making. We aimed to demonstrate how this type of analysis can be used in medical decision making to assess whether psychiatrists were influenced in their treatment recommendations by information on the genotype of a patient, despite knowing the patient's response to treatment as measured by the Positive and Negative Syndrome Scale. A patient's genetic information may be used to predict their response to therapy; such information, however, becomes redundant, and should not influence decisions, once a clinician knows the patient's actual response to treatment.

METHODS

Sixty-seven psychiatrists were presented with patients' pre- or post-treatment scores on the Positive and Negative Syndrome Scale for two hypothetical treatments for schizophrenia. Psychiatrists were also informed whether the patient possessed a genotype linked to hyper-responsiveness to one of the treatments, and were asked to recommend one of these two treatments. Attribute nonattendance assessed whether the information on genotype influenced psychiatrists' treatment recommendations.

RESULTS

Years of experience predicted whether psychiatrists were influenced by the genetic information. Psychiatrists with 1 year or less of experience had a 46% probability of considering genetic information, whereas psychiatrists with at least 15 years of experience had a lower probability (7%).

CONCLUSIONS

Psychiatrists and other clinicians should be cautious about allowing a patient's genetic information to carry unnecessary weight in their clinical decision making.

Integration of pharmacometabolomics with pharmacokinetics and pharmacodynamics: towards personalized drug therapy

Personalized medicine, in modern drug therapy, aims at a tailored drug treatment accounting for inter-individual variations in drug pharmacology to treat individuals effectively and safely. The inter-individual variability in drug response upon drug administration is caused by the interplay between drug pharmacology and the patients' (patho)physiological status. Individual variations in (patho)physiological status may result from genetic polymorphisms, environmental factors (including current/past treatments), demographic characteristics, and disease related factors. Identification and quantification of predictors of inter-individual variability in drug pharmacology is necessary to achieve personalized medicine. Here, we highlight the potential of pharmacometabolomics in prospectively informing on the inter-individual differences in drug pharmacology, including both pharmacokinetic (PK) and pharmacodynamic (PD) processes, and thereby guiding drug selection and drug dosing. This review focusses on the pharmacometabolomics studies that have additional value on top of the conventional covariates in predicting drug PK. Additionally, employing pharmacometabolomics to predict drug PD is highlighted, and we suggest not only considering the endogenous metabolites as static variables but to include also drug dose and temporal changes in drug concentration in these studies. Although there are many endogenous metabolite biomarkers identified to predict PK and more often to predict PD, validation of these biomarkers in terms of specificity, sensitivity, reproducibility and clinical relevance is highly important. Furthermore, the application of these identified biomarkers in routine clinical practice deserves notable attention to truly personalize drug treatment in the near future.

How Do Psychiatrists Apply the Minimum Clinically Important Difference to Assess Patient Responses to Treatment?

Symptom report scales are used in clinical practice to monitor patient outcomes. Using them permits the definition of a minimum clinically important difference (MCID) beyond which a patient may be judged as having responded to treatment. Despite recommendations that clinicians routinely use MCIDs in clinical practice, statisticians disagree about how MCIDs should be used to evaluate individual patient outcomes and responses to treatment. To address this issue, we asked how clinicians actually use MCIDs to evaluate patient outcomes in response to treatment. Sixty-eight psychiatrists made judgments about whether hypothetical patients had responded to treatment based on their pre- and posttreatment change scores on the widely used Positive and Negative Syndrome Scale. Psychiatrists were provided with the scale's MCID on which to base their judgments. Our secondary objective was to assess whether knowledge of the patient's genotype influenced psychiatrists' responder judgments. Thus, psychiatrists were also informed of whether patients possessed a genotype indicating hyperresponsiveness to treatment. While many psychiatrists appropriately used the MCID, others accepted a far lower posttreatment change as indicative of a response to treatment. When psychiatrists accepted a lower posttreatment change than the MCID, they were less confident in such judgments compared to when a patient's posttreatment change exceeded the scale's MCID. Psychiatrists were also less likely to identify patients as responders to treatment if they possessed a hyperresponsiveness genotype. Clinicians should recognize that when judging patient responses to treatment, they often tolerate lower response thresholds than warranted. At least some conflate their judgments with information, such as the patient's genotype, that is irrelevant to a post hoc response-to-treatment assessment. Consequently, clinicians may be at risk of persisting with treatments that have failed to demonstrate patient benefits.

Low protein diets in patients with chronic kidney disease: a bridge between mainstream and complementary-alternative medicines?

Dietary therapy represents an important tool in the management of chronic kidney disease (CKD), mainly through a balanced reduction of protein intake aimed at giving the remnant nephrons in damaged kidneys a "functional rest". While dialysis, transplantation, and pharmacological therapies are usually seen as "high tech" medicine, non pharmacological interventions, including diets, are frequently considered lifestyle-complementary treatments. Diet is one of the oldest CKD treatments, and it is usually considered a part of "mainstream" management. In this narrative review we discuss how the lessons of complementary alternative medicines (CAMs) can be useful for the implementation and study of low-protein diets in CKD. While high tech medicine is mainly prescriptive, prescribing a "good" life-style change is usually not enough and comprehensive counselling is required; the empathic educational approach, on which CAMs are mainly, though not exclusively based, may support a successful personalized nutritional intervention.There is no gold-standard, low-protein diet for all CKD patients: from among a relatively vast choice, the best compliance is probably obtained by personalization. This approach interferes with the traditional RCT-based analyses which are grounded upon an assumption of equal preference of treatments (ideally blinded). Whole system approaches and narrative medicine, that are widely used in the study of CAMs, may offer ways to integrate EBM and personalised medicine in the search for innovative solutions respecting individualization, but gaining sound data, such as with partially-randomised patient preference trials.

Evidence for Clinical Implementation of Pharmacogenomics in Cardiac Drugs

OBJECTIVE

To comprehensively assess the pharmacogenomic evidence of routinely used drugs for clinical utility.

METHODS

Between January 2, 2011, and May 31, 2013, we assessed 71 drugs by identifying all drug/genetic variant combinations with published clinical pharmacogenomic evidence. Literature supporting each drug/variant pair was assessed for study design and methods, outcomes, statistical significance, and clinical relevance. Proposed clinical summaries were formally scored using a modified AGREE (Appraisal of Guidelines for Research and Evaluation) II instrument, including recommendation for or against guideline implementation.

RESULTS

Positive pharmacogenomic findings were identified for 51 of 71 cardiovascular drugs (71.8%), representing 884 unique drug/variant pairs from 597 publications. After analysis for quality and clinical relevance, 92 drug/variant pairs were proposed for translation into clinical summaries, encompassing 23 drugs (32.4% of drugs reviewed). All were recommended for clinical implementation using AGREE II, with mean ± SD overall quality scores of 5.18±0.91 (of 7.0; range, 3.67-7.0). Drug guidelines had highest mean ± SD scores in AGREE II domain 1 (Scope) (91.9±6.1 of 100) and moderate but still robust mean ± SD scores in domain 3 (Rigor) (73.1±11.1), domain 4 (Clarity) (67.8±12.5), and domain 5 (Applicability) (65.8±10.0). Clopidogrel (CYP2C19), metoprolol (CYP2D6), simvastatin (rs4149056), dabigatran (rs2244613), hydralazine (rs1799983, rs1799998), and warfarin (CYP2C9/VKORC1) were distinguished by the highest scores. Seven of the 9 most commonly prescribed drugs warranted translation guidelines summarizing clinical pharmacogenomic information.

CONCLUSION

Considerable clinically actionable pharmacogenomic information for cardiovascular drugs exists, supporting the idea that consideration of such information when prescribing is warranted.

Emerging roles of frailty and inflammaging in risk assessment of age-related chronic diseases in older adults: the intersection between aging biology and personalized medicine

A chronic disease in older adults usually runs a course that is less predictable than in younger individuals. Unexplained variations in disease incidence, prognosis, therapeutic responses, and toxicity are frequently observed among older adults. This heterogeneity poses huge challenges to the current one-size-fits-all health care systems, and calls for more personalized managements of chronic diseases in older adults. Aging is characterized by progressive deterioration of bodily functions with increasing risk of failure over time. The entire process is hierarchically organized, and progresses from intracellular events to changes at systemic and ultimately organism levels at different rates among different individuals. Aging biology exerts great influences on the development and progression of most age-related chronic diseases. Thus, aging biology could contribute to the complexity of illnesses that increase with age, and aging biomarkers possess a great potential to enable personalized health risk assessment and health care. We review evidences supporting the roles of aging biomarkers in risk assessment of prevalent age-related diseases. Frailty phenotype is an objectively measured indicator of advanced-stage aging that is characterized by organism-level dysfunction. In contrast, altered inflammation markers level signifies an earlier stage between cellular abnormalities and systems dysfunction. Results of human observational studies and randomized controlled trials indicate that these measures, albeit simple, greatly facilitate classification of older patients with cancer, chronic kidney disease, cardiovascular diseases and type 2 diabetes mellitus into groups that vary in disease incidence, prognosis and therapeutic response/toxicity. As the detailed mechanisms underlying the complex biologic process of aging are unraveled in the future, a larger array of biomarkers that correlate with biologic aging at different stages will be discovered. Following the translational research framework described in this article, these research efforts would result in innovations in disease prevention and management that address the huge unmet health needs of aging populations.

Low-protein diets in CKD: how can we achieve them? A narrative, pragmatic review

Low-protein diets (LPDs) have encountered various fortunes, and several questions remain open. No single study, including the famous Modification of Diet in Renal Disease, was conclusive and even if systematic reviews are in favour of protein restriction, at least in non-diabetic adults, implementation is lagging. LPDs are considered difficult, malnutrition is a threat and compliance is poor. LPDs have been reappraised in this era of reconsideration of dialysis indications and timing. The definition of a normal-adequate protein diet has shifted in the overall population from 1 to 1.2 to 0.8 g/kg/day. Vegan-vegetarian diets are increasingly widespread, thus setting the groundwork for easier integration of moderate protein restriction in Chronic Kidney Disease. There are four main moderately restricted LPDs (0.6 g/kg/day). Two of them require careful planning of quantity and quality of food: a 'traditional' one, with mixed proteins that works on the quantity and quality of food and a vegan one, which integrates grains and legumes. Two further options may be seen as a way to simplify LPDs while being on the safe side for malnutrition: adding supplements of essential amino and keto acids (various doses) allows an easier shift from omnivorous to vegan diets, while protein-free food intake allows for an increase in calories. Very-low-protein diets (vLPDs: 0.3 g/kg/day) combine both approaches and usually require higher doses of supplements. Moderately restricted LPDs may be adapted to virtually any cuisine and should be tailored to the patients' preferences, while vLPDs usually require trained, compliant patients; a broader offer of diet options may lead to more widespread use of LPDs, without competition among the various schemas.

Individualisation of drug treatments for patients with long-term conditions: a review of concepts

OBJECTIVES

Patients and policy makers advocate that drug treatments should be individualised. However, the term is used in a variety of ways. We set out to identify the range of related terminology and concepts in the general field of individualisation, map out the relationships between these concepts and explore how patients' perspectives are considered.

DESIGN

We consulted members of an established patient and public involvement group about their experience of medicine taking for long-term conditions and their ideas about individualisation. We then conducted a scoping review of the literature to explore how terms surrounding individualisation of drug treatment are used and defined in the literature, and to explore the extent to which patients' perspectives are represented, with a view to informing future recommendations as to how individualisation can be operationalised.

METHODS

We identified relevant literature using a range of search strategies. Two researchers independently extracted definitions of terms using a template. Inductive and deductive methods were used to explore the data.

RESULTS

Definitions were categorised according to the following themes: medical management; pharmacogenetics, the patient's perspective; interactions between the healthcare provider and patient and management of long-term conditions.

CONCLUSIONS

Within the literature reviewed, the involvement of patients in the ongoing management of drug treatment was largely absent. We propose the use of a new term 'mutually agreed tailoring' (MAT). This describes the ongoing pharmacological management of conditions that incorporates patients' specific needs, experiences and existing strategies for using their medications, and the professionals' clinical judgement. This usually includes patients monitoring their symptoms and, with the support of the professional, making appropriate product, dose or timing adjustments as necessary. Our previous work suggests that many patients and doctors are successfully practising MAT, so we suggest that a formal description may facilitate wider utilisation of strategies that will improve patient outcomes.

Which low-protein diet for which CKD patient? An observational, personalized approach

OBJECTIVES

Low protein diets (LPDs) are milestones in chronic kidney disease (CKD). Concerns over compliance and safety limit their use. The aim of this study was to test the feasibility and main results of a multiple-choice approach to LPDs, adapted to patient preferences.

METHODS

From December 2007 to January 2013, all CKD patients (stages 4/5; progressive stage 3) without contraindications (malnutrition, short life expectancy), were offered two main LPDs (proteins 0.6 g/kg daily): Vegan supplemented (LPD-KA) or with "aproteic" commercial food (LPD-ACF). LPDs followed a qualitative approach based on forbidden and allowed food; one to three free meals per week, and flexible control policy (1-3 mo). Start of dialysis, death, and combined outcome (death-dialysis) were analyzed by Kaplan-Meier curves and Cox model. Comparison with dialysis in patients with glomerular filtration rate (GRF) <15 mL/min, (corresponding to "early" dialysis start) employed standardized mortality rates, with respect to the Italian and the United States Dialysis Registry.

RESULTS

One hundred eighty-five patients (222 patient-years) started at least a trial of LPD-KA, 122 (177 patients-years) LPD-ACF; only 3 patients with GFR <30 mL/min denied an LPD trial. Patients who chose LPD-KA were younger than those on LPD-ACF (63 versus 74 y), had less comorbidity (82% versus 93%), higher proteinuria (1.4 versus 0.7 g/d) and lower GFR (17 versus 23 mL/min) (P < 0.001). Median daily protein intake was 0.7 g/kg on both diets (Maroni-Mitch formula). The combined outcome (death or dialysis) was not influenced by the diet chosen (Cox analysis). Relative risk for death on the diet (patients with GFR <15 mL/min) was 0.5 with respect to the Italian Registry and 0.3 to the United States Dialysis Registry. The diets had comparable costs (1 y on dialysis: 50 patient-years on LPD).

CONCLUSIONS

The choice of diet is strictly linked to patient characteristics, thus supporting a multiple-choice offer. Once corrected for baseline data, both LPDs led to similar results, suggesting at least survival equivalence with dialysis, at lesser cost.

Pharmacometabonomics and personalized medicine

BACKGROUND

Pharmacometabonomics is a new branch of science, first described in 2006 and defined as 'the prediction of the effects of a drug on the basis of a mathematical model of pre-dose metabolite profiles'. Pharmacometabonomics has been used to predict drug metabolism, pharmacokinetics (PK), drug safety and drug efficacy in both animals and humans and is complementary to both pharmacogenomics (PGx) and pharmacoproteomics.

METHODS

A literature review using the search terms pharmacometabonomics, pharmacometabolomics, pharmaco-metabonomics, pharmaco-metabolomics and the singular form of all those terms was conducted in October 2012 using PubMed and Web of Science. The review was updated until mid April 2013.

RESULTS

Since the original description of pharmacometabonomics in 2006, 21 original publications and eight reviews have emerged, covering a broad range of applications from the prediction of PK to the prediction of drug metabolism, efficacy and safety in humans and animals.

CONCLUSIONS

Pharmacometabonomics promises to be an important new approach to the delivery of personalized medicine to improve both drug efficacy and safety for patients in the future. Pharmacometabonomics is particularly powerful as it is sensitive to both genetic and environmental factors such as diet, drug intake and most importantly, a person's microbiome. PGx is now over 50 years old and although it has not achieved as much as some hoped, it is starting to have important applications in personalized medicine. We predict that pharmacometabonomics will be equally important in the next few decades and will be both valuable in its own right and complementary to pharmacoproteomics and PGx.

Personalized medicine - where do we stand? Pouring some water into wine: a realistic perspective

Abstract

Reviewing the past and the present status of personalized medicine, the hope and promise from several years ago was critically compared to what is really achieved to tailor the drug treatment according to the patient’s individuality. The basis for consideration is what we know about the variant of the disease the patient is suffering from, and about the mechanisms influencing the plasma concentration-time profile, such as activity of metabolizing enzymes and transporters. In cancer treatment, drugs are currently selected regarding molecular properties of the cancer tissue, eg, expressing receptors such as HER2 receptor. Currently diagnostic tests are available allowing to detect somatic cell mutations that can be used to guide drug selection. Unfortunately, tumor heterogeneity and developing resistance by further mutations may limit the success of the therapy determined by molecular diagnostics. The present status can be described that in drug kinetics we know the influencing factors and we understand the mechanisms. However, only in a few cases the genetic background is the main determinant of kinetic variability, and environmental and other factors have an additional important role. Therefore, much more has to be done before we can translate the accumulating knowledge into a benefit for the patient. Only then, we can speak about personalized medicine.