Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

European Neuropsychopharmacology (2017) 27, 205–216

www.elsevier.com/locate/euroneuro

REVIEW

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring Lisa C. Martiala,b, Rob E. Aarnoutsea,b, Martina Mulderc, Arnt Schellekensc,d,e, Roger J.M. Brüggemanna,b, David M. Burgera,b, Aart H. Schenec,d, Albert Batallac,e,n a

Radboud University Medical Center, Department of Pharmacy, Nijmegen, The Netherlands Radboud Institute for Health Sciences, Nijmegen, The Netherlands c Radboud University Medical Center, Department of Psychiatry, Reinier Postlaan 10, Route 966, 6500 HB Nijmegen, The Netherlands d Donders Institute for Brain, Cognition and Behavior, Radboud University Nijmegen, Nijmegen, The Netherlands e Radboud University, Nijmegen Institute for Scientist-Practitioners in Addiction, Nijmegen, The Netherlands b

Received 8 August 2016; received in revised form 2 December 2016; accepted 5 January 2017

KEYWORDS

Abstract

TDM; Dried blood spot; Finger prick; Antipsychotics; Mood stabilizers; Antidepressants

Assessment of drug concentrations is indicated to guide dosing of a selected number of drugs used in psychiatry. Conventionally this is done by vena puncture. Novel sampling strategies such as dried blood spot (DBS) sampling have been developed for various drugs, including antipsychotics, antidepressants and mood-stabilizers. DBS sampling is typically performed by means of a finger prick. This method allows for remote sampling, which means that patients are not required to travel to a health care facility. The number of DBS assays for drugs used in psychiatry has increased over the last decade and includes antidepressants (tricyclic and serotonin and/or norepinephrine reuptake inhibitors), mood stabilizers and first- and secondgeneration antipsychotics. Available assays often comply with analytical validation criteria but are seldom used in routine clinical care. Little attention has been paid to the clinical validation

n Corresponding author at: Radboud University Medical Center, Department of Psychiatry, Reinier Postlaan 10, Route 966, 6500 HB Nijmegen, The Netherlands. Fax: +31 243 610 304. E-mail address: [email protected] (A. Batalla).

http://dx.doi.org/10.1016/j.euroneuro.2017.01.009 0924-977X/& 2017 Elsevier B.V. and ECNP. All rights reserved.

206

L.C. Martial et al. and implementation processes of home sampling. Ideally, not only medicines but also clinical chemistry parameters should be measured within the same sample. This article reflects on the position of DBS remote sampling in psychiatry and provides insight in the requisites of making such a sampling tool successful. & 2017 Elsevier B.V. and ECNP. All rights reserved.

1.

Introduction

Monitoring of plasma drug concentrations and clinical chemistry tests is an important component of modern personalized pharmacological therapy in psychiatry. Traditionally, plasma or serum are used for these analyses, as these matrices are easy to handle and stable upon storage. Repeated sampling can be challenging, however, as venous sampling can only be performed in equipped health-care facilities and as such poses a burden for the patient. Dried blood spot (DBS) sampling provides an elegant alternative. It was first introduced more than fifty years ago as a sampling method in newborn screening ‘heelprick’ for phenylketonuria (Newman and Starr, 1971). A DBS sample can be obtained by means of a finger prick by spotting a drop of blood directly onto filter paper. The sample can be taken by patients themselves at home and can be sent to the laboratory using regular post mail. Over the past decade, there has been an increase in the development of bioanalytical assays to measure samples collected by means of DBS sampling (Stove and Spooner, 2015). Currently, its use has expanded to a wide range of purposes such as for preclinical and clinical research, large epidemiologic studies, HIV or Hepatitis C screening or analysis of laboratory tests (e.g. Vitamin D) (Chang et al., 2015; Hoeller et al., 2015; Soulier et al., 2016). However, despite the fact that novel strategies for (remote) monitoring are warranted in psychiatry (e.g. in order to improve medication compliance) (Millan et al., 2015) the potential applicability of DBS in this field remains largely unexplored. One of the most promising clinical applications of DBS is for therapeutic drug monitoring (TDM). TDM allows individualizing of drug doses, guided by measurement of plasma drug concentrations. The aim is to keep a patient's plasma concentrations within a therapeutic range (Aarnoutse et al., 2003; Bruggemann and Aarnoutse, 2015; Lempers et al., 2015). Several criteria can be identified that are required in order for TDM to be useful. First, the patient's response cannot be assessed by clinical observations or predicted by the administered dose. Second, there exists a large inter-individual variation in pharmacokinetic parameters. Third, the relationship between the drug concentration and the pharmacological effect (or toxicity) must be known. Fourth, there is a narrow range of concentrations that are effective and well tolerated. Fifth, a validated laboratory assay must be available with a fast turn around time (Aarnoutse et al., 2003). Especially in case of long-term treatment in outpatient settings, DBS can be a very useful tool for TDM, providing a novel methodology that might help clinicians obtaining adequate responses, preventing concentration-related adverse events, monitoring interactions and improving treatment compliance.

Excellent candidate drugs for TDM are immunosuppressants, antimicrobial drugs (HIV, HepC and antifungal drugs) and antiepileptics but also some antidepressants, antipsychotics and mood stabilizers as they are used for long periods in outpatient settings, display wide inter-patient variability in their pharmacokinetics, have small therapeutic windows and require repetitive concentration measurements (Barraclough et al., 2011; Bruggemann and Aarnoutse, 2015; Martial et al., 2016; Wilhelm et al., 2014). In psychiatry, TDM is routinely indicated for a selection of drugs, including tricyclic antidepressants (TCAs), mood stabilizers (lithium, valproic acid, carbamazepine) and clozapine (Hiemke, 2016). Despite the fact that DBS sampling has gained more attention over the past years and that it has a clear potential for application in routine clinical care, this sampling method still plays a marginal role in today's clinical practice in psychiatry. The aim of this manuscript is to describe the current progress of DBS in psychiatry, to discuss the benefits and drawbacks of this sampling method and to gain insight in the future potential applications of DBS in monitoring psychiatric treatment. To this end, we performed a review of the literature on (1) DBS assays measuring drug concentrations, (2) their clinical validation and (3) implementation studies for drugs commonly used in psychiatry. In addition, we provide a broader view of DBS in the context of TDM for psychiatric drugs and discuss the steps that are required in order to make DBS successful in this field of medicine.

2. 2.1.

Experimental procedures Search strategy

The electronic search was performed using PubMed. A combination of the following terms was used: “dried blood spot” [All Fields] AND “psychiatry” [MeSH Terms], “antipsychotic” [All Fields], “antidepressant” [All Fields], “mood stabilizer” [All Fields], “antiepileptic” [All Fields]; “DBS (not deep brain stimulation)” [All Fields] AND “psychiatry” [MeSH Terms], “antipsychotic” [All Fields], “antidepressant” [All Fields], “mood stabilizer” [All Fields], “antiepileptic” [All Fields]. All studies published in any language up to April 2016 and indexed in the mentioned database were considered. The data was extracted by two reviewers (LM and AB). In addition, as to gather a broader view of DBS sampling and place this in the context of psychiatric treatment, we searched the literature for general review articles on DBS and pharmacological therapy.

2.2.

Inclusion and exclusion criteria

Included were DBS studies involving antipsychotics, antidepressants, mood stabilizers and antiepileptic drugs: (1) investigating an analytical method to measure drug concentrations, (2) clinical validation studies and (3) implementation studies. Excluded were

DBS assays

Antipsychotics Clozapine

Therapeutic drug monitoring General items*

DBS-specific items#

Stability tested

Extreme weather conditions (1C)

Clinical validation studies

Recommendations for TDM (Hiemke et al., 2011; National Institute for Health and Care Excellence, 2014a, b, 2016; Nederlandse Vereniging voor Psychiatrie, 2012, 2013, 2015)

Recommendations for regular laboratory tests (National Institute for Health and Care Excellence, 2014a, b, 2016; Nederlandse Vereniging voor Psychiatrie, 2012, 2013, 2015)

(Patteet et al., 2015b)

+

Spot volume: Hematocrit: -

+

N/A

(Saracino et al., 2011)

+

Spot volume: -

+

N/A

(Patteet et al., 2015a) Passing-Bablock 1. Titration regression did not show any systemic or 2. After dose adjustments proportional bias. Agreement not assessed. 3. For special indication: Predictive performance not quantified. DBS * Start/stop smoking or excesresults led to identical clinical interpretasive caffeine use tion compared to serum concentration * Fever as a result of an inflam(sensitivity of 92%). (n =10) matory reaction * Suspected toxicity or adverse events N/A * Potential drug interaction * Treatment adherence

1. Weekly during 18 weeks, subsequently every month: cell blood count (leucocytes and differential blood count) 2. 1st, 2nd, 3rd, 6th month and subsequently every 12 months: glucose and glycated hemoglobin 3. 3rd month and subsequently every 12 months: lipid profile

+

N/A

N/A

N/A

Hematocrit: -

Other antipsychotics

Mood stabilizers Lithium

(Patteet et al., 2015b) + (amisulpride, aripiprazole, bromperidol, haloperidol, olanzapine, paliperidone, pipamperone, quetiapine, risperidone, zuclopenthixol) (Mercolini et al., 2014) + (ziprasidone)

N/A

Spot volume: Hematocrit: -

Spot volume: Hematocrit: -

(Patteet et al., 2015a) Passing-Bablock showed proportional bias for amisulpride, quetiapine and metabolite and paliperidone. Agreement not assessed. Predictive performance not quantified. DBS results led to identical clinical interpretation compared to serum concentration. (n = 8-36, depending on compound) (Mercolini et al., 2014) Neither regression analysis nor agreement assessed. Predicted and observed plasma concentration ratios quantified, very close to 1. (n =10)

Not routinely advised, only for After 6 weeks, 3rd month and special indication: subsequently every 12 months: * Treatment adherence glucose and lipid profile * Suspected toxicity or adverse events

N/A

1. Titration 2. Routinelyevery 3-6 months 3. After dose adjustments 4. For special indication:

207

1. Every 3-6 months: cell blood count (including differential), electrolytes, renal function, thyroid function * Suspected toxicity or adverse 2. At least every 6 months: renal function events * Potential drug interaction 3. At least every 12 * Treatment adherence months: above mentioned parameters plus glucose and lipid profile

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

Table 1 Overview of drugs commonly measured for therapeutic drug monitoring (TDM) purposes and related clinical chemistry tests and the analytical and clinical validation.

208

Table 1 (continued ) DBS assays Valproic acid

(Linder et al., 2015)

Therapeutic drug monitoring +

(Pohanka et al., 2014)

+

(Rhoden et al., 2014)

+

Spot volume: + Hematocrit: +

-20 - 45

N/A

+

N/A

(Pohanka et al., 2014)

+

45

Spot volume: + Hematocrit: +

+

-20 - 45

Linear regression. * Treatment adherence Agreement assessed with Bland-Altman plot: average bias 4-20%. Predictive performance not quantified. (n = 32) (Rhoden et al., 2014) No regression analysis, agreement not assessed. Predictive performance in terms of MAPE or MPPE was not quantified but serum concentration fairly predicted from DBS. (n = 17) (Kong et al., 2014) Deming regression showed systematic error. Agreement moderate (Bland-Altman). Predictive performance not quantified. (n = 81) N/A 1. Titration

Spot volume: + Hematocrit: +

Spot volume: Hematocrit: +

Carbamazepine

Lamotrigine

(Linder et al., 2015)

+

1. In the 3rd month and subsequently at least every 12 months: cell blood count, electro* Suspected toxicity or adverse lytes, liver function, events glucose and lipid profile * Potential drug interaction

+

1. Titration 2. After dose adjustments 3. For special indication:

1. After 4-6 weeks (enzyme induction): cell blood count, electrolytes, renal function, liver function rd month and N/A * Suspected toxicity or adverse 2. In the 3 subsequently every 12 events months: above para* Potential drug interaction meters plus glucose (Shokry et al., 2015) * Treatment adherence Linear regression. Potential systematic or and lipid profile 2. After 4-6 weeks (enzyme induction) 3. After dose adjustments 4. For special indication:

(Shah et al., 2013)

+

Spot volume: +

+

40

(Shokry et al., 2015)

+

Hematocrit: + Spot volume: + Hematocrit: +

+

-20 - 37

(Aburuz et al., 2010)

+

Spot volume: + -Hematocrit: +

N/A

(Shah et al., 2013)

+

Spot volume: + Hematocrit: -

+

40

N/A

Not routinely advised, only for Not routinely advised special indication: * Lack of effect (after 4-6 weeks) * Suspected toxicity or adverse events * Potential drug interaction

(Linder et al., 2015)

+

Spot volume: + Hematocrit: +

+

-20 -45

N/A

* Treatment adherence

L.C. Martial et al.

proportional bias not assessed. Agreement not assessed. Predictive performance not quantified. (n = 19) (Kong et al., 2014) Deming regression showed no proportional or systematic error. Poor agreement, assessed with Bland-Altman. Predictive performance not quantified. (n =101) (Aburuz et al., 2010) Neither regression analysis nor agreement by means of Bland-Altman plot. DBS concentrations close to plasma concentration but predictive performance quantified. (n = 12)

Spot volume: + Hematocrit: 7

+

N/A

Serotonin Norepi- (Berm et al., 2014) (ven- + nephrine Reup- lafaxine, take Inhibitors O-desmethylvenlafaxine) (SNRIs)

Spot volume: + Hematocrit: +

+

N/A

Selective Seroto- (Deglon et al., 2010) + nin Reuptake (fluoxetine,norfluoxetine, reboxetine, Inhibitors (SSRIs)and Nor- paroxetine) epinephrine reuptake inhibitors (NRI)

Spot volume: Hematocrit: -

+

-20 - 40

(Berm et al., 2016) Passing-Bablock regression: proportional error for amitriptyline and nortryptiline. Agreement of DBS vs. plasma of nortryptiline assessed by Bland-Altman: poor agreement (430% bias). Predictive performance not quantified. (n = 29-61, depending on compound)

(Berm et al., 2016) Passing-Bablock regression: proportional error for venlafaxine, O-desmethylvenlafaxine . Agreement not assessed. Predictive performance not quantified. DBS results led to identical clinical interpretation compared to serum concentration (sensitivity of 100%). (n =28) N/A

1. Titration (imipramine, Not routinely advised amitriptyline, nortriptyline, clomipramine) 2. For special indication: * Lack of effect (after 4-6 weeks) * Suspected toxicity or adverse events * Potential drug interaction * Treatment adherence Not routinely advised, only for Not routinely advised special indication: * Lack of effect (after 4-6 weeks) * Suspected toxicity or adverse events * Potential drug interaction * Treatment adherence Not routinely advised, only for Not routinely advised special indication: * Lack of effect (after 4-6 weeks) * Suspected toxicity or adverse events * Potential drug interaction * Treatment adherence

TDM =Therapeutic drug monitoring; DBS=Dried blood spot; N/A=Not available. * General items include accuracy and precision. #DBS-specific items include influence of hematocrit effect and influence of blood spot volume. + = assessed with good results, 7 = assessed with poor results.

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

Antidepressants Tricyclic antide- (Berm et al., 2015) (imi- + pressants pramine, amitriptyline, (TCAs) nortriptyline, clomipramine)

209

210

L.C. Martial et al.

studies involving drugs without indication for treating the following psychiatric disorders according to the Diagnostic and Statistical Manual for Mental Disorders ‐ Fifth Edition (DSM-5): schizophrenia spectrum and other psychotic disorders, bipolar and related disorders and depressive disorders (American Psychiatric Association, 2013).

2.3.

Data collation

2.3.1. DBS assays We screened if the published assays complied with the Food and Drug Administration (FDA) or European Medicines Agency (EMA) international guidelines on method validation (European Medicines Agency, 2011; Food and Drug Administration, 2001). These documents provide guidance on bioanalytical method validation of biological matrices with recommendations and acceptance criteria of critical parameters including accuracy, precision, selectivity, sensitivity, and stability. Despite the absence of DBS-specific recommendations on method validation in FDA or EMA-guided guidelines, there is good agreement on which parameters to test (i.e. accuracy, precision and influence of hematocrit or spot volume), and an excellent review guiding the validation of DBS (Jager et al., 2014a). General items of the analytical DBS assays include accuracy and precision as well as stability of DBS samples upon storage. Details on DBS specific items including influence of hematocrit and spot volume were also gathered. As DBS sampling is ultimately suitable for home sampling with subsequent transport by normal postal services, extreme weather conditions likely to occur in mail boxes should ideally be evaluated. It is known that in summer, temperatures of 20 1C above outside temperature can occur in mailboxes. Pertinent details from each study were recorded and are presented in Table 1.

mean absolute difference (or mean ratio or mean relative difference) and the corresponding 95% levels of agreement. It is left to the discretion of the clinician whether the agreement of two methods is acceptable or not (Bland and Altman, 1986). An example is shown in Figure 2. The ultimate goal of DBS concentration measurement is to predict the corresponding venous plasma concentration. In case of inadequate agreement several correction factors can be used to improve the prediction of plasma concentrations, such as conversion factors based on the ratio [DBS concentration: plasma concentration], linear regression or by taking into account the hematocrit or the fraction of the drug that is bound to blood cells (fBC). Hematocrit is an important parameter as it can influence the concentration measured in DBS. Formulas exist to predict the plasma concentration using DBS concentrations, hematocrit and fBC such as the one developed by Li and Tse, (2010): (DBS[drug a]/ [1 hematocrit])x(1 fBC)=plasma[drug A] (Li and Tse, 2010). The predictive performance can be quantified by calculating the median percentage prediction error (MPPE), given by the median [100%x(Predicted concentration Observed concentration)/Observed concentration]. In addition, the median absolute percentage prediction error (MAPE) can be calculated, by [100%x|(Predicted concentration Observed concentration)|/Observed concentration]. MMPE is a measure of bias while MAPE is a measure of precision (Barraclough et al., 2011; Sheiner and Beal, 1981). Acceptance criteria can vary but often values of MMPE and MAPE o15% are applied (Jager et al., 2014b; Ting et al., 2006).

3. 3.1.

2.3.2. Clinical validation studies A clinical validation study consists of a comparison of the venous plasma concentration (gold standard) with dried blood spots concentrations obtained by a finger prick in the same patient taken at the same time. Ideally, the whole concentration range is validated based on a large enough sample size (n=40) (Clinical and Laboratory Standards Institute, 2013). Clinical validation studies were evaluated for three different measures of agreement: (1) Passing-Bablock/Deming regression analysis, (2) Bland-Altman analysis displaying agreement and (3) predictive performance. With Passing-Bablock regression analysis, concentrations of the new method (DBS) are plotted against the gold standard (plasma concentration). Passing-Bablock regression analysis allows for measurement errors (residual errors) on both the x and y-axis and is not hampered by a few outliers. Deming regression can be an alternative but is more influenced by outliers. In regression analysis for method comparison it is not the regression coefficient (R2) that counts, as this correlation only tells how well two methods correlate but not how well they agree, but the slope and intercept (Bland and Altman, 1986; Sheiner and Beal, 1981). A slope significantly different from 1 indicates a proportional bias; an intercept significantly different from 0 indicates a systematic bias. For good interpretation, both the estimates of the slope and intercept as well as their 95% confidence intervals (CI) are shown. An example of a Passing-Bablock regression plot is displayed in Figure 1. Bland–Altman plots provide insight in the agreement between two methods. The plots give insight in the variability in agreement over the concentration range and help to identify if the DBS assay is interchangeable with the gold standard and whether it can really guide clinical decision making with consequent dose-adaptations. The x-axis usually displays the mean of the two matched concentrations and the y-axis displays their absolute difference, their ratio or the relative difference in %. Horizontal trendlines include the

Results Study selection

One hundred and thirty three studies were identified. 121 did not meet the a priori selection criteria or met the exclusion criteria. Twelve studies were initially included. The references and authorship of the selected papers were also screened for relevant articles, yielding three additional papers. From the fifteen papers included, twelve were DBS assays (Aburuz et al., 2010; Berm et al., 2014, 2015; Deglon et al., 2010; Linder et al., 2015; Mercolini et al., 2014; Patteet et al., 2015b; Pohanka et al., 2014; Rhoden et al., 2014; Saracino et al., 2011; Shah et al., 2013; Shokry et al., 2015), from which five also reported a clinical validation study (Aburuz et al., 2010; Mercolini et al., 2014; Pohanka et al., 2014; Rhoden et al., 2014; Shokry et al., 2015). Three articles were clinical validation studies (Berm et al., 2016; Kong et al., 2014; Patteet et al., 2015a). No implementation studies were found. Most publications date from 2013 onwards showing that the development of DBS assays measuring drugs used in psychiatry is relatively new. DBS assays that have been developed for drugs acting on the central nervous system include antidepressants (tricyclic antidepressants [TCAs], Selective Serotonin Reuptake Inhibitors [SSRIs], Norepinephrine Reuptake Inhibitors [NRI] and Serotonin Norepinephrine Reuptake Inhibitors [SNRIs]) (Berm et al., 2014, 2016, 2015; Deglon et al., 2010), mood stabilizers (Aburuz et al., 2010; Kong et al., 2014; Linder et al., 2015; Pohanka et al., 2014; Rhoden et al., 2014; Shah et al., 2013; Shokry et al., 2015) and first- and second-generation antipsychotics (Mercolini et al., 2014; Patteet et al., 2015a,b; Saracino et al., 2011).

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

211

Figure 1 Passing-Bablock regression plot. Plot was adapted from an R-script from on the mcr package (Holmes, 2015). Both the regression line, the 95% CI and the line of identity are displayed.

Figure 2 Bland-Altman plot displaying the average concentration against the difference in concentration with the two sampling methods. Upper limit of agreement is [mean + 2 SD] and lower limit of agreement is [mean-2 SD].

3.2.

DBS assays and clinical validation studies

The following section provides a qualitative description of the included studies, grouped by type of medication (see Table 1 for details).

3.2.1. Antipsychotics Two DBS assays for clozapine were found (Patteet et al., 2015b; Saracino et al., 2011), one of them included 14 other antipsychotics and (partially) complied with international guidelines (FDA and EMA) for analytical method validation (Patteet et al., 2015b). The validation was performed according to their in-house guidelines based on those international guidelines (Patteet et al., 2015b). In terms of stability it has been reported that 10–60% of the inactive

metabolite clozapine-N-oxide is converted back to clozapine upon drying and storage, depending on the pretreatment of the DBS filter paper (Temesi et al., 2013). This might result in an overestimation of the blood concentration and can have major effects on the interpretation of the results. For the two published DBS-assays on clozapine (Patteet et al., 2015b; Saracino et al., 2011) it is unclear to which extent clozapine-N-oxide was converted back to clozapine. The group of Patteet et al., (2015a) also performed a clinical validation, in which both serum and DBS samples from the same patients were analysed (Patteet et al., 2015a), obtaining a high sensitivity (Patteet et al., 2015a) (Table 1). One DBS assay of ziprasidone has been published, showing a good agreement between DBS and plasma concentrations (Mercolini et al., 2014) (Table 1).

212

L.C. Martial et al.

3.2.2. Mood stabilizers Three DBS assays have been published for valproic acid (Linder et al., 2015; Pohanka et al., 2014; Rhoden et al., 2014), with one fully complying (Linder et al., 2015) and another partially complying with EMA guidelines (Pohanka et al., 2014). Of the three clinical validation studies published (Kong et al., 2014; Pohanka et al., 2014; Rhoden et al., 2014), only one quantified the relative predictive performance (Rhoden et al., 2014). Three assays have been published reporting a DBS method for carbamazepine (Linder et al., 2015; Shah et al., 2013; Shokry et al., 2015), two of them combined the measurement of both carbamazepine and the active metabolite carbamazepine 10,11-epoxide (Shah et al., 2013; Shokry et al., 2015). Only one assay complied with the EMA guideline on method validation (Linder et al., 2015). Two clinical validation studies have been published (Kong et al., 2014; Shokry et al., 2015), with only one showing relevant data on the differences between DBS and plasma (Kong et al., 2014) (Table 1). Three DBS assays have been developed for lamotrigine (Aburuz et al., 2010; Linder et al., 2015; Shah et al., 2013), of which one study complied with the EMA international guideline for method validation (Linder et al., 2015). One study reported a clinical validation with good agreement between DBS and plasma concentrations (Aburuz et al., 2010) (Table 1). 3.2.3. Antidepressants Berm et al., (2015) developed an assay for several TCAs and their active metabolites (Berm et al., 2015), which complied with the FDA international guideline (Table 1). A subsequent clinical validation study was performed using Passing-Bablock regression, which showed a proportional error for all components but no systematic bias. A Bland– Altman plot of nortriptyline showed a relative bias of 730% (Berm et al., 2016). More than 75% of paired samples resulted in the same clinical interpretation (Table 1). In terms of stability upon storage of DBS samples, it is known that the metabolite N-oxide amitriptyline is converted back to amitriptyline depending on sampling material (varying between 10–88%), which means that an overestimation of the parent compound might occur with DBS measurement (Temesi et al., 2013). This effect was not considered in the published DBS-assay for amitriptyline (Berm et al., 2015). Deglon et al. (2010) developed an assay for SSRIs (fluoxetine, norfluoxetine and paroxetine) and one NRI (reboxetine) (Deglon et al., 2010), and the study of Berm et al., (2014) included one assay for the SNRI venlafaxine and the metabolite O-desmethylvenlafaxine (Berm et al., 2014). Only the last complied with the FDA guideline (Berm et al., 2014). One clinical validation study has been published, investigating venlafaxine and his metabolite, reporting a high sensitivity (Berm et al., 2016) (Table 1).

4.

Discussion

This is the first review exploring the potential applicability of a novel method for remote sampling in psychiatry. We found 15 studies involving DBS assays and clinical validation studies of several antipsychotics, mood stabilizers and antidepressants. Overall, the included studies showed some

heterogeneity in the parameters used for bioanalytical method validation (i.e. accuracy, precision and influence of hematocrit or spot volume) and often a poor agreement (or lacking agreement measures) between plasma and DBS methods. No implementation studies were found.

4.1.

DBS for TDM of psychiatric drugs

One of the most interesting clinical applications of DBS in psychiatry is TDM. Psychiatric medications are frequently used for long-term treatment in outpatient settings and some require periodically concentration measurements. However, it is important to notice that TDM is only routinely indicated for a selection of drugs in clinical practice (Hiemke, 2016) (Table 1). Considering the group of antipsychotics, routine sampling for TDM is indicated only for clozapine during the full course of treatment (Table 1). Clozapine complies with the requirements for TDM given the well-defined relationship between concentration and effect, which cannot be predicted from the dose, and given the important interindividual variability in exposure. Changes in clinical status such as infections, starting or stopping smoking or dosages above 600 mg prompt for concentration measurement. The importance of regular monitoring of other antipsychotics than clozapine for dose titration is a question of debate (Hiemke, 2016). TDM guidelines highly recommend monitoring of olanzapine, amisulpiride, haloperidol, fluphenazine, perazine, perphenazine and thioridazine and recommend or consider it useful for almost all current antipsychotics [see (Hiemke et al., 2011) for details]. In treatment guidelines however, monitoring of other antipsychotics than clozapine is restricted to special indications, such as for assessment of adherence or in case of remarkable adverse effects using standard or low doses, which may point to interactions or poor metabolizer phenotypes (Table 1). This disagreement might be explained by the lack of well-designed clinical trials on medical and economic benefits of TDM in this group of drugs (Hiemke, 2016). The implementation of novel techniques such as DBS might help to expand the use of plasma concentration determination in daily clinical practice and help to explore the potential benefits of TDM, as such benefits can only be gained if the method is adequately integrated into the clinical treatment process (Hiemke, 2016). Regarding mood stabilizers, lithium is the compound for which TDM is definitely mandatory. It complies with all requisites for TDM. Lithium has a small therapeutic window, and changes in clinical status such as fever, changes in renal function or diarrhea might have a big impact on excretion and require immediate concentration measurement (Table 1). Although lithium is an interesting candidate for DBS, no assays have been published yet. Other frequently used mood stabilizers, such as valproic acid and carbamazepine, are also potential candidates for DBS sampling as both drugs are frequently subjected to TDM. For lamotrigine regular TDM is not routinely recommended (National Institute for Health and Care Excellence, 2014a; Nederlandse Vereniging voor Psychiatrie, 2015) (Table 1). Considering the different groups of antidepressants, routine TDM is only highly recommended for TCAs (Hiemke et al., 2011) (Table 1). TDM of SSRIs, NRI and SNRIs are not

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

Figure 3 DBS sampling process. These photos were kindly supplied by Dried Blood Spot Laboratories (DBSL), Geleen, The Netherlands.

routinely recommended and plasma concentrations are exceptionally monitored in daily practice during titration. Monitoring is confined to lack of effect, treatment adherence or in case of adverse events presumably related to high concentrations (Table 1). Therefore, remote sampling might be less applicable to those groups of antidepressants.

4.2. Current advantages and disadvantages of DBS sampling DBS samples can be obtained by a minimally invasive procedure. Figure 3 shows how the sampling procedure looks like. This method has several advantages over conventional sampling such as (1) the small amount of blood required; (2) sampling can be performed at home making DBS an attractive option in case of remote areas but also when regular or lifelong sampling is required; (3) the possibility to draw trough concentrations or 12 h concentrations on which reference values are based but that are difficult to ‘capture’ with hospital-based sampling; (4) low costs are involved in the transport and storage of samples as compounds are often very stable in their dried form; and (5) there is no contamination risk for the environment. Moreover, DBS sampling can be cost-saving as compared to conventional blood sampling (Jager et al., 2014a). In contrast to the obvious advantages of DBS in terms of convenience in sampling, DBS has some important disadvantages worth mentioning. First, the development of a novel method of analysis is costly as it usually takes several months of full-time work for one laboratory technician, given the criteria to comply with as stated by regulatory agents such as the EMA and FDA (European Medicines Agency, 2011; Food and Drug Administration, 2001). Some specific challenges exist when developing a DBS assay that are absent for plasma assays. Probably the main challenge is to quantify the concentration of the compound of interest based on an uncertain amount (volume) of blood spotted on the filter paper. When a fixed part of the spot is analysed, hematocrit determines the volume of blood present in the sample (Capiau et al., 2013). The so-called ‘hematocrit effect’ influences the spreading of blood onto the filter paper. Not taking this effect into account severely biases the concentration measurement of the compound of

213

interest. To overcome this issue, Patteet et al., (2015a) analysed the whole spot in their assay of antipsychotics instead of a fixed part of the spot (Patteet et al., 2015a). Although this might be an option when fixed amounts of blood are spotted under controlled circumstances in a clinical setting, for patients sampling at home it remains a challenge to deliver reproducible and good quality spots of uniform size. Several innovative solutions have been developed in an attempt to overcome the hematocrit effect. DBS assessments of hematocrit have been published such as oncard UV analysis or measurement of potassium as a ‘marker’ for hematocrit (Capiau et al., 2013; De Kesel et al., 2014; Miller et al., 2013). Potassium measurement requires an additional development step and dedicated analytical equipment not readily present in all routine pharmacy laboratories. Another elegant option is a device allowing for a fixed amount of blood to be applied onto the filter paper and allowing for whole spot analysis (DBS Technology, http://dbs-system.ch/technology/). With dried plasma spots (DPS) the drop of blood is filtered on the paper by ‘trapping’ the blood cells on one layer only allowing the plasma to pass onto the second layer which is used for analysis (Baldelli et al., 2015; Ikeda et al., 2014; Li et al., 2012). Qyntest™ designed a sample card with woven polymer layers allowing for a homogenous blood flow with predictable blood volumes (Mengerink et al., 2015). Mitra™ completely abandoned the filter paper and designed a device that allows for sampling exactly 10 microliters of blood (Spooner et al., 2015). The main disadvantages of those devices are their high costs as compared to the conventional DBS filter paper (Whatman903™). Second, after developing the method, whole blood DBS concentrations have to be recalculated towards plasmabased references values for clinical interpretation. In case the hematocrit effect is large, actual hematocrit values are required for this calculation. In homogenous populations hematocrit values can be assumed constant, so a population value can be used as proposed by Jager et al. (2014b). Third, one of the main disadvantages often stated for DBS is risk for poor quality of sampling by the patient. It should be acknowledged that good quality spots can only be obtained after careful instruction and regular practice. For drugs or parameters only monitored for special indications, the advantages of DBS home sampling may not be appreciated (e.g. antipsychotics other than clozapine, antidepressants other than TCAs or lamotrigine).

4.3.

Clinical chemistry tests

For an optimal and efficient monitoring, DBS analysis should ideally also measure laboratory blood tests that are frequently assessed in psychiatry (see Table 1 for details). For instance, apart from the assessment of plasma concentrations of clozapine, a cell blood count must be regularly performed given the risk of agranulocytosis (Table 1). Herein DBS sampling might potentially play an important role, as frequent venous sampling is often reported as cause of burden and as an important drawback for the use of clozapine (Nielsen et al., 2010). As it stands, no DBS assays on white cell blood count have been published yet.

214

L.C. Martial et al.

In the same line, laboratory tests are also regularly monitored in patients undergoing treatment with mood stabilizers (such as renal and thyroid function for lithium) (National Institute for Health and Care Excellence, 2014a; Nederlandse Vereniging voor Psychiatrie, 2015) and could therefore also benefit from home sampling by means of DBS (Table 1). To the best of our knowledge, only DBS assays for creatinine have been published (Koop et al., 2013; Koster et al., 2015). The precision of those assays is usually about 10–15% (coefficient of variation [CV]). In comparison, routine plasma assays for creatinine reach 2% CV. This difference in precision is related to analytical issues intrinsic to whole blood inherent to DBS analysis. Consequently, current DBS assays for creatinine can be used for assessment of either very high or very low concentrations. These assays are not yet suitable for a reliable follow-up of renal function and there are no published assays measuring other laboratory tests regularly monitored in psychiatry (such as electrolytes, thyroid function, lipid profile or liver function).

4.4.

Future directions

Three steps have been identified for a successful development and implementation of remote sampling by means of DBS: [1] analytical method development and validation, [2] clinical validation in the patient population directly comparing finger prick sampling with regular blood drawing and applying appropriate statistical methods to assess agreement of methods and [3] implementation of DBS sampling in routine care with careful attention to training and feedback of spotting technique. From the above, it will be clear that DBS for psychiatric drugs is currently in steps 1 and 2, still with a relatively limited amount of reports on this approach. The actual use of DBS in clinical practice (step 3) remains very low. This can be explained by some prerequisites essential for DBS to be clinically valuable that are not yet fulfilled in psychiatry. First, for DBS to be successful in psychiatry, as in every other field, careful selection of the population and compounds is essential. In light of the monitoring characteristics described in Table 1, the best drug candidates for DBS home sampling are clozapine, lithium, valproic acid and carbamazepine. As TCA concentrations are usually only measured upon initiation of therapy, current plasma assays following regular venous sampling are probably adequate. These compounds might not benefit from DBS sampling given the investments required when developing new assays and the necessity of proper instructions and feed-back to the patient. The same holds for other antidepressants (e.g. SSRIs, NRI, SNRI), antipsychotics other than clozapine and the mood stabilizer lamotrigine, which are even less frequent subject of TDM. Second, In order to really sort out the disadvantages related to venous sampling (e.g. patients travelling to a health-care facility, costs), the frequently monitored laboratory blood tests should also be determined from DBS samples. For instance, renal function (e.g. creatinine, urea) is ideally measured along with lithium every 3–6 months (Table 1). In the case of treatment with clozapine, white cell blood count is actually measured more frequently than clozapine concentrations and patients would probably

benefit from this less-invasive method of blood extraction. Yet, no adequate DBS assays exist for these clinical chemistry tests. Future work should focus on developing such assays along with drug concentration measurement. Third, attention could be paid to the clinical validation of remote sampling. This ensures the clinician that DBS sampling actually works for their patients and is a reliable alternative to conventional sampling. Clinical validation with proper statistical assessment of agreement of methods is very often lacking in the published literature (Milosheska et al., 2015). Fourth, after the clinical validation, implementation of the method into clinical practice requires explicit attention. Patients performing the finger prick themselves are then trained for adequate and systematic sampling. Our experience in the case of tacrolimus home sampling is that with repetitive feed-back on spot quality, the proportion of adequate spots delivered by the patients is about 90% (data not published). For routine plasma measurement this would be unacceptably low. Implementation of DBS sampling requires time investment and a change of routines. Only if all three requisites are met DBS can be successfully used in clinical practice.

5.

Conclusions

In conclusion, for a selected number of psychiatric drugs DBS home sampling can offer a great alternative over conventional sampling for plasma measurement. In order to maximize DBS benefits, clinical chemistry tests are ideally measured in the same matrix. As it stands, DBS drug assays are now becoming available but assays on laboratory tests are still lacking. As investments on development of DBS assays increase, future research should focus on those drugs (e.g. clozapine, lithium, valproic acid and carbamazepine) and blood parameters (e.g. cell blood count, electrolytes, renal and thyroid function) subjected to regular sampling in clinical practice in psychiatry.

Role of funding source This review was not funded by any party.

Contributors Authors LM, RA and AB designed the review. Authors LM and AB carried out the literature search. Authors LM and AB wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.

Conflict of interest The authors declare that they have no conflict of interest.

Acknowledgment We thank Mr. Jac van der Heijden for providing us with the photos of DBS sampling.

Dried Blood Spot sampling in psychiatry: Perspectives for improving therapeutic drug monitoring

References Aarnoutse, R.E., Schapiro, J.M., Boucher, C.A., Hekster, Y.A., Burger, D.M., 2003. Therapeutic drug monitoring: an aid to optimising response to antiretroviral drugs? Drugs 63, 741–753. Aburuz, S., Al-Ghazawi, M., Al-Hiari, Y., 2010. A simple dried blood spot assay for therapeutic drug monitoring of lamotrigine. Chromatographia 71, 1093–1099. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders: DSM-5, fifth ed., Washington, DC. Baldelli, S., Cattaneo, D., Giodini, L., Baietto, L., Di Perri, G., D'Avolio, A., Clementi, E., 2015. Development and validation of a HPLC-UV method for the quantification of antiepileptic drugs in dried plasma spots. Clin. Chem. Lab. Med. 53, 435–444. Barraclough, K.A., Isbel, N.M., Kirkpatrick, C.M., Lee, K.J., Taylor, P.J., Johnson, D.W., Campbell, S.B., Leary, D.R., Staatz, C.E., 2011. Evaluation of limited sampling methods for estimation of tacrolimus exposure in adult kidney transplant recipients. Br. J. Clin. Pharmacol. 71, 207–223. Berm, E.J., Brummel-Mulder, E., Paardekooper, J., Hak, E., Wilffert, B., Maring, J.G., 2014. Determination of venlafaxine and O-desmethylvenlafaxine in dried blood spots for TDM purposes, using LC-MS/MS. Anal. Bioanal. Chem. 406, 2349–2353. Berm, E.J., Odigie, B., Bijlsma, M.J., Wilffert, B., Touw, D.J., Maring, J.G., 2016. A clinical validation study for application of DBS in therapeutic drug monitoring of antidepressants. Bioanalysis 8, 413–424. Berm, E.J., Paardekooper, J., Brummel-Mulder, E., Hak, E., Wilffert, B., Maring, J.G., 2015. A simple dried blood spot method for therapeutic drug monitoring of the tricyclic antidepressants amitriptyline, nortriptyline, imipramine, clomipramine, and their active metabolites using LC–MS/MS. Talanta 134, 165–172. Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1, 307–310. Bruggemann, R.J., Aarnoutse, R.E., 2015. Fundament and prerequisites for the application of an antifungal TDM service. Curr. Fungal Infect. Rep. 9, 122–129. Capiau, S., Stove, V.V., Lambert, W.E., Stove, C.P., 2013. Prediction of the hematocrit of dried blood spots via potassium measurement on a routine clinical chemistry analyzer. Anal. Chem. 85, 404–410. Chang, J., Tarasova, T., Shanmugam, V., Azarskova, M., Nguyen, S., Hurlston, M., Sabatier, J., Zhang, G., Osmanov, S., Ellenberger, D., Yang, C., Vitek, C., Liulchuk, M., Nizova, N., 2015. Performance of an early infant diagnostic test, AmpliSens DNA-HIVFRT, using dried blood spots collected from children born to human immunodeficiency virus-infected mothers in Ukraine. J. Clin. Microbiol. 53, 3853–3858. Clinical and Laboratory Standards Institute, 2013. Measurement Procedure Comparison and Bias Estimation Using Patient Samples. EP09-A3. DBS Technology, 2015. 〈http://dbs-system.ch/technology/〉. (Date accessed: 23-11). De Kesel, P.M., Capiau, S., Stove, V.V., Lambert, W.E., Stove, C.P., 2014. Potassium-based algorithm allows correction for the hematocrit bias in quantitative analysis of caffeine and its major metabolite in dried blood spots. Anal. Bioanal. Chem. 406, 6749–6755. Deglon, J., Lauer, E., Thomas, A., Mangin, P., Staub, C., 2010. Use of the dried blood spot sampling process coupled with fast gas chromatography and negative-ion chemical ionization tandem mass spectrometry: application to fluoxetine, norfluoxetine, reboxetine, and paroxetine analysis. Anal. Bioanal. Chem. 396, 2523–2532.

215

European Medicines Agency, 2011. Guideline on bioanalytical method validation. EMEA/CHMP/EWP/192217/2009. Food and Drug Administration, 2001. Guidance for Industry. Bioanalytical Method Validation. U.S. Department of Health and Human Services. Hiemke, C., 2016. Consensus guideline based therapeutic drug monitoring (TDM) in psychiatry and neurology. Curr. drug Deliv. 13, 353–361. Hiemke, C., Baumann, P., Bergemann, N., Conca, A., Dietmaier, O., Egberts, K., Fric, M., Gerlach, M., Greiner, C., Grunder, G., Haen, E., Havemann-Reinecke, U., Jaquenoud Sirot, E., Kirchherr, H., Laux, G., Lutz, U.C., Messer, T., Muller, M.J., Pfuhlmann, B., Rambeck, B., Riederer, P., Schoppek, B., Stingl, J., Uhr, M., Ulrich, S., Waschgler, R., Zernig, G., 2011. AGNP consensus guidelines for therapeutic drug monitoring in psychiatry: update 2011. Pharmacopsychiatry 44, 195–235. Hoeller, U., Baur, M., Roos, F.F., Brennan, L., Daniel, H., Fallaize, R., Forster, H., Gibney, E.R., Gibney, M., Godlewska, M., Hartwig, K., Kolossa, S., Lambrinou, C.P., Livingstone, K.M., Lovegrove, J.A., Macready, A.L., Manios, Y., Marsaux, C.F., Martinez, J.A., Celis-Morales, C., Moschonis, G., NavasCarretero, S., O'Donovan, C.B., San-Cristobal, R., Saris, W.H., Surwillo, A., Traczyk, I., Tsirigoti, L., Walsh, M.C., Woolhead, C., Mathers, J.C., Weber, P., 2015. Application of dried blood spots to determine vitamin D status in a large nutritional study with unsupervised sampling: the Food4Me project. Br. J. Nutr., 1–10. Holmes, D.T., 2015. Deming and Passing Bablok Regression in R. R-Bloggers.com. 〈http://www.r-bloggers.com/deming-andpassing-bablok-regression-in-r/〉 (Date accessed: 18-07-2016). Ikeda, K., Ikawa, K., Yokoshige, S., Yoshikawa, S., Morikawa, N., 2014. Gas chromatography–electron ionization–mass spectrometry quantitation of valproic acid and gabapentin, using dried plasma spots, for therapeutic drug monitoring in in-home medical care. Biomed. Chromatogr.: BMC 28, 1756–1762. Jager, N.G., Rosing, H., Schellens, J.H., Beijnen, J.H., 2014a. Procedures and practices for the validation of bioanalytical methods using dried blood spots: a review. Bioanalysis 6, 2481–2514. Jager, N.G., Rosing, H., Schellens, J.H., Beijnen, J.H., Linn, S.C., 2014b. Use of dried blood spots for the determination of serum concentrations of tamoxifen and endoxifen. Breast cancer Res. Treat. 146, 137–144. Kong, S.T., Lim, S.H., Lee, W.B., Kumar, P.K., Wang, H.Y., Ng, Y.L., Wong, P.S., Ho, P.C., 2014. Clinical validation and implications of dried blood spot sampling of carbamazepine, valproic acid and phenytoin in patients with epilepsy. PLoS One 9, e108190. Koop, D.R., Bleyle, L.A., Munar, M., Cherala, G., Al-Uzri, A., 2013. Analysis of tacrolimus and creatinine from a single dried blood spot using liquid chromatography tandem mass spectrometry. J. Chromatogr. B, Anal. Technol. Biomed. Life Sci. 926, 54–61. Koster, R.A., Greijdanus, B., Alffenaar, J.W., Touw, D.J., 2015. Dried blood spot analysis of creatinine with LC–MS/MS in addition to immunosuppressants analysis. Anal. Bioanal. Chem. 407, 1585–1594. Lempers, V.J., Martial, L.C., Schreuder, M.F., Blijlevens, N.M., Burger, D.M., Aarnoutse, R.E., Bruggemann, R.J., 2015. Druginteractions of azole antifungals with selected immunosuppressants in transplant patients: strategies for optimal management in clinical practice. Curr. Opin. Pharmacol. 24, 38–44. Li, W., Tse, F.L., 2010. Dried blood spot sampling in combination with LC–MS/MS for quantitative analysis of small molecules. Biomed. Chromatogr.: BMC 24, 49–65. Li, Y., Henion, J., Abbott, R., Wang, P., 2012. The use of a membrane filtration device to form dried plasma spots for the quantitative determination of guanfacine in whole blood. Rapid Commun. Mass Spectrom.: RCM 26, 1208–1212. Linder, C., Andersson, M., Wide, K., Beck, O., Pohanka, A., 2015. A LC-MS/MS method for therapeutic drug monitoring of

216 carbamazepine, lamotrigine and valproic acid in DBS. Bioanalysis 7, 2031–2039. Martial, L.C., Jacobs, B.A., Cornelissen, E.A., de Haan, A.F., Koch, B.C., Burger, D.M., Aarnoutse, R.E., Schreuder, M.F., Bruggemann, R.J., 2016. Pharmacokinetics and target attainment of mycophenolate in pediatric renal transplant patients. Pediatric Transplantation. Mengerink, Y., Mommers, J., Qiu, J., Mengerink, J., Steijger, O., Honing, M., 2015. A new DBS card with spot sizes independent of the hematocrit value of blood. Bioanalysis 7, 2095–2104. Mercolini, L., Mandrioli, R., Protti, M., Conca, A., Albers, L.J., Raggi, M.A., 2014. Dried blood spot testing: a novel approach for the therapeutic drug monitoring of ziprasidone-treated patients. Bioanalysis 6, 1487–1495. Millan, M.J., Goodwin, G.M., Meyer-Lindenberg, A., Ove Ogren, S., 2015. Learning from the past and looking to the future: emerging perspectives for improving the treatment of psychiatric disorders. Eur. Neuropsychopharmacol.: J. Eur. Coll. Neuropsychopharmacol. 25, 599–656. Miller, J.H., Poston, P.A., Rutan, S.C., Karnes, T.H., 2013. An oncard approach for assessment of hematocrit on dried blood spots which allows for correction of sample volume. J Anal. Bioanal. Tech. 4, 162. Milosheska, D., Grabnar, I., Vovk, T., 2015. Dried blood spots for monitoring and individualization of antiepileptic drug treatment. Eur. J. Pharm. Sci.: Off. J. Eur. Fed. Pharm. Sci. 75, 25–39. National Institute for Health and Care Excellence, 2014a. Bipolar disorder: assessment and management. 〈http://nice.org.uk/ guidance/cg185〉 (Data accessed: 05-2016). National Institute for Health and Care Excellence, 2014b. Psychosis and schizophrenia in adults: prevention and management. 〈http://nice.org.uk/guidance/cg178〉 (Data accessed: 05.2016). National Institute for Health and Care Excellence, 2016. Depression in adults: recognition and management. 〈http://nice.org.uk/ guidance/cg90〉 (Data accessed: 05.2015). Nederlandse Vereniging voor Psychiatrie, 2012. Multidisciplinaire richtlijn schizofrenie. 〈http://www.nvvp.net/website/richtlij nen/overzicht-richtlijnen〉 (Data accessed: 05.2016). Nederlandse Vereniging voor Psychiatrie, 2013. Multidisciplinaire richtlijn depressie. 〈http://www.nvvp.net/website/richtlijnen/ overzicht-richtlijnen〉 (Data accessed: 05.2016). Nederlandse Vereniging voor Psychiatrie, 2015. Multidisciplinaire richtlijn bipolaire stoornissen. 〈http://www.nvvp.net/website/ richtlijnen/overzicht-richtlijnen〉 (Data accessed: 05.2016). Newman, R.L., Starr, D.J., 1971. Estimation of blood phenylalanine from a dried blood spot using the Guthrie test. J. Clin. Pathol. 24, 576–578. Nielsen, J., Dahm, M., Lublin, H., Taylor, D., 2010. Psychiatrists' attitude towards and knowledge of clozapine treatment. J. Psychopharmacol. 24, 965–971. Patteet, L., Maudens, K.E., Stove, C.P., Lambert, W.E., Morrens, M., Sabbe, B., Neels, H., 2015a. Are capillary DBS applicable for

L.C. Martial et al. therapeutic drug monitoring of common antipsychotics? A proof of concept. Bioanalysis 7, 2119–2130. Patteet, L., Maudens, K.E., Stove, C.P., Lambert, W.E., Morrens, M., Sabbe, B., Neels, H., 2015b. The use of dried blood spots for quantification of 15 antipsychotics and 7 metabolites with ultrahigh performance liquid chromatography–tandem mass spectrometry. Drug Test. Anal. 7, 502–511. Pohanka, A., Mahindi, M., Masquelier, M., Gustafsson, L.L., Beck, O., 2014. Quantification of valproic acid in dried blood spots. Scand. J. Clin. Lab. Investig. 74, 648–652. Rhoden, L., Antunes, M.V., Hidalgo, P., Alvares da Silva, C., Linden, R., 2014. Simple procedure for determination of valproic acid in dried blood spots by gas chromatography–mass spectrometry. J. Pharm. Biomed. Anal. 96, 207–212. Saracino, M.A., Lazzara, G., Prugnoli, B., Raggi, M.A., 2011. Rapid assays of clozapine and its metabolites in dried blood spots by liquid chromatography and microextraction by packed sorbent procedure. J. Chromatogr. A 1218, 2153–2159. Shah, N.M., Hawwa, A.F., Millership, J.S., Collier, P.S., McElnay, J. C., 2013. A simple bioanalytical method for the quantification of antiepileptic drugs in dried blood spots. J. Chromatogr. B, Anal. Technol. Biomed. Life Sci. 923–924, 65–73. Sheiner, L.B., Beal, S.L., 1981. Some suggestions for measuring predictive performance. J. Pharmacokinet. Biopharm. 9, 503–512. Shokry, E., Villanelli, F., Malvagia, S., Rosati, A., Forni, G., Funghini, S., Ombrone, D., Della Bona, M., Guerrini, R., la Marca, G., 2015. Therapeutic drug monitoring of carbamazepine and its metabolite in children from dried blood spots using liquid chromatography and tandem mass spectrometry. J. Pharm. Biomed. Anal. 109, 164–170. Soulier, A., Poiteau, L., Rosa, I., Hezode, C., Roudot-Thoraval, F., Pawlotsky, J.M., Chevaliez, S., 2016. Dried blood spots: a tool to ensure broad access to hepatitis C screening, diagnosis, and treatment monitoring. J. Infect. Dis. 213, 1087–1095. Spooner, N., Denniff, P., Michielsen, L., De Vries, R., Ji, Q.C., Arnold, M.E., Woods, K., Woolf, E.J., Xu, Y., Boutet, V., Zane, P., Kushon, S., Rudge, J.B., 2015. A device for dried blood microsampling in quantitative bioanalysis: overcoming the issues associated blood hematocrit. Bioanalysis 7, 653–659. Stove, C., Spooner, N., 2015. DBS and beyond. Bioanalysis 7, 1961–1962. Temesi, D., Swales, J., Keene, W., Dick, S., 2013. The stability of amitriptyline N-oxide and clozapine N-oxide on treated and untreated dry blood spot cards. J. Pharm. Biomed. Anal. 76, 164–168. Ting, L.S., Villeneuve, E., Ensom, M.H., 2006. Beyond cyclosporine: a systematic review of limited sampling strategies for other immunosuppressants. Ther. drug Monit. 28, 419–430. Wilhelm, A.J., den Burger, J.C., Swart, E.L., 2014. Therapeutic drug monitoring by dried blood spot: progress to date and future directions. Clin. Pharmacokinet. 53, 961–973.