Neuropharmacology xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Neuropharmacology
journal homepage: www.elsevier.com/locate/neuropharm

Invited review

Personalizing dosing of risperidone, paliperidone and clozapine using

therapeutic drug monitoring and pharmacogenetics
Jose de Leona,b,c,∗
a
University of Kentucky Mental Health Research Center at Eastern State Hospital, Lexington, KY, USA
b
Psychiatry and Neurosciences Research Group (CTS-549), Institute of Neurosciences, University of Granada, Granada, Spain
c
Biomedical Research Centre in Mental Health Net (CIBERSAM), Santiago Apostol Hospital, University of the Basque Country, Vitoria, Spain

HIGHLIGHTS

• Dividing serum concentration by daily dose gives a concentration-to-dose(C/D) ratio.

• The C/D ratio under steady-state and trough conditions measures drug clearance.
• Total risperidone C/D ratio for the oral formulation is around 7 ng/ml per mg/day.
• Long-acting paliperidone provides a C/D ratio around 8 ng/ml per mg/day.
• In the US, clozapine C/D ratios typically range from 0.6 to 1.2 ng/ml per mg/day.

ARTICLE INFO ABSTRACT

Keywords: By combining knowledge of pharmacogenetics, therapeutic drug monitoring (TDM) and drug-drug interactions
Antipsychotic agents/administration & dosage (DDIs) the author developed a model for personalizing antipsychotic dosing, which is applied to risperidone, 9-
Clozapine hydroxyrisperidone or paliperidone, and clozapine.
Delayed-action preparations Drugs are approved using an average dose for an ideal average patient, but pharmacologists have described
Drug monitoring
outliers: genetic poor metabolizers (PMs) and ultrarapid metabolizers (UMs). Environmental and personal
Paliperidone palmitate
variables can also make patients behave as PMs or UMs. Drug clearance is represented by the concentration-to-
Risperidone
dose (C/D) ratio under steady-state and trough conditions. A very low C/D ratio indicates a UM, while a very
high C/D ratio indicates a PM.
Total risperidone C/D ratio for the oral formulation is around 7 ng/ml per mg/day and can be influenced by
CYP2D6 polymorphism, DDIs with inducers and inhibitors, and renal function. Oral paliperidone has low
availability; its C/D ratio is around 4.1 ng/ml per mg/d and can be influenced by inducers and renal impairment.
Once-a-month long-acting paliperidone provides a C/D ratio around 7.7 ng/ml per mg/day at steady state, which
is expected to be in the 8th month (before the 9th injection). TDM is particularly important for long-acting
paliperidone formulations that may accumulate once steady state is reached (after years for the 3- and 6-month
formulations). In the US, clozapine C/D ratios typically range from 0.6 (male smokers) to 1.2 (female non-
smokers) ng/ml per mg/day. East Asians’ clozapine C/D ratios appear to be twice as high. Inhibitors (including
fluvoxamine and oral contraceptives) and inflammation can also increase clozapine C/D ratios.

1. Introduction In 1996 he moved to Lexington, Kentucky, to run a unit for treat-

In 1994 at a scientific meeting the author learned that the cyto-
chrome P450 (CYP) 1A2 (CYP1A2) metabolized clozapine and caffeine
(Bertilsson et al., 1994), which later explained a clozapine intoxication
due to a drug-drug interaction (DDI) with caffeine (Odom-White and de
Leon, 1996). This led him to start studying CYP science and pharma-
cokinetics in pharmacological journals.
ment-refractory patients in a state hospital; to face this challenge he
planned to use CYP 2D6 (CYP2D6) genotyping and measurement of
antipsychotic blood levels, which is called therapeutic drug monitoring
(TDM) by pharmacologists. Then he realized that he needed to combine
knowledge of pharmacogenetics (de Leon, 2016), TDM (de Leon, 2018)
and DDIs (de Leon, 2019).
The Food and Drug Administration (FDA) did not worry about CYPs

∗
UK Mental Health Research Center at Eastern State Hospital, 1350 Bull Lea Road, Lexington, KY, 40511, USA.
E-mail address: jdeleon@uky.edu.

https://doi.org/10.1016/j.neuropharm.2019.05.033
Received 3 January 2019; Received in revised form 17 May 2019; Accepted 27 May 2019
0028-3908/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Jose de Leon, Neuropharmacology, https://doi.org/10.1016/j.neuropharm.2019.05.033

J. de Leon
until 1996 (Flockhart, 1996). Oral risperidone was introduced in 1993
despite very limited pharmacokinetic data (Huang et al., 1993). The
company proposed that: 1) risperidone pharmacokinetics was “well
understood” and 2) CYP2D6 poor metabolizer (PMs) status was irrele-
vant because risperidone and 9-hyroxyrisperidone have the same
pharmacological activity since adding their serum concentrations pro-
vides the active moiety (Heykants et al., 1994; Van Peer et al., 1996).
The author's review of his patients' TDM results taught him that these
two statements are not correct. First, the author found a risperidone
DDI that suggested CYP3A4 was also important for risperidone meta-
bolism (de Leon and Bork, 1997) and then that CYP2D6 PMs were
prone to developing toxicity (Bork et al., 1999). Later on, the author
hypothesized that a lower threshold for crossing the blood brain barrier
(BBB) makes 9-hydroxyrisperidone less potent. It is interesting that,
when the company marketed 9-hydroxyrisperidone as paliperidone, it
recommended approximately twice the daily risperidone dosage, thus
in agreement with the author's hypothesis that 9-hydroxyrisperidone
may be less potent (de Leon et al., 2010).
After 20 years, the author has developed a model for personalizing
the dosing of risperidone, 9-hydroxyrisperidone, which was marketed
under the name paliperidone, and clozapine. He will need to increase
his limited experience with TDM for aripiprazole, olanzapine and
quetiapine to be able to provide similar comprehensive models for their
personalized dosing in future articles. He has no experience or very
limited TDM experience with other more recently marketed second-
generation antipsychotics, but they appear to follow similar pharma-
cokinetic principles, including linear kinetics.
2. Personalized dosing
The randomized clinical trials (RCTs) used to approve drugs are
designed to provide an average dose for an ideal average patient, but
clinicians know that not all of their patients are average. Moreover,
pharmacologists developed pharmacogenetic science (Meyer, 2004)
after identifying patients who need unusually low or high doses, called
outliers by statisticians (de Leon, 2012). Outliers are not well re-
presented by the mean and are ignored by companies that conduct RCTs
and by researchers who use an evidence-based medicine (EBM) ap-
proach. From the pharmacological and statistical points of view, it is
nonsensical to ask which clozapine dosage is best (Subramanian et al.,
2017). There is no best clozapine dose for an ideal patient because he/
she does not represent all clozapine patients well. DDI, sex, smoking,
ethnicity and possibly genetics and various genetic combinations are
factors that require different average clozapine doses. In the end, the
best individual dose for a patient is better established by TDM. PMs
need low doses to avoid adverse drug reactions (ADRs), and ultrarapid
metabolizers (UMs) need high doses to maintain efficacy.
Articles traditionally state that approximately 7% of Caucasians are
believed to be cytochrome P450 2D6 (CYP2D6) PMs, compared to
1–3% of individuals of other races. A recent article estimated a
worldwide prevalence of CYP2D6 PMs from 0.4 to 5.4% (Gaedigk et al.,
2017). CYP2D6 PMs have two null alleles and they have no CYP2D6 in
their bodies (activity score = 0). Most commercial tests do a reasonable
job of identifying CYP2D6 PMs (de Leon, 2017). In the US population,
approximately 1.5% are CYP2D6 UMs. Gaedigk et al. (2017) estimated
the world prevalence at 1–21%. CYP2D6 UMs typically have at least 3
active copies of the gene (activity score = 3). Most commercial tests
misclassify CYP2D6 UMs (de Leon, 2017). Between the CYP2D6 PMs
and UMs, one finds the intermediate metabolizers (IMs) with decreased
activity, and the normal metabolizers (NMs) which were called ex-
tensive metabolizers in the past. These are the “genetic” PMs or UMs,
but environmental (DDIs and smoking) and personal variables (aging,
pregnancy or medical illness) can make patients behave as if they are
PMs or UMs (de Leon, 2009).
2
Neuropharmacology xxx (xxxx) xxxx
3. Drug clearance and TDM
Pharmacologists define the elimination of a drug from the body as
drug clearance. Clearance includes not only drug metabolism but renal
and hepatobiliary clearance; the relevance of these 3 systems varies
from drug to drug (Smith et al., 2019).
Antipsychotics follow linear kinetics, which means that in each
patient taking typical clinical doses (and no changes in inhibitors or
inhibitors) there is a linear relationship between the dose (D) in mg/day
and the concentration (C) (ng/ml). This linear relationship is re-
presented by the concentration-to-dose (C/D) ratio under steady-state
and trough conditions. When comparing individuals, a very low C/D
ratio indicates a UM, while a very high C/D ratio indicates a PM. In the
experience of the author, pharmacists have been exposed to the concept
of antipsychotic C/D ratios or at least can easily grasp it due to their
quality education in pharmacokinetics during pharmacy school. On the
other hand, physicians, including psychiatrists, frequently have diffi-
culty with the concept of antipsychotic C/D ratio since they received a
poor education in pharmacokinetics during medical school. If the
reader has problems with this concept, the author recommends that he/
she review some of the lectures describing real cases using C/D ratios in
a free psychopharmacology course available on the web (de Leon,
2015–2016).
The therapeutic reference ranges used in this article were obtained
from the third version of TDM guidelines (Hiemke et al., 2018) devel-
oped by expert consensus, which have been summarized for clinicians
(Schoretsanitis et al., 2018a). To calculate the ranges, these experts
reviewed studies in which therapeutically effective dosages were used.
By their definitions, “the lower limit of a drug's therapeutic range re-
presents the threshold value in blood below which therapeutic effects
are not significantly different from placebo” and “the upper limit of the
therapeutic range is determined on the basis of maximal efficacy and
increased risk of adverse drug reactions (ADRs), assuming that ADRs
are related to a drug's concentration in the body”. They also explain
that “therapeutic reference range is an orienting, population-based
range which may not necessarily be applicable to all patients” (Hiemke
et al., 2018).
Clinicians need to remember that these therapeutic reference ranges
are for concentrations that are trough and steady-state concentrations.
Trough concentration refers to the lowest point of the concentration
within a day and is typically collected early in the morning before
taking any medication. Steady-state concentration requires at least 5
half-lives after the last time the dosage of the TDM drug was changed.
For oral risperidone, paliperidone and clozapine most patients are at
steady state one week after the last dosage change, but PM patients may
have longer half-lives and it may be better, if possible, to wait addi-
tional days. Similarly, a recent change in inhibitors or inducers of any of
these three antipsychotics with no corresponding dosage change in the
antipsychotic may mean that TDM is no longer at steady state (Spina
et al., 2016).
4. Personalizing risperidone dosing
In the United States (US), risperidone is marketed in oral and long-
acting injectable (LAI) formulations (Schoretsanitis et al., 2017). This
risperidone section describes: 1) clearance, 2) risperidone/9-hydro-
xyrisperidone ratio, 3) total risperidone C/D ratio, 4) therapeutic range
and therapeutic index, 5) pharmacodynamic variability and dosing, 6) a
proposal for personalizing dosing, 7) other types of genotyping, 8)
personalizing LAI dosing, and 9) major pending issues related to per-
sonalizing dosing.
4.1. Risperidone clearance
This review of risperidone clearance is based on articles written by:
1) the marketer (Huang et al., 1993; Mannens et al., 1993; Heykants
J. de Leon
Table 1
Risperidone clearance after stratification by patient characteristics.
Metabolism to 9-hydroxyrisperidone
CYP2D6 CYP3A4 Risperidone/9-hydroxyrisperidone ratio
Induced NM important ↑↑ role 0.2
Induced PM absent ↑↑ role around 2
NM main auxiliary 0.2
CYP2D6 inhibitor inhibited auxiliary >1
PM absent only enzyme around 2
Geriatric NM main auxiliary 0.2
C/D: concentration/dose; CYP2D6: cytochrome P450 2D6; CYP3A4: cytochrome P450 3A4; NM: normal metabolizer; PM: poor metabolizer.
et al., 1994; Snoeck et al., 1995; Van Peer et al., 1996), 2) the author
(de Leon and Bork, 1997; Bork et al., 1999; Barnhill et al., 2005; de
Leon et al., 2005, 2006, 2008b, 2009; de Leon, 2014; Leon et al., 2007;
Spina et al., 2016; Schoretsanitis et al., 2017) and 3) other authors
(Ereshefsky, 1996; Fang et al., 1999; Güzey et al., 2000; Spina et al.,
2000; Yasui-Furukori et al., 2001; Hefner et al., 2015, 2016).
In the average patient, risperidone is mainly metabolized by hy-
droxylation to 9-hydroxyrisperidone by CYP2D6 and CYP3A4, and 9-
hydroxyrisperidone is mainly eliminated by renal clearance (glo-
merular filtration and tubular secretion). Table 1 explains the relevance
of each clearance method varying across patients.
4.2. Risperidone/9-hydroxyrisperidone ratio
Ereshefsky (1996) proposed that an inverted risperidone/9-hydro-
xyrisperidone ratio, having more risperidone than 9-hydro-
xyrisperidone in serum, is a sign of being a CYP2D6 PM or taking a
powerful CYP2D6 inhibitor. CYP2D6 NMs have risperidone/9-hydro-
xyrisperidone ratios around 0.2. Box 1’s upper panel describes how to
calculate this ratio and its meaning.
4.3. Total risperidone C/D ratio
The normal total risperidone C/D ratio is around 7 ng/ml per mg/
day (de Leon et al., 2008b). A higher C/D ratio, particularly when it
gets close to 14 ng/ml per mg/day (twice the normal ratio of 7), is a
sign of poor clearance, characteristic of genetic PM status, taking a
powerful CYP2D6 inhibitor, or renal impairment. When the total C/D
ratio is < 3.5 ng/ml per mg/day (half the normal ratio of 7), this can
be a sign of rapid clearance, such as taking a powerful CYP3A4 inducer,
but usually is explained merely by lack of compliance. Box 1's middle
panel describes how to calculate this ratio and its meaning.
4.4. R therapeutic range and therapeutic index
The risperidone therapeutic reference range is 20–60 ng/mL, ob-
tained by adding risperidone and 9-hydroxyrisperidone concentrations
(Hiemke et al., 2018). Box 1's lower panel explains that risperidone has
a narrow therapeutic window or index and is prone to cause ADRs
(Simpson and Lindenmayer, 1997; Regenthal et al., 2005; Migliardi
et al., 2009; Leucht et al., 2013; Schoretsanitis et al., 2017).
4.5. Pharmacodynamic variability and dosing
Four pharmacodynamic factors contributing to lower pharmacolo-
gical tolerance are important in R dosing for: 1) any patient before ti-
tration, 2) naïve patients, 3) geriatric patients, and 4) those with ob-
vious neuropathological brain changes (de Leon et al., 2008a, 2008b).
The company recommends titrating risperidone, probably due to the
need to develop tolerance at α peripheral receptors, which can cause
orthostatic changes, and at brain H1 receptors, which can cause seda-
tion. Naïve patients need to be treated with lower risperidone doses
3
Neuropharmacology xxx (xxxx) xxxx
Renal clearance Total C/D
Risperidone 9-hydroxyrisperidone ratio
little ↑↑ < 3.5
important ↑↑ <7
little important 7
↑ ↓ 7–14
↑ ↓ close to14
little ↓ by 30% >7
(Williams, 2001); exposure to antipsychotics is associated with a not-
well-understood development of tolerance in D2 receptors. Geriatric age
is associated with reduced R clearance due to decreased glomerular
flirtation (GFR), but aging is probably also associated with some
pharmacodynamic factors: a decrease in dopamine neurons and in-
creased risk of EPS (extrapyramidal symptoms). Lower risperidone
doses are recommended in patients with abnormal brains such as those
with dementing illnesses and intellectual disability (de Leon et al.,
2008b).
4.6. A complex proposal for personalizing risperidone dosing
The best way of personalizing risperidone dosing is to pay attention
to the clinical response (look for efficacy in the absence of ADRs),
conduct risperidone TDM, and keep the patient at a total risperidone
serum concentration of 20–60 ng/ml (Hiemke et al., 2018). In the
average patient with a total C/D ratio of 7 ng/ml per mg/day, it would
be reached using a risperidone dose between approximately 3 and
8 mg/day (20/7 = 2.9 and 60/7 = 8.6). As risperidone was approved
when higher antipsychotics were the norm and risperidone tends to
have a narrow therapeutic range, the author usually recommends a
lower upper range of 45 ng/ml, which corresponds to a risperidone dose
of approximately 6 mg/day in the average patient.
In the absence of TDM, Table 2 provides guidelines for starting,
target and maximum R doses for adult patients taking no other anti-
psychotic; it upgrades the excellent clinical guideline proposed by
Williams (2001) by adding pharmacokinetic and pharmacodynamic
concepts. Risperidone should not be used in patients with renal im-
pairment; other second-generation antipsychotics with limited renal
clearance (Spina et al., 2016) should be used.
4.7. Other types of genotyping need further development to aid personalized
R dosing
Commercial pharmacogenetic tests can be marketed in North
America and Europe with almost no supervision by any regulatory
agency and do not have to demonstrate clinical validity or clinical
utility (de Leon, 2016; de Leon and Spina, 2016). Pharmacodynamic
genes such as gene variants in dopamine or serotonin receptors are not
ready for clinical use. Similarly, the p-glycoprotein gene variants are
not ready for clinical practice (de Leon, 2015b).
4.8. Personalizing LAI risperidone dosing
In 2003, based on the hypothesis that its pharmacology is similar to
oral risperidone, the FDA approved the LAI risperidone formulation
developed by risperidone's marketer using a complex microsphere
technology (Schoretsanitis et al., 2017). Based on company RCTs, Kane
et al. (2003) considered 25, 50 and 75 mg of risperidone LAI every 2
weeks to be respectively equivalent to oral daily doses of 2, 4 and 6 mg/
day. This equivalence has never been systematically tested. The author
prefers the recommendation from Castberg and Spigset (2005), who
J. de Leon Neuropharmacology xxx (xxxx) xxxx

Box 1
Interpreting oral R TDM to personalize R dosing

Collect at trough (early morning before meds) and at steady state (≥5 days after dose change)

Risperidone/9-hydroxyrisperidone ratio
Calculation Divide serum risperidone by 9-hydroxyrisperidone concentration
Meaning Index of CYP2D6 activity in most patients
In CYP2D6 PMs it may be an index of CYP3A4 activity
Genetic N Median1 (25th, 75th Percentiles) Mean ± SD
UM (3 active alleles) 7 0.03 (0.02, 0.04) 0.04 ± 0.03
NM (1–2 active alleles) 179 0.08 (0.04, 0.017) 0.17 ± 0.28
IM (> 0 but < 1) 21 0.56 (0.29, 1.1) 0.87 ± 0.92
PM (0) 14 2.5 (1.8, 4.1) 4.2 ± 4.5
Environment Inducers do not change this ratio.
Powerful CYP2D6 inhibitors can lead to risperidone/9-hydroxyrisperidone ratios > 1.
Clinical Summary A risperidone/9-hydroxyrisperidone ratio > 1 (called an inverted ratio since risperidone > 9-hydroxyrisperidone levels)
indicates a CYP2D6 PM or the intake of a powerful inhibitor and poor risperidone elimination. Some IMs2 may have a
risperidone/9-hydroxyrisperidone ratio close to 1 in the absence of inhibitors.
Total C/D Ratio
Calculation Divide the total concentration (risperidone + 9-hydroxyrisperidone) in ng/ml by risperidone dose in mg/day.
Meaning Index of total ability to clear risperidone (CYP2D6, CYP3A and renal excretion)
Normal Values 7 in USA research and clinical laboratories and in German TDM studies
10 in Italian and Spanish research laboratories
Differences may be due to different analytic techniques and/or calibrations
Each laboratory needs to establish its normal C/D
C/D ratio < half of normal (< 3.5 in USA or German TDM or < 5 in Italian research laboratory)
Indicates ↑ to eliminate risperidone (if the patient is compliant)
After excluding non-compliance, it can be explained by CYP3A4 inducers.
C/D ratio > almost twice the normal value (close to 14 in USA or German TDM or close to 20 in Italian research laboratory)
Indicates poor ability to clear risperidone (if TDM was measured at trough)
Can be explained by:
- CYP2D6 PM genotype
- use of CYP2D6 inhibitors or CYP3A4 inhibitors
- infections or inflammations may inhibit CYP3A4
- renal insufficiency (limited clinical experience by author)
Risperidone has a narrow therapeutic window or index
- Risperidone's therapeutic window or index is 3 (dividing the upper range by the lower: 60/20 = 3) which indicates risperidone is an antipsychotic with a relatively narrow
therapeutic index (Schoretsanitis et al., 2017).
- This narrow therapeutic index explains why in the SGA meta-analyses risperidone has high propensity to cause some dose-related ADRs, such as extrapyramidal symptoms (EPS)
and hyperprolactinemia (Leucht et al., 2013). As a matter of fact, there is a linear relationship between EPS and risperidone dose (Simpson and Lindenmayer, 1997). Although
it has never been well studied (Regenthal et al., 2005), pharmacokinetic science suggests that the dose-related EPS are better described as concentration-related EPS, meaning
that, in each patient, as concentration increases EPS risk increases. Migliardi et al. (2009) support hyperprolactinemia as definitively a concentration-related ADR, after
correcting for confounding factors.

ADR: adverse drug reaction; C/D: concentration/dose; CYP2D6: cytochrome P450 2D6; CYP3A4: cytochrome P450 3A4; EPS: extrapyramidal symptoms; IM:
intermediate metabolizer; NM: normal metabolizer; PM: poor metabolizer; SD: standard deviation; TDM: therapeutic drug monitoring; UM: ultrarapid metabolizer;
USA: United States of America.
1
In a naturalistic sample of 221 risperidone patients in Kentucky (Leon et al., 2007), not contaminated by CYP inhibitors, nor by hepatic/renal impairment, these
are the medians (25th, 75th percentiles) of the risperidone/9-hydroxyrisperidone ratio according to genetic profile (de Leon et al., 2009). The comparison of
median with mean ± SD demonstrated that the data did not follow a normal distribution. The 221 patients had a median of 0.12 and mean ± SD of
0.49 ± 1.2 ng/ml per mg/day. After excluding the 14 CYP2D6 PMs, the median was 0.10 and the mean ± SD of 0.24 ± 0.44 ng/ml per mg/day.
2
Some IMs have very low activity. The most typical example is East Asians with CYP2D6 *10/*10, which accounts for approximately 10% of Chinese, Koreans and
Japanese. These IMs, with very low activity, may also have problems eliminating risperidone and a risperidone/9-hydroxyrisperidone ratio around 1.

ADR: adverse drug reaction; C/D: concentration/dose; CYP2D6: cytochrome P450 2D6; CYP3A4: cytochrome P450 3A4; EPS: extrapyramidal
symptoms; IM: intermediate metabolizer; NM: normal metabolizer; PM: poor metabolizer; SD: standard deviation; TDM: therapeutic drug
monitoring; UM: ultrarapid metabolizer; USA: United States of America.1In a naturalistic sample of 221 risperidone patients in Kentucky (Leon et al.,
2007), not contaminated by CYP inhibitors, nor by hepatic/renal impairment, these are the medians (25th, 75th percentiles) of the risperidone/9-
hydroxyrisperidone ratio according to genetic profile (de Leon et al., 2009). The comparison of median with mean ± SD demonstrated that the data
did not follow a normal distribution. The 221 patients had a median of 0.12 and mean ± SD of 0.49 ± 1.2 ng/ml per mg/day. After excluding the
14 CYP2D6 PMs, the median was 0.10 and the mean ± SD of 0.24 ± 0.44 ng/ml per mg/day.2Some IMs have very low activity. The most typical
example is East Asians with CYP2D6 *10/*10, which accounts for approximately 10% of Chinese, Koreans and Japanese. These IMs, with very low
activity, may also have problems eliminating risperidone and a risperidone/9-hydroxyrisperidone ratio around 1.

divided the applied LAI dose by the number of days. This RCT provided
respective equivalents of 1.8 mg/day, 3.6 mg/day and 5.4 mg/day,
slightly lower than the manufacturer's equivalents. In a systematic re-
view of the LAI formulation (Schoretsanitis et al., 2017), 6 studies had
total risperidone C/D ratios ranging from 7.4 to 9.7 ng/ml per mg/day.
The 6 samples included 297 patients, whose combined weighted mean
total risperidone C/D ratio was 8.8 ng/ml per mg/day, not far from the
7 ng/ml per mg/day that the author considers normal for oral risper-
idone. Thirteen CYP2D6 PMs had approximately double values,
15.4 ng/ml per mg/day (Hendset et al., 2009). In a later German TDM
study, 63 patients on LAI R had a median total C/D ratio which was
slightly higher than in 851 patients on oral risperidone, 7.6 versus
6.3 ng/ml per mg/day (Schoretsanitis et al., 2018b).
In a risperidone LAI systematic review (Schoretsanitis et al., 2017),
the weighted mean risperidone/9-hydroxyrisperidone ratio was 0.48 in
329 patients, which is approximately twice what the author found in his
oral TDM study with 221 patients (Box 1). Similarly, in the recent
German TDM study, the median was higher, 0.51, in 63 patients on LAI

4
J. de Leon Neuropharmacology xxx (xxxx) xxxx

Box 2
Interpreting clozapine TDM to personalize clozapine dosing based on examples

Patient examples that relate clozapine C/D ratios and therapeutic doses

1) A patient requires 900 mg/day to reach 350 ng/ml of clozapine concentration. The clozapine C/D ratio = 0.4 (350/900 = 0.4).
a) The author has seen a few patients with clozapine C/D ratios around 0.4: Caucasian ♂ smokers taking potent inducers such as phenytoin.
2) A patient requires 600 mg/day to reach 350 ng/ml of clozapine concentration. The clozapine C/D ratio = 0.6 (350/600 = 0.6).
a) Caucasian ♂ smokers typically have C/D ratios around 0.6.
3) A patient requires 300 mg/day to reach 350 ng/ml of clozapine concentration. The clozapine C/D ratio = 1.2 (350/300 = 1.2).
a) Caucasian ♀ non-smokers typically have C/D ratios around 1.2.
b) Caucasian ♂ taking a clinically relevant inhibitor may have C/D ratios around 1.2.
c) East Asian ♂ smokers have C/D ratios around 1.2.
4) A patient requires 150 mg/day to reach 350 ng/ml of clozapine concentration. The clozapine C/D ratio = 2.3 (350/150 = 2).
a) Caucasian ♀ non-smokers taking a clinically relevant inhibitor may have C/D ratios around 2.3.
b) East Asian ♀ non-smokers may have C/D ratios around 2.3.
Genetic, environmental or personal factors leading to the need for high or low clozapine doses
1) Patients requiring higher doses:
- Individuals with extremely high CYP1A2 activity are probably very rare (< 1%).
- Smokers need higher doses. Smoking ↓ C/D ratio by 0.80. Therefore, smoking patients need ↑in dose (approximately 1.25 times higher) to reach therapeutic clozapine
concentrations.
- Inducers such as rifampin, phenytoin, and phenobarbital ↓C/D ratio by a factor of approximately 1.5.Therefore, patients taking inducers need ↑ in dose (approximately 1.5
times higher) to reach therapeutic clozapine concentrations.
- Mild inducers of clozapine may include valproate, charbroiled food, roasted coffee, omeprazole and cruciferous vegetables (including broccoli). Mild inducers do not usually
require dosing changes.
2) Subjects requiring lower doses:
- The author thought individuals with extremely low CYP1A2 activity were probably rare (< 1%). More recently he has defined PMs in each ethnic group as those
having > mean + 2 SD of total clozapine concentrations for their sex and smoking status group. He has found that 1–10% of Chinese and Caucasians are PMs who need
clozapine doses < 125 mg/day to reach 350 ng/ml
- Fluvoxamine is a very powerful inhibitor. On average, fluvoxamine ↑ the C/D ratio by 3 times, which requires ↓ clozapine doses to 1/3. Published cases of 5–10-fold increases
in C/D ratios exist. Do not prescribe fluvoxamine without clozapine TDM.
- Ciprofloxacin (and some similar antibiotics) and high caffeine doses are clinically relevant inhibitors and may require ↓ clozapine dose by half.
- Major inflammations or infections (pneumonia, upper respiratory infections with fever, or appendicitis) may decrease clozapine metabolism and may require ↓ clozapine dose
by half.
- Females have lower capacity to metabolize clozapine, probably due to the presence of estrogens.
- Oral contraceptives and estrogen elevations during pregnancy can inhibit clozapine metabolism.
- Fluoxetine and paroxetine are mild inhibitors of clozapine metabolism (dosing changes are not usually required). Sertraline in high doses can occasionally inhibit clozapine
metabolism.
- Patients with renal or hepatic impairment need lower clozapine doses; there are no dosing studies.
- With advanced age, particularly after the age of 65 years, there is a progressive diminution of the ability to clear clozapine. The ↑ C/D ratio associated with aging is small.

C/D ratio: concentration/dose ratio; CYP1A2: cytochrome P450 1A2; East Asians include Chinese, Koreans and possibly Japanese; PM: poor metabolizer; SD:
standard deviation; TDM: therapeutic drug monitoring.

C/D ratio: concentration/dose ratio; CYP1A2: cytochrome P450 1A2; East Asians include Chinese, Koreans and possibly Japanese; PM: poor
metabolizer; SD: standard deviation; TDM: therapeutic drug monitoring.

risperidone than 0.26 in 851 patients on oral risperidone (Schoretsanitis
et al., 2018b). This increased mean risperidone/9-hydroxyrisperidone
ratio probably reflects a relative decrease in first-pass metabolism on
the LAI risperidone formulation (Castberg and Spigset, 2005). The re-
lative CYP2D6 contribution to risperidone metabolism may decrease
when moving from the oral to the LAI microsphere formulation.
In a systematic review of the LAI microsphere risperidone for-
mulation, most studies had drug concentrations very close to the lower
limit of the range for oral risperidone, 20 ng/ml. At steady state, LAI
antipsychotics may be therapeutic in lower concentrations than the oral
antipsychotics, since they tend to have more stable concentrations
within the same day and from day to day (Schoretsanitis et al., 2018c).
Clinicians using TDM for the risperidone LAI microsphere for-
mulation need to: 1) consider steady state to be reached ≥6 weeks after
the first injection; 2) pay attention to a) co-medications with inducers/
inhibitors, b) severe inflammations/infections, and c) hepatic/renal
impairment; and 3) use Castberg's recommendation to calculate ris-
peridone dosing (Schoretsanitis et al., 2017).
A new once-a-month subcutaneous risperidone LAI formulation
called RBP-7000 was developed by another company and recently
marketed in the US (Citrome, 2018) without any published TDM study
(Schoretsanitis et al., 2017). Laffont et al. (2015) proposed that patients
on 2, 3 or 4 mg/day of oral risperidone should receive RBP-7000
monthly SC injections of 60, 90 or 120 mg respectively. In the absence
of any published pharmacokinetic studies (Schoretsanitis et al., 2017)
and due to statements about half-life by the marketer that are extremely
likely to be wrong, the author recommends that all clinicians obtain
risperidone TDM before switching any patient to this RBP-7000 for-
mulation and repeat TDM at least every 3 months in all patients.
4.9. Major pending issues related to personalizing risperidone dosing
There are 3 major unresolved issues related to TDM and persona-
lizing dosing for LAI risperidone: 1) the need for half-life studies
(Schoretsanitis et al., 2019a), as the data provided by the marketer does
not appear to be correct for RBP-7000 and is unverified by independent
investigators for the microsphere formulation; 2) the possibility that
lower total risperidone concentrations between 20 and 30 ng/ml may
be enough for the risperidone microsphere formulation; and 3) limited
understanding of the relevance of the increase in risperidone/9-hy-
droxyrisperidone ratio when moving from the oral to the LAI micro-
sphere formulation. The author recommends that any clinicians using
LAI risperidone should use TDM to be sure that the concentrations are
reasonable (20–60 mg/ml). CYP2D6 UM status and pregnancy are
probably associated with increased risperidone clearance and may re-
quire TDM to verify that the risperidone dose is sufficient for that pa-
tient. Two issues need future study by pharmacologists: 1) the need to
correct total risperidone concentrations if 9-hydroxyrisperidone is de-
finitively demonstrated to be less potent, and 2) the role of the 9-hy-
droxyrisperidone enantiomers (Yasui-Furukori et al., 2001).

5
J. de Leon Neuropharmacology xxx (xxxx) xxxx

Table 2
Approximations for personalizing risperidone dosing in the absence of TDM for average adults not taking any other antipsychotica by combining pharmacodynamic
and pharmacokinetic influences.
Normal Carbamazepine and other inducersb CYP2D6 PMs or paroxetinec Non-potent inhibitorsd

(multiply x 2) (divide by 2) (divide by 1.3)

DOSES (mg/day)
Average adult Startinge 1–2 2–4 0.5–1 0.75–1.5
Targetf 4 8 2 3
Maximumg 6 12 3 4
First psychotic episodeh Startinge 1 2 0.5 0.75
Targetf 2 4 1 1.5
Maximumg 4 8 2 3
Geriatric, liveri insufficiency or organicity Startinge 0.25 0.5 0.12 0.12
Targetf 2 4 1 1.5
Maximumg 4 8 2 3
Demented Startinge 0.5 1 0.25 0.25
Targetf 1 2 0.5 0.75
Maximumg 1.5 3 0.75 1
Adult ID behaviorsj Startinge 1–2 2–4 0.5–1 0.75–1.5
Targetf 2 4 1 1.5
Maximumg 4 8 2 3

Do not use risperidone if there is renal impairment, or with fluoxetine,k or if a CYP2D6 PM is taking a CYP3A4 inhibitor.
CYP2D6: cytochrome 2D6; CYP3A4: cytochrome 3A4; ID: intellectual disability; PM: poor metabolizer; TDM: therapeutic drug monitoring. This table includes
modifications from prior versions (de Leon et al., 2008b; Spina and de Leon, 2015).
a
If the patient is taking another antipsychotic, lower risperidone doses than those described in this table should be used as long as the other antipsychotic is
prescribed. If the patient has taken risperidone, prior doses can be used for orientation as long there are no changes in co-prescription of inhibitors or inducers, no
development of renal or hepatic insufficiency, and the patient has not reached geriatric age. TDM can provide a better approximation for an individual than these
average recommendations.
b
These indications are based mainly on carbamazepine data. Other CYP3A inducers have not been well studied. Clinically relevant inducers that may have similar
or greater effects than carbamazepine are rifampin, phenobarbital, primidone, phenytoin, nonnucleoside reverse transcript inhibitors (efavirenz and delavirdine).
Mild inducers that are probably less potent than carbamazepine are oxcarbazepine, topiramate, dexamethasone, prednisone and St. John's wort. Mild inducers have
not been studied.
c
Paroxetine is a potent CYP2D6 inhibitor; it frequently inhibits CYP2D6 completely.
d
Non-potent inhibitors include moderate CYP2D6 inhibitors such as duloxetine and bupropion, and any potent CYP3A inhibitor: fluvoxamine, cimetidine,
ketoconazole, erythromycin, clarithromycin, protease inhibitors, grapefruit juice, and diltiazem. High doses of sertraline may also be an inhibitor of risperidone
metabolism. Disregard inhibitor effects if the patient is taking potent inducers.
e
Starting dose refers to the first dose given to the patient. This dose has to be slowly titrated over a few days based on tolerance for reaching the target dose. Lower
starting doses are recommended to avoid orthostatic hypotension and other adverse drug reactions to the first risperidone doses.
f
Target dose refers to the dose expected to be therapeutic in the average patient with these specific characteristics.
g
Maximum dose refers to the maximum recommended dose expected to be therapeutic in the average patient with these specific characteristics.
h
First-episode patients are probably less tolerant of risperidone and other high potent antipsychotics, such as haloperidol (de Leon et al., 2008b).
i
There is no published information on hepatic insufficiency in clinical samples but conservative dosing appears reasonable.
j
These doses are for challenging behaviors that did not respond to non-pharmacological interventions. Psychotic episodes in adults with IDs may require higher
doses; see average doses for adults.
k
If a patient has any renal impairment, select another antipsychotic besides amisulpiride or paliperidone. Fluoxetine is a powerful CYP2D6 inhibitor and a mild-to-
moderate CYP3A4 inhibitor; thus, it should be avoided in patients on risperidone. Similarly, when a CYP2D6 PM is taking any CYP3A4 inhibitor, another anti-
psychotic not metabolized by CYP2D6 and CYP3A4 should be used. Do not use with risperidone: aripiprazole, brexpriprazole or iloperidone (Spina et al., 2016).

5. Personalizing paliperidone dosing
Dopheide (2008) proposed that paliperidone is more amenable for
developing LAI formulations than risperidone, which was approved in
the US in 2009 as a monthly paliperidone palmitate injection (PP1M)
for schizophrenia. The three-month LAI paliperidone formulation (or
PP3M) was approved in 2015. A 6-month LAI formulation (or PP6M) is
currently being developed.
This paliperidone section describes: 1) bioavailability, 2) clearance,
3) C/D ratio, 4) therapeutic range, 5) DDIs related to personalizing
dosing, 6) personal variables related to personalizing dosing, 7) LAI
TDM for personalizing dosing, and 8) personalized treatment com-
bining oral risperidone and PP1M.
5.1. Paliperidone bioavailability
Surprisingly, the FDA approved an extended-release (ER) formula-
tion that only provides 28% bioavailability and allowed the company to
complete bioavailability studies with an immediate-release formulation
with 80% bioavailability (Sheehan et al., 2010). On the other hand, the
FDA forced the recall of a generic formulation of paliperidone's ER
formulation with even lower bioavailability (Cassels, 2017).
5.2. Paliperidone clearance
Extrapolating data from the immediate-release formulation to the
oral ER formulation, approximately 25% of oral paliperidone absorbed
is metabolized mainly by CYP3A4 while the other 75% is eliminated in
urine as paliperidone, half by GFR and the other half excreted by an
unknown renal transporter (Schoretsanitis et al., 2018d).
Carbamazepine, an inducer of CYP3A4 (and P-gp), enhances the
clearance of oral paliperidone. In patients under carbamazepine in-
duction, ¾ of the paliperidone may be metabolized by CYP3A4 and
only ¼ may be eliminated in the urine unchanged (Schoretsanitis et al.,
2018d). As pregnancy is associated with increased GFR and CYP3A4
activity, pregnancy may increase paliperidone clearance (Schoretsanitis
et al., 2018d).
Oral paliperidone goes through first-pass metabolism, while LAI
paliperidone does not. In fact, the increased clearance of carbamaze-
pine in oral paliperidone is 3 times as great while it is 2 times as great

6
J. de Leon
for PP1M (Schoretsanitis et al., 2018c); this is compatible with a greater
role of CYP3A4 (or P-gp) in clearance due to first-pass metabolism for
oral paliperidone.
5.3. Paliperidone C/D ratio
In the TDM systematic review of oral paliperidone (6 studies with a
total sample of 221 non-Korean adult patients), non-geriatric patients
had a weighted mean C/D ratio of 4.1 ng/ml per mg/d (Schoretsanitis
et al., 2018d). For PP1M (3 studies with a total sample of 69 adult
patients), it was almost twice as high, 7.7 ng/ml per mg/day
(Schoretsanitis et al., 2018c).
Two studies including 100 Korean adult patients provided a
weighted mean C/D ratio of 2.59 ng/ml per mg/day. This extremely
low C/D ratio is most probably due to the low bioavailability of the
Korean ER formulation; another less likely possibility is that Koreans
have greater activity in the unknown renal transporter that eliminates
paliperidone (Schoretsanitis et al., 2018d).
5.4. Paliperidone's therapeutic range
For paliperidone, Hiemke et al. (2018) in their consensus guidelines,
adopted the therapeutic reference range of oral risperidone (20–60 ng/
mL).
5.5. DDIs related to personalizing paliperidone dosing
When adding a potent CYP3A4 inducer such as carbamazepine,
phenytoin, or phenobarbital, the prescriber should multiply the pali-
peridone oral dose by 3 and the PP1M dose by 2 to maintain stable
paliperidone concentrations. A better alternative is to conduct pali-
peridone TDM after maximal induction, which would be after 5 weeks
for carbamazepine due to its auto-induction, or after 3 weeks for other
inducers (de Leon, 2015a). Paliperidone TDM should be used when
adding mild CYP3A4 inducers such as oxcarbazepine, topiramate or St.
John's wort (de Leon, 2015a).
In a company study, single-dose valproate was an inhibitor which
would require halving the oral paliperidone dose (Remmerie et al.,
2016), but in repeated dosing valproate appeared to be an inducer. The
author recommends avoiding co-prescription of valproate with oral
paliperidone, or PP1M. If co-prescribed, clinicians should perform re-
peated TDMs as valproate may act as an inhibitor, an inducer, or both,
at different times in the same patient.
5.6. Personal variables related to personalizing paliperidone dosing
The C/D ratio of 6.9 ng/ml/mg/day in one study of 15 elderly
Japanese patients suggested that geriatric patients may only need ap-
proximately half of the paliperidone oral dose (Schoretsanitis et al.,
2018d). As geriatric age may increase PP1M C/D ratios (by a factor of
1.26), TDM should guide dosing in these patients (Schoretsanitis et al.,
2018d). Regardless of the severity level of renal impairment, another
antipsychotic should be selected instead of paliperidone.
5.7. LAI TDM and personalizing paliperidone dosing
There is an urgent need to investigate real-world PP1M half-life.
Currently, clinicians should consider steady state to be reached no
earlier than after 9 injections of PP1M (8 months), but it may be longer
for obese patients and possibly for female patients. A TDM level ob-
tained 3 or 4 months after PP1M onset may provide a C/D ratio that is
half the 7.7 ng/ml per mg/day expected at steady state, around 3–4 ng/
ml per mg/day. If the concentration after 3 or 4 months is already in the
therapeutic range of 20–60 ng/ml, the clinician may need to consider
decreasing the dose since the patient may tolerate the first injections
well, but after steady state is reached beyond the 9th injection,
7
Neuropharmacology xxx (xxxx) xxxx
concentration levels may double, increasing the risk of dose-related
ADRs (Schoretsanitis et al., 2018c).
Deltoid injections provide better PP1M absorption than gluteal in-
jections, which may provide particularly low concentrations in obese
patients. Steady state may take more than one year for PP3M and
several years for PP6M. For that reason, Lee et al. (2015) proposed that
PP3M steady state is achieved in one and one-half years for deltoid
injections and almost two years for gluteal injections. Due to these
extremely long half-lives, the author recommends establishing baselines
before moving from PP1M to PP3M and from PP3M to PP6M by: 1)
examining the patient for dyskinetic movements; 2) completing an
electrocardiogram (ECG); and 3) collecting blood for baseline TDM,
prolactin and creatine kinase. In the unlucky situation that ADRs de-
velop after reaching steady state, years after moving from PP1M to
PP3M or from PP3M to PP6M, the clinician can compare the later va-
lues with baseline values (Schoretsanitis et al., 2018c). Clinicians also
need to remember that long-term hyperprolactinemia may be a risk
factor for osteoporosis (Montejo et al., 2016).
5.8. Personalized treatment combining oral risperidone and PP1M
Taking into account the limited bioavailability of oral paliperidone,
there is no good reason to use this compound; oral risperidone appears
to be a much more reliable choice. There is no published TDM data with
RBP-7000 but data on the risperidone microsphere formulation is
problematic when compared with LAI paliperidone due to: 1) unreliable
absorption in obese patients, 2) need for initial oral supplementation,
and 3) need for injections every 2 weeks instead of every 4 weeks
(Schoretsanitis et al., 2017). Thus, the clinician may consider going
from oral risperidone to PP1M.
According to the marketer's simulation study (Russu et al., 2018),
patients on oral risperidone should be switched to the following PP1M
equivalent doses: 1) 1 mg/day to 39 mg (or 25 equivalents of active
paliperidone) every month, 2) 2 mg/day to 78 mg (or 50 equivalents)
every month, 3) 4 mg/day to 156 mg (or 100 equivalents) every month,
and 4) 6 mg/day to 234 mg (or 150 equivalents) every month.
As the above marketer's guidelines are derived from simulations and
not real-world measurements, it is a good idea to complete TDM in all
patients moving from oral risperidone to PP1M, although in the average
patient, oral risperidone and PP1M may have similar C/D ratios.
CYP2D6 PM status due to genetic variations or use of CYP2D6 in-
hibitors may require relatively higher PP1M doses; this is described in
Table 8 of a prior article (Schoretsanitis et al., 2018c).
In conclusion, paliperidone dosing does not benefit from genotyping
in the manner that risperidone dosing does, but can greatly benefit from
TDM particularly when: 1) using LAI formulations due to their long
half-lives; 2) treating older patients; and 3) co-prescribing inducers,
including carbamazepine, or valproate, which may be an inducer or an
inhibitor at different times.
6. Personalizing clozapine dosing
This clozapine section describes: 1) clearance; 2) genetics; 3) the
influence of sex, smoking and ethnicity on dosing; 4) DDIs related to
personalizing dosing; 5) other personal variables related to persona-
lizing dosing; 6) therapeutic range; 7) clozapine and total clozapine C/
D ratios; 8) normal variations on clozapine TDM; and 9) personalizing
dosing.
6.1. Clozapine clearance
Clozapine has 3 metabolic pathways: 1) N-demethylation of cloza-
pine to norclozapine; 2) clozapine-N-oxide, which is reversible and
partially conducted by flavin-containing monooxygenase (FMO); and 3)
glucuronidation by the UDP glucuronosyltransferases (UGTs).
Demethylation is mainly accomplished by CYP1A2 which may
J. de Leon
account for 70% of clozapine metabolism (Bertilsson et al., 1994).
CYP2C19, CYP3A4 and CYP2D6 have lesser roles in demethylation
(Olesen and Linnet, 2001). Norclozapine is further metabolized by
conjugation with sulfates and glucuronides, but Schaber et al. (1998)
estimated that 71% of free serum norclozapine is eliminated through
tubular secretion by an unknown transporter (Barclay et al., 2019).
6.2. Genetics
Most known CYP1A2 alleles have no clear effects on function (de
Leon, 2015a). It has been proposed that the CYP1A2*1F allele is asso-
ciated with the increased effects of inducibility from smoking in Cau-
casians, but it is not clear that influences clozapine clearance (Kootstra-
Ros et al., 2005). A unique French clozapine PM had a very rare allele,
CYP1A2*7 in heterozygous state, which has never been identified again
(Allorge et al., 2003). In a Chinese inpatient sample, this allele was not
identified in any of the PMs who accounted for 4% (5/129) of the
sample and needed very low clozapine doses of 75–115 mg/day to get
clozapine therapeutic concentrations (Ruan et al., 2019).
In a recent systematic review of the literature on clozapine TDM, 13
cases were identified as being published because overly high clozapine
doses were needed (Ruan et al., 2019). Nine of these 13 cases appear to
be compatible with clozapine UM status in US or European patients,
defined as needing > 900 mg/day to get a clozapine concentration of
350 ng/ml. Several authors proposed that these cases were actually
genetic UMs but they did not provide enough information to clarify that
lack of compliance or inducers other than smoking were excluded
(Ruan et al., 2019). The high clozapine dose of a US case taking up to
1300 mg/day appeared to be explained because the patient was a male
smoker who also was prescribed valproate, which acted as an inducer
(Riesselman et al., 2013).
6.3. The influence of sex, smoking and ethnicity on clozapine dosing
Genetic studies on clozapine metabolism should consider smoking
and sex major known determinants of CYP1A2 activity. Smoking, by
stimulating the aryl hydrocarbon receptor, is an inducer of CYP1A2.
Females have slightly lower CYP1A2 activity, which is probably ex-
plained by the fact that estrogens are CYP1A2 inhibitors; in fact, oral
contraceptives are clinically relevant inhibitors of clozapine metabo-
lism (de Leon 2015a).
In 1997, two studies (Chang et al., 1997; Chong et al., 1997), fre-
quently ignored by Western psychiatrists, described how Chinese pa-
tients received half the clozapine dosage used in Western countries, but
had clozapine concentrations similar to Westerners. This is compatible
with a very well-controlled caffeine study demonstrating lower caffeine
clearance in Koreans than Swedes (Ghotbi et al., 2007). Caffeine is also
mainly metabolized by CYP1A2. A recent systematic review (Ruan
et al., 2019) found significantly lower clearance in East Asian patients
than in Caucasians. The East Asian samples included Korean and Chi-
nese patients, who are known to be closely related. Japanese are usually
defined as East Asians but the systematic review identified no Japanese
studies. After the systematic review search, a limited study has pro-
vided some limited data on clozapine clearance (Kikuchi et al., 2018)
which is compatible with the idea that Japanese are likely to have
clearance similar to Chinese and Koreans. The concept of East Asian is a
rough approximation of a complex genetic reality in Asia and is usually
applied to Chinese, Koreans, Japanese and Mongolians, who are closely
related from the genetic point of view (Singh and Teo, 2019). The
limited published clozapine data (Ruan et al., 2019) suggest that other
Asians may also have lower clozapine clearance than Caucasians. As a
matter of fact, other groups with clozapine clearance lower than Cau-
casians may include a substantial part of the world, possibly including
patients from India (Rajkumar et al., 2013) and possibly even non-
Caucasian patients with American ancestry (González-Esquivel et al.,
2011), who are descendants of East Asian immigrants to the Americas
8
Neuropharmacology xxx (xxxx) xxxx
(Nielsen et al., 2017).
6.4. DDIs related to personalizing clozapine dosing
Other potent clozapine inhibitors are fluvoxamine and cipro-
floxacin. Caffeine is metabolized by CYP1A2 and is probably a com-
petitive inhibitor. Other inducers of clozapine metabolism are ome-
prazole, rifampin and the antiepileptic inducers, such as
carbamazepine, phenytoin and phenobarbital (Spina et al., 2016).
6.5. Other personal variables related to personalizing clozapine dosing
There are no studies on the effect of renal clearance on clozapine
elimination, but renal impairment and geriatric age are likely to have
effects on clozapine clearance. Severe infections (Ruan and de Leon,
2019) or inflammations (Ruan et al., 2018), probably by releasing cy-
tokines, can also inhibit clozapine metabolism. Pneumonia appears to
be particularly lethal in clozapine patients, possibly because the in-
flammation can contribute to clozapine intoxications, creating positive
feedback and further worsening the symptoms (Cicala et al., 2019; de
Leon et al., 2019; Ruan and de Leon, 2019).
Clozapine is lipophilic and probably deposits in the fat tissue;
therefore, clozapine-induced weight gain may decrease clozapine
clearance, but the effects are probably small in most patients. However,
it cannot be ruled out that females with extreme obesity may behave as
clozapine PMs (Diaz et al., 2018). During pregnancy, CYP1A2 activity
decreases remarkably, probably due to the massive increase in estro-
gens, but there are no clozapine TDM studies in pregnancy.
6.6. Clozapine's therapeutic range
Hiemke et al. (2018), in their consensus guideline, recommend a
therapeutic reference range of 350–600 ng/ml for clozapine without
including norclozapine. As norclozapine has no antipsychotic efficacy,
it makes sense to focus only on clozapine concentration for the efficacy
of clozapine therapy in treatment-resistant schizophrenia
(Schoretsanitis et al., 2019b). Although it cannot be ruled out that some
unusual patients may need clozapine concentrations > 600 ng/ml, the
author has never seen any in his clinical practice and recommends that
if any patient has repeated and consistent trough steady-state con-
centrations in the therapeutic range and does not show even a partial
response to clozapine, the clozapine trial should be considered a failure.
He usually tries to keep these concentrations in the therapeutic range
(between 400 and 500 ng/ml) for 2–3 months unless clozapine ADRs
cannot be tolerated. For other indications besides treatment-resistant
schizophrenia, the author has anecdotal experience that some patients
may respond to low clozapine doses with concentrations lower than
350 ng/ml. They include those with symptomatic polydipsia (de Leon
et al., 1995) and patients with self- or hetero-aggressive behavior in the
context of intellectual disability and the absence of psychosis (Sabaawi
et al., 2006).
Hiemke et al. (2018), in their consensus guideline, describe >
1000 ng/ml for clozapine as a laboratory alert level. Although the
author has seen many patients with trough steady-state clozapine
concentrations > 1000 ng/ml and no clozapine ADRs, he usually re-
commends that clinicians consider decreasing the clozapine dosage if
concentrations are consistently > 1000 ng/ml (de Leon and Simpson,
2004).
Serum norclozapine concentrations appear to contribute to some
clozapine ADRs including constipation (de Leon et al., 2003) and sei-
zures (McCollum et al., 2018). Moreover, a review of the pharmaco-
logical and clinical literature suggests that using total serum clozapine
concentration (by adding clozapine and norclozapine) may be better
than clozapine concentrations for: 1) reflecting clozapine clearance and
2) possibly predicting some dose-related ADRs, such as hypersalivation,
constipation, sedation and seizures (Schoretsanitis et al., 2019b).
J. de Leon Neuropharmacology xxx (xxxx) xxxx

Table 3
Current approximations for initial clozapine dosing and subsequent use of TDM for average adults.
DO NOT START when infection/inflammation is present: before starting, verify that CRP and ESR are normal. Valproate: divide titration in half. Other inhibitors (fluvoxamine or
oral contraceptives): stop them.

First week

- Typicala Caucasian:
- Typicalc East Asian:
a first dose of 25 mg at night to avoid sedation and orthostatic hypotensionb If tolerated, ↑ daily dose by 25 mg, keeping approximately 2/3 at
night Schedule ↑, avoiding major increases during times with less supervision (weekends). Target 100 mg/day at the end of the first week. Draw
CRP/ESR (stop until normal); if clozapine TDM > 250 ng/ml, dose may be enough.
a first dose of 12.5 mg at night to avoid sedation and orthostatic hypotensionb If tolerated, ↑ daily dose by 12.5 mg, keeping approximately 2/
3 at night Schedule ↑, avoiding major increases during times with less supervision (weekends). Target 50 mg/day at the end of the first week.
Draw CRP/ESR (stop until normal); if clozapine TDM > 250 ng/ml, dose may be enough.

Second week
- Typicala Caucasian:
- Typicalc East Asian:
if tolerated, ↑ twice 50 mg/day each time, keeping approximately 2/3 at night Target 200 mg/day at the end of the second week. Draw CRP/
ESR (stop until normal); if clozapine TDM > 250 ng/ml, dose may be enough.
if tolerated, ↑ twice 25 mg/day each time, keeping approximately 2/3 at night. Target 100 mg/day at the end of the second week.
Draw CRP/ESR (stop until normal); if clozapine > 250 ng/ml, dose may be enough.

Third week
- Typicala Caucasian: if tolerated, ↑ twice 50 mg/day each time, keeping approximately 2/3 at night. Target 300 mg/day at the end of the third week.
- Typicalc East Asian: if tolerated, ↑ twice 25 mg/day each time, keeping approximately 2/3 at night. Target 150 mg/day at the end of the third week. Target
150 mg/day at the end of the third week.

Fourth week TDMd

- Caucasians on 300 mg/day:
East Asians on 150 mg/day:
C = 350 ng/ml; C/D ratio = 350/300 = 1.2. Needs ≥300 mg/day (bettere 300–400 mg/day).
C = 175 ng/ml; C/D ratio = 175/300 = 0.6. Needs ≥600 mg/day (bettere 600–700 mg/day).
C = 87 ng/ml; C/D ratio = 87/300 = 0.3.f May need up to 900 mg/day
C = 350 ng/ml; C/D ratio = 350/150 = 2.4. Needs ≥150 mg/day (bettere 150–200 mg/day).
C = 175 ng/ml; C/D ratio = 175/150 = 1.2. Needs ≥300 mg/day (bettere 300–400 mg/day).

Further increases in patients needing higher doses

- In Caucasians, ↑twice 50 mg/day each time with a target ↑ 100 mg/week.
- In East Asians, ↑twice 25 mg/day each time with a target ↑ 50 mg/week.

Target doses
- One week after you reach your target dose, repeat TDM to verify > 350 ng/ml (betterf if in 400 s ng/ml).
- Whenever clozapine C reaches > 1000 ng/ml, look for the reason and consider cutting the clozapine dose in half.
- Make the administration as easy as possible, including rounding doses and administering twice a day.g
- If you do not have access to clozapine TDM, best approximated recommendations:
Caucasians: ♀ non-smokers: around 300 mg/day (100 in the early morning and 200 at night).
♂non-smokers and ♀ smokers: 300–600 mg/day.
♂ smokers: around 600 mg/day (200 in the early morning and 400 at night).
East Asians: ♀ non-smokers: around 150 mg/day (50 in the early morning and 100 at night).
♂ non-smokers and ♀ smokers: 150–300 mg/day.
♂ smokers: around 300 mg/day (100 in the early morning and 200 at night).

Further need for TDM

- If clinically relevant inducers or inhibitors are added or discontinued, repeat TDM after steady state is reached.
- Development of major inflammations or major infectionsh requires close monitoring for ADRs. If TDM results take several days to arrive, you may want to cut the clozapine dose in
half after collecting TDM.
- Remember that major changes in smoking may have relevant effects on clozapine C after a few (1–3) weeks. Smoking cessation may ↑ clozapine C (approximately by a factor of
1.25).
- Watch for dramatici changes in caffeine intake.

ADR: adverse drug reaction; C: concentration; C/D ratio: clozapine concentration/dose ratio; CRP: C-reactive protein; ESR: erythrocyte sedimentation rate; TDM:
therapeutic drug monitoring.
a
These recommendations are not for Caucasians taking clinically relevant inhibitors, of geriatric age or taking medication that increases clozapine risk due to
pharmacodynamic drug interactions. For them, consider the titration recommended for East Asians.
b
During titration, orthostatic changes in blood pressure and pulse need to be measured. Cut scores for abnormal values and a scale for documentation are provided
by Sabaawi et al. (2006). Orthostatic changes should be assessed before increasing any clozapine dose for at least the first 3 weeks. Orthostatic abnormalities or other
ADRs should signal slower titration. Fever should prompt stopping titration and drawing CRP and ESR and waiting until they normalize to proceed with titration.
c
Consider slowing the titration for East Asians taking clinically relevant inhibitors, of geriatric age or taking medication that increases clozapine risk due to
pharmacodynamic drug interactions.
d
Draw trough steady-state clozapine C in the early AM at least one week after the last clozapine dose increase.
e
Fluctuations of 10% are not unusual in the same patient taking the same dose and using the same laboratory. A clozapine C result of around 350 ng/ml may be
associated with risk of subtherapeutic concentrations on other days. Keeping a clozapine C in the 400 s ng/ml appears easy and safe.
f
Double-check as much as you can that lack of adherence does not explain low clozapine C. Repeat TDM if you have any suspicion that low clozapine C could be
explained by lack of adherence.
g
Occasionally a single dose at night facilitates compliance and can ↓ risks for ADRs unless the patient gets up in the middle of the night.
h
Major infections include pneumonia, upper respiratory infections with fever, pyelonephritis and appendicitis.
i
Although no prospective study defines what a “dramatic” change is, caution is recommended with increases or decreases of daily caffeine intake of > 1 cup of
coffee (or 2 cans of caffeinated soda) in non-smokers and > 3 cups (or 6 cans of caffeinated soda) in smokers. For example, when a smoker taking clozapine increases
caffeine intake by three cups of coffee (e.g., from 2 to 5 cups per day), clinicians should watch for increased side effects due to ↑ clozapine C.

9
J. de Leon
Therefore, when these dose-related ADRs are present, the author re-
commends decreasing clozapine dose but keeping clozapine con-
centrations close to 350 ng/ml (in the low 400s ng/ml) and trying to
decrease the total concentrations as much as possible. Gemfibrozil may
inhibit the renal transporter that excretes norclozapine and can cause
elevations limited to norclozapine without increasing clozapine con-
centrations (Barclay et al., 2019).
6.7. Clozapine and total clozapine C/D ratio
In the US, the average patient usually needs from 300 to 600 mg/
day to reach the lowest values in the therapeutic range (350 ng/ml) for
schizophrenia. US male smokers usually have a therapeutic con-
centration ≥350 ng/ml on a dose of 600 mg/day; this produces a C/D
ratio of 0.60 (600/350). US female non-smokers usually have a con-
centration ≥350 ng/ml with a dose of 300 mg/day; this produces a C/D
ratio of 1.17 (300/350). Therefore, in the US, average clozapine C/D
ratios typically range between 0.6 and 1.2 ng/ml per mg/day. On the
other hand, for Chinese patients, clozapine C/D ratios appear to be
approximately double (Spina and de Leon, 2015). In a small study
comparing 20 Caucasians from Australia and 20 Asian patients from
Singapore without matching for sex and smoking, Ng et al. (2005)
described a clozapine C/D ratio of 1.01 vs 2.65 ng/ml per mg/day. In a
systematic review, the clozapine C/D ratio was significantly higher
when comparing the weighted mean values of 1.57 in 876 East Asians
vs. 1.07 in 1147 Caucasians (Ruan et al., 2019).
While the clozapine C/D ratio can be used by clinicians for dosing,
the total clozapine C/D ratio, calculated by adding clozapine and nor-
clozapine, better reflects clozapine clearance and may better predict
some of the clozapine dose-related ADRs (Schoretsanitis et al., 2019b).
6.8. Normal variations on clozapine TDM
Clinicians are not always aware that a single clozapine level must be
viewed with caution and that a pattern change after measuring several
levels is much easier and more reliable to interpret. Laboratory, tech-
nical and natural variations can cause some day-to-day variations in
clozapine levels, even after assuming stability of all possible con-
founding factors such as timing of collection, dose and schedule,
smoking variations and DDIs. It seems reasonable to suggest that only a
change by a factor of two is probably meaningful from the clinician's
perspective This means that if an individual has a clozapine level of
500 ng/ml, the next one in the same stable conditions should not
be > 1000 or < 250 ng/ml. However, a change from 500 to 400 ng/ml
is probably not significant (Sabaawi et al., 2006).
When a patient with repeated normal clozapine or total clozapine
C/D ratios exhibits a dramatic decrease in these ratios, the clinician
should first consider lack of compliance unless this decrease can be
explained by the addition of an inducer, such as smoking, omeprazole
or an antiepileptic with inductive properties. As previously described,
clozapine UMs are very rare so when a patient with no prior levels has a
very low clozapine or total clozapine C/D ratio, the first suspicion
should be lack of compliance.
6.9. Personalizing clozapine dosing
Box 2 provides patient examples and data on how to personalize
clozapine dosing. The best way of personalizing clozapine dosing is to
pay attention to the clinical response (look for efficacy in the absence of
ADRs), conduct clozapine TDM and keep the patient's serum clozapine
concentration at 350–600 ng/ml (Hiemke et al., 2018).
A titration scale for minimizing clozapine ADRs has been described
(Sabaawi et al., 2006). Table 3 provides starting and target clozapine
doses for adult patients, which can be better implemented by adding
TDM. Table 3 updates a prior version (Spina and de Leon, 2015) after
taking into account the risk of rapid titration, which can cause
10
Neuropharmacology xxx (xxxx) xxxx
clozapine-induced myocarditis (Chopra and de Leon, 2016).
7. Conclusions
After 20 years of combining pharmacogenetics and TDM to perso-
nalize dosing for antipsychotics, the author has developed models for
risperidone, paliperidone, and clozapine based on the concept of C/D
ratios at trough and steady-state conditions. Risperidone dosing is in-
fluenced by CYP2D6 polymorphism, DDIs and renal function.
Paliperidone dosing is influenced by inducers and renal function. TDM
is particularly important for the LAI formulations because the medica-
tion may accumulate once steady state is reached (after years for PP3M
and PP6M). Clozapine dosing is influenced by smoking, sex, DDIs and
ethnicity.
Declaration of interest
No commercial organizations had any role in the writing of this
paper for publication. J. de Leon personally develops his presentations
for lecturing, has never lectured using any pharmaceutical or pharma-
cogenetic company presentations, and has never been a consultant for
pharmacogenetic or pharmaceutical companies. In the past, J. de Leon
received researcher-initiated grants from Eli Lilly (one ended in 2003
and the other, as co-investigator, ended in 2007); from Roche Molecular
Systems, Inc. (ended in 2007); and, in a collaboration with Genomas,
Inc., from the NIH Small Business Innovation Research program (ended
in 2010). J. de Leon has been on the advisory boards of Bristol-Myers
Squibb (2003/04) and AstraZeneca (2003). Roche Molecular Systems
supported one of his educational presentations, which was published in
a peer-reviewed journal (2005). His lectures were supported once by
Sandoz (1997), twice by Lundbeck (1999, 1999), twice by Pfizer (2001,
2001), three times by Eli Lilly (2003, 2006, and 2006), twice by Janssen
(2000, 2006), once by Bristol-Myers Squibb (2006), and seven times by
Roche Molecular Systems, Inc. (once in 2005 and six times in 2006).
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Acknowledgments
The author acknowledges Lorraine Maw, M.A., from the University
of Kentucky Mental Health Research Center at Eastern State Hospital,
who helped in editing the article and the reviewers who provided im-
portant suggestions for improving the article.
References
Allorge, D., Chevalier, D., Lo-Guidice, J.M., Cauffiez, C., Suard, F., Baumann, P., Eap,
C.B., Broly, F., 2003. Identification of a novel splice-site mutation in the CYP1A2
gene. Br. J. Clin. Pharmacol. 56, 341–344.
Barclay, J., McCollum, B., Schoretsanitis, G., de Leon, J., 2019. Gemfibrozil may decrease
norclozapine elimination: a case report. J. Clin. Psychopharmacol (in press).
Barnhill, J., Susce, M.T., Diaz, F.J., de Leon, J., 2005. Risperidone half-life in a patient
taking paroxetine – a case report. Pharmacopsychiatry 38, 223–225.
Bertilsson, L., Carrillo, J.A., Dahl, M.L., Llerena, A., Alm, C., Bondesson, U., Lindström, L.,
Rodriguez de la Rubia, I., Ramos, S., Benitez, J., 1994. Clozapine disposition covaries
with CYP1A2 activity determined by a caffeine test. Br. J. Clin. Pharmacol. 38,
471–473.
Bork, J.A., Rogers, T., Wedlund, P.J., de Leon, J., 1999. A pilot study on risperidone
metabolism: the role of cytochromes P450 2D6 and 3A. J. Clin. Psychiatry 60,
469–476.
Castberg, I., Spigset, O., 2005. Serum concentrations of risperidone and 9-hydro-
xyrisperidone after administration of the long-acting injectable form of risperidone:
evidence from a routine therapeutic drug monitoring service. Ther. Drug Monit. 27,
103–106.
Cassels, C., 2017. Paliperidone manufacturer issues urgent recall. Medscape - Jun 15,
2017. https://www.medscape.com/viewarticle/881708.
Chang, W.H., Lin, S.K., Lane, H.Y., Hu, W.H., Jann, M.W., Lin, H.N., 1997. Clozapine
dosages and plasma drug concentrations. J. Formos. Med. Assoc. 96, 599–605.
J. de Leon
Chong, S.A., Tan, C.H., Khoo, Y.M., Lee, H.S., Wong, K.E., Ngui, F., Winslow, M., 1997.
Clinical evaluation and plasma clozapine concentrations in Chinese patients with
schizophrenia. Ther. Drug Monit. 19, 219–223.
Chopra, N., de Leon, J., 2016. Clozapine-induced myocarditis may be associated with
rapid titration: a case report verified with autopsy. Int. J. Psychiatry Med. 51,
104–115.
Cicala, G., Barbieri, M.A., Spina, E., de Leon, J., 2019. A comprehensive review of the
swallowing difficulties and dysphagia in adults associated with antipsychotic. Expert
Rev. Clin. Pharmacol. 12, 219–234.
Citrome, L., 2018. Sustained-release risperidone via subcutaneous injection: a systematic
review of RBP-7000 (PERSERIS (™)) for the treatment of schizophrenia. Clin.
Schizophrenia Relat. Psychoses 12, 130–141.
de Leon, J., 2009. The future (or lack of future) of personalized prescription in psychiatry.
Pharmacol. Res. 59, 81–89.
de Leon, J., 2012. Evidence-based medicine versus personalized medicine: are they
enemies? J. Clin. Psychopharmacol. 32, 153–164.
de Leon, J., 2014. False-negative studies may systematically contaminate the literature on
the effects of inducers in neuropsychopharmacology: part II: focus on bipolar dis-
order. J. Clin. Psychopharmacol. 34, 291–296.
de Leon, J., 2015b. The effects of antiepileptic inducers in neuropsychopharmacology, a
neglected issue. Part I: a summary of the current state for clinicians. Rev. Psiquiatía
Salud Ment. 8, 97–115.
de Leon, J., 2015a. The effects of antiepileptic inducers in neuropsychopharmacology, a
neglected issue. Part II: pharmacological issues and further understanding. Rev.
Psiquiatía Salud Ment. 8, 167–188.
de Leon, J., 2015-2016. Training Psychiatrists to Think like Pharmacologists. http://
inhn.org/home/courses/jose-de-leon-training-psychiatrists-to-think-like-
pharmacologists.html Last, Accessed date: 23 December 2018.
de Leon, J., 2016. Pharmacogenetic tests in psychiatry: from fear to failure to hype. J.
Clin. Psychopharmacol. 36, 299–304.
de Leon, J., 2017. Have we successfully implemented CYP2D6 genotyping in psychiatry?
Expert Opin. Drug Metabol. Toxicol. 13, 1201–1203.
de Leon, J., 2018. A critical commentary on the 2017 AGNP consensus guidelines for
therapeutic drug monitoring in neuropsychopharmacology. Pharmacopsychiatry 51,
63–68.
de Leon, J., 2019. Why do you keep telling me that drug-drug interactions are important
in psychopharmacology when I do not see them in my clinical practice? My failure to
convince clinicians. J. Clin. Psychopharmacol. 39, 1–4.
de Leon, J., Armstrong, S.C., Cozza, K.L., 2006. Clinical guidelines for psychiatrists for the
use of pharmacogenetic testing for CYP450 2D6 and CYP450 2C19. Psychosomatics
47, 75–85.
de Leon, J., Bork, J., 1997. Risperidone and cytochrome P450 3A. J. Clin. Psychiatry 58,
450.
de Leon, J., Odom-White, A., Josiassen, R.C., Diaz, F.J., Cooper, T.B., Simpson, G.M.,
2003. Serum antimuscarinic activity during clozapine treatment. J. Clin.
Psychopharmacol. 23, 336–341.
de Leon, J., Sandson, N.B., Cozza, K.L., 2008a. A preliminary attempt to personalize
risperidone dosing using drug-drug interactions and genetics: part I. Psychosomatics
49, 258–270.
de Leon, J., Sandson, N.B., Cozza, K.L., 2008b. A preliminary attempt to personalize
risperidone dosing using drug-drug interactions and genetics: part II. Psychosomatics
49, 347–361.
de Leon, J., Sanz, E.J., Norén, G.N., De las Cuevas, C., 2019. Pneumonia may be more
frequent and have more fatal outcomes in clozapine than in other second-generation
antipsychotics. World Psychiatr (in press).
de Leon, J., Simpson, G.M., 2004. Postmortem clozapine levels. J. Clin. Psychopharmacol.
24, 100–101.
de Leon, J., Spina, E., 2016. What is needed to incorporate clinical pharmacogenetics tests
into the practice of psychopharmacotherapy? Expert Rev. Clin. Pharmacol. 9,
351–354.
de Leon, J., Susce, M.T., Pan, R.M., Fairchild, M., Koch, W.H., Wedlund, P.J., 2005. The
CYP2D6 poor metabolizer phenotype may be associated with risperidone adverse
drug reactions and discontinuation. J. Clin. Psychiatry 66, 15–27.
de Leon, J., Susce, M.T., Johnson, M., Hardin, M., Maw, L., Shao, A., Allen, A.C., Chiafari,
F.A., Hillman, G., Nikoloff, D.M., 2009. DNA microarray technology in the clinical
environment: the AmpliChip CYP450 test for CYP2D6 and CYP2C19 genotyping. CNS
Spectr. 14, 19–34.
de Leon, J., Verghese, C., Stanilla, J.K., Lawrence, T., Simpson, G.M., 1995. Treatment of
polydipsia and hyponatremia in psychiatric patients. Can clozapine be a new option?
Neuropsychopharmacology 12, 133–138.
de Leon, J., Wynn, G., Sandson, N.B., 2010. The pharmacokinetics of paliperidone versus
risperidone. Psychosomatics 51, 80–88.
Diaz, F.J., Josiassen, R.C., de Leon, J., 2018. The effect of body weight changes on total
plasma clozapine concentrations determined by applying a statistical model to the
data from a double-blind trial. J. Clin. Psychopharmacol. 38, 442–446.
Dopheide, J.A., 2008. Paliperidone: an improvement over risperidone? Am. J. Health
Syst. Pharm. 65, 401.
Ereshefsky, L., 1996. Pharmacokinetics and drug interactions: update for new anti-
psychotics. J. Clin. Psychiatry 57 (Suppl. 11), 12–25.
Fang, J., Bourin, M., Baker, G.B., 1999. Metabolism of risperidone to 9-hydro-
xyrisperidone by human cytochromes P450 2D6 and 3A4. Naunyn-Schmiedeberg’s
Arch. Pharmacol. 359 147-15.1.
Flockhart, D.A., 1996. Drug interactions, cardiac toxicity, and terfenadine: from bench to
clinic? J. Clin. Psychopharmacol. 16, 101–103 1996.
Gaedigk, A., Sangkuhl, K., Whirl-Carrillo, M., Klein, T., Leeder, J.S., 2017. Prediction of
CYP2D6 phenotype from genotype across world populations. Genet. Med. 19, 69–76.
11
Neuropharmacology xxx (xxxx) xxxx
Ghotbi, R., Christensen, M., Roh, H.K., Ingelman-Sundberg, M., Aklillu, E., Bertilsson, L.,
2007. Comparisons of CYP1A2 genetic polymorphisms, enzyme activity and the
genotype-phenotype relationship in Swedes and Koreans. Eur. J. Clin. Pharmacol. 63,
537–546.
González-Esquivel, D.F., Castro, N., Ramírez-Bermúdez, J., Custodio, V., Rojas-Tomé, S.,
Castro-Román, R., Jung-Cook, H., 2011. Plasma levels of clozapine and norclozapine
in Mexican schizophrenia patients. Arzneimittelforschung 61, 335–339.
Güzey, C., Aamo, T., Spigset, O., 2000. Risperidone metabolism and the impact of being a
cytochrome P450 2D6 ultrarapid metabolizer. J. Clin. Psychiatry 61, 600–601.
Hefner, G., Falter, T., Bruns, K., Hiemke, C., 2015. Elevated risperidone serum con-
centrations during acute inflammation, two cases. Int. J. Psychiatry Med. 50,
335–344.
Hefner, G., Shams, M.E., Unterecker, S., Falter, T., Hiemke, C., 2016. Inflammation and
psychotropic drugs: the relationship between C-reactive protein and antipsychotic
drug levels. Psychopharmacology (Berlin) 233, 1695–1705.
Hendset, M., Molden, E., Refsum, H., Hermann, M., 2009. Impact of CYP2D6 genotype on
steady-state serum concentrations of risperidone and 9-hydroxyrisperidone in pa-
tients using long-acting injectable risperidone. J. Clin. Psychopharmacol. 29,
537–541.
Hiemke, C., Bergemann, N., Clement, H.W., Conca, A., Deckert, J., Domschke, K.,
Eckermann, G., Egberts, K., Gerlach, M., Greiner, C., Gründer, G., Haen, E.,
Havemann-Reinecke, U., Hefner, G., Helmer, R., Janssen, G., Jaquenoud, E., Laux, G.,
Messer, T., Mössner, R., Müller, M.J., Paulzen, M., Pfuhlmann, B., Riederer, P., Saria,
A., Schoppek, B., Schoretsanitis, G., Schwarz, M., Gracia, M.S., Stegmann, B., Steimer,
W., Stingl, J.C., Uhr, M., Ulrich, S., Unterecker, S., Waschgler, R., Zernig, G., Zurek,
G., Baumann, P., 2018. Consensus guidelines for therapeutic drug monitoring in
neuropsychopharmacology: update 2017. Pharmacopsychiatry 51, 9–62.
Heykants, J., Huang, M.L., Mannens, G., Meuldermans, W., Snoeck, E., Van Beijsterveldt,
L., Van Peer, A., Woestenborghs, R., 1994. The pharmacokinetics of risperidone in
humans: a summary. J. Clin. Psychiatry 55 (Suppl. 13), 17.
Huang, M.L., Van Peer, A., Woestenborghs, R., De Coster, R., Heykants, J., Jansen, A.A.,
Zylicz, Z., Visscher, H.W., Jonkman, J.H., 1993. Pharmacokinetics of the novel an-
tipsychotic agent risperidone and the prolactin response in healthy subjects. Clin.
Pharmacol. Ther. 54, 257–268.
Kane, J.M., Eerdekens, M., Lindenmayer, J.P., Keith, S.J., Lesem, M., Karcher, K., 2003.
Long-acting injectable risperidone: efficacy and safety of the first long-acting atypical
antipsychotic. Am. J. Psychiatry 160, 1125–1132.
Kikuchi, Y.S., Kanbayashi, T., Shimizu, T., 2018. Relationship between clozapine-induced
electroencephalogram abnormalities and serum concentration of clozapine in
Japanese patients with treatment-resistant schizophrenia. Psychiatr. Investig. 16,
279–284.
Kootstra-Ros, J.E., Smallegoor, W., van der Weide, J., 2005. The cytochrome P450
CYP1A2 genetic polymorphisms *1F and *1D do not affect clozapine clearance in a
group of schizophrenic patients. Ann. Clin. Biochem. 42, 216–219.
Laffont, C.M., Gomeni, R., Zheng, B., Heidbreder, C., Fudala, P.J., Nasser, A.F., 2015.
Population pharmacokinetic modeling and simulation to guide dose selection for
RBP-7000, a new sustained-release formulation of risperidone. J. Clin. Pharmacol.
55, 93–103.
Lee, L.H., Choi, C., Collier, A.C., Barr, A.M., Honer, W.G., Procyshyn, R.M., 2015. The
pharmacokinetics of second-generation long-acting injectable antipsychotics: lim-
itations of monograph values. CNS Drugs 29, 975–983.
Leon, J.d., Susce, M.T., Pan, R.M., Wedlund, P.J., Orrego, M.L., Diaz, F.J., 2007. A study
of genetic (CYP2D6 and ABCB1) and environmental (drug inhibitors and inducers)
variables that may influence plasma risperidone levels. Pharmacopsychiatry 40,
93–102.
Leucht, S., Cipriani, A., Spineli, L., Mavridis, D., Orey, D., Richter, F., Samara, M., Barbui,
C., Engel, R.R., Geddes, J.R., Kissling, W., Stapf, M.P., Lässig, B., Salanti, G., Davis,
J.M., 2013. Comparative efficacy and tolerability of 15 antipsychotic drugs in schi-
zophrenia: a multiple-treatments meta-analysis. Lancet 382, 951–962.
Mannens, G., Huang, M.L., Meuldermans, W., Hendrickx, J., Woestenborghs, R.,
Heykants, J., 1993. Absorption, metabolism, and excretion of risperidone in humans.
Drug Metab. Dispos. 21, 1134–1141.
McCollum, B., Barclay, J., de Leon, J., 2018. Unexpected falls during clozapine treatment
explained by myoclonus. Prim. Care Companion CNS Disord. 20 (1) pii:17l02151.
Meyer, U.A., 2004. Pharmacogenetics – five decades of therapeutic lessons from genetic
diversity. Nat. Rev. Genet. 5, 669–676.
Migliardi, G., Spina, E., D'Arrigo, C., Gagliano, A., Germanò, E., Siracusano, R., Diaz, F.J.,
de Leon, J., 2009. Short- and long-term effects on prolactin of risperidone and
olanzapine treatments in children and adolescents. Prog. Neuro-Psychopharmacol.
Biol. Psychiatry 33, 1496–1501.
Montejo, Á.L., Arango, C., Bernardo, M., Carrasco, J.L., Crespo-Facorro, B., Cruz, J.J., Del
Pino, J., García Escudero, M.A., García Rizo, C., González-Pinto, A., Hernández, A.I.,
Martín Carrasco, M., Mayoral Cleries, F., Mayoral van Son, J., Mories, M.T.,
Pachiarotti, I., Ros, S., Vieta, E., 2016. Spanish consensus on the risks and detection of
antipsychotic drug-related hyperprolactinaemia. Rev. Psiquiatía Salud Ment. 9,
158–173.
Ng, C.H., Chong, S.A., Lambert, T., Fan, A., Hackett, L.P., Mahendran, R., Subramaniam,
M., Schweitzer, I., 2005. An inter-ethnic comparison study of clozapine dosage,
clinical response and plasma levels. Int. Clin. Psychopharmacol. 20, 163–168.
Nielsen, R., Akey, J.M., Jakobsson, M., Pritchard, J.K., Tishkoff, S., Willerslev, E., 2017.
Tracing the peopling of the world through genomics. Nature 541, 302–310.
Olesen, O.V., Linnet, K., 2001. Contributions of five human cytochrome P450 isoforms to
the N-demethylation of clozapine in vitro at low and high concentrations. J. Clin.
Pharmacol. 41, 823–832.
Odom-White, A., de Leon, J., 1996. Clozapine levels and caffeine. J. Clin. Psychiatry 57,
175–176.
J. de Leon
Rajkumar, A.P., Poonkuzhali, B., Kuruvilla, A., Jacob, M., Jacob, K.S., 2013. Clinical
predictors of serum clozapine levels in patients with treatment-resistant schizo-
phrenia. Int. Clin. Psychopharmacol. 28, 50–56.
Regenthal, R., Kunstler, U., Hesse, S., Sabri, O., Preiss, R., 2005. D2 dopamine receptor
occupancy, risperidone plasma level and extrapyramidal motor symptoms in pre-
viously drug-free schizophrenic patients. Int. J. Clin. Pharmacol. Ther. 43, 370–378.
Remmerie, B., Ariyawansa, J., De Meulder, M., Coppola, D., Berwaerts, J., 2016. Drug-
drug interaction studies of paliperidone and divalproex sodium extended-release
tablets in healthy participants and patients with psychiatric disorders. J. Clin.
Pharmacol. 56, 683–692.
Riesselman, A., Strobl, B., Cooley, A.T., de Leon, J., 2013. A case report that suggested
that aspirin's effects on valproic acid metabolism may contribute to valproic acid's
inducer effects on clozapine metabolism. J. Clin. Psychopharmacol. 33, 812–814.
Ruan, C.J., de Leon, J., 2019. Thirty years of both ignorance and clinical experience
suggest that clozapine intoxication during co-occurring infections and inflammation
may have higher morbidity and mortality than is currently believed. Psychosomatics
60, 221–222.
Ruan, C.J., Zhang, X.L., Guo, W., Li, W.B., Zhuang, H.Y., Li, Y.Q., Wang, C.Y., Tang, Y.L.,
Zhou, F.C., de Leon, J., 2018. Two cases of high serum clozapine concentrations
occurring during inflammation in Chinese patients. Int. J. Psychiatry Med. 53,
292–305.
Ruan, C.J., Zang, Y.N., Wang, C.Y., Cheng, Y.H., Sun, C., Spina, E., de Leon, J., 2019.
Clozapine metabolism in East Asians and Caucasians: a pilot exploration of the pre-
valence of poor metabolizers and a systematic review. J. Clin. Psychopharmacol. 39,
135–144.
Russu, A., Kern Sliwa, J., Ravenstijn, P., Singh, A., Mathews, M., Kim, E., Gopal, S., 2018.
Maintenance dose conversion between oral risperidone and paliperidone palmitate 1
month: practical guidance based on pharmacokinetic simulations. Int. J. Clin. Pract.
72, e13089.
Sabaawi, M., Singh, N.N., de Leon, J., 2006. Guidelines for the use of clozapine in in-
dividuals with developmental disabilities. Res. Dev. Disabil. 27, 309–336.
Schaber, G., Stevens, I., Gaertner, H.J., Dietz, K., Breyer-Pfaff, U., 1998. Pharmacokinetics
of clozapine and its metabolites in psychiatric patients: plasma protein binding and
renal clearance. Br. J. Clin. Pharmacol. 46, 453–459.
Schoretsanitis, G., de Leon, J., Haen, E., Stegmann, B., Hiemke, C., Gründer, G., Paulzen,
M., 2018b. Pharmacokinetics of risperidone in different application forms – com-
paring long-acting injectable and oral formulations. Eur. Neuropsychopharmacol. 28,
130–137.
Schoretsanitis, G., Kane, J.M., Ruan, C.J., Spina, E., Hiemke, C., de Leon, J., 2019b. A
comprehensive review of the clinical utility of and a combined analysis of the clo-
zapine/norclozapine ratio in therapeutic drug monitoring for adult patients. Expert
Rev. Clin. Pharmacol. https://doi.org/10.1080/17512433.2019.1617695. Epub
ahead of print.
Schoretsanitis, G., Paulzen, M., Unterecker, S., Schwarz, M., Conca, A., Zernig, G.,
Gründer, G., Haen, E., Baumann, P., Bergemann, N., Clement, H.W., Domschke, K.,
Eckermann, G., Egberts, K., Gerlach, M., Greiner, C., Havemann-Reinecke, U., Hefner,
G., Helmer, R., Janssen, G., Jaquenoud-Sirot, E., Laux, G., Messer, T., Mössner, R.,
Müller, M.J., Pfuhlmann, B., Riederer, P., Saria, A., Schoppek, B., Silva Gracia, M.,
12
Neuropharmacology xxx (xxxx) xxxx
Stegmann, B., Steimer, W., Stingl, J.C., Uhr, M., Ulrich, S., Waschgler, R., Zurek, G.,
Hiemke, C., 2018a. TDM in psychiatry and neurology: a comprehensive summary of
the consensus guidelines for therapeutic drug monitoring in neuropsychopharma-
cology, update 2017; a tool for clinicians. World J. Biol. Psychiatr. 19, 162–174.
Schoretsanitis, G., Spina, E., Hiemke, C., de Leon, J., 2017. A systematic review and
combined analysis of therapeutic drug monitoring studies for long-acting risperidone.
Expert Rev. Clin. Pharmacol. 10, 965–981.
Schoretsanitis, G., Spina, E., Hiemke, C., de Leon, J., 2018d. A systematic review and
combined analysis of therapeutic drug monitoring studies for oral paliperidone.
Expert Rev. Clin. Pharmacol. 11, 625–639.
Schoretsanitis, G., Spina, E., Hiemke, C., de Leon, J., 2018c. A systematic review and
combined analysis of therapeutic drug monitoring studies for long-acting paliper-
idone. Expert Rev. Clin. Pharmacol. 11, 1237–1253.
Schoretsanitis, G., Villasante-Tezanos, A.G., de Leon, J., 2019a. Studies of half-lives of
long-acting antipsychotics are needed. Pharmacopsychiatry 52, 45–46 Oct 22.
Sheehan, J.J., Sliwa, J.K., Amatniek, J.C., Grinspan, A., Canuso, C.M., 2010. Atypical
antipsychotic metabolism and excretion. Curr. Drug Metabol. 11, 516–525.
Simpson, G.M., Lindenmayer, J.P., 1997. Extrapyramidal symptoms in patients treated
with risperidone. J. Clin. Psychopharmacol. 17, 194–201.
Singh, R., Teo, Y.Y., 2019. "Asian" phenotype underestimates the genetic diversity of Asia
yet overstates its impact on variability in drug disposition and pharmacodynamics.
Clin. Pharmacol. Ther. 105, 802–805.
Smith, D., Beaumont, K., Maurer, T.S., Di, L., 2019. Clearance in drug design. J. Med.
Chem. 62, 2245–2255.
Snoeck, E., Van Peer, A., Sack, M., Horton, M., Mannens, G., Woestenborghs, R., Meibach,
R., Heykants, J., 1995. Influence of age, renal and liver impairment on the phar-
macokinetics of risperidone in man. Psychopharmacology (Berlin) 122, 223–229.
Spina, E., Avenoso, A., Facciolà, G., Salemi, M., Scordo, M.G., Giacobello, T., Madia, A.G.,
Perucca, E., 2000. Plasma concentrations of risperidone and 9-hydroxyrisperidone:
effect of comedication with carbamazepine or valproate. Ther. Drug Monit. 22,
481–485.
Spina, E., de Leon, J., 2015. Clinical applications of CYP genotyping in psychiatry. J.
Neural. Transm. (Vienna) 122, 5–28.
Spina, E., Hiemke, C., de Leon, J., 2016. Assessing drug-drug interactions through ther-
apeutic drug monitoring when administering oral second-generation antipsychotics.
Expert Opin. Drug Metabol. Toxicol. 12, 407–422.
Subramanian, S., Völlm, B.A., Huband, N., 2017. Clozapine dose for schizophrenia.
Cochrane Database Syst. Rev. 6, CD009555 2017.
Van Peer, A., Meuldermans, W., Woestenborghs, R., Heykants, J., 1996. Clinical phar-
macokinetics of risperidone. In: Kane, J.M., Moller, H.J., Fawouter, S. (Eds.),
Serotonin in Antipsychotic Treatment: Mechanism and Clinical Practice. Marcel
Dekker, Inc., New York, pp. 277–291.
Williams, R., 2001. Optimal dosing with risperidone: updated recommendations. J. Clin.
Psychiatry 62, 282–289.
Yasui-Furukori, N., Hidestrand, M., Spina, E., Facciolá, G., Scordo, M.G., Tybring, G.,
2001. Different enantioselective 9-hydroxylation of risperidone by the two human
CYP2D6 and CYP3A4 enzymes. Drug Metab. Dispos. 29, 1263–1268.