Variability in Metabolism of Imipramine and Desipramine Using Urinary Excretion Data†

Kelley Ramey1, Joseph D. Ma1,2, Brookie M. Best1,3, Rabia S. Atayee1,2 and Candis M. Morello1,4*

1Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego (UCSD), Pharmaceutical Sciences Building

(PSB), Dean’s Suite, Room 1121, 9500 Gilman Drive, MC 0657, La Jolla, CA 92093-0657, USA, 2Doris A. Howell Palliative Care Service, San

Diego, CA, USA, 3UCSD Department of Pediatrics, Rady Children’s Hospital, San Diego, CA, USA and 4Diabetes Intense Medical

Management Clinic, Veterans Affairs San Diego Healthcare System, La Jolla, CA, USA

*Author to whom correspondence should be addressed. Email: candismorello@ucsd.edu

Variability in imipramine and desipramine metabolism was evaluatedusing urinary excretion data from patients with pain. Liquid chroma-tography–tandem mass spectrometry was used to quantitate con-centrations in urine specimens. Interpatient population contained600 unique imipramine specimens, whereas intrapatient populationhad 137 patients with two or more specimens. Normal concentrationranges of imipramine, desipramine and the desipramine/imipraminemetabolic ratio (MR) were established, and various factors were test-ed for MR impact. Geometric mean of imipramine urine concentrationwas 0.46 mg/g of creatinine, and desipramine was 0.67 mg/g ofcreatinine. Gender, concomitant known CYP2C19 inhibitor use andurine pH did not affect MR. However, proton-pump inhibitor (PPI)users had a significantly lower mean MR than those without a listedPPI. Early age group (18–36 years) had a significantly higher meanMR than middle (37–66 years) and late (67–90 years) age groups.Approximately one-third were positive for one or more of hydroco-done, oxycodone, hydromorphone or oxymorphone. Patients withno opioids reported in the medication list had a significantly lowergeometric mean MR than those with prescribed opioids (1.03 vs.1.54, P 5 0.004). Patients with only one prescribed opioid had alower MR than those with two or more prescribed opioids. Patientswith younger age, prescribed opioids and no listed PPI weremore likely to have a higher geometric mean urinary desipramine/imipramine MR.

Introduction

Imipramine is a tricyclic antidepressant (TCA) that can be pre-

scribed for the treatment of neuropathic pain (NP). Although

the primary use of TCAs for the treatment of depression has de-

creased with the introduction of many other antidepressants,

which have a more tolerable side-effect profile compared with

the TCAs, the TCAs remain an effective therapeutic option for cer-

tain types of depressive illnesses and NP. TCAs are considered first-

line treatment for NP, and are effective for central poststroke NP,

painful polyneuropathy and painful diabetic neuropathy (1, 2).

Both norepinephrine and serotonin reuptake inhibition are impor-

tant in the treatment of NP (2), yet TCAs have also been shown to

have direct opioid receptor activity (3).

All TCAs have low affinity for the m-opioid receptor, and imip-

ramine and desipramine have higher affinity for the k-opioid re-

ceptor over other receptor subtypes (3). Stimulation of these

various opioid receptors may be the means through which the

TCAs elicit their analgesic effects, which have been documented

in both humans and rats (3). Also, stimulation of d-opioid

receptors may help with depression (3). Several studies have

found that opioid receptor antagonists block the antinociceptive

effects of TCAs (3, 4) and when administered with morphine,

TCAs provide synergistic analgesia (3). The effects of TCAs on

opioid receptors are not fully understood, but these interactions

may be vital to their efficacy in the treatment of NP and

depression.

Imipramine’s affinity for muscarinic acetylcholine receptors is

the cause of adverse effects such as dry mouth, blurry vision,

urine retention, constipation, memory loss and tachycardia (5).

Strong binding to histamine H1-receptors can cause moderate

sedation and lower the seizure threshold (6). These effects are

less common with desipramine due to its secondary amine struc-

ture (6). Finally, overdoses of both imipramine and desipramine

can have cardiotoxic effects, where imipramine can cause QT

prolongation and desipramine can induce changes in the QRS

complex (6). Understanding imipramine’s metabolism and sourc-

es of its variability in and among patients will help clinicians rec-

ognizewhy certain patients are less able to tolerate the drug than

others.

Imipramine is extensivelymetabolized in the liver by cytochrome

P450 (CYP) enzymes. Imipramine undergoesN-demethylation to its

main, active metabolite desipramine via CYP2C19, CYP1A2 and

CYP3A4. CYP2C19 contributes the most to this process, while

CYP1A2 and CYP3A4 have smaller contributions, with wide fluctu-

ations in CYP2C19 and CYP3A4 activity among poor and extensive

metabolizers (7–9). CYP2D6 is responsible for hydroxylation of

both imipramine and desipramine into 2-, 10-hydroxyimipramine

and 2-, 10-hydroxydesipramine, respectively. Desipramine does

not undergo metabolism by any other CYP enzymes and thus is

more sensitive to CYP2D6 polymorphisms. For example, the plasma

metabolic ratio (MR) of desipramine/imipramine is able to discrim-

inate between CYP2D6 extensive and poor metabolizers (10, 11).

This is of clinical significance in that response to desipramine use

may differ based on CYP2D6 metabolizer status. Whether the

same MR in urine is able to reproduce findings in plasma is un-

known. A large majority of imipramine metabolites are recovered

in the urine (.75%), while up to �20% is excreted via the bile

and feces (12). Imipramine metabolites in the urine are found in

the following amounts: imipramine þ desipramine (1–4%), other

non-conjugated metabolites (15–35%), glucuronide metabolites

(40–60%) and non-extractable polar metabolites (20–30%) (13).

In the population of patients with pain, where often one drug

is not sufficient to treat the pain, it is essential to understand the

potential interactions between imipramine and medications

taken concomitantly, especially opioids. In addition, variability

in drug response and metabolism is due to genetic, non-genetic

and environmental factors. This analysis of urinary excretion data

examined factors such as age, sex, urine pH and concomitant

†Presented as a poster at: PAINWeek: National Conference on Pain for

Frontline Practitioners, 5–8 September 2012, Las Vegas, NV.

# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal of Analytical Toxicology 2014;38:368–374

doi:10.1093/jat/bku034 Advance Access publication April 29, 2014 Article

medications to determine their effect on imipramine and desipr-

amine metabolism.

Methods

Urine specimens for patients being treated for chronic pain were

analyzed for routine clinical care using LC–MS-MS at Millennium

Laboratories (San Diego, CA, USA). For this study, a retrospective

data analysis was conducted using de-identified data already col-

lected from clinical specimens to determine the urinary concen-

trations of imipramine, desipramine and creatinine. The dataset

included subject identification number, specimen identification

number, date of urine sample collection, subject’s date of birth,

sex, physician’s practice code, subject’s current medication list,

urine-creatinine concentration, urine pH and urine drug concen-

trations of imipramine, desipramine and other pain medications.

Institutional Review Board-exempt status was granted by the

University of California–San Diego Human Research Protection

Program.

Subjects and specimens

Between January 2011 and April 2012, 715,651 urine specimens

were tested for imipramine and desipramine. Selected specimens

were positive for both imipramine and desipramine, defined as

concentrations of �50 ng/mL. Specimens with urine-creatinine

concentration of ,20 mg/dL were excluded to limit for subject

variability due to hydration status. Patients whomet the inclusion

criteria with one visit or the first of several visits were defined as

the intersubject population. The intrasubject population was de-

fined as subjects who met the inclusion criteria with two or

more visits.

LC–MS-MS analysis

An Agilent 1200 series binary pump SL LC system, well-plate sam-

pler and thermostatted column compartment paired with an

Agilent Triple Quadrupole mass spectrometer and an Agilent

Mass Hunter software were used for analysis of imipramine

and desipramine. Chromatographic separation was performed

using an acetonitrile– formic acid–water gradient running at

0.4 mL/min and a 2.1 � 50 mm2, 1.8-mm Zorbax SB-C18 column.

Mobile phase A ¼ þ0.1% formic acid in water, B ¼ 0.1% formic

acid in acetonitrile and column temperature was set to 508C.Samples were prepared for injection by incubating 25 mL of

urine with 50 units of b-glucuronidase Type L-II from Patella

vulgata (keyhole limpet) Sigma Product number G 8132

(Sigma-Aldrich Corp., St. Louis, MO, USA) in 50 mL of 0.4 M ace-

tate buffer (pH 4.5) for 3 h at 458C. Five microliters of the solu-

tion was injected for each sample.

All spectra were collected using positive electrospray ioniza-

tion. The optimized instrumental parameters were as follows:

gas temperature, 3508C; drying gas, 12 L/min; nebulizer gas

(nitrogen), 35 psi (�24,100 Pa); capillary voltage, 3,000 V and

fragmenter voltage, 60 V. Multiple reaction monitoring (MRM)

mode was used for quantitation. Scan time was set to 500 ms.

In the MRM mode, two transitions were used to identify and

quantitate a single compound. Data were acquired running

the QQQ in MRM mode, using transitions imipramine: 281.1!86, 281.1! 58 and desipramine: 267.1! 72, 267.1! 44.

A quantitative transition was used to calculate concentration

based on the quantifier ion, and a second transition was used

to ensure accurate identification of the target compound based

on the ratio of the quantifier ion to the quantifier ion.

HPLC-grade water, acetonitrile, methanol and formic acid

were obtained from VWR (Westchester, PA, USA). Imipramine

and desipramine were obtained from Cerrilliant Corp. (Round

Rock, TX, USA). The deuterated internal standards were diluted

to 1,000 ng/mL by adding them to synthetic urine (Microgenics

Corp., Fremont, CA, USA).

Quantitative analysis was performed using the Agilent Mass

Hunter Quantitative Analysis software. A four-point calibration

curve was created by using a linear fit and forcing the line to

go through the origin. Accepted accuracy for calibrators was

+20% of the target value and the coefficient of determination

(R2) was required to be �0.99 as verification of linearity and

goodness of fit. The lower limit of quantitation for both the imip-

ramine and desipramine was 50 ng/mL. The upper limit of line-

arity for both the imipramine and desipramine assays was

50,000 ng/mL.

Calculations, statistical methods and graphical analyses

Imipramine and desipramine concentrations were normalized

using creatinine concentration to account for hydration status

(14). The analyte concentration (ng/mL) was divided by the

urine-creatinine level (mg/dL), and divided by 10 to correct

the volume, with the final concentration reported as milligram

analyte per gram creatinine (mg/g cr). The MR was defined as

the concentration ratio of desipramine to imipramine, a unitless

value. Dosage information was not originally collected as part of

the routine clinical urine drug monitoring, so doses were not

available for this retrospective analysis. In the absence of dosage

information, the MR is the best estimate of imipramine metabo-

lism to desipramine. The imipramine and desipramine creatinine-

corrected urine concentrations were also summed together by

converting from milligrams of drug to millimoles.

Descriptive statistics and graphical analyses were performed

using Microsoft Excelw 2010 (Microsoft Corp., Redmond, WA,

USA) and OriginPro 8.6 (OriginLab Corp., Northampton, MA,

USA). To achieve normal distributions of analyte concentrations,

the creatinine-corrected concentrations were log transformed.

Log-transformed data were used to calculate descriptive statis-

tics. The log-transformed data were back-transformed to deter-

mine the geometric mean and geometric 95% confidence

intervals (CIs).

Concomitant medications

Each concomitant drug that was detected in the urine specimen

was tested individually for its effect on the log MR. For each case,

two-sample populations were created: samples that were positive

for an opioid, benzodiazepine or other pain medications versus

those that were negative for that concomitant medication.

Differences in log MR were then compared between populations

of reported and no reported opioid use, using provider-reported

medication lists. Differences in log MR were also compared be-

tween the population reporting two or more opioids with the

population reporting only one opioid. Similarly, the effect of

CYP2C19 inhibitors on the log MR was tested between groups

Metabolism of Imipramine and Its Metabolite Desipramine 369

with reported use and no reported use of a CYP2C19 inhibitor.

CYP2C19 inhibitors were determined from the ‘P450 Drug

Interaction Table’ from the Indiana University’s School of

Medicine website (15). For these comparisons above, two-sided,

two sample t-tests were performed, and a significance level of

0.05 was used.

Evaluation of other factors

Age, sex and urinary pHwere also evaluated for their effect on the

MR and on the sum of imipramine plus desipramine (Iþ D). Based

on the age groups from a previous study (16), the interpatient

population was divided into three age categories: young (18–36

years), middle (37–66 years) and old (67–90 years). The corre-

sponding log MR and I þ D values for the three age groups were

compared with an one-way ANOVA test. A two-sample t-test com-

pared MR and I þ D by sex. Linear regression analyses using the

least-squares method was performed using urine pH; values

ranged from 4.6 to 9.3.

Results

Of the 715,651 patients tested for imipramine, 882 tested posi-

tive for imipramine in the urine and satisfied our inclusion crite-

ria. After removing duplicate specimens from the same patient

ID, 600 unique specimens remained, characterizing the interpa-

tient population. Of the 600 patients, only 105 had imipramine

reported in the medications list. The intrapatient population con-

sisted of 137 patients with two or more positive imipramine

urine specimens from separate visits. The number of visits

in the intrapatient population ranged from 3 to 13 visits.

Imipramine and desipramine concentrations did not exceed

the upper limit of linearity (50,000 ng/mL). No subjects report-

ed desipramine in the medication list.

Imipramine and desipramine concentrations are shown in

Figure 1. A Gaussian distribution was observed for the log-

transformed concentrations of imipramine, desipramine and

MR. The imipramine geometric mean was 0.461 mg/g cr and

the 95% CI of the mean was 0.417–0.508. The geometric mean

for desipramine concentration was 0.674 mg/g cr and the 95% CI

of the mean was 0.606–0.750. The observed range of desipr-

amine concentrations was slightly higher than imipramine. The

geometric mean for the sum of imipramine plus desipramine

was 4.605 nmoles/g cr and the 95% CI of the mean was 4.181–

5.072.

For the intrapatient population, the imipramine geometric

mean urine concentration was 0.600 mg/g cr and the 95% CI

of the mean was 0.502–0.716. The desipramine geometric

mean concentration was 0.810 mg/g cr and the 95% CI of the

mean was 0.665–0.987. Once again, urine desipramine concen-

trations were slightly higher than imipramine concentrations.

For the intrasubject population, the geometric mean for the

sum of imipramine plus desipramine was 6.219 nmoles/g cr

and the 95% CI of the mean was 4.981–7.764.

Linear regression analysis was used to determine the relation-

ship between imipramine and desipramine concentrations, as

well as imipramine concentration and MR. A positive correlation

(y ¼ 0.79x þ 0.098, R2 ¼ 0.52; Figure 2) was observed between

imipramine and desipramine concentrations.

The relationship between parent drug and metabolite was fur-

ther assessed by examining MR. The MRs were similar between

the two populations; the geometric mean for the interpatient

population was 1.47 with a 95% CI of 1.36–1.59, whereas the

intrapatient geometric mean was 1.35 with a 95% CI of 1.17–

1.57. Linear regression with imipramine concentration and MR

produced a poor linear fit (R2 ¼ 0.068; Figure 1d), showing

that a correlation between the two was quite small.

Other factors and MR variability

The young age group had a significantly higher mean log MR than

the middle and older age groups, as shown in Table I, with geo-

metric means of 1.95, 1.45 and 1.22, respectively. Sex, CYP2C19

inhibitors and urine pHwere not significantly related to the imip-

ramine MR; all relevant two-sample t-tests produced results of

P . 0.05. However, proton-pump inhibitors (PPIs), when tested

alone without the other CYP2C19 inhibitors, were significantly

associated with lower imipramine MR (1.12 vs. 1.50, P ¼ 0.02).

The patients had the following PPIs in their medication list:

omeprazole (20), esomeprazole (17), pantoprazole (4), rabepra-

zole (3) and lansoprazole (1).

The young group also had a significantly lower sum of I þ D in

the urine compared with the middle age group, with geometric

means in the young, middle and older age groups of 3.24, 4.98

and 4.35, respectively (P ¼ 0.018). When assessing the sum of

I þ D, females had higher values than males, 5.06+ 3.26 vs.

3.81+3.43 nmoles/g cr (P ¼ 0.0062). Urine pH showed a signif-

icant negative relationship with I þ D, but the correlation was

weak (R2 ¼ 0.013).

Concomitant medications

About 30% of patients were positive for one or more of the fol-

lowing: hydrocodone, oxycodone, hydromorphone or oxymor-

phone. Fourteen percent were positive for morphine, and 9%

were positive for methadone. The concomitant opioid frequen-

cies are shown in Table II.

Evaluation of the individual opioids for their impact on the

imipramine MR produced no significant results. However, imip-

ramine patients with no reported opioids in the medication list

had a significantly lower geometric mean MR than the

opioid-reporting population, as shown in Table II. Furthermore,

patients with two or more reported opioids had a significantly

higher MR than those with only one reported opioid (1.71 vs.

1.45, P , 0.05). A slightly, but not significantly, higher mean

MR was found in subjects positive for oxymorphone (P ¼ 0.07),

with geometric mean MRs of 1.59 and 1.40, respectively. A slightly

higher MR was also seen for patients positive for hydromorphone

(P ¼ 0.09), where the respective means were 1.57 and 1.41. Other

common concomitant drugs, including gabapentin, carisoprodol

and benzodiazepines, were not significantly related to the MR

(P . 0.05).

Discussion

The results of this study help establish observed normal ranges

for urinary concentrations of imipramine and desipramine. The

mean desipramine urine concentration was slightly higher than

370 Ramey et al.

the mean imipramine urine concentration. Desipramine plasma

half-life has a wide reported range of 7–60 h, and imipramine

plasma half-life is 8–16 h (6). The wider range of desipramine

plasma half-life may be due to the variability in hepatic metabo-

lism (6). Imipramine also undergoes hepatic metabolism; but has

multiple CYP enzymes involved, leading to a wide bioavailability

range after oral administration of 22–77%. Enterohepatic circula-

tion may occur for both drugs. Elimination rates from the plasma

will include metabolism and urinary excretion, whereas elimina-

tion rates in the urine represent only the fraction of the drug and

drug metabolites that get excreted via the urine. Less than 5% of

imipramine is excreted unchanged in the urine (6) and desipr-

amine fraction excreted unchanged is ,3%, although this varies

between different ethnic groups (17).

The scatterplot in Figure 2 shows a positive relationship be-

tween desipramine and imipramine urine concentrations. The

parent drug andmetabolitewere moderately positively correlated,

with an R2-value of 0.52. Besides demethylation to desipramine by

CYP2C19, CYP1A2 and CYP3A4, imipramine undergoes hydroxyl-

ation by CYP2D6 to 2- and 10-hydroxyimipramine. No evidence of

saturation of metabolism was noted in this study. This is not sur-

prising for two main reasons. First, imipramine is metabolized by

multiple CYP enzymes and secondly desipramine concentrations,

not imipramine concentrations, have been found to increase with

an increasing imipramine dose (18).

Age

The young age group had a significantly higher mean log MR than

the middle and older age groups, which agrees with previous

studies on concentrations in plasma. Studies in children have re-

ported higher concentrations of TCAmetabolites compared with

adults (19–21). The young also had the lowest total amount of

active drug in urine, imipramine plus desipramine.

Steady-state plasma concentrations of imipramine were found

to be higher in an older population, ranging from 50 to 65 year

Figure 1. (a) Log creatinine-corrected imipramine concentrations. The median (interquartile range) imipramine urine concentrations were 0.484 mg/g creatinine (0.187–1.11). (b)Log creatinine-corrected desipramine concentrations. The median (interquartile range) desipramine urine concentrations were 0.712 mg/g creatinine (0.252–1.82). (c) Log creatinine-corrected MRs. The median (interquartile range) urine MRs were 1.43 (0.754–2.77). (d) Relationship between imipramine concentration and MR. The MR showed a weak negativerelationship with the imipramine urine concentration (y ¼ 20.21x ¼ 0.098, R2 ¼ 0.068, P , 0.0001). GM, geometric mean.

Metabolism of Imipramine and Its Metabolite Desipramine 371

old, than a younger population ranging from 30 to 39 year old (22).

One explanation for this finding is that younger people have a

more rapid demethylation pathway than older people, whereas

the hydroxylation pathways are comparable. Possible explanations

for this include age-related changes in enzyme function and

changes in hepatic blood flow (22). Enzyme function may be de-

creased for several reasons. Decreased quantities of hepatic

smooth endoplasmic reticulum with age may contribute to re-

duced activity of NADPH-cytochrome P450 reductase, or de-

creased quantity of cytochrome P450 enzymes, or both (23).

The decrease in metabolism appears to affect demethylation of

imipramine more than hydroxylation of desipramine, leading to

higher concentrations of the parent drug and a lower MR.

Studies have also found a decline in imipramine clearance in

the elderly with no change in bioavailability, suggesting that

this is due to impaired demethylation to desipramine (23).

Another study, however, has shown a disproportionate increase

in desipramine levels when imipramine dose was increased (18),

suggesting that hydroxylation, and not demethylation, was the

saturable pathway. While both pathways are variable between pa-

tients, the overall consensus on imipramine metabolism is that it

is impaired in the elderly, which is consistent with the findings in

this study.

No significant difference in the MR was found between men

and women, which is consistent with previous studies (19, 22).

This study did find a higher sum of imipramine plus desipramine

in women compared with men. Some studies have found higher

TCA concentrations in women, but these findings have been at-

tributed to concomitant medications such as oral contraceptives,

which can inhibit hepatic enzymes such as CYP3A (17). The lack

of MR difference may also be due to CYP2D6 inhibition, which

would result in expected increases in urinary concentrations of

both imipramine and desipramine, thus resulting in no change in

the overall MR.

Proton-pump inhibitors

The PPIs, that is omeprazole, esomeprazole, lansoprazole, panto-

prazole and rabeprazole, are known inhibitors of CYP2C19 (13).

All PPIs are extensively metabolized in the liver, with the

CYP2C19 enzyme being a major metabolic pathway that is in

turn inhibited from metabolizing other drugs (24). Lansoprazole

and omeprazole are the most potent inhibitors of CYP2C19,

while rabeprazole and pantoprazole are the least potent inhibitors

(24). Inhibition constants (Ki) for the PPIs are as follows: lansopra-

zole 0.4–1.5 mM, omeprazole 2–6 mM, esomeprazole 8 mM, pan-

toprazole 14–69 mM and rabeprazole 17–21 mM (25). A

decrease in MR for patients with a PPI in the medication list can

be explained by inhibition of demethylation of imipramine, lead-

ing to increased imipramine and thus lower MR. One would ex-

pect other CYP2C19 inhibitors to lower MR, but this was not

the case in the present study. One possible explanation is that uri-

nary MRs are not robust to detect subtle changes via CYP2C19

inhibition.

Opioids

Patients with opioids contained in their medication list had a sig-

nificantly higher MR than those without concomitant opioids

(1.54 vs. 1.03, P ¼ 0.004). While tests assessing each individual

Figure 2. Relationship between imipramine and desipramine. Urine desipramine ispositively correlated with urine imipramine concentrations, but the slope of 0.79 showsthat they are not directly proportional (y¼ 0.79xþ 0.098, R2¼ 0.52, P , 0.0001).

Table IEffect of Various Factors on MR

N Geometric mean MR Significance

AgeYoung (Y, 18–36 years) 70 1.95 Y–M: P ¼ 0.047;Middle (M, 37–66 years) 417 1.45 Y–O: P ¼ 0.005;Old (O, 67–90 years) 97 1.22 M–O: P ¼ 0.241

GenderMale 201 1.38 P ¼ 0.125Female 397 1.51

CYP2C19 inhibitionInhibition 71 1.31 P ¼ 0.145No inhibition 529 1.49

PPIsPPI prescribed 45 1.12 P ¼ 0.024No PPI prescribed 555 1.50

OpioidsOpioids prescribed 476 1.54 P ¼ 0.004No opioids prescribed 46 1.03

2þ opioidsTwo or more opioids prescribed 173 1.71 P ¼ 0.039One opioid prescribed 303 1.45

Table IIType and Frequency of Concomitant Opioids

Opioid # Subjectswith drug in themedication list

# Subjectswith drugin urinea

% Subjectswith drug in themedication list

% Subjectswith drugin urinea

Oxycodone 208 186 35 31Hydrocodone 200 190 33 32Morphine 73 81 12 14Fentanyl 45 41 8 7Tramadol 34 27 6 5Methadone 33 56 6 9Oxymorphone 23 196 4 33Buprenorphine 18 15 3 3Hydromorphone 12 205 2 34Codeine 10 10 2 2Propoxyphene 4 1 0.7 0.2Meperidine 1 2 0.2 3Opium 1 N/A 0.2 0

aConcentration of drug is above the cutoff value.

372 Ramey et al.

opioid were not significant in their effect on imipramine MR, opi-

oids in general may increase it. It should be noted that inaccura-

cies in the physician-reported medication list were a possibility,

and some lists (78) were not filled out. One possible explanation

for this increased MR is the inhibition of desipramine metabolism

by CYP2D6. Tramadol, codeine, oxycodone and hydrocodone are

all substrates of CYP2D6 (23), and methadone is a known com-

petitive inhibitor of CYP2D6 (26). Also, drugs metabolized by

CYP2D6 often follow nonlinear pharmacokinetics due to the en-

zyme’s saturable nature (27, 28). In addition, genetic polymor-

phisms of CYP2D6 exist, with .80 known variant alleles.

Several CYP2D6 variant alleles are known to result in enzyme ac-

tivity that may be normal, increased, decreased or even absent

(27, 29). The prevalence of these mutations differs among ethnic

groups, with 7% of Caucasians and 1–3% of other ethnic groups

known to have limited CYP2D6 function (28). While this is only

one possible explanation that was seen as a class effect, there is a

need for further research in this matter.

Physicians may want to proceed with caution when prescrib-

ing opioids concurrently with imipramine. Impairment of metab-

olism due to either impaired CYP2D6 function or concomitant

use of medications that are CYP2D6 substrates or inhibitors

may result in increased levels of desipramine. Elevated TCA plas-

ma concentrations lead to increased adverse effects and can lead

to serious events such as prolonged cardiac QT interval, central

nervous system effects, coma and even death (30, 31). The de-

creased availability of CYP2D6 may also cause a slight increase

in imipramine plasma concentration, as this is also a substrate

of the enzyme. Patients taking the aforementioned opioids may

need a lower imipramine dose to prevent increased plasma con-

centrations of both the parent drug and the metabolite.

Limitations

A key limitation of this study is that dose amount and time of dose

administration were not known. Changes in MR over time could

be due to differences in postdose sampling times and differences

in half-lives of the two compounds. Additional limitations include

the lack of CYP 2D6 genotype information, which could help in-

terpret MR findings. Finally, the lack of plasma concentration data

precludes any direct comparisons between urine findings from

this study and expected plasma concentrations. Although these

limitations must be taken into consideration when interpreting

these results, advantages of this study were a large sample size

and robust quantitative drug and metabolite assay results within

each subject.

Conclusion

Urinary concentrations of imipramine and desipramine display

large variability both in and among patients. The MR (desipr-

amine/imipramine) was increased in young patients (18–36 year

old) compared with both middle (37–66 year old) and older

(67–90 year old) populations. At the same time, the sum of imip-

ramine plus desipramine in urine was significantly lower for young

patients. Sex and CYP2C19 inhibitors were not significantly related

to the MR, but PPIs alone were associated with lower MR, possibly

due to decreased imipramine metabolism. Females had signifi-

cantly higher sums of imipramine plus desipramine in urine.

While individual opioids alone were not significantly associated

with the MR, patients positive for oxymorphone or hydromor-

phone had a slightly increased MR. Patients with no opioids in

their medications list had a significantly lower MR than those

with prescribed opioids. Furthermore, patients with only one pre-

scribed opioid had a lower MR than those with two or more pre-

scribed opioids. This suggests that opioids may inhibit CYP2D6

metabolism of desipramine or induce N-demethylation of imipra-

mine, leading to higher concentrations of this active metabolite.

This analysis of urinary excretion ranges of imipramine and

desipramine establishes typical urinary concentrations for pain

patients and discusses variables that can alter those ranges.

Physicians may want to closely monitor imipramine doses in

younger patients, female patients and patients taking concomi-

tant opioids or PPIs. Adverse effects have been shown to occur

over very wide ranges of both imipramine and desipramine plas-

ma concentrations, but the ranges listed here may be useful to

clinical providers as a reference.

Acknowledgments

K.R. was awarded a student fellowship stipend to participate in this

project, supported by the UC San Diego Skaggs School of Pharmacy

and Pharmaceutical Sciences and an unrestricted gift to the UC San

Diego Skaggs School of Pharmacy and Pharmaceutical Sciences

from Millennium Research Institute. The authors thank Dr

Amadeo J. Pesce for his expert advice and guidance in this project.

Conflict of interest

J.D.M. is a paid consultant for Millennium Laboratories.

References

1. Dharmshaktu, P., Tayal, V., Kalra, B.S. (2012) Efficacy of antidepres-

sants as analgesics: a review. Journal of Clinical Pharmacology,

52, 6–17.2. Dworkin, R.H., O’Connor, A.B., Audette, J., Baron, R., Gourlay, G.K.,

Haanpaa, M.L. et al. (2010) Recommendations for the pharmacolog-

ical management of neuropathic pain: an overview and literature up-

date. Mayo Clinic Proceedings, 85, S3–S14.3. Onali, P., Dedoni, S., Olianas, M.C. (2010) Direct agonist activity of tri-

cyclic antidepressants at distinct opioid receptor subtypes. Journal

of Pharmacology and Experimental Therapeutics, 332, 255–265.4. Biegon, A., Samuel, D. (1980) Interaction of tricyclic antidepressants

with opiate receptors. Biochemical Pharmacology, 29, 460–462.5. Gillman, P.K. (2007) Tricyclic antidepressant pharmacology and

therapeutic drug interactions updated. British Journal of

Pharmacology, 151, 737–748.6. Clinical Pharmacology. Imipramine Monograph. 2012. http://www.

clinicalpharmacology-ip.com/Forms/Monograph/monograph.

aspx?cpnum=309&sec=monadve&t=0 (accessed Aug 15, 2012).

7. Su, P., Coutts, R.T., Baker, G.B., Daneshtalab, M. (1993) Analysis of

imipramine and three metabolites produced by isozyme CYP2D6 ex-

pressed in a human cell line. Xenobiotica, 23, 1289–1298.8. Lemoine, A., Gautier, J.C., Azoulay, D., Kiffel, L., Belloc, C.,

Guengerich, F.P. et al. (1993) Major pathway of imipramine metabo-

lism is catalyzed by cytochromes P-450 1A2 and P-450 3A4 in human

liver. Molecular Pharmacology, 43, 827–832.9. Brøsen, K., Klysner, R., Gram, L.F., Otton, S.V., Bech, P., Bertilsson, L.

(1986) Steady-state concentrations of imipramine and its metabolites

in relation to the sparteine/debrisoquine polymorphism. European

Journal of Clinical Pharmacology, 30, 679–684.

Metabolism of Imipramine and Its Metabolite Desipramine 373

10. Koyama, E., Tanaka, T., Chiba, K., Kawakatsu, S., Morinobu, S., Totsuka,

S. et al. (1996) Steady-state plasma concentrations of imipramine

and desipramine in relation to S-mephenytoin 40-hydroxylationstatus in Japanese depressive patients. Journal of Clinical

Psychopharmacology, 16, 286–293.11. Koyama, E., Sohn, D.R., Shin, S.G., Chiba, K., Shin, J.G., Kim, Y.H. et al.

(1994) Metabolic disposition of imipramine in oriental subjects: rela-

tion to metoprolol alpha-hydroxylation and S-mephenytoin

40-hydroxylation phenotypes. Journal of Pharmacology and

Experimental Therapeutics, 271, 860–867.12. Crammer, J.L., Scott, B., Woods, H., Rolfe, B. (1968) Metabolism

of 14C-imipramine. I. Excretion in the rat and in man.

Psychopharmacologia, 12, 263–277.13. Gram, L.F. (1974) Metabolism of tricyclic antidepressants. A review.

Danish Medical Bulletin, 21, 218–231.14. Cone, E.J., Caplan, Y.H., Moser, F., Robert, T., Shelby, M.K., Black, D.L.

(2009) Normalization of urinary drug concentrations with specific

gravity and creatinine. Journal of Analytical Toxicology, 33, 1–7.15. Indiana University/Division of Clinical Pharmacology. P450 Drug

Interaction Table. 2012. http://medicine.iupui.edu/clinpharm/ddis/table.aspx (accessed June 25, 2012).

16. Ishizawa, Y., Yasui-Furukori, N., Takahata, T., Sasaki, M., Tateishi, T.

(2005) The effect of aging on the relationship between the cyto-

chrome P450 2C19 genotype and omeprazole pharmacokinetics.

Clinical Pharmacokinetics, 44, 1179–1189.17. Rudorfer, M.V., Lane, E.A., Chang, W.H., Zhang, M.D., Potter, W.Z.

(1984) Desipramine pharmacokinetics in Chinese and Caucasian vol-

unteers. British Journal of Clinical Pharmacology, 17, 433–440.18. Bjerre, M., Gram, L.F., Kragh-Sorensen, P., Kristensen, C.B., Pedersen,

O.L., Moller, M. et al. (1981) Dose-dependent kinetics of imipramine

in elderly patients. Psychopharmacology, 75, 354–357.19. Rudorfer, M.V., Potter, W.Z. (1999) Metabolism of tricyclic antide-

pressants. Cellular and Molecular Neurobiology, 19, 373–409.20. Potter, W.Z., Calil, H.M., Sutfin, T.A., Zavadil, A.P., Jusko, W.J., Rapoport,

J. et al. (1982) Active metabolites of imipramine and desipramine in

man. Clinical Pharmacology and Therapeutics, 31, 393–401.

21. Preskorn, S.H., Bupp, S.J., Weller, E.B., Weller, R.A. (1989) Plasma lev-

els of imipramine and metabolites in 68 hospitalized children.

Journal of the American Academy of Child and Adolescent

Psychiatry, 28, 373–375.22. Gram, L.F., Sondergaard, I., Christiansen, J., Petersen, G.O., Bech, P.,

Reisby, N. et al. (1977) Steady-state kinetics of imipramine in pa-

tients. Psychopharmacology, 54, 255–261.23. Abernethy, D.R., Greenblatt, D.J., Shader, R.I. (1985) Imipramine and

desipramine disposition in the elderly. Journal of Pharmacology

and Experimental Therapeutics, 232, 183–188.24. Lettino, M. (2010) Inhibition of the antithrombotic effects of clopi-

dogrel by proton pump inhibitors: facts of fancies? European

Journal of Internal Medicine, 21, 484–489.25. Li, X.Q., Andersson, T.B., Ahlstrom, M., Weidolf, L. (2004) Comparison

of inhibitory effects of the proton pump-inhibiting drugs omepra-

zole, esomeprazole, lansoprazole, pantoprazole, and rabeprazole on

human cytochrome P450 activities. Drug Metabolism and

Disposition, 32, 821–827.26. Wu, D., Otton, S.V., Sproule, B.A., Busto, U., Inaba, T., Kalow, W. et al.

(1993) Inhibition of human cytochrome P450 2D6 (CYP2D6)

by methadone. British Journal of Clinical Pharmacology, 35,30–34.

27. Kroemer, H.K., Eichelbaum, M. (1995) ‘It’s the genes, stupid’.

Molecular bases and clinical consequences of genetic cytochrome

P450 2D6 polymorphism. Life Sciences, 56, 2285–2298.28. Leppert, W. (2011) CYP2D6 in the metabolism of opioids for mild to

moderate pain. Pharmacology, 87, 274–285.29. Stamer, U.M., Stuber, F. (2007) Genetic factors in pain and its treat-

ment. Current Opinion in Anesthesiology, 20, 478–484.30. Billups, S.J., Delate, T., Dugan, D. (2009) Evaluation of risk factors for

elevated tricyclic antidepressant plasma concentrations.

Pharmacoepidemiology and Drug Safety, 18, 253–257.31. Labianca, R., Sarzi-Puttini, P., Zuccaro, S.M., Cherubino, P., Vellucci, R.,

Fornasari, D. (2012) Adverse effects associated with non-opioid and

opioid treatment in patients with chronic pain. Clinical Drug

Investigation, 32, 53–63.

374 Ramey et al.