Characterization of the methemoglobin forming metabolites of benzocaine and lidocaine

Article abstract of DOI:10.1080/00498254.2016.1196402

1.Topical anesthesia with benzocaine or lidocaine occasionally causes methemoglobinemia, an uncommon but potentially fatal disorder where the blood has a reduced ability to transport oxygen. Previous in vitro studies using human whole blood have shown tha

Full text of DOI:10.1080/00498254.2016.1196402

Xenobiotica
the fate of foreign compounds in biological systems
ISSN: 0049-8254 (Print) 1366-5928 (Online) Journal homepage: http://www.tandfonline.com/loi/ixen20
Characterization of the methemoglobin forming
metabolites of benzocaine and lidocaine
Neil Hartman, Hongfei Zhou, Jinzhe Mao, Daniel Mans, Michael Boyne II,
Vikram Patel & Thomas Colatsky
To cite this article: Neil Hartman, Hongfei Zhou, Jinzhe Mao, Daniel Mans, Michael Boyne
II, Vikram Patel & Thomas Colatsky (2016): Characterization of the methemoglobin forming
metabolites of benzocaine and lidocaine, Xenobiotica, DOI: 10.1080/00498254.2016.1196402
To link to this article: http://dx.doi.org/10.1080/00498254.2016.1196402
Published online: 20 Jun 2016.
Submit your article to this journal
Article views: 85
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ixen20
Download by: [FU Berlin]
Date: 13 March 2017, At: 05:28

http://tandfonline.com/ixen
ISSN: 0049-8254 (print), 1366-5928 (electronic)
Xenobiotica, Early Online: 1–8
2016 Informa UK Limited, trading as Taylor & Francis Group. DOI: 10.1080/00498254.2016.1196402
!
RESEARCH ARTICLE
Characterization of the methemoglobin forming metabolites of
benzocaine and lidocaine
Neil Hartman1, Hongfei Zhou1, Jinzhe Mao1, Daniel Mans2, Michael Boyne II2, Vikram Patel1, and Thomas Colatsky1
1Division of Applied Regulatory Science, FDA/CDER/OTS/OCP, Silver Spring, MD, USA and 2Division of Pharmaceutical Analysis, FDA/CDER/OPS/OTR,
Silver Spring, MD, USA
Abstract
Keywords
1. Topical anesthesia with benzocaine or lidocaine occasionally causes methemoglobinemia,
an uncommon but potentially fatal disorder where the blood has a reduced ability to
transport oxygen. Previous in vitro studies using human whole blood have shown that
benzocaine causes more methemoglobin (MetHb) formation than lidocaine, and that both
compounds require metabolic transformation to form the MetHb producing species. In the
current investigation, the active species of benzocaine forming the MetHb was investigated.
2. HPLC analysis of benzocaine samples incubated with human hepatic S9 showed the
formation of a peak with the same UV spectrum and retention time as benzocaine
hydroxylamine (BenzNOH). To confirm the activity of BenzNOH, MetHb production following
exposure to the compound was determined in whole human blood using an Avoximeter
4000 CO-oximeter.
Carboxylesterase, cytochrome P450,
methemoglobinemia, N-methyl transferase
History
Received 6 April 2016
Revised 27 May 2016
Accepted 28 May 2016
Published online 17 June 2016
3. BenzNOH produced MetHb in a concentration dependent manner without the need for
metabolic activation. Benzocaine in the presence of metabolic activation required a
concentration of 500 mM to produce a similar degree of MetHb formation as 20 mM
BenzNOH without activation. Previous work suggested that two metabolites of lidocaine
may also form MetHb; N-hydroxyxylidine and 4-hydroxyxylidine. Of these two metabolites
4-hydroxyxylidine produced the most MetHb in whole blood in vitro in the absence of
metabolic activation, however BenzNOH produced up to 14.2 times more MetHb than
4-hydroxyxylidine at a similar concentration.
4. These results suggest that the ability of benzocaine to form MetHb is likely to be mediated
through its hydroxylamine metabolite and that this metabolite is inherently more active
than the potentially MetHb-forming metabolites of lidocaine.
Introduction
MetHba from benzocaine has been estimated in retrospective
studies to be in the range of 0.005%–0.25% when used to
anesthetize the airway for various diagnostic procedures
(Chowdhary et al., 2013; Kane et al., 2007). Lidocaine is also
used for anesthesia in these procedures and has been reported
to cause less MetHba than benzocaine (Chowdhary et al.,
2013); however, benzocaine is also used more frequently than
lidocaine for these procedures so relying on the number of
reported cases may not be a true assessment of the actual
incidence (IMS Health, 2013).
Benzocaine is a topical local anesthetic frequently used to
anesthetize the oropharynx for various surgical procedures as
well as being used over-the-counter for various mouth and
skin discomforts. It is generally well tolerated, however it can
occasionally
cause
methemoglobinemia
(MetHba)
(Chowdhary et al., 2013; Guay, 2009; Kane et al., 2007); an
uncommon but potentially fatal disorder in which the
oxidation state of the iron in hemoglobin is changed from
Fe(II) to Fe(III). The resulting ‘‘methemoglobin’’ (MetHb)
while nontoxic, is incapable of transporting oxygen; levels of
MetHb exceeding 60% of total hemoglobin are often fatal,
with patients compromised by other disorders (such as anemia
or bacterial infections) being at greater risk (Coleman &
Coleman, 1996; Skold et al., 2011). The clinical incidence of
The FDA has been evaluating the risk of MetHba with
benzocaine and lidocaine use with the objective of recom-
mending the safest procedure for airway anesthesia. Because
there was limited comparative data on the clinical propensity
of benzocaine and lidocaine to cause MetHba, this laboratory
performed a comparison of the ability of benzocaine and
lidocaine to produce MetHb in an in vitro system (Hartman
et al., 2014). Most compounds which cause MetHba are only
active upon metabolic activation (Bradberry, 2003).
The Hartman investigation therefore used human liver
Address for correspondence: Neil Hartman, U.S. Food and Drug
Administration, Building 64, Room 2020, 10903 New Hampshire Ave.,
Silver Spring, MD 20993, USA. E-mail: Neil.hartman@fda.hhs.gov

2
N. Hartman et al.
Xenobiotica, Early Online: 1–8
post-mitochondrial homogenates (S9) to provide metabolic
activation, as well as human whole blood to detect MetHb
formation. In these studies, it was found that high concentra-
tions (500 mM) of benzocaine produce large amounts (40%) of
MetHb in these systems, whereas no MetHb was formed in
similar systems containing 500 mM lidocaine. A metabolite of
lidocaine, 2,6-xylidine, which is formed extensively in the
body (Keenaghan & Boyes, 1972) but formed only very
slowly in the in vitro system used in the Hartman investigation
(Parker et al., 1996) produced very small amounts (4%) of
MetHb. Studies with recombinant cytochrome P450 revealed
that 1A2 produced the most MetHb when incubated with
benzocaine, with several other isoforms producing lesser
amounts. While the exact nature of the active metabolite was
not established from the Hartman study, a potential metabolite
of benzocaine, benzocaine hydroxylamine (BenzNOH), was
synthesized and was found to have many times the activity of
benzocaine and which could demonstrate this without further
metabolic activation. The present investigation seeks to more
definitely identify the active metabolite of benzocaine and to
compare that to active metabolites of lidocaine as well as to
further determine the metabolic fate of benzocaine to allow a
better understanding of when MetHba may be expected.
N-Hydroxyxylidine was synthesized based on the method
of Nelson with modifications (Nelson et al., 1978). Briefly,
1-nitro-2,6 xylene was dissolved in ethanol and treated with
an aqueous solution of ammonium chloride. Twice the molar
amount of zinc powder was added in small increments
keeping the temperature below 70 ꢀC. The solution was
filtered and the filtrate was extracted with dichloromethane.
The extract was dried and the resulting oil was triturated with
pentane to give a white solid. This was recrystallized twice
from hexane to give NOH xylidine as white needles. Purity
was 98.1% by HPLC. Identity was confirmed by NMR and
mass spectrometry.
Determining the stability of BenzNOH
One hundred micromolar BenzNOH was incubated under the
following conditions: pH 2.7 at 23 ꢀC (10 mM ammonium
phosphate buffer), pH 7.4 at 23 ꢀC (phosphate buffered saline)
or pH 7.4 at 37 ꢀC with 1 mM NADPH. One hundred
microliter samples were taken every 30 min and analyzed
directly by HPLC. Incubations were run in triplicate. An
Agilent 1200 Series HPLC (Santa Clara, CA) was used for the
analysis with quantification by UV absorbance at 290 nm
using a model G1315D diode array detector. A Phenomenex
Luna 3 mm C18(2) 100 ꢁ 4.6 mm column (Torrence, CA) was
used for the analysis. The mobile phase was 10 mM ammo-
nium phosphate pH 2.7 progressing to 50% acetonitrile over
15 min at a flow rate of 1 ml/min.
Materials and methods
Materials
Human material was obtained from commercial sources using
anonymized donors. Human whole blood was obtained from
Lampire Biologicals (Pipersville, PA) with Citrate Phosphate
Dextrose Adenine (CPDA-1) used for anticoagulation.
Subjects were of both sexes and were asked to abstain from
prescription and over-the-counter medications for 48 h before
the blood was drawn. Human liver postmitochondrial fraction
(S9, lot SUW) was obtained from BioreclamationIVT
(Baltimore, MD). This preparation was made from livers
unsuitable for transplantation pooled from 10 male subjects.
Recombinant human cytochrome P450 (CYP) containing
microsomes (Supersomes) were obtained from BD
Biosciences (San Jose, CA). Benzocaine, isopropyl-4-amino-
benzoate (isobenzocaine), 2,6-xylidine (xylidine), 4-amino-
3,5-dimethylphenol (4-hydroxyxylidine), 1-nitro-2,6 xylene,
zinc powder, 4-aminobenzoic acid, nicotinamide adenine
dinucleotide phosphate (NADP), glucose-6-phosphate,
glucose-6-phosphate dehydrogenase, phenylmethanesulfonyl
fluoride (PMSF), acetyl-coenzyme A (acetyl-CoA), acetyl-
DL-carnitine HCl and carnitine acetyltransferase were
obtained from Sigma-Aldrich (St. Louis, MO). Phosphate
buffered saline was obtained from Life Technologies (Grand
Island, NY). All other chemicals were obtained from com-
mercial sources.
Identification of benzocaine active metabolites
Metabolites of benzocaine were identified in S9 incubations
with separation by HPLC. Incubations consisted of 2.64 ml
100 mM potassium phosphate buffer, pH 7.4, with 3.3 mM
magnesium chloride, 300 ml S9 (final concentration 2 mg S9
protein/ml), 30 ml NADPH generating system (282 mg
glucose-6-phosphate, 84 mg NADP, 120 units glucose-6-
phosphate dehydrogenase per ml; final concentration 10 mM
glucose-6-phosphate, 1 mM NADP per ml) and 30 ml benzo-
caine in ethanol (final ethanol concentration 1%). Benzocaine
was added at 500 mM final concentration. Incubations were
maintained at 37 ꢀC on a shaking water bath; incubations were
started by the addition of drug. Aliquots of 500 ml were added
to 2 ml acetonitrile with 0.1% acetic acid which contained
50 nmol isobenzocaine as an internal standard. The recovery
of isobenzocaine was used to adjust the quantitation of both
benzocaine and BenzNOH. Aliquots were taken every 30 min
for 120 min. Any precipitate was removed by centrifugation
and the acetonitrile was evaporated under a stream of dry
nitrogen to approximately 200 ml. The concentrated samples
were diluted with 1 volume HPLC eluent and 100 ml samples
were analyzed by HPLC. HPLC conditions were as described
above. In incubations using recombinant human microsomes,
the S9 was replaced with recombinant human microsomes
representing a single isoform to a final concentration of
15 pmol P450/ml.
Synthesis of hydroxylamine metabolites
BenzNOH was synthesized by the method of Shintani & Fu
(2003) with modifications as previously described (Hartman
et al., 2014). Purity and identity were checked by mass
spectrometry and NMR. The product was approximately 95%
pure by NMR with the major impurity being benzocaine
present at a level of about 3%.
Mass spectral analysis
Fractions of the above HPLC analyses were collected and the
fractions of interest were evaporated to dryness under dry

DOI: 10.1080/00498254.2016.1196402
Methemoglobin from benzocaine and lidocaine
3
nitrogen and analyzed by mass spectrometry. Samples were
taken up in 200 ml 0.1% formic acid and introduced by direct
infusion. Mass spectra were obtained using a Thermo
QExactive mass spectrometer (Grand Island, NY) under
positive ion mode. Data were acquired in full scan mode
(50–400 m/z) at an operating resolution of 70 000 at m/z 200.
Final spectra were composed of 50 averaged MS scans. HCD
MS/MS spectra were also acquired. Data analysis was
performed using Thermo XCalibur QualBrowser software.
MetHb incubations and measurement
Incubations to assay the extent of MetHb formation in whole
blood were based on the work of Coleman (Coleman &
Taylor, 1997). Incubations consisted of 2.2 ml whole blood
with 250 ml phosphate buffered saline or S9 (final concentra-
tion 2 mg S9 protein/ml), 25 ml NADPH generating system (as
described above), and 25 ml benzocaine, xylidine, or specific
benzocaine or lidocaine metabolites in ethanol (final ethanol
concentration 1%). Study compounds were used at various
concentrations as specified. Incubations were run in 20 ml
glass scintillation vials; blood was maintained at 37 ꢀC on a
shaking waterbath for 60 min before adding other reagents to
allow for full oxygenation. Incubations were performed in
triplicate and were started by the addition of drug. Samples of
100 ml were taken at time zero and periodically thereafter,
generally every 30 min for 5 h. Samples were analyzed for
MetHb immediately after collection using an Avoximeter
4000 whole blood CO-oximeter (ITC, Edison, NJ) following
the manufacturer’s instructions. Some incubations also con-
tained 100 mM phenylmethanesulfonyl fluoride (PMSF) and/
or an acetyl-CoA generating system consisting of acetyl-CoA
(final concentration 0.1 mM), acetyl-D,L-carnitine HCl (final
concentration 4.6 mM) and carnitine acetyltransferees (final
concentration 0.06 units/ml).
Figure 1. Stability of BenzNOH under various incubation conditions.
mixtures proved to be problematic. Analysis of synthetic
standard BenzNOH showed that the amino OH group
fragmented in the source, producing a mass spectrum very
similar to that of benzocaine itself (m/z 166.086). While the
parent ion (m/z 154.050) could be detected in standards, it was
not detectable in the S9 incubation samples. To aid in the
detection of BenzNOH recombinant 1A2 was used in place of
S9 (data not shown). The BenzNOH peak from a 1A2
incubation was collected and used to obtain a mass spectrum
which supported the identity of BenzNOH by exact mass
(observed 154.05042, theoretical 154.04987).
The lidocaine metabolites 4-hydroxyxylidine and N-
hydroxyxylidine were tested for their ability to produce
MetHb in human whole blood. In the absence of S9 100 mM
4-hydroxyxylidine resulted in MetHb levels of about
22.3 0.6% which peak at about 210 min (Figure 4). The
results were similar in the presence of S9 up to 120 min, after
which there is a gradual loss of MetHb. When 100 mM N-
hydroxyxylidine was used in the absence of S9 there was an
immediate increase of MetHb which peaked at 13.1 0.1% at
30 min and decreased relatively rapidly thereafter. In the
presence of S9 peak, MetHb levels of 28.7 2.6% were seen
at 60 min with N-hydroxyxylidine which decreased at an
intermediate rate afterward. In the presence of S9, 500 mM
xylidine produced 3.5 0.6% MetHb at 5 h consistent with
previous studies (Hartman et al., 2014).
Data presentation
All data except the chromatograms are presented as the mean
of three determinations. Error bars represent standard devi-
ation. Some error bars do not extend beyond the graphing
symbols.
Results
When blood was incubated with increasing concentrations
of BenzNOH, a dose-dependent increase in MetHb was found
which peaked at 120–150 min and which generally main-
tained this concentration over the remaining time of the
incubation except for the lower concentrations which showed
a slight reduction at later time points (Figure 5a). The
maximum observed concentration ranged from 28.6 0.4% to
79.3 0.3% for 5 to 100 mM BenzNOH, respectively. The
relationship was less than linear, as higher concentrations
showed less than an additive increase when compared to
lower ones. This is as compared to the control incubation of
benzocaine 500 mM in the presence of S9, which reached peak
MetHb concentrations similar to that of 20 mM BenzNOH
without S9, however this peak was reached at a later time
and there was a more prominent decrease thereafter.
When 4-hydroxyxylidine was substituted in the same
Because phenylhydroxylamines are inherently unstable, the
stability of BenzNOH was ascertained. At 23 ꢀC and pH 7.4,
BenzNOH had a half-life of 126 1.0 min (Figure 1). At
23 ꢀC and pH 2.7, the half-life was 57.8 5.1 h which was
confirmed in studies lasting 72 h (data not shown). At 37 ꢀC
and pH 7.4 in the presence of 1 mM NADPH, no significant
loss of BenzNOH could be detected over the length of the
3.5 h incubation.
In metabolic studies involving benzocaine in the presence
of S9 two metabolites could be discerned; a peak eluting at
5.5 min with a retention time and UV spectrum similar to
4-aminobenzoic acid and a peak eluting at 12.2 min with a
retention time and UV spectrum similar to BenzNOH
(Figure 2). BenzNOH reached levels of 20.5 1.6 mM in
30 min and remained basically unchanged thereafter (Figure
3). Mass spectral analysis of BenzNOH in the reaction

4
N. Hartman et al.
Xenobiotica, Early Online: 1–8
Figure 2. (a) HPLC separation of benzocaine and BenzNOH. The dashed trace shows the absorbance at 290 nm of an incubation containing S9 and
100 mM benzocaine at time zero. The solid trace shows the same incubation at 120 min. The double dotted trace is synthetically produced BenzNOH.
(b) UV spectrum of benzocaine and BenzNOH. The solid trace shows the UV spectrum of the metabolite identified in the panel on the left, the long
dash trace shows the spectrum of benzocaine and the short dash trace shows the spectrum of BenzNOH. The spectra are normalized to 100%
absorbance.
system, a dose-dependent increase of MetHb was also found,
which ranged from 2.5 0.3% for 10 mM to 14.5 0.4% for
100 mM (Figure 5b). The timeframe of the increase varied
subtly for the different concentrations, with the lower
concentrations increasing approximately linearly throughout
the incubation and the higher concentrations reaching a peak
at about 150 min and decreasing slightly thereafter. When
measured at 150 min, the relationship scaled nearly linearly
up to 50 mM, with the 100 mM measurement being less than
the additive relationship expected. The control incubation of
500 mM xylidine in the presence of S9 was practically
indistinguishable from 10 mM 4-hydroxyxylidine without S9.
To evaluate the metabolic activity of S9 on benzocaine in
detail, either PMSF, to inhibit carboxylesterase, or an
acetylCoA generating system, to activate acetyltransferase,
or both were added to incubations of benzocaine, S9 and
Figure 3. Benzocaine consumption and BenzNOH formation by human
S9 in the presence of 500 mM benzocaine.
whole blood. Benzocaine at 200 mM alone showed the curve
typical of other benzocaine incubations, peaking at
25.8 2.0% at approximately 120 min and decreasing grad-
ually thereafter (Figure 6). Upon addition of AcCoA, a curve
of similar shape to the benzocaine alone curve resulted, with
each point being 52 2% less than the value of the
benzocaine alone curve. Addition of PMSF caused a curve
which was coincident with the benzocaine alone curve for
earlier time points, but which increased continuously over the
duration of the incubation, reaching 63.1 3.0% at 300 min.
Addition of both PMSF and AcCoA produced a curve which
was similar to the PMSF + benzocaine curve, but which was
again 51 4% lower at each point.
Discussion
BenzNOH was found to be unstable at pH 7.4, but was
stabilized by low pH, thus HPLC analysis could be completed
without breakdown. It was also found that the presence of
NADPH stabilized BenzNOH, which indicated that
BenzNOH formed during the incubation with S9 would be
unlikely to degrade during the length of the metabolic
incubation. The nature of this breakdown product was not
Figure 4. 4-OH-Xylidine and N-OH-xylidine were incubated in the
presence of whole blood with or without S9 and the formation of MetHb
was observed. 500 mM benzocaine and 500 mM xylidine, both with S9,
were used as positive controls.

DOI: 10.1080/00498254.2016.1196402
Methemoglobin from benzocaine and lidocaine
5
Figure 5. BenzNOH or 4-OH-xylidine was incubated at various concentrations in the presence of whole blood and the extent of MetHb formation was
determined. 500 mM benzocaine and 500 mM xylidine, both with S9, were used as positive controls.
incubations, amounting to only 4% of the concentration of the
original benzocaine. This amount of metabolism would still
account for the total amount of MetHb formed by benzocaine
however, because 20 mM BenzNOH in the absence of S9
produced a very similar MetHb profile to 500 mM benzocaine
with S9. The differences in the MetHb formation profiles of
500 mM benzocaine + S9 and 20 mM BenzNOH alone are
likely to be due to the amount of time (presumably 30 min) it
takes to generate BenzNOH from benzocaine, which would
account for the slower increase of MetHb seen with benzo-
caine, as well as due to the presence of other metabolic
enzymes present in the S9 which could degrade the BenzNOH
formed in the benzocaine incubation. Earlier work has shown
benzocaine to be extensively metabolized by carboxylesterase
and acetyltransferase (Henrikus & Kampffmeyer, 1992;
Kraeling et al., 1996; Meinertz et al., 1991). In the current
Figure 6. Benzocaine at 200 mM was incubated with whole blood and S9
with or without PMSF, to inactivate carboxylesterase, or an acetyl-CoA
generating system, to activate N-acetyltransferase, and the extent of
MetHb formation was measured.
studies using S9, carboxylesterase was active but no cofactor
was provided for the acetyltransferase reactions. Studies were
therefore performed to investigate the significance of these
reactions (Figure 6). When N-acetyltransferase was activated
by the addition of an acetyl-CoA generating system, the
degree of MetHb formation decreased by a constant amount
when compared to the similar incubation without acetyl-CoA.
This suggests that BenzNOH may be acetylated directly,
possibly preferentially. If the decline in MetHb formation was
due to loss of benzocaine due to acetylation, the difference
between incubations with and without acetyl Co-A should be
small at first and increase with time. It is not known if
acetylated BenzNOH produces MetHb, but it is unlikely due
to the presence of the acetyl group. It has been reported that
O-acetylated hydroxylamines can form adducts with macro-
molecules and thus may be mutagenic (Colvin et al., 1998).
This may be the case for acetylated BenzNOH as well.
In the presence of PMSF to inhibit carboxylesterase
(Figure 6), incubations of benzocaine in whole blood and
S9 increased continuously with time as opposed to decreasing
determined, however it is likely to be in the form of
nitrosobenzocaine, because hydroxylamines are known to
form nitrosoamines in the presence of oxygen and reducing
agents such as NADPH can inhibit this oxidation or allow the
back reduction (Keise, 1966). When benzocaine was incu-
bated with S9 without blood, BenzNOH reached a plateau at
30 min and remained relatively constant thereafter; it is
unclear whether that is due to the cessation of production of
BenzNOH or the balancing of production with degradation. In
the presence of blood, MetHb continued to increase for
150 min suggesting that continued production of BenzNOH
may not be necessary for levels of MetHb seen in the whole
blood incubations because MetHb levels continue to increase
at time points after 30 min when BenzNOH is not increasing.
BenzNOH was a relatively minor metabolite in the S9

6
N. Hartman et al.
Xenobiotica, Early Online: 1–8
O
at about 150 min as was seen in the absence of PMSF
(Figure 6). This observation could be due to increased
formation of BenzNOH due to higher levels of un-
metabolized benzocaine or due to decreased hydrolysis of
BenzNOH. Because the earlier time points of the curves with
and without PMSF are essentially coincident, it appears that
increased formation of BenzNOH is unlikely. The increased
formation of MetHb in the presence of PMSF suggests that
patients with compromised carboxylesterase activity, either
through congenital lack of enzyme or inhibition by other
compounds, might be particularly sensitive to MetHba upon
treatment with benzocaine.
OH
N-Acetyltransferase
Esterase
NH2
O
O
OH
O
4-Amino
benzoic acid
O
N-Acetyl
transferase
NH
NH2
O
O
Benzocaine
Esterase
The action of carboxylesterase on BenzNOH should result
in the formation of 4-hydroxylaminobenzoic acid. While
4-aminobenzoic acid does not form MetHb in the whole
blood/S9 incubations used in this study (Hartman et al.,
2014), the ability of 4-hydroxylaminobenzoic acid to form
MetHb is unknown, however this compound is likely to form
MetHb in a similar manner to BenzNOH. Given this, it is still
entirely possible that 4-hydroxylaminobenzoic acid may be
less active than BenzNOH. 4-Hydroxylaminobenzoic acid has
been detected in the blood of a patient who developed
MetHba subsequent to benzocaine administration (Spiller
et al., 2014). No BenzNOH was found in that study. It is not
clear if 4-hydroxylaminobenzoic acid detected was respon-
sible for the development of MetHba or if it was the result of
enzymatic degradation of BenzNOH.
4-Acetamino
benzoic acid
O
NH
CYP450
Acetylbenzocaine
MetHb
Hb
O
O
O
O
O
2
HN
NADP
N
O
NADPH
OH
When incubated with whole blood, BenzNOH caused
increasing levels MetHb in a dose dependent but not strictly
proportional manner. Part of this lack of proportionality is
due to the obvious constraint that it is not possible to go
beyond total conversion of hemoglobin to MetHb; however
total conversion still does not occur even at the highest
levels of BenzNOH tested, indicating that there must be
other factors to be considered. The formation of MetHb by
BenzNOH is likely to be similar to the formation of
phenylhydroxylamine and dapsone hydroxylamine, which
have been extensively investigated (Keise, 1966; Kramer
et al., 1972). Phenylhydroxylamine (PhenylNOH) and
hemoglobin both react with molecular oxygen to form
phenylnitrosoamine (PhenylNO), MetHb and an incom-
pletely characterized activated oxygen species. PhenylNO
can then be reduced back to PhenylNOH by either NADH,
NADPH or glutathione, however NADPH appears to be the
preferred co-factor. The resulting PhenylNOH can then
participate in another cycle; it has been estimated that each
molecule of PhenylNOH can produce up to 50 molecules of
MetHb in this manner (Kramer et al., 1972). MetHb formed
naturally in the erythrocyte can be reduced back to
hemoglobin by two endogenous enzyme systems, one
using NADH and the other using NADPH for the reduction
(Coleman & Coleman, 1996; Mansouri, 1985; Skold et al.,
2011). The same cofactors which convert MetHb to
hemoglobin can thus convert PhenylNO to PhenylNOH,
therefore an equilibrium is set up with increasing amounts of
PhenylNOH leading to higher concentrations of MetHb. This
same increase of MetHb increases its own conversion back
to hemoglobin, therefore there is never 100% formation of
MetHb. This is likely to be the mechanism responsible for
MetHb formation from BenzNOH (Figure 7).
Benzocaine hydroxylamine
Nitrosobenzocaine
N-Acetyl
transferase
Esterase
O
O
O
OH
O
HN
HN
OH
O
O-Acetyl benzocaine
hydroxylamine
4-Hydroxylamino
benzoic acid
Figure 7. Benzocaine metabolism.
There are two metabolites of lidocaine that might be
responsible for the production of MetHb, N-hydroxyxylidine
and 4-hydroxyxylidine (Figure 8) (Nelson et al., 1978). Both
of these metabolites are formed from the actions of
cytochrome P450 on xylidine, which is itself formed from
lidocaine by hydrolysis (Parker et al., 1996). In addition, 4-
hydroxyxylidine can be produced from N-hydroxyxylidine by
CYP 2E1 in a reaction that does not appear to require
NADPH (Gan et al., 2001). N-Hydroxyxylidine probably
produces MetHb through a mechanism similar to that of
BenzNOH, with the formation of nitrosoxylidine from
N-hydroxyxylidine, while 4-hydroxyxylidine would follow a
related pathway where the aminophenol cycles with the
quinone imine form (Keise, 1966). In this investigation, while
N-hydroxyxylidine produces a rapid initial increase in

DOI: 10.1080/00498254.2016.1196402
Methemoglobin from benzocaine and lidocaine
7
Figure 8. Lidocaine metabolism.
MetHb, the MetHb level decreases again after only 30 min
when S9 is not present. This observation may indicate that the
back reaction of nitrosoxylidine to N-hydroxyxylidine is
either not efficient or is not energetically favorable, resulting
in limited cycling. The increased activity of N-hydroxyx-
ylidine in the presence of S9 suggests that it may be converted
to another species, probably 4-hydroxyxylidine. The activity
of 4-hydroxyxylidine in the presence of whole blood was
roughly the same in the presence or absence of S9, suggesting
that no further metabolism is necessary to form MetHb and
that S9 does not significantly metabolize 4-hydroxyxylidine
further. 4-Hydroxyxylidine glucuronide is a known metabolite
of lidocaine (Keenaghan & Boyes, 1972), however no
cofactors for glucuronidation were present in the incubation.
While this investigation focused mostly on the identifica-
tion of MetHb-forming metabolites of local anesthetics using
synthetically produced metabolites, there is some evidence of
other metabolites not featured in this study. In early studies
other peaks consistent with hydroxylated benzocaine deriva-
tives could be detected mass spectrometrically in S9 studies
with benzocaine (data not shown). The amount of these
metabolites present was too small to allow further character-
ization. In light of this, it is possible that other benzocaine
metabolites could form MetHb, however BenzNOH is the
most likely candidate from a mechanistic approach. In all of
these studies, it would be beneficial to get at least mass
spectral identification and quantification of the metabolites
involved, however it should be remembered that some of these
species are unstable and may be bound to macromolecules in
the incubation medium. Not all incubation conditions were
investigated either; the presence of the NADPH generating
system present to support the S9 incubations could have
affected the reducing equivalents present in the erythrocytes.
It should also be mentioned that while these studies were
concerned with hepatic metabolism these anesthetics are
being applied to skin and mucous membranes which can lead
to different metabolic profiles (D’Alvise et al., 2013; Nathan
et al., 1990), however in the case of administration to the
oropharynx it can be expected that a significant amount of the
dose is swallowed and therefore absorbed from the intestines.
When compared at the same concentrations, BenzNOH
causes 5.5–14.2 times the formation of MetHb as
4-hydroxyxylidine, with the lower, and probably more
clinically relevant, concentrations showing the greatest
effect. Clinical plasma levels for benzocaine are rarely
determined, however a level of 31 mM was recorded in a
patient treated with benzocaine who had MetHba (Spiller
et al., 2014). This same patient had a lidocaine level of
1.8 mM; typical plasma levels for patients receiving lidocaine
for pharyngeal anesthesia are less than 10 mM (Xue et al.,
2009). Thus it seems that benzocaine plasma levels are likely
to be higher than lidocaine plasma levels for this indication.
Lidocaine also requires several metabolic steps to be
converted to 4-hydroxyxylidine while BenzNOH can be
formed directly from benzocaine, suggesting that the active
species for lidocaine should be present at a level substantially
less than that for benzocaine. All of these factors taken
together strongly suggest that regarding pharyngeal anesthe-
sia, MetHba should be considerably more frequent with
benzocaine administration than with lidocaine.
Acknowledgements
The authors would like to thank Dr. Sarah Rogstad for
helpful discussions on the mass spectral analysis of
BenzNOH.
Declaration of interest
This investigation was supported by a CDER Directors Grant.
The authors report no declarations of interest.

8
N. Hartman et al.
Xenobiotica, Early Online: 1–8
experience from 28 478 transesophageal echocardiograms: incidence,
outcomes, and predisposing factors. Arch Intern Med 167:1977–82.
Keenaghan JB, Boyes RN. (1972). The tissue distribution, metabolism
and excretion of lidocaine in rats, guinea pigs, dogs and man.
J Pharmacol Exp Ther 180:459–63.
Keise M. (1966). The biochemical production of ferrihemoglobin
forming derivatives from aromatic amines, and mechanisms of
ferrihemoglobin formation. Pharmacol Rev 18:1091–161.
Kraeling MEK, Lipicky RJ, Bronaugh RL. (1996). Metabolism of benzo-
caine during percutaneous absorption in the hairless guinea
pig: acetylbenzocaine formation and activity. Skin Pharmacol 9:
221–30.
Disclaimer
The findings and conclusions in this article have not been
formally disseminated by the Food and Drug Administration
and should not be construed to represent any Agency
determination or policy.
References
Bradberry SM. (2003). Occupational methaemoglobinaemia.
Mechanisms of production, features, diagnosis and management
including the use of methylene blue. Toxicol Rev 22:13–27.
Chowdhary S, Bukoye B, Bhansali AM, et al (2013). Risk of topical
anesthetic-induced methemoglobinemia: a 10-year retrospective case-
control study. JAMA Int Med 173:771–6.
Kramer PA, Glader BE, Li T-K. (1972). Mechanism of methemoglobin
formation by diphenylsulfones. Effect of 4-amino-40-hydroxyamino-
diphenylsulfone and other P-substituted derivatives. Biochem
Pharmacol 21:1265–74.
Mansouri A. (1985). Review: methemoglobinemia. Am J Med Sci 289:
200–9.
Coleman MD, Coleman NA. (1996). Drug-induced methaemoglobin-
aemia. Treatment issues. Drug Saf 14:394–405.
Coleman MD, Taylor CH. (1997). Bioactivation of benzocaine to a
methaemoglobin-forming metabolite by rat and human microsomes in
vitro. Environ Toxicol Pharmacol 3:47–52.
Meinertz JR, Gingerich WH, Allen JL. (1991). Metabolism and
elimination of benzocaine by rainbow trout, Oncorhynchus mykiss.
Xenobiotica 21:525–33.
Nathan D, Sakr A, Lichtin JL, Bronaugh RL. (1990). In
vitro skin absorption and metabolism of benzoic acid,
p-aminobenzoic acid, and benzocaine in the hairless guinea pig.
Pharm Res 7:1147–51.
Colvin ME, Hatch FT, Felton JS. (1998). Chemical and biological factors
affecting mutagen potency. Mutat Res 400:479–92.
D’Alvise J, Mortensen R, Hansen SH, Janfelt C. (2013). Detection of
follicular transport of lidocaine and metabolism in adipose tissue in
pig ear skin by DESI mass spectrometry imaging. Anal Bioanal Chem
406:3735–42.
Gan J, Skipper PL, Tannenbaum SR. (2001). Oxidation of 2,6-
dimethylaniline by recombinant human cytochrome p450s and
human liver microsomes. Chem Res Toxicol 14:672–7.
Guay J. (2009). Methemoglobinemia related to local anesthetics: a
summary of 242 episodes. Anesth Analg 108:837–45.
Hartman NR, Mao JJ, Zhou H, et al (2014). More methemoglobin is
produced by benzocaine treatment than lidocaine treatment in human
in vitro systems. Regul Toxicol Pharmacol 70:182–8.
Nelson SD, Nelson WL, Trager WF. (1978). N-hydroxyamide metab-
olites of lidocaine. Synthesis, characterization, quantitation, and
mutagenic potential. J Med Chem 21:721–5.
Parker RJ, Collins JM, Strong JM. (1996). Identification of 2,6-xylidine
as a major lidocaine metabolite in human liver slices. Drug Metab
Dispos 24:1167–73.
Shintani R, Fu GC. (2003). The catalytic enantioselective synthesis of
b-lactams: the first intramolecular Kinugasa reactions and the
interception of an intermediate in the reaction cascade. Angew
Chem Int Ed Engl 42:4082–5.
Skold A, Cosco DL, Klein R. (2011). Methemoglobinemia: pathogen-
esis, diagnosis, and management. South Med J 104:757–61.
Spiller HA, Russell JL, Casavant MJ, et al. (2014). Identification of
N-Hydroxy-para-aminobenzoic acid in a cyanotic child after benzo-
caine exposure. Clin Toxicol 52:976–9.
Henrikus BM, Kampffmeyer HG. (1992). Ester hydrolysis and conju-
gation reactions in intact skin and skin homogenate, and by liver
esterase of rabbits. Xenobiotica 22:1357–66.
IMS Health. Inpatient Healthcare Utilization System (IHCARUS). Years
2008-2012. Data obtained June. 2013.
Kane GC, Hoehn SM, Behrenbeck TR, Mulvagh SL. (2007).
Benzocaine-induced methemoglobinemia based on the Mayo Clinic
Xue FS, Liu HP, He N, et al. (2009). Spray-as-you-go airway
topical anesthesia in patients with a difficult airway: a randomized,
double-blind comparison of 2% and 4% lidocaine. Anesth Analg 108:
536–43.