American Journal of Therapeutics 21, e106–e109 (2014)

Severe Propylene Glycol Toxicity Secondary

to Use of Anti-Epileptics

Unnikrishnan Pillai, MD,* Jatinder C. Hothi, MD, and Zeenat Y. Bhat, MD

Propylene glycol toxicity presenting as high anion gap metabolic acidosis and osmolar gap has been
extensively reported in literature, and most of them are secondary to intravenous lorazepam
infusion. However, propylene glycol is used as a solvent in a number of medications that are
frequently utilized in critical care setting, and hence one should be aware that the toxicity is possible
from a variety of medication. Phenobarbital and phenytoin are one of those, and we hereby report
a novel case of propylene glycol toxicity secondary to phenobarbital and phenytoin infusion in
a patient with refractory status epilepticus. Furthermore, our patient had end-stage renal disease,
which we think could have been an important precipitating factor for the toxicity. Because most of
the symptoms from propylene glycol toxicity can mimic sepsis—which is very common in critical
care unit patients—this life threatening scenario could be easily missed. Regular monitoring of
osmolar gap is an easily available intervention in the at risk patients.

Keywords: propylene glycol, phenytoin, phenobarbiturate

CASE PRESENTATION medications—intravenous lacasamide and propofol.

Our patient was a 57-year-old female with end-stage
renal disease (ESRD) with past medical history of
hypertension, seizure disorder, multiple strokes,
diastolic heart failure, and chronic hepatitis C was
admitted for elective right cranioplasty. Her past
surgical history was relevant for craniotomy (for
evacuation of subdural hematoma) 3 months before
the present admission. Her home medications included
phenytoin, levetiracetam, amlodipine, hydralazine,
lisinopril, and minoxidil.
The cranioplasty procedure was complicated by
large extra dural hemorrhages, which required multi-
ple blood transfusions and removal of bone flap. The
patient developed generalized tonic clonic seizures
after the procedure, which did not respond to the
first-line antiepileptic medications—phenytoin and
levetiracetam, and subsequently even second-line
Division of Nephrology, Department of Internal Medicine, Wayne
State University, Detroit, MI.
The authors have no conflicts of interest to disclose.
*Address for correspondence: Fellow, Division of Nephrology,
Department of Internal Medicine, Wayne State University,
Detroit, MI 48201. E-mail: unnikrishnanpk@yahoo.com
1075-2765 Ó 2012 Lippincott Williams & Wilkins
Phenobarbitol infusion was thus started to induce
barbiturate coma. The patient was maintained on
her regular dialysis schedule (thrice a week; with
3.5 hours for each session). On the eighth day of
phenobarbitol infusion, the patient went into shock
and was thus started on vasopressors. The laborotory
analyses were suggestive of a high anion gap
metabolic acidosis (anion gap of 22) due to lactic
acidosis—12 gm/dL (reference range: 0.6–2.4 g/dL)
(Table 1). The serum albumin was 3.5 g/dL (3.5–
5.0 g/dL). An extensive evaluation to ascertain the
cause of lactic acidosis, including computerized
tomographic scan of the abdomen, blood cultures,
and toxic alcohol screen were negative.1 Because the
seizures had subsided after phenobarbitol infusion,
seizures as the cause of lactic acidosis were ruled out.
Finally after extensive searching, the cause of high
lactic acid was attributed to propylene glycol, which
we noted, is a constituent of phenobarbital. The
infusion was immediately stopped, and the patient
was initiated on continuous renal replacement ther-
apy (CRRT). Patient responded by resolution of lactic
acidosis and normalization of her anion gap.
Unfortunately, the patient had recurrence of seizure
activity due to discontinuation of phenobarbital. Pheno-
barbital was restarted at a lower dose along with
www.americantherapeutics.com
Propylene Glycol Toxicity e107

intravenous phenytoin and valproic acid. Four days

(off CRRT)
Day 15

3.7

1.6

1.5
after restarting phenobarbital and intravenous phe-

140

105

147
24
20

11
nytoin, the patient redeveloped severe high anion
gap lactic acidosis with lactic acid of 15 g/dL (0.6–
2.4 g/dL) and hypotension (Table 1). Evaluation
showed high free valproic acid level of 19.1 mg/mL
phenobarbiturate,
Day 13 (stopped

restarted CRRT)
(normal reference range: 5–10 mg/mL) along with
elevated liver enzymes [ALT: 311 U/L (normal
4.4

2.0
range: 0–65 U/L) and AST: 331 U/L (normal range:

18

70
23
15
7
130

100 0–37 U/L)]. Phenobarbital and valproic acid were

discontinued, and she was restarted on CRRT in view
of hypotension and lactic acidosis. Her measured
serum osmolality was 315 mOsmol/kg, with calcu-
lated osmolality of 270 mOsmol/kg and osmolar
phenobarbital at
Day 9 (restarted

gap of 45 mOsmol/kg. Unfortunately her serum

lower dose)

osmolarity was measured 1 hour after the initiation

4.0

1.6

0.9

of CRRT, which may have decreased the actual

138

104

126
22
25

12

measured osmolarity due to the clearance of pro-

pylene glycol and its metabolite. One day after
reinitiation of CRRT, the patient became hemody-
namically stable (off vasopressors), her lactic acido-
sis resolved (Table 1).
stopped phenobarbital)
Day 8 (CRRT initiated,

The reason for high anion gap lactic acidosis in our

patient can be primarily attributable to accumulation
of propylene glycol, which is a constituent of both
4.8

6.0

phenobarbital and phenytoin. Administration of

130

100

30

75
22
12
8

higher doses of phenobarbital and phenytoin in the

setting impaired renal function and impaired hepatic
function (chronic hepatitis C and valproic acid
induced liver impairment) further aggravated the
lactic acidosis. The diagnosis of propylene glycol
toxicity should be primarily clinical, with reversible
phenobarbital)
Day 5 (started

lactic acidosis in the setting of culprit drug usage

(Table 2). Our patient got a total of 3900 mg of
4.7

5.0

1.5
143

106
25
30

90
12

intravenous phenobarbital and 8900 mg of intrave-

nous phenytoin in a 7-day period.

DISCUSSION
Day 1

4.8

5.4

0.6
26
28

92
12
138

100

Propylene glycol is a commonly used solvent in

Table 1. Relevant laboratory findings.

intravenous and topical drug preparations (Table 2). It

Serum carbon dioxide, mmol/L

Table 2. Medications containing propylene glycol

as solvent.
Blood urea nitrogen, mg/dL
Serum potassium, mmol/L

Serum creatinine, mg/dL

Serum chloride, mmol/L
Serum sodium, mmol/L

Medications
Laboratory parameters

Serum glucose, mg/dL

Anion gap, mmol/L

Lorazepam
L-lactic acid, g/dL

Phenobarbital
Diazepam
Phenytoin
Trimethoprim–sulfamethoxazole
Etomidate
Nitroglycerin

www.americantherapeutics.com American Journal of Therapeutics (2014) 21(4)

e108
is used as a solvent for relatively insoluble material.
Propylene glycol, also called 1,2-propanediol or pro-
pane-1,2-diol,(molecular weight: 76) is an organic com-
pound with formula C3H8O2 or HO-CH2-CHOH-CH3. It
is a colorless, nearly odorless, clear, viscous liquid with
a faintly sweet taste and miscible with water, acetone,
and chloroform.
If given in large quantities over a short period,
propylene glycol can cause high anion gap lactic
acidosis. Of the medications that contain propylene
glycol as solvent, lorazepam contained the largest
proportion of the compound. Usual strength of
intravenous lorazepam used is 2 mg/mL, and each
milliliter contains 0.8 mL of propylene glycol which
comes to 830 mg.2 Propylene glycol toxicity associated
with lorazepam in fusion has been widely reported in
literature as it agent of choice for sedation in critically
ill patients.3 Other therapeutic agents associated with
propylene glycol toxicity includes etomidate, nitro-
glycerin, and diazepam.4,5
There have been very few reported cases of pheno-
barbiturate infusion with propylene glycol toxicity.6,7
There was no reported case of phenobarbiturate induced
propylene glycol toxicity in ESRD, which will further
complicate due to the fact that patient already had
impaired renal clearance and they may not be able to
handle the propylene glycol in high doses of pheno-
barbiturate. Each milliliter of phenobarbiturate injection
(dosage strength 130 mg of phenobarbiturate per mL)
contains 0.75 mL of propylene glycol. Whereas each
milliliter of phenytoin injection (dosage strength 50 mg
of phenytoin per mL) contains 0.40 mL of propylene
glycol. Our patient got a cumulative dose of 3970 mg
phenobarbiturate and 8870 mg of phenytoin in a 7-day
period, both of which combined contains a total of
96 gm of propylene glycol. Maximum daily recom-
mended intake of propylene glycol by WHO is
25 mg/kg/day.8 Weight of our patient was 50 kg,
so her maximum allowed dose per day is 1.25 gm
(1250 mg), whereas our patient got a total of 13.7 gm
(13,700 mg) per day for 7 days, which is 10 times the
maximum recommended dose. The fact that our patient
had ESRD with no residual renal function along with
chronic hepatitis C definitely has played a role in the
impaired metabolism of propylene glycol.
Exact mechanism of propylene glycol toxicity is not
that clear. In presence of normal hepatic and renal
function, the half-life of propylene glycol is 1.5–
3.3 hours.9 Commonly used propylene glycol is
a 50:50 mixture of D and L isomer. Of the administered
dose, 40% is excreted unchanged in the urine and 60%
undergoes hepatic metabolism to L-lactaldehyde and
D-lactaldehyde, respectively. Further metabolism of
lactaldehyde is by aldehyde dehydrogenase to D and L
American Journal of Therapeutics (2014) 21(4)
Pillai et al
lactic acid. D-lactaldehyde is not a good substrate
for aldehyde dehydrogenase, so in propylene glycol
toxicity, D-lactaldehyde accumulates causing many of
the toxic effects. Further metabolism of D-lactaldehyde
to D-lactic acid is by a slow pathway in liver. Metab-
olism of L-Lactate is much faster than D-lactate
through the gluconeogenesis and glycogen generation.
Whereas D-lactic acid is metabolized by D-2 hydrox-
yacid dehydoxyacid dehydrogenase. In our patient, it
would have been useful to measure the D-Lactic acid
level, but it is a send out test and is not immediately
available, its importance in clinical decision making is
debatable. Because there is a preferential metabolism of
L-isomer of propylene glycol by alcohol and aldehyde
dehydrogenase, several people have postulated that
much of the side effects are due to D-isomeric forms.10
Characteristic feature of propylene glycol toxicity
includes increased osmolality and high anion gap lactic
acidosis. Increased serum Propylene glycol concen-
tration has been reported even within 2 hours of
lorazepam infusion.11 Major clinical features of pro-
pylene glycol toxicity includes progressive renal
failure, cardiac dysfunction, central nervous system
depression, and seizures. Renal toxicity of ethylene
glycol is thought to be direct toxicity to the proximal
tubular cells. Morshed et al12 has shown direct toxic
effect of propylene glycol in cultured human proximal
tubular cells. Toxic effect of propylene glycol
independent of kidney injury has also been reported
in literature.13 CNS symptoms include confusion,
lethargy, drowsiness, coma, and seizure. Other signif-
icant features include refractory hypotension, hypona-
tremia, hemolysis, hemoglobinuria bradyarrythmias,
and multiorgan dysfunction.14–16 Sometime propylene
glycol may mimic sepsis with systemic inflammatory
response. If the patient has pre-existing renal and
hepatic impairment or had to receive large doses
of medications or receives multiple medications con-
taining propylene glycol, chances of developing
toxicity are high.
Our patient got very high doses of intravenous
phenobarbital along with phenytoin, both of which
contain propylene glycol, and she had impaired
hepatic and renal clearance because of her chronic
hepatitis C and ESRD. Her hepatic function was further
deranged by her valproic acid induced hepatic
dysfunction. Also alcohol abuse predisposes patient
to propylene glycol toxicity as both of them have the
same pathway for metabolism.
There is no clear data about the correlation of
propylene glycol level and toxicity. So the regular
surveillance of serum propylene glycol level and its
correlation to toxicity is unclear. Most reported cases
toxicity was seen at levels more than 100 mg/dL.17 In
www.americantherapeutics.com
Propylene Glycol Toxicity
most hospitals propylene glycol levels are not immedi-
ately available to make clinical decisions, but osmolar
gap can be used as a surrogate marker for toxicity.18
Arroliga et al11 showed that there is significant cor-
relation between serum propylene glycol level and
osmolarity. Critically ill patients receiving medica-
tions containing propylene glycol regular monitoring
of osmolar gap should be routine particularly in the
setting of impaired renal clearance. If an explained
osmolar gap of more than 10–15 develops, we should
consider stopping the culprit medication even before
a propylene glycol levels are available as there may be
a delay in getting the serum levels in most hospitals.
Because propylene glycol is a low molecular weight
substance with minimal protein binding, it can be
dialysed very effectively. If the patient is on regular
intermittent hemodialysis, we should definitely
consider intensifying dialysis regimen to daily or
increase the duration of dialysis. Our patient was
started on CRRT. We also did not consider using
fomipezole an inhibitor of alcohol dehydrogenase
as the patient was already on regular intermittent
hemodialysis, and there is no clear evidence in
literature regarding the use of this agent in propylene
glycol toxicity.
CONCLUSIONS
Barbiturate coma is a widely used treatment in
neurocritical care units in management of refractory
seizures. Physicians should be cognizant about the fact
that patients on very high doses of phenobarbiturate,
particularly those with impaired clearance like ESRD
patients and with hepatic impairment, are prone to
develop high anion gap lactic acidosis along with
osmolar gap. We should have a high index of suspicion
for propylene glycol toxicity as most of the symptoms
mimic sepsis/systemic inflammatory response, which
is very common in critical care unit patients. Regular
monitoring of osmolar gap is an easily available
intervention to recognize this life threatening iatrogenic
complication.
REFERENCES
1. Htyte N, White L, Sandhu G, et al. An extreme and life-
threatening case of recurrent D-lactate encephalopathy.
Nephrol Dial Transplant. 2011;26:1432–1435.
www.americantherapeutics.com
e109
2. Mullins ME, Barnes BJ. Hyperosmolar metabolic acidosis
and intravenous lorazepam. N Engl J Med. 2003;347:
857–858.
3. Jacobi J, Fraser GL, Coursin DB, et al. Clinical practice
guide lines for the sustained use of sedatives and
analgesics in the critically ill adult. Crit Care Med. 2002;
30:119–141.
4. Demey HE, Bossaert LL. Propylene glycol intoxication
and nitroglycerin therapy. Crit Care Med. 1987;15:540.
5. Bedichek E, Kirschbaum B. A case of propylene glycol
toxic reaction associated with etomidate infusion. Arch
Intern Med. 1991;151:2297–2298.
6. Yorgin PD, Theodorou AA, Al-Uzri A, et al. Propylene
glycol-induced proximal renal tubular cell injury. Am
J Kidney Dis. 1997;30:134–139.
7. Bledsoe KA, Kramer AH, Propylene glycol toxicity
complicating use of barbiturate coma. Neurocrit Care.
2008;9:122–124.
8. Cawley MJ. Short-term lorazepam infusion and concern
for propylene glycol toxicity: case report and review.
Pharmacotherapy. 2001;21:1140–1144.
9. Yip L. Other alcohols, glycols, and glycol ethers. In: Dart
RC, et al, eds. Medical Toxicology, 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins;2004:1230–1242.
10. Zar T, Yusufzai I, Sullivan A, et al. Acute kidney injury,
hyperosmolality and metabolic acidosis associated with
lorazepam. Nat Clin Pract Nephrol. 2007;3:515–520.
11. Arroliga AC, Shehab N, McCarthy K, et al. Relationship
of continuous infusion lorazepam to serum propylene
glycol concentration in critically ill adults. Crit Care Med.
2004;32:1709–1714.
12. Morshed KM, Jain SK, McMartin KE. Acute toxicity of
propylene glycol: an assessment using cultured proxi-
mal tubule cells of human origin. Fundam Appl Toxicol.
1994;23:38–43.
13. Neale BW, Mesler EL, Young M, et al. Propylene glycol-
induced lactic acidosis in a patient with normal renal
function: a proposed mechanism and monitoring, recom-
mendations. Ann Pharmacother. 2005;39:1732–1735.
14. Demey HE, Bossaert LL. Propylene glycol intoxication
and nitroglycerin therapy. Crit Care Med. 1987;15:540.
15. Wilson KC, Reardon C, Farber HW. Propylene glycol
toxicity. N Engl J Med. 2000;343:815.
16. Zar T, Graeber C, Perazella MA. Recognition, treatment,
and prevention of propylene glycol toxicity. Semin Dial.
2007;20:217–219.
17. Wilson KC, Reardon C, Theodore AC, et al. Propylene-
glycol toxicity: a severe iatrogenic illness in ICU patients
receiving IV benzodiazepines: a case series an pro-
spective, observational pilot study. Chest. 2005;128:
1674–1681.
18. Yaucher NE, Fish JT, Smith HW, et al. Propylene glycol
associated renal toxicity from lorazepam infusion.
Pharmacotherapy. 2003;23:1094–1099.
American Journal of Therapeutics (2014) 21(4)