a b s t r a c t

Background: Delirium tremens (DT) is the severest form of alcohol withdrawal syndrome, frequently after alcohol withdrawal seizures. Delirium tremens occurs in a small proportion of patients with alcohol withdrawal seizures; nevertheless, early identification of high-risk patients is important for intensive preventive management of un- expected episodes due to agitation and its associated increased mortality. However, there are limited studies on clinical predictors of the development of DT in patients with alcohol withdrawal seizures.

Methods: Patients who visited the emergency department with acute seizures were included in the study when alcohol withdrawal was the only or the strongest precipitating factor for seizures. All patients were carefully ob- served for at least 48 hours in the intensive care unit after the initial assessment to closely monitor vital signs and development of DT. Clinical and laboratory findings were analyzed for predicting the development of DT. Results: Of the 97 patients (82 males; mean age, 48.6 +- 13.3 years) with alcohol withdrawal seizures, 34 (35.1%) developed DT. low platelet count, high blood level of homocysteine, and low blood level of pyridoxine were as- sociated with the subsequent development of DT. Low platelet count and high blood level of homocysteine were independent risk factors with high diagnostic sensitivity and specificity for the development of DT. Conclusions: The study indicated that some easily determined parameters are potential clinical predictors for the development of DT in patients with alcohol withdrawal seizures. These findings would be helpful in clinical iden- tification and management patients at high risk for DT.

(C) 2015

Introduction

Alcohol is associated with seizures and epilepsy in several ways. Alcohol is one of the most common seizure-precipitating factors in epi- lepsy patients [1], and abrupt cessation of prolonged alcohol use can re- sult in alcohol withdrawal seizures [2]. Seizures in Alcohol-dependent patients may also be due to some diseases commonly associated with alcoholism, such as cerebral trauma, infections, and stroke. Studies showed a strong and consistent association between alcohol consump- tion and epilepsy and unprovoked seizures as well as a dose-response relationship between the level of alcohol consumption and the risk of epilepsy [3]. Clinically, alcohol abuse is a common comorbidity in

? This work was supported by Konkuk University.

* Corresponding author at: Department of Neurology, Dongguk University Ilsan Hospi- tal, 814 Siksa-dong, Ilsan dong-gu, Goyang-si, Gyeonggi-do, 411-773, Korea. Tel.: +82 31

961 7215; fax: +82 31 961 7977.

E-mail addresses: [email protected] (D.W. Kim), [email protected] (H.K. Kim), [email protected] (E.-K. Bae), [email protected] (S.-H. Park), [email protected] (K.K. Kim).

patients with seizures or epilepsy, including first seizures, recurrent seizures in patients with known epilepsy, and even status epilepticus [4].

alcohol withdrawal syndrome consists of early withdrawal symp- toms of tremor; irritability; seizures and agitation; and delayed Withdrawal symptoms, such as disorientation and delirium with auto- nomic hyperactivity. Delirium tremens (DT) is the severest form of alco- hol withdrawal syndrome, which is characterized by a fluctuating disturbance of consciousness and a change in cognition that develops over a short period and tends to fluctuate throughout the day. Delirium tremens occurs in a small proportion of patients with alcohol withdraw- al syndrome; nevertheless, early identification of high-risk patients is important for clinical management by appropriate detection and prompt treatment and prevention of unexpected accident due to agita- tion [5]. In addition, delayed identification of DT is potentially associated with serious morbidity and mortality resulting from hyperthermia, cardiac arrhythmia, complications of seizures, or concomitant medical disorders [6].

In clinical practice, it is often difficult to evaluate the risk of DT in pa- tients with alcohol withdrawal seizures because they frequently present in a sedated, postictal, or poor general condition. However, there are

http://dx.doi.org/10.1016/j.ajem.2015.02.030

0735-6757/(C) 2015

702 D.W. Kim et al. / American Journal of Emergency Medicine 33 (2015) 701-704

only limited studies on clinical predictors for DT development, in patients with alcohol withdrawal seizures. We identified clinical or laboratory predictors of the development of DT in patients with alcohol withdrawal seizures.

Number of Total Enrollment (n = 97)

Number of Excluded (n = 16)

Other causes of acute symptomatic seizures (n = 7) IV therapy before blood sample collection (n = 4) Insufficient duration for hospitalization (n = 3)

Other concomitant medical disease (n = 2)

Total Patients with Alcohol Withdrawal Seizures (n = 113)

Number of Patients without DT (n = 63)

Number of Patients with DT (n = 34)

Methods

Patients

Consecutive patients who visited the emergency department at 4 tertiary referral hospitals for acute seizures from January 2013 to Octo- ber 2014 were retrospectively included in the study on physician as- sessment of alcohol withdrawal as the only or strongest precipitating factor for seizures. Patients were not included when they had other causes of acute symptomatic seizures such as electrolyte imbalance, hy- poglycemia, hyperglycemia, traumatic brain injury, and Ischemic or hemorrhagic stroke. After the initial assessment, all patients received intravenous fluids with supplemental thiamine and multivitamins as well as as-needed doses of benzodiazepines for Rescue therapy of the seizures. Patients were carefully observed for at least 48 hours in the in- tensive care unit to closely monitor their vital signs and DT develop- ment. Diagnosis of DT was made by the attending physician with the Diagnostic and Statistical Manual of Mental disorders, Fourth Edition, criteria, which includes disturbance of consciousness, change in cogni- tion or perceptual disturbance developing in a short period, and the emergence of symptoms during or shortly after withdrawal from heavy Alcohol intake. Patients diagnosed with DT were treated with scheduled and as-needed doses of benzodiazepines according to the recommendation regimen [7-9]. Patients were excluded from the analysis when they were hospitalized for an insufficient duration for the observation, were previously diagnosed with epilepsy, or had a history of DT. We also excluded patients with chronic renal disease or concomitant treatment with folate or vitamin supplements. The local ethics committees approved the study, and all participants or caregivers gave informed consent.

Laboratory tests and statistical analysis

Venous blood samples were collected as soon as the patients arrived at the emergency department before treatment initiation except for fluid therapy with normal saline and the intravenous injection of benzo- diazepine as rescue therapy of seizures. Laboratory tests included com- plete blood count, liver function test, renal function test, low-density lipoprotein cholesterol, homocysteine, folate, vitamin B12, and pyridox- ine as a form of pyridoxal 5?-phosphate (PLP).

Patients were divided into DT and non-DT groups. We compared the clinical characteristics and Laboratory test results between the 2 groups using the Student t test for continuous variables and the ?² test for cat- egorical variables. Variables with a P value b .1 were included in a mul- tiple logistic regression model to identify the independent predictors of DT. In the multiple logistic regression analysis, backward elimination was used to select variables maintained in the final model. Parametric receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to measure the effectiveness of clinical parame- ters in predicting DT. Correlation between the results of the laboratory tests was analyzed by the Pearson correlation coefficient. All the analy- ses were conducted with SPSS version 21.0 (SPSS Inc, Chicago, IL). A P value b .05 was considered statistically significant.

Results

Of the 113 patients with alcohol withdrawal seizures, 16 were ex- cluded for various reasons, and a total of 97 patients (82 males, 15 fe- males) were included in the study (Fig. 1). The mean age of patients at presentation was 48.6 +- 13.3 years (range, 26-84 years). Twenty- one patients had a previous history of seizures, and 20 patients had

Fig. 1. Algorithm for enrollment of patients with alcohol withdrawal seizures; IV, intravenous.

potentially epileptogenic lesions on neuroimaging such as old stroke and trauma. The proportion of patients who developed DT was 35.1% (34/97), whereas 65.9 % of patients did not develop DT. There was no correlation between the results of laboratory tests including homocys- teine, folate, vitamin B12, and pyridoxine. Low platelet count (P b

.001), high blood level of homocysteine (P b .001), and low blood level of pyridoxine (P = .011) were associated with the development of DT (Table 1). Multiple logistic regression analysis showed that hyperhomo- cysteinemia and thrombocytopenia were independent risk factors for the development of DT (Table 2). Both hyperhomocysteinemia and

Table 1

Comparisons of clinical and laboratory factors for the development of DT in patients with alcohol withdrawal seizures

Comparisons of clinical and laboratory results

Factors DT group (n = 34) Non-DT group (n = 63) P

Age	49.5 +- 12.6	48.1 +- 13.7	.621
Sex (male: female)	32:2	50:13	.077
History of seizures (n = 21)	8	13	.799
Structural lesion (n = 20)	7	13	1.000
WBC (x103/uL)	8.21 +- 3.34	8.55 +- 3.87	.670
Hemoglobin level (g/dL)	12.8 +- 2.19	13.5 +- 2.35	.178
Platelet (x103/uL)	124.8 +- 71.0	193.5 +- 91.4	b.001
Total bilirubin (mg/dL)	1.58 +- 1.3	1.40 +- 1.72	.610
AST (IU/L)	198.7 +- 164.2	538.4 +- 335.8	.564
ALT (IU/L)	63.4 +- 55.0	106.6 +- 42.8	.566
Total cholesterol (mg/dL)	194.2 +- 55.0	174.2 +- 48.3	.069
LDL cholesterol (mg/dL)	92.3 +- 39.8	78.7 +- 37.8	.155
Albumin (g/dL)	3.86 +- 0.68	3.85 +- 0.59	.968
Sodium (mmol/L)	136.3 +- 4.30	136.8 +- 5.71	.588
Potassium (mmol/L)	3.93 +- 0.80	3.77 +- 0.55	.263
Chloride (mmol/L)	98.2 +- 5.75	101.1 +- 7.48	.058
BUN (mg/dL)	14.1 +- 10.3	14.5 +- 13.0	.872
Creatinine (mg/dL)	0.95 +- 0.44	0.85 +- 0.29	.206
Ammonia (ug/dL)	126.7 +- 87.7	94.7 +- 94.2	.138
Lactic acid (mmol/L)	14.7 +- 31.5	6.98 +- 5.71	.078
Homocysteine (umol/L)	23.4 +- 13.3	13.4 +- 5.41	b.001
Vitamin B12 (pg/ml)	970.5 +- 450.0	947.6 +- 590.8	.862
Folate (ng/ml)	10.0 +- 5.76	8.83 +- 3.86	.265
PLP (ug/L)	7.46 +- 5.32	13.0 +- 8.68	.011

Abbreviations: WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; LDL, low-density lipoprotein; BUN, serum urea nitrogen.

D.W. Kim et al. / American Journal of Emergency Medicine 33 (2015) 701-704 703

Table 2

Multivariate analysis for the independent risk factors of DT development in patients with alcohol withdrawal seizures

Multivariate analysis by logistic regression

	Factors	Odds ratio	95% CI	P
	Platelet (x103/uL)	0.987	0.979-0.994	.0006
	Homocysteine (umol/L)	1.235	1.098-1.389	.0005

Abbreviation: CI, confidence interval.

thrombocytopenia showed significant classification accuracy in ROC curves (for homocysteine, AUC: 0.788; SE: 0.058; P = .0001; 95% confi- dence interval: 0.681-0.872; for platelet, AUC: 0.716; SE: 0.052; P =

.0001; 95% confidence interval: 0.615-0.804), and a comparison of the area under the ROC curve did not show statistical difference of the 2 pa- rameters (P = .500, Fig. 2). The diagnostic sensitivity and specificity of homocysteine were 81.48% and 67.31%, respectively, at a cutoff value of 14.5 mg/dL. The diagnostic sensitivity and specificity of platelet count were 72.72% and 69.84%, respectively, at a cutoff point of 137 K.

Discussion

As alcohol withdrawal seizures and DT are considered as Severe forms of alcohol withdrawal syndrome, several previous studies evalu- ated the predictors of alcohol withdrawal seizures and/or DT in patients with alcohol dependence. Several markers were found to be associated with the subsequent development of alcohol withdrawal seizures and/ or DT, including low platelet count, low blood level of potassium, and high blood level of homocysteine [10-12]. Our study also showed that low platelet count and high blood level of homocysteine were asso- ciated with the development of DT. The results suggested that pa- tients with thrombocytopenia and hyperhomocysteinemia require intensive treatment for prevention and management of DT because

Fig. 2. Receiver operating characteristic curve for blood homocysteine and platelet count in patients with and without DT (for homocysteine, AUC: 0.788; SE: 0.058; P = .0001; 95% confidence interval: 0.681-0.872; for platelet, AUC: 0.716; SE: 0.052; P = .0001; 95% confidence interval: 0.615-0.804). A comparison of the area under the ROC curve did not show statistical difference of the 2 parameters (P = .500).

delayed identification of DT is potentially associated with unexpected accident and increased mortality.

Homocysteine is metabolized by 2 major pathways. The remethylation process is primarily responsible for the basal homocysteine levels and is known as the folate and vitamin B12-dependent metabolism. Homocyste- ine is also metabolized by transsulfuration, in the presence of higher blood methionine levels as well as when cysteine synthesis is required. Pyrido- xine acts as a cofactor in this metabolic process [13]. In alcoholics, there is a strong correlation between blood homocysteine levels and alcohol concentrations, with homocysteine levels steadily decreasing during alco- hol withdrawal. Hyperhomocysteinemia in alcoholics is associated with an impaired homocysteine metabolism, due to deranged remethylation that is partly caused by an alcohol-induced deficiency of vitamin B12 and folate [14]. However, we found no significant difference in blood levels of vitamin B12 and folate between the DT and non-DT groups; and the blood level of homocysteine was not correlated with the blood levels of vitamin B12, folate, and pyridoxine. The exact mecha- nisms of hyperhomocysteinemia causing seizures and DT are still unknown, but a suggested mechanism involves chronic alcohol consumption-induced up-regulation of the N-methyl-D-aspartate re- ceptor system. Withdrawal of alcohol may lead to an overstimulation of N-methyl-D-aspartate receptors by excitatory amino acids such as glutamate and homocysteine [15].

Pyridoxine currently refers to a collective of 6 biologically intercon- vertible compounds. Among them, PLP is the primary biologically active compound of pyridoxine and acts as a coenzyme in the catabolism of ho- mocysteine. Low PLP status is associated with an increased risk of stroke and transient ischemic attack, irrespective of homocysteine levels [16]. The most outstanding symptoms caused by pyridoxine deficiency are re- lated to the central nervous system and include hyperacousis, hyperirrita- bility, impaired alertness, and seizures, which share many common features with DT. Pyridoxine has long been suggested as a therapeutic tar- get in DT [17]. Our study confirmed that pyridoxine deficiency might be associated with the development of DT. However, low blood level of pyridoxine was not an independent risk factor for DT; this might be due to high prevalence of pyridoxine deficiency found in patients with hyper- homocysteinemia, although we were unable to detect a correlation between the serum levels of homocysteine and pyridoxine. Interestingly, agitation is one of the common adverse events that occur with levetira- cetam treatment reportedly caused by pyridoxine deficiency [18,19]. Considering the link between homocysteine and pyridoxine metabolisms and their association in the development of DT, they might have an additive effect in the development of DT.

We found that low platelet count was an independent risk factor for the development of DT in patients with alcohol withdrawal seizures. The mechanisms for this finding are unknown. One possibility is that thrombocytopenia and the development of DT are parallel phenomena, that is, they both reflect effects of long-term and heavy alcohol con- sumption. Excessive alcohol consumption reportedly associated with thrombocytopenia, due to possible bone marrow toxicity resulting in reduced platelet production [10,11].

Our study has several limitations. The first limitation is that only a small number of patients were included. The correlation of several markers with the development of DT could be a matter of chance. Neglecting to measure the clinical markers for DT, including the initial vital signs and blood thiamine level, was also another limitation. In ad- dition, because the diagnosis of DT was solely made by attending physi- cians who were not blinded to the clinical or laboratory findings, it is possible that the classification of patients with and without DT could be biased by the clinical information. Finally, the enrolled patients are not representative of general patients with alcohol withdrawal seizures because all patients were recruited from the emergency departments of tertiary Referral hospitals. Despite these limitations, the present study has clinical implications because it demonstrated several circulatory markers that are involved in the development of DT in patients with al- cohol withdrawal seizures. Furthermore, the data suggested that active

704 D.W. Kim et al. / American Journal of Emergency Medicine 33 (2015) 701-704

administration of vitamin B12, folate, and pyridoxine to lower blood homocysteine level should be considered for the prevention and management of DT, whereas current management recommendations for DT are focused on the rapid and adequate control of agitation with sedative hypnotic drugs [8,9].

References

1. Nakken KO, Solaas MH, Kjeldsen MJ, Friis ML, Pellock JM, Corey LA. Which seizure- precipitating factors do patients with epilepsy most frequently report? Epilepsy Behav 2005;6:85-9.
2. Hillbom M, Pieninkeroinen I, Leone M. Seizures in alcohol-dependent patients: epidemiology, pathophysiology and management. CNS Drugs 2003;17:1013-30.
3. Samokhvalov AV, Irving H, Mohapatra S, Rehm J. Alcohol consumption, unprovoked sei- zures, and epilepsy: a systematic review and meta-analysis. Epilepsia 2010;51:1177-84.
4. Hansen CK, Fisher J, Joyce N, Edlow JA. Emergency department consultations for patients with neurological emergencies. Eur J Neurol 2011;18:1317-22.
5. Turner RC, Lichstein PR, Peden Jr JG, Busher JT, Waivers LE. Alcohol withdrawal syndromes: a review of pathophysiology, clinical presentation, and treatment. J Gen Intern Med 1989;4:432-44.
6. Khan A, Levy P, DeHorn S, Miller W, Compton S. Predictors of mortality in patients with delirium tremens. Acad Emerg Med 2008;15:788-90.
7. Jaeger TM, Lohr RH, Pankratz VS. Symptom-triggered therapy for alcohol withdrawal syndrome in medical inpatients. Mayo Clin Proc 2001;76:695-701.
8. Mayo-Smith MF, Beecher LH, Fischer TL, Gorelick DA, Guillaume JL, Hill A, et al. Management of alcohol withdrawal delirium. An evidence-based practice guideline. Arch Intern Med 2004;164:1405-12.
9. Schuckit MA. Recognition and management of withdrawal delirium (delirium tremens). N Engl J Med 2014;371:2109-13.
10. Berggren U, Fahlke C, Berglund KJ, Blennow K, Zetterberg H, Balldin J. Thrombocyto- penia in early alcohol withdrawal is associated with development of delirium tremens or seizures. Alcohol Alcohol 2009;44:382-6.
11. Latvala J, Parkkila S, Niemela O. Excess alcohol consumption is common in patients with cytopenia: studies in blood and bone marrow cells. Alcohol Clin Exp Res 2004;28:619-24.
12. Eyer F, Schuster T, Felgenhauer N, Pfab R, Strubel T, Saugel B, et al. Risk assessment of moderate to severe alcohol withdrawal-predictors for seizures and delirium tremens in the course of withdrawal. Alcohol Alcohol 2011;46:427-33.
13. Siniscalchi A, Mancuso F, Gallelli L, Ferreri Ibbadu G, Biagio Mercuri N, De Sarro G. Increase in plasma homocysteine levels induced by drug treatments in neurologic patients. Pharmacol Res 2005;52:367-75.
14. Bleich S, Carl M, Bayerlein K, Reulbach U, Biermann T, Hillemacher T, et al. Evidence of increased homocysteine levels in alcoholism: the Franconian alcoholism research studies (FARS). Alcohol Clin Exp Res 2005;29:334-6.
15. Bleich S, Bayerlein K, Hillemacher T, Degner D, Kornhuber J, Frieling H. An assess- ment of the potential value of elevated homocysteine in predicting alcohol- withdrawal seizures. Epilepsia 2006;47:934-8.
16. Kelly PJ, Shih VE, Kistler JP, Barron M, Lee H, Mandell R, et al. Low Vitamin B6 but not homocyst(e)ine is associated with increased risk of stroke and transient ischemic attack in the era of folic acid grain fortification. Stroke 2003;34:e51-4.
17. Palmer EJ. Pyridoxine hydrochloride in the treatment of acute alcoholism and delirium tremens. Va Med Mon (1918) 1955;82:15-6.
18. Major P, Greenberg E, Khan A, Thiele EA. Pyridoxine supplementation for the treat- ment of levetiracetam-induced behavior side effects in children: preliminary results. Epilepsy Behav 2008;13:557-9.
19. Davis GP, McCarthy JT, Magill DB, Coffey B. Behavioral effects of levetiracetam mitigated by pyridoxine. J Child Adolesc Psychopharmacol 2009;19:209-11.