Baclofen Neurotoxicity in Uremic Patients: Is Continuous Ambulatory Peritoneal Dialysis Less Effective Than Intermittent Hemodialysis?

Abstract

Objectives. To compare the efficacy between continuous ambulatory peritoneal dialysis (CAPD) and intermittent hemodialysis in the treatment of baclofen-associated neurotoxicity. Methods. Three uremic patients suffering baclofen-associated neurotoxicity were treated by CAPD at our hospital. We obtain 12 cases with detailed clinical course of baclofen-associated neurotoxicity treated by hemodialysis from a literature review using the Medline and Science Citation Index, six of these patients were treated by early hemodialysis which was defined as hemodialysis intervention within 48 h of the onset of clinical toxicity. Results. Our cases regain full consciousness within 2–3 days after the onset of neurotoxicity. Clinical characteristics of our cases including age, dialysis time, preexisting central nervous system (CNS) lesion, concomitant use of CNS depressant, total baclofen dose, onset of neurotoxicity, and duration of neurotoxicity are not significantly different from reported cases that treated by either early or routine hemodialysis. Conclusions. Baclofen should be avoided in uremic patients. When consciousness disturbance occurs in uremic patients, baclofen-associated neurotoxicity should be included in the list of differential diagnosis. According to experiences of our patients, we conclude that neither early nor routine hemodialysis is more effective than CAPD in shortening the recovery time of baclofen-associated neurotoxicity.

• Baclofen
• Neurotoxicity
• Continuous ambulatory peritoneal dialysis

Introduction

Baclofen, a β-(p-chlorphenyl) derivative of the γ-aminobutyric acid, is the most effective agent for treating spasticity of spinal origin.[[1]] When baclofen is used in therapeutic doses, spinal effects predominate; neurotoxicity becomes more notably evident in patients taking increased doses and in patients with renal failure. Hemodialysis, especially early hemodialysis, has been reported to be effective in alleviating the clinical symptoms and shortening the recovery time for patients with baclofen-associated neurotoxicity.[[2]] Until now, only one case of baclofen-associated neurotoxicity has been reported in end stage renal disease (ESRD) patients undergoing CAPD in the Lit.,[[3]] and CAPD was considered to be a less effectively extracorporeal method for such patients.[[2]], [[3]] We report three cases of CAPD patients suffering baclofen-associated neurotoxicity. Another 12 cases obtained from the lit.[[2]], [[4]], [[5]], [[6]], [[7]], [[8]] with detailed clinical course of baclofen-associated neurotoxicity treated by hemodialysis are reviewed. Comparison of the clinical characteristics between CAPD patients and hemodialysis patients with baclofen-associated neurotoxicity and evaluation the efficacy of CAPD in treating baclofen-associated neurotoxicity were also performed.

Methods

Between June 1993 and March 2001, three uremic patients suffering baclofen-associated neurotoxicity were treated with discontinuation of the drug and performance of routine CAPD program that was previously used for their ESRD. We obtain another 12 cases[[2]], [[4]], [[5]], [[6]], [[7]], [[8]] with detailed clinical course of baclofen-associated neurotoxicity treated by hemodialysis from a literature review using the Medline and Science Citation Index; six of these patients were treated by early hemodialysis. Clinical characteristics of patients treated by CAPD including age, dialysis time, preexisting CNS lesion, concomitant use of CNS depressant, total baclofen dose, onset of neurotoxicity, and duration of neurotoxicity were compared to those treated by hemodialysis. Since the various data sets were not normally distributed, results are expressed as medians (minimum, maximum). Comparisons of numerical data between groups were performed by the Mann–Whitney U test, and of categorical data by a Fisher's exact probability test, when appropriate. Statistical significance was set at the p<0.05 level. All data were entered into a database and analyzed using the Statistical Package for the Social Sciences (SPSS Ver. 10.0; Chicago, IL) for Windows.

Case Report

Case 1

A 50-year-old man with ESRD secondary to diabetic nephropathy who had been undergoing CAPD for 1.5 years was admitted to our hospital suffering severe low back pain and numbness of bilateral low extremities for months. Upon admission, MRI revealed large disk herniation with marked nerve root compression above the L4–L5 level. The patient received laminectomy and foraminotomy for this problem. Owing to a wound pain, baclofen 5 mg four times daily was administered after the operation. He also took triazolam 0.25 mg nightly for insomnia. Meanwhile, this patient pursued his routine CAPD program (four 2L-dialysate exchanges/day).

Four days after initiation of baclofen the man presented with acute confusion and agitation. Neurological examination was unremarkable except for altered consciousness (Glasgow Coma Scale E₃V₃M₅). The man's blood pressure was 144/80 mmHg and his temperature was 36.9°C. Neither arrhythmia nor respiratory distress was found. Laboratory data showed hemoglobin 7.7 g/dL, white blood cell 8900/mm³ with normal differential count, and platelet 238000/mm³. Meanwhile, serum sodium was 144 meq/L, potassium 4.1 meq/L, bicarbonate 23.3 mmol/L, urea 113 mg/dL, creatinine 16.2 mg/dL, glucose 223 mg/dL, and calcium 10.5 mg/dL. Baclofen-associated encephalopathy was considered the most likely explanation for his acute neurological condition, and baclofen was discontinued. Peritoneal dialysis continued to be performed as usual by the patient's family. Two days later, the man regained full consciousness.

Case 2

A 58-year-old man with ESRD secondary to chronic glomerulonephritis who had been undergoing CAPD for three years was sent to the emergency department for consciousness disturbance lasting since the previous day. Blood pressure was 180/92 mmHg, the temperature was 36.2°C, and no arrhythmia or respiratory distress was noted. Besides altered consciousness (Glasgow Coma Scale E₄V₁M₅), the neurological examination was normal. Laboratory datashowed hemoglobin 9.1 g/dL, white blood cell 6100/mm³ with normal differential count, and platelet 169000/mm³. Serum sodium was 141 meq/L, potassium 4.3 meq/L, bicarbonate 25.7 mmol/L, urea 77 mg/dL, creatinine 18.1 mg/dL, glucose 145 mg/dL, and calcium 10.4 mg/dL. Finally, a brain computerized tomography only revealed signs of cortical atrophy.

After admission, the man's family brought in a list of medications that included baclofen 5 mg three times daily, prescribed for arthralgia at an orthological clinic one day before the onset of consciousness disturbance. Accordingly, baclofen-associated encephalopathy was considered to be the most likely etiology for the patient's acute neurological picture, and baclofen was stopped. Meanwhile, the patient pursued his routine CAPD program (four 2L-dialysate exchanges/day). Complete recovery of the man's neurological status occurred three days later.

Case 3

A 76-year-old man with ESRD owing to hypertension who had been undergoing CAPD for three months was sent to the emergency department for consciousness disturbance (Glasgow Coma Scale E₄V₃M₅) lasting since the previous day. Blood pressure was 188/102 mmHg, the temperature was 36.5°C, and no arrhythmia or respiratory distress was noted. Laboratory data showed hemoglobin 12.2 g/dL, white blood cell 8900/mm³ with normal differential count, and platelet 137000/mm³. Serum sodium was 136 meq/L, potassium 4.0 meq/L, urea 71 mg/dL, creatinine 8.3 mg/dL, glucose 98 mg/dL, and calcium 9.3 mg/dL.

After admission, we found that baclofen 10 mg three times daily was given to him by a neurologist for low back pain one day before the onset of consciousness disturbance. We discontinued Baclofen and performed his routine CAPD program (four 2.5L-dialysate exchanges/day) to shorten the recovery time of baclofen-associated neurotoxicity. He regains full consciousness one day later.

Previous three cases are summarized in Table 1.

Table 1. Summary of uremic patients with Baclofen-associated neurotoxicity treated by CAPD

	Case 1	Case 2	Case 3
Age (y)	50	58	76
Sex	Male	Male	Male
Underlying disease of uremia	Diabetic nephropathy	Chronic glomerulonephritis	Hypertension
Dialysis modalitya	CAPD	CAPD	CAPD
Dialysis timeb (y)	1.5	3.0	0.25
CNS lesionc	Nil	Nil	Nil
CNS depressantd	(+)	Nil	(+)
Daily baclofen dose	5 mg qid	5 mg tid	10 mg tid
Total baclofen dose (mg)	85	25	20
Onset of neurotoxicitye (d)	4.5	1.5	1.5
Manifestations of neurotoxicity	Consciousness disturbance agitation	Consciousness disturbance	Consciousness disturbance
Duration of neurotoxicity (d)	2.0	3.0	2.0
Treatment modalityf	CAPD	CAPD	CAPD

CAPD: continuous ambulatory peritoneal dialysis; CNS: central nervous system; y: year; d: day.

^aMaintenance dialysis mode for ESRD.

^bDuration of maintenance dialysis.

^cPreexisting CNS lesion.

^dConcomitant CNS depressant.

^eTime from the initiation of baclofen to the onset of clinical intoxication.

^fDialysis mode during neurotoxicity.

Result

To determine the efficacy of CAPD in the treatment of baclofen-associated neurotoxicity, we compare the clinical parameters, total baclofen dose and the course of neurotoxicity between the CAPD group and the hemodialysis group that includes early and routine hemodialysis. As shown in the Table 2, the age, dialysis time, preexisting CNS lesion, concomitant use of CNS depressant, and total baclofen dose are not significantly different between the two groups. The clinical course of neurotoxicity includes onset and duration neurotoxicity are also not significantly different between the two groups.

Table 2. Comparison of the clinical variables between the CAPD and the hemodialysis groups

Patient	Age (y)	Dialysis modalitya	Treatment dialysis modeb	Dialysis time (y)c	CNS lesiond	CNS depressante	Total baclofen dose (mg)	Onsetf (d)	Neurotoxicityg (d)
Current report (n = 3)
1	50	CAPD	CAPD	1.5	−	+	85	4.5	2
2	58	CAPD	CAPD	3	−	−	25	1.5	3
3	76	CAPD	CAPD	0.25	−	+	20	1.5	2
Literature review (n = 12)
4^2	65	HD	Early HD	2	−	−	40	1	3
5^2	55	HD	Early HD	3	−	−	45	3	3
6^2	39	HD	Early HD	1	+	−	35	2.5	2
7^2	39	HD	Early HD	1.5	−	+	15	0.8	2
8^2	66	HD	Early HD	4	−	−	20	3.3	1
9^2	47	HD	Early HD	16	−	+	20	0.8	4
10^4	57	HD	HD	?	+	−	40	1	4
11^5	29	HD	HD	?	−	−	70	5	1
12^5	48	HD	HD	?	−	−	40	4	1
13^6	64	HD	HD	0.6	−	−	12.5	3	5
14^7	69	HD	HD	1.6	+	−	20	1.5	2
15^8	65	HD	HD	0.5	+	−	20	4	0.25
p*	0.386				0.516	0.154	0.555	0.771	0.882

p*: Comparison between the CAPD group and the hemodialysis (early and routine) group.

CAPD: continuous ambulatory peritoneal dialysis; CNS: central nervous system; HD: hemodialysis; Early HD: hemodialysis intervention within 48 h of the onset of clinical toxicity; d: day; y: year.

^aMaintenance dialysis mode for ESRD.

^bDialysis mode during neurotoxicity.

^cDuration of maintenance dialysis.

^dPreexisting CNS lesion.

^eConcomitant CNS depressant.

^fTime from the initiation of baclofen to the onset of clinical intoxication.

^gDuration of neurotoxicity.

Furthermore, we compare the efficacy of CAPD and early hemodialysis in treating baclofen-associated neurotoxicity. As shown in the Table 3, the age, dialysis time, pre-existing CNS lesion, concomitant use of CNS depressant, total baclofen dose, and the clinical course of neurotoxicity are not significantly different between the two groups.

Table 3. Comparison of the clinical variables between the CAPD and the early hemodialysis groups

Variablesa	CAPD (n = 3)	Early HD (n = 6)	p
Age (y)	58 (50, 76)	51 (39, 66)	0.300
Dialysis timeb (y)	1.5 (0.25, 3.0)	2.5 (1, 16)	0.298
% of CNS lesionc	0	16.7	1.000
% of CNS depressantd	66.7	33.3	0.524
Total baclofen dose (mg)	25 (20, 85)	27.5 (15, 45)	0.599
Onsete(d)	1.5 (1.5, 4.5)	1.7 (0.8, 3.3)	0.435
Neurotoxicityf (d)	2 (2, 3)	2.5 (1, 4)	0.784

CAPD: continuous ambulatory peritoneal dialysis; CNS: central nervous system; Early HD: hemodialysis intervention within 48 h of the onset of clinical toxicity; y: year; d: day.

^aData was presented as median (minimum, maximum) except for percentage of CNS lesion and CNS depressant.

^bDuration of maintenance dialysis.

^cPreexisting CNS lesion.

^dConcomitant CNS depressant.

^eTime from the initiation of baclofen to the onset of clinical intoxication.

^fDuration of neurotoxicity.

Discussion

Baclofen, an agonist of γ-aminobutyric acid, is the most effective agent in treating spasticity of spinal origin,[[1]] and has also been used to treat intractable hiccups,[[9]] trigeminal neuralgia,[[10]] hemifacial spasm,[[11]] and tardive dyskinesia in psychiatric patients.[[12]]

The therapeutic dosage range of baclofen is between 15 and 80 mg/day, and the therapeutic blood level is considered to be 80 to 400 ng/mL. When baclofen is administered in therapeutic doses, spinal effects predominate, while neurotoxicity becomes increasingly evident as the dosage is increased. In single doses, baclofen is rapidly and almost completely absorbed from the gastrointestinal tract, while absorption becomes prolonged and less complete with higher doses. Baclofen is moderately lipophilic, and approximately 30% are bound to serum proteins at therapeutic level. It has a molecular weight of 213 Da and a volume of distribution of 0.83 or 2.4 L/kg.[[13]], [[14]] Animal studies have found baclofen to be widely distributed throughout the body, but with only small amounts crossing the blood–brain barrier.[[1]] On average, without cerebral lesions, the concentration of baclofen in human plasma is approximately 8.5 times that of the simultaneous cerebrospinal fluid concentration following several months of therapy.[[15]] Entry of baclofen into animal brain tissue is gradual and elimination is very slow.[[16]] Baclofen is primarily excreted by glomerular filtration, with a clearance proportional to creatinine clearance. Seventy to eighty percent of the administered dose is recovered unchanged in the urine within 72 h. The half-life of baclofen is between 4.5 and 6.8 h in healthy subjects, but increases in patients with renal failure.[[17]] Thus, baclofen accumulation and neurotoxicity may occur when normal doses are administered to patients with impaired renal function. Several reports have also suggested that baclofen sensitivity and baclofen-associated neurotoxicity may be increased in the elderly or in patients with cerebral lesions.[[18]], [[19]]

Baclofen-associated neurotoxicity can be classified into two types.[[6]] Patients with acute intoxication present with four major clinical manifestations: encephalopathy, respiratory depression, muscular hypotonia, and generalized hyporeflexia. Meanwhile, patients with chronic intoxication present with hallucination, impaired memory, catatonia or acute mania. Acute intoxication has a faster onset, shorter duration, more severe clinical manifestations, and higher incidence of seizures than chronic intoxication. Most reported cases involving patients with impaired renal function display clinical manifestations of acute intoxication.

Owing to the lack of an effective antagonist of baclofen, treatment is generally supportive. Extracorporeal techniques for accelerating the elimination of the drug from the body have been reported to shorten the recovery time of baclofen-associated neurotoxicity in patients with advanced renal failure, especially hemodialysis.[[2]] Peritoneal dialysis has been considered to be ineffective in removing baclofen,[[3]] and less effective than early hemodialysis in treating baclofen-associated neurotoxicity.[[2]]

As shown in Table 1, our cases demonstrate acute baclofen intoxication syndrome as most reported cases in the literature. No significant difference between our cases and those treated by hemodialysis in the clinical parameters include age, dialysis time, preexisting CNS lesion, concomitant use of CNS depressant, total baclofen dose exist as shown in Table 2. The clinical course of neurotoxicity is also similar in both groups. Furthermore, our cases also demonstrate similar clinical profiles as those treated by early hemodialysis reported by Chen et al.[[2]] as shown in Table 3. Most importantly, the duration of neurotoxicity is not significantly different between the two groups. Since baclofen is only partially dialyzable and crosses the blood–brain barrier poorly,[[16]] intermittently short-term hemodialysis only can remove limited amount of baclofen from the brain. In most reported cases treated by hemodialysis, there is always a delay in consciousness recovery[[2]], [[7]] due to the major determinant of baclofen-associated neurotoxicity is the concentration of baclofen in the central nervous system and one session of short-term hemodialysis can result in little decrease of the baclofen level in the central nervous system.

However, there is a continuous removal of baclofen in the plasma that accelerates elimination of the drug from brain in our cases treated with CAPD, although less effective than hemodialysis. Based on the pharmacokinetics of baclofen and clinical observations of the patients herein, we concluded that intermittent hemodialysis, early or routine is not more effective than CAPD in shortening the recovery time of baclofen-associated neurotoxicity in uremic patients.

In conclusion, we report baclofen-associated neurotoxicity in three uremic patients undergoing CAPD. Baclofen should be avoided in uremic patients. In patients with advanced renal failure who develop severe baclofen-associated neurotoxicity, dialysis may be the most appropriate treatment for alleviating clinical symptoms and shortening recovery time. CAPD and intermittent hemodialysis have similar efficacy in treating baclofen-associated neurotoxicity, especially in uremic patients.