Extracorporeal treatments for isoniazid poisoning: Systematic review and recommendations from the EXTRIP workgroup

Isoniazid toxicity from self-poisoning or dosing errors remains common in regions of the world where tuberculosis is prevalent. Although the treatment of isoniazid poisoning is centered on supportive care and pyridoxine administration, extracorporeal treatments (ECTRs), such as hemodialysis, have been advocated to enhance elimination of isoniazid. No systematic reviews or evidence-based recommendations currently exist on the benefit of ECTRs for isoniazid poisoning. The Extracorporeal Treatments in Poisoning (EXTRIP) workgroup systematically collected and rated the available evidence on the effect of and indications for ECTRs in cases of isoniazid poisoning. We conducted a systematic review of the literature, screened studies, extracted data on study characteristics, outcomes, and measurement characteristics, summarized findings, and formulated recommendations following published EXTRIP methods. Forty-three studies (two animal studies, 34 patient reports or patient series, and seven pharmacokinetic studies) met inclusion criteria. Toxicokinetic or pharmacokinetic analysis was available for 60 patients, most treated with hemodialysis (n = 38). The workgroup assessed isoniazid as “Moderately Dialyzable” by hemodialysis for patients with normal kidney function (quality of evidence = C) and “Dialyzable” by hemodialysis for patients with impaired kidney function (quality of evidence = A). Clinical data for ECTR in isoniazid poisoning were available for 40 patients. Mortality of the cohort was 12.5%. Historical controls who received modern standard care including appropriately dosed pyridoxine generally had excellent outcomes. No benefit could be extrapolated from ECTR, although there was evidence of added costs and harms related to the double lumen catheter insertion, the extracorporeal procedure itself, and the extracorporeal removal of pyridoxine. The EXTRIP workgroup suggests against performing ECTR in addition to standard care (weak recommendation, very low quality of evidence) in patients with isoniazid poisoning. If standard dose pyridoxine cannot be administered, we suggest performing ECTR only in patients with seizures refractory to GABA_A receptor agonists (weak recommendation, very low quality of evidence).     

Treatment of acute isoniazid toxicity.

The clinical symptoms and treatment of acute isoniazid toxicity are presented. The use of supportive measures and chemotherapy are discussed in detail. The pharmacology and biochemistry underlying the symptons of isoniazid poisoning are aslo presented. It is concluded that diazepam in combination with pyridoxine is the treatment of choice for the management of convulsions associated with isoniazid toxicity. Pyridoxine should be administered intravenously in amounts equal to the estimated quantity of isoniazid ingested, even if seizures have not occurred.     

Seizures induced by theophylline and isoniazid in mice.

Isoniazid-induced seizures respond poorly to anticonvulsants but well to pyridoxine (Vitamin B6); theophylline produces difficult-to-treat seizures with substantial morbidity and mortality. Theophylline therapy depresses plasma pyridoxal-5'-phosphate (PLP), the active metabolite of pyridoxine, suggesting that theophylline-induced seizures might be amenable to treatment with pyridoxine. Our study established the dose-response relationship for convulsions due to isoniazid and theophylline in mice and determined if pyridoxine antagonized such seizures. Female CD-1 outbred mice weighing 25 to 30 g were used. Clonic seizures had clonic activity lasting 5 sec; tonic seizures had loss of the righting reflex with tonic hindlimb extension. Groups of 10 mice received single doses of 50, 100, 150, 200, 250 or 300 mg aminophylline/kg i.p. or 100, 150, 200, 250, 300 or 350 mg isoniazid/kg i.p. and were observed for seizures or death. Pyridoxine or saline with aminophylline or isoniazid were administered simultaneously. The LD50 for aminophylline was 266 mg/kg; for isoniazid it was 160 mg/kg. Doses of 150 mg aminophylline/kg or 100 mg isoniazid/kg did not induce seizures. Pyridoxine with aminophylline or isoniazid did not alter the frequency or time of onset of seizures or death. This was unexpected because pyridoxine antagonizes theophylline-induced seizures in mice and reverses isoniazid-induced seizures in humans. We found no evidence that PLP depletion in mice is a mechanism for seizures induced by isoniazid or aminophylline in a fashion similar to isoniazid in humans.     

Potentiation of pyridoxine by depressants and anticonvulsants in the treatment of acute isoniazid intoxication in dogs

Treatments for acute isoniazid (INH) intoxication have included, singly and in various combinations, a great variety of drugs. As a consequence it is difficult to evaluate the efficacy of these antidotes, except for pyridoxine, the most commonly recommended one. In some cases of INH poisoning evaluation is further complicated because of concurrent alcohol ingestion. The objectives of this investigation were to determine whether ethanol enhances the toxic effects of acute INH overdose, as suggested by some clinical reports, and to evaluate the antidotal efficacy of phenobarbital, pentobarbital, phenytoin, ethanol, or diazepam when each is administered in combination with pyridoxine. Male dogs were either pretreated with iv ethanol and challenged 1 hr later with po INH, 50 or 75 mg/kg, or they were given INH, 75 mg/kg, and injected iv 30 min later with the test drugs, alone or in combination with pyridoxine. Ethanol pretreatment not only did not enhance the toxicity of INH but, in fact, it reduced the severity of convulsions, although it did not change the mortality rate. In the antidotal study, none of the five CNS depressants or anticonvulsants protected against clonic-tonic seizures or death. Pyridoxine, however, reduced the severity of the seizures and prevented death, although it did not completely block convulsions. The combination antidotal treatments (pyridoxine plus each of the CNS drugs) were the most effective; they prevented both convulsions and lethality. It is suggested that pyridoxine is the basic antidote for treatment of acute INH poisoning, and that the addition of an anticonvulsant or a CNS depressant to the therapy enhances effectiveness.     

急性重度异烟肼中毒9例救治分析

目的探讨重度异烟肼中毒救治的有效方法。方法对9例重度异烟肼中毒病例进行回顾性分析。对于吞服异烟肼剂量明确患者,均给予大剂量维生素B6（与中毒异烟肼比1：1）静脉注射,同时联合地西泮静脉注射,控制抽搐。结果 9例患者中,8例痊愈出院,死亡1例。结论对于重度异烟肼中毒,大剂量维生素B6静脉注射,可有效拮抗异烟肼。     

Acute isoniazid toxicity and the need for adequate pyridoxine supplies

A 25-year-old, 54-kg Hispanic man who had recently started multidrug therapy for pulmonary tuberculosis presented in status epilepticus after ingesting 9 g of isoniazid in a suicide attempt. Successful management of this patient required collaboration between several institutions to provide the large amount of necessary intravenous pyridoxine. Ultimately, this single overdose depleted the supply of intravenous pyridoxine for a significant region of the state of Nebraska. Isoniazid is commonly used to treat tuberculosis, but it is encountered relatively infrequently as the cause of an acute overdose. Severe isoniazid overdoses may present as seizure activity that is refractory to conventional antiepileptic therapy. Although intravenous pyridoxine is an effective antidote for isoniazid overdoses in patients presenting with status epilepticus, this agent has few indications and is typically stocked in limited quantities. In regions with large populations of patients who receive antituberculosis therapy, collaborative networks must be created to ensure that adequate supplies of intravenous pyridoxine (> or = 20 g) are available for effective treatment of isoniazid poisonings.     

Rhabdomyolysis in isoniazid poisoning.

BACKGROUND: Rhabdomyolysis is one of the reported complications of isoniazid poisoning, but relevant data are limited. METHODS: A retrospective study was conducted on isoniazid poisoning cases seen at the Philippine General Hospital over 5 years (1992-1997). Patients excluded from the study were those who coingested other substances, including hepatotoxic and nephrotoxic drugs, those with underlying medical illnesses, and those without creatine phosphokinase muscle fraction determinations. RESULTS: Out of the 270 cases of isoniazid poisoning, 52 patient records were reviewed. Common clinical manifestations were seizures (100%), depressed sensorium (53%), and vomiting (45%). Laboratory results showed leukocytosis (74.5%), metabolic acidosis (29%), and impaired liver function tests (21%). Creatine phosphokinase muscle fraction was elevated in 59.6% of cases, beginning at a dose of 2.4 g. Values peaked on days 5 and 6 and declined on days 7 and 8. Statistically significant correlations were observed for the elevation of creatine phosphokinase muscle fraction with the duration/amount of drug ingested and the frequency of seizure. No correlation was observed between the frequency of seizures and elevated creatine phosphokinase muscle fraction nor between the time delay in consultation and elevation of creatine phosphokinase muscle fraction. CONCLUSIONS: The incidence of rhabdomyolysis in isoniazid poisoning was 3/100 cases a year. The findings suggesting its direct toxic effect on the muscles may not be clinically relevant.     

Evolving epidemiology of drug-induced seizures reported to a poison control center system

Introduction

We sought to determine whether or not the causes and consequences of drug-induced seizures have changed in the last decade.

Methods

We conducted a retrospective review of all calls to the California Poison Control System in 2003 in which seizures occurred in association with poisoning or drug intoxication. We reviewed the poison center chart of each case to determine the drug(s) involved, the type of seizures, and the medical outcome. We compared the cause of reported seizures to that found in previous investigations.

Results

386 cases were evaluated and related to poisoning or drug intoxication. The leading causes of seizures were bupropion (89 cases, 23%), diphenhydramine (32 cases, 8.3%), tricyclic antidepressants (30 cases,7.7%), tramadol (29 cases, 7.5%), amphetamines (27 cases, 6.9%), isoniazid (23 cases, 5.9%), and venlafaxine (23 cases, 5.9%). Since 1993, there was a statistically significant increase in antidepressant related seizures but a decrease in TCA and cocaine related seizures. In 265 patients (68.6%) only a single seizure was reported, while 3.6% (14 cases) reported status epilepticus. Two-thirds (65.5%) of the cases involved suicide attempts and 14.8% the direct result of drug abuse. There were 7 deaths. Of the 7 deaths, 4 people had significant hyperthermia. There was a statistically significant increased risk of death associated with stimulant exposure.

Conclusion

While tricyclic antidepressants, antihistamines, stimulants, and isoniazid remain common causes of drug induced seizures, bupropion, tramadol, and venlafaxine have emerged as common causes of drug-induced seizures for which poison center consultation is requested.     

Interaction with pyridoxal as a possible mechanism of hydralazine hypotension.

The mechanism by which the antihypertensive vasodilator hydralazine relaxes vascular smooth muscle is unknown. The drug interacts with pyridoxal and can produce B6 deficiency; it also inhibits a number of enzymes requiring pyridoxal as a cofactor, but there is no apparent relation between its enzymatic and blood pressure effects. To explore the possibility of a hydralazine-pyridoxal interaction at a nonenzymatic site, the acute hypotensive response to hydralazine was determined by tail cuff blood pressure (BP) measurements in conscious normotensive rats pretreated or not pretreated with pyridoxine. Other animals were pretreated with isoniazid, a drug also capable of reacting with pyridoxal. Responses to hydralazine were diminished by pyridoxine and enhanced by isoniazid; those to the vasodilator diazoxide or to the alpha-adrenergic blocker zolertine were unaffected by such pretreatments. The inhibitory effect of pyridoxine was absent when rats were pretreated with the calcium antagonists verapamil or cinnarizine. Hydralazine hypotension in anesthetized rats was also reduced by pyridoxal pretreatment. These results suggest that at least part of hydralazine-induced hypotension may be related to interaction with pyridoxal, possibly through interference with an effect of the vitamer on calcium and/or sodium transport into vascular smooth muscle.     

Pharmacodynamic interactions between isoniazid and theophylline in mice and rats, and the influence of pyridoxine.

Convulsions induced by acute administration of isoniazid (1), theophylline (2) as well as combinations of 1 and 2 were evaluated in male albino mice and male Wistar rats. The effect of pyridoxine (3) on these seizures was tested. Serum and brain levels of 1 after coadministration with 2 and caffeine (4) were assessed. The relevance of the observed pharmacokinetic phenomena in serum is questionable for the CNS processes because animals convulsed late (starting 90 min) and no significant changes of brain levels of 1 were observed. In conclusion, interactions of 1 and 2 may not occur through common mechanisms and exist only if the dose of 1 is toxic suggesting toxicological rather than therapeutic relevance.     

Effect of isoniazid on the pharmacodynamics of cefazolin-induced seizures in rats

Both isoniazid (INH) and cefazolin (CEZ) can have serious adverse effects on the central nervous system (CNS), causing seizures. In this study, we investigated the effect of INH on the pharmacodynamics of CEZ-induced seizures in rats. Male Wistar rats pretreated with INH (150 mg/kg i.p.) or saline received an intravenous infusion of CEZ at 3.2 g/h/rat until the onset of seizures, then samples of cerebrospinal fluid (CSF), blood (for serum), and brain were obtained immediately. The administration of INH was associated with a reduction in the total dose of CEZ required to produce seizures. The concentrations of CEZ in serum, brain, and CSF in INH-treated rats at the onset of seizures were significantly lower than those in control rats. In rats coadministered with pyridoxine (150 mg/kg s.c.), the concentration of CEZ in CSF at the onset of seizures was significantly higher than that in rats administered INH only. These results suggest that INH potentiates the sensitivity of the CNS to CEZ-induced seizures, and that the increased sensitivity is associated with the inhibition of vitamin B(6) metabolism by INH.     

Recurrent seizures in tramadol intoxication: implications for therapy based on 100 patients

Tramadol is an atypical opioid analgesic used in the treatment of mild to moderate pain. Despite being a GABA(A) agonist, seizures are a prominent complication with its therapeutic use, abuse or overdose. For patients who have had a tramadol-induced seizure, the likelihood of recurrent seizures and the need for emergent anticonvulsant prophylaxis is unknown. However, treatment of patients with anticonvulsants prophylactically may cause adverse effects and increased morbidity in tramadol poisoning. We studied the outcome and frequency of recurrent seizures in tramadol-intoxicated patients in an attempt to determine the need for prophylactic anticonvulsant therapy. This was a retrospective cohort study of tramadol-intoxicated patients who had at least one seizure. Patients' age, sex, cause(s) of intoxication, route of poisoning, dose or number of capsules or tablets taken, vital signs, other signs or symptoms, numbers of seizures, length of stay, co-ingestions and past medical history were ascertained. Exactly 100 patients met the inclusion criteria. Eighty-two per cent were men, and 50% were between 21 and 30 years old. By our standard clinical protocol, none were treated with seizure prophylaxis after their first seizure. Only 7% had recurrent seizures and all patients recovered without sequelae. Because of the low risk of multiple seizures in tramadol poisoning and the lack of morbidity in patients who do seize, it appears to be unnecessary to administer prophylactic anticonvulsant therapy in patients with tramadol poisoning, even if they have an initial seizure.     

Convulsions as the etiology of lactic acidosis in acute isoniazid toxicity in dogs.

The mechanism by which acute isoniazid (INH) overdose causes lactic acidosis is unknown. This study examines the role of convulsion in the development of lactic acidosis in dogs given po lethal doses of INH (75 mg/kg). Following INH, dogs did not develop acidosis until after they had experienced clonic-tonic convulsions. Acidosis, which became more pronounced with successive convulsions, was associated with marked increase of serum lactate (immediate postconvulsive level, 12.3 meq/liter, compared to control of 1.83 meq/liter). Diazepam, 0.5 mg/kg, plus pyridoxine HCl, 150 mg/kg, injected iv immediately following po INH, prevented both convulsions and changes in pH and lactate. When the antidotes were administered after the second convulsion of INH toxicity no further convulsions occurred and blood pH and lactate returned to control levels in 2 hr. “Curarization” of INH-treated dogs prevented both motor seizures and marked increase of lactate. In animals treated with INH and allowed to convulse two times to develop severe acidosis, correction of the acidosis with NaHCO₃ failed to prevent further convulsions and death. Diazepam and pyridoxine combinations, in doses that were ineffective for terminating INH-induced convulsions, became effective when given after correction of acidosis with NaHCO₃. It is concluded that convulsion is the main cause of lactic acidosis in acute INH toxicity and that correction of acidosis does not terminate acute INH toxicity. Although correction of acidosis increases the antidotal effectiveness of diazepam plus pyridoxine, increasing the dose of either or both of these two drugs achieves the same antidotal advantage without the potential problems of metabolic alkalosis which may result from NaHCO₃ administration.     

Acute human self-poisoning with the N-phenylpyrazole insecticide fipronil--a GABAA-gated chloride channel blocker

OBJECTIVE: Fipronil, a broad spectrum N-phenylpyrazole insecticide that inhibits GABAA-gated chloride channels, has been in use since the mid-1990s. A high affinity for insect compared to mammalian GABA receptors results in lower animal toxicity than other insecticides blocking this channel. To date, only two accidental cases of fipronil poisoning in humans have been published. CASE SERIES: We report seven patients with fipronil self-poisoning seen prospectively in Sri Lanka together with pharmacokinetics for four patients. Non-sustained generalized tonic-clonic seizures were seen in two patients (peak measured plasma fipronil concentrations 1600 and 3744 microg/L); both were managed with diazepam without complications. A patient with a peak measured plasma concentration of 1040 microg/L was asymptomatic throughout his stay. Plasma concentration was still high at discharge 3-4 days post-ingestion when the patients were well. Retrospective review of >1000 pesticide poisoning deaths since 1995 found only one death from fipronil-based products. In contrast to the good outcome of the above cases, this patient required intubation and ventilation and had continuous fits despite therapy with barbiturates and benzodiazepines. CONCLUSIONS: Our experience with prospectively observed patients suggests that fipronil poisoning is characterized by vomiting, agitation, and seizures, and normally has a favorable outcome. Management should concentrate on supportive care and early treatment of seizures. However, further experience is needed to determine whether increased susceptibility to fipronil or larger doses can produce status epilepticus.     

Acute metaldehyde poisoning in Taiwan

Metaldehyde, a cyclic tetramer of acetaldehyde, is a widely used molluscicide. Although cases with acute metaldehyde poisoning have been reported, the occurrence of severe poisoning is uncommon. To provide more information on human metaldehyde poisoning, we reviewed 15 cases of metaldehyde exposure reported to the Taiwan National Poison Control Center at the Taipei Veterans General Hospital between 1991 and 2002. While 7 patients were asymptomatic, the other 8 patients, including 4 who coingested alcohol or other poisons, exhibited toxic manifestations of abdominal pain, dizziness, nausea, irritation of oral mucosa, and seizures after oral exposure. One patient died after ingesting 12 g (or 258.6 mg/kg) of metaldehyde. Although the toxicity from metaldehyde is largely mild, the clinical course of metaldehyde poisoning may be rapidly deteriorating and fatal on rare occasions. Physicians should therefore be cautious in managing patients with metaldehyde poisoning, and vigorous supportive measures should be promptly instituted in patients who manifest severe toxicity.     

Histamine reactions due to ingestion of tuna fish (Thunnus argentivittatus) in patients on anti-tuberculosis therapy

An outbreak of fish poisoning affecting 56 patients in the tuberculosis ward is reported. The fish responsible was tuna, a tropical sea fish with a high histamine content. Accumulation of histamine due to inhibition of histaminase by isoniazid and other nicotinamide derivatives used in antituberculosis therapy is suggested as the underlying mechanism of poisoning.     

维库溴铵治疗硫丹中毒13例临床观察

目的 探讨维库溴铵治疗硫丹中毒所致剧烈抽搐的疗效.方法 13例硫丹中毒患者均出现全身频繁持续性抽搐,常规给予安定、苯巴比妥,如不能控制发作,静脉给予维库溴铵4 mg,继以1～2mg/h速度持续静脉泵注,并常规给予洗胃、综合及对症支持治疗.结果 给予维库溴铵后,5例于40min内抽搐得到有效控制,其余8例于2～5h内抽搐基本被控制;维库溴铵维持时间2d～1周;平均机械通气时间3d;本组无死亡病例.结论 维库溴铵可有效控制硫丹中毒所致的顽固性抽搐,提高抢救成功率.