There is an extensive list of products containing theophylline. An important distinction to make in the toxicology of theophylline is whether the patient has taken a sustained release preparation or not.
Theophylline poisoning is a toxicological emergency. Complicated poisonings have a high morbidity. After dose, the most clinically important distinctions to make are whether the preparation is sustained release and whether the patient has toxicity from an acute single ingestion or from chronic overmedication.
Management decisions should be based on both clinical assessment and laboratory information (particularly theophylline concentrations).
The management of theophylline toxicity is compounded by clinical differences between chronic (overmedication) and acute (large ingestion) intoxication, inter- and intra- individual variability in theophylline metabolism and dose dependent kinetics in the poisoned patient.
Theophylline poisoning requires frequent observation and aggressive efforts toward achieving successful detoxification. The mainstay of initial management should be adequate and repeated doses of activated charcoal. Every effort should be made to ensure the early success of these more conservative but very effective measures as failure of this management strategy is an indication for haemodialysis. In patients with clinically severe toxicity haemodialysis (high efficiency if available) should be undertaken. Patients with at risk concentrations but with moderate toxicity should (where practical) be transferred to centers where these techniques can be performedp
At toxic concentrations theophylline causes inhibition of phosphodiesterase with resultant cAMP accumulation. In addition there may be alterations in intracellular calcium translocation 8.
Dose related increases in catecholamine concentrations occur with both therapeutic and toxic concentrations of theophylline 10,12 and much of theophylline's cardiovascular and metabolic toxicity has been attributed to this catecholamine excess 10.
In animals it has been shown that the rise in catecholamines precedes but is proportional to the eventual peak theophylline concentration. This rise in catecholamines is thought to cause the hypokalaemia and hyperglycaemia seen in acute poisonings 10,15 which can also precede and predict eventual theophylline toxicity 3,15,16.
In addition, adenosine is known to be responsible for negative feedback to the heart in situations of sympathetic overstimulation 13. The blockade of adenosine receptors and loss of negative feedback may therefore compound the effect of excess catecholamines.p
Conventional preparations exhibit virtually complete and rapid absorption (peak concentrations 0.5-2 h).
Therapeutic doses of sustained release preparations vary in the total extent of absorption and in the time to peak concentration (4-18 h).
In acute poisoning with sustained release preparations the peak concentration usually occurs between 2 and 18 hours after admission 1 but can occur up to 24 hours 3,4.
Factors contributing to this include delayed gastric emptying and tablet aggregation 5-7. Suppositories have erratic absorption and may cause chronic toxicity.
The mean apparent volume of distribution for theophylline is 0.5 L/kg and in normal adults the clearance is 40-45 mL/kg/hr giving a half-life of approximately 8 hours.
In overdose, hepatic metabolism of theophylline is frequently saturated & the apparent half-life can be as long as 30 hours.
The pharmacokinetics of theophylline may be further affected by intercurrent hepatic, cardiac or renal disease and numerous medications. In addition intercurrent illness also changes the individual patient's susceptibility to the various complications of theophylline toxicity. The scene, therefore, is set for an extremely variable response to any given dose of theophylline and even some variability in response to a given plasma concentration of theophylline depending on the type of poisoning (acute or chronic) and any underlying medical conditions.
Controlled release medication
The onset of major toxicity following the ingestion of sustained release preparations may be delayed by up to 24 hours.
Patients with overdoses of sustained release preparations need repeated clinical and laboratory assessment.
At concentrations within the therapeutic range, theophylline is known to:
The incidence of toxic symptoms and signs increases with the concentration of theophylline 16.
Although symptoms tend to evolve in a sequential fashion, individual variation is such that signs of major toxicity may not be preceded by symptoms of mild toxicity 17,18. This is especially so in chronic overmedication where, in addition, toxicity occurs at lower theophylline concentrations than in acute large ingestions 1,3.
The cardiac effects are complex and include both positive inotropic effects and a net decrease in peripheral vascular resistance leading to an increase in cardiac output and organ perfusion. The effect on individual organs is variable. In the renal vascular bed, vasodilatation and increased perfusion results in a natriuresis. In contrast, there is an increase in cerebrovascular vascular resistance which may, in part, be responsible for the lowered seizure threshold 8.
Sinus tachycardia is the most common cardiac manifestation of theophylline toxicity. Characteristically, toxic patients and laboratory animals have a high cardiac output with decreased peripheral resistance due to beta 2-mediated vasodilatation, often associated with a fall in mean arterial pressure 10,11,25,26. This can be exacerbated by the hypovolaemia which occurs as a consequence of protracted emesis and diarrhoea. Hypotension is more common in acute than chronic intoxication 3.
Supraventricular (paroxysmal atrial tachycardia, multifocal atrial tachycardia and atrial fibrillation/flutter) and ventricular arrhythmias may occur particularly with theophylline concentrations of > 100 mg/L (550 micromol/L) in acute poisonings 13. In chronic overmedication, arrhythmias, particularly atrial, occur at concentrations of around 40 mg/L (220 micromol/L) 3. Theophylline has been shown to cause abnormal atrial automaticity in isolated human atrial muscle which can be suppressed with the calcium antagonist diltiazem 27. Diltiazem does not appear to reduce theophylline induced increased myocardial contractility which is thought to be secondary to the translocation of intracellular calcium stores 27.
Therapeutic concentrations produce an increase in gastrin release and gastric acid production and decrease in lower oesophageal sphincter pressure 9. Nausea and vomiting are frequent symptoms in both acute and chronic poisonings but are more common in acute poisonings 1,3. These symptoms are due to a central emetic effect combined with local effects such as decreased lower oesophageal tone and increased gastric acid production 9. The duration and amount of vomiting correlates with the peak theophylline concentration and the duration of toxic concentrations 19. Diarrhoea has also been noted and is thought to be due to increased gastrointestinal secretions. Gastrointestinal haemorrhage has been reported 19.
Direct stimulation of the respiratory centre may cause the patient to hyperventilate.
The patient may appear agitated secondary to cerebral excitation and hyperreflexia is common. Seizures (general or focal with secondary generalisation) may occur 17,18,22,23 and tend to occur at lower theophylline concentrations in chronic toxicity 1,3. They are a poor prognostic sign with death reported in 50% of patients who had seizures in one series (although these patients may not have received adequate detoxification) 17. Cerebral injury is common after seizures.
Postulated mechanisms have included both cerebral vasoconstriction (related to adenosine blockade) and rises in cerebral concentrations of cyclic AMP which have been shown to be epileptogenic in rats 23. It has been suggested that patients with pre-existing brain injury are more prone to focal seizures 22, however EEGs of patients with no pre-existing cerebral lesions have been shown to have an increase in paroxysmal activity when theophylline concentrations were within the accepted therapeutic range 24. In the majority of patients there is a good correlation between theophylline concentrations and the likelihood of seizures 1,3,17 although in the paediatric age group seizures may occur with serum concentrations just above the therapeutic range 18. Hallucinosis and psychosis has been reported 11,20,21.
This is most likely to occur with sustained release preparations and the clinical effects will be similar. If the patient is asymptomatic and more than 24 hours have elapsed then no treatment is indicated. In all other circumstances the treatment, including gastrointestinal decontamination, should be done as usual.
In acute poisonings hypokalaemia, hyperglycaemia, hypercalcaemia, hypophosphataemia and lactic acidosis may occur 3,10,15,28-30. All of these abnormalities have been attributed to catecholamine excess with intracellular movement of potassium and catecholamine-stimulated gluconeogenesis. In acute poisonings, hypokalaemia may predict serious toxicity before serum concentrations reach their peak. Respiratory alkalosis may occur secondary to stimulation of the respiratory centre. A raised creatinine kinase can occur with or without seizures 11,31.
Rhabdomyolysis induced renal failure has been reported 32.
Chronic toxicity tends to be associated with less hypokalaemia and higher bicarbonate concentrations than acute toxicity 3.
As there is a long and variable absorption following an acute ingestion of sustained release preparations theophylline concentrations need to be taken 2nd hourly until the concentration has clearly reached a plateau or is falling. Once the theophylline concentration has begun to decline, concentrations should still be taken 4th hourly to ensure that no secondary peak occurs from ongoing absorption of controlled release formulations. There is a correlation between the peak concentration in both acute and chronic poisonings and the development of major toxicity and death 13.
In adult patients with acute single ingestions the incidence of seizures and arrhythmias is increased when the serum concentrations are greater than 100 mg/L (550 micromol/L) and especially so if the concentrations are greater than 150 mg/L (825 micromol/L). However major toxicity in children has been documented at lower concentrations 21.
The threshold for chronic toxicity is less well defined but there is a significant risk of major complications occurring with concentrations greater than 40 mg/L (220 micromol/L). This may in part be due to these patients tending to be either very young or old with chronic illness. Individual variations including age and pre-existing illness need to be taken into account when planning treatment of toxicity.
To this end the clinical assessment of toxicity has an independent value in defining treatment.
The major difference in toxicity relate to the type of formulation as controlled release preparations give a prolonged and delayed toxicity.
All patients require frequent clinical assessment of their severity. The history should establish:
(e.g. cardiac or neurological disease) or alter theophylline pharmacokinetics (e.g. hepatic disease).
There is often an overlap in clinical features of severity. The most serious category should be assumed.
|Nausea||Vomiting but tolerates decontamination||Vomiting & not tolerating decontamination|
|Pulse < 120||Pulse < 140||pulse >140|
|Systolic BP > 120 mmHg||Systolic BP > 100 mmHg||Systolic BP < 100 mmHg|
|No arrhythmias||Atrial or ventricular ectopics||SVT or Ventricular Tachycardia|
|Agitation or hyperreflexia||Seizures|
|Potassium < 3.0 mmol/L||Potassium < 3.0 mmol/L|
|Glucose > 10 mmol/L||Glucose > 10 mmol/L|
|Rising 2nd hourly theophylline concentrations in the presence of apparently effective decontamination|
Potentially significant toxicity includes all chronic overmedication, acute ingestions of > 10 mg/kg, and acute ingestions with more than mild toxicity regardless of stated amount ingested.
Criteria for consideration of intensive care unit admission:
IV fluids are essential because of beta-mediated vasodilatation. ECG monitoring is mandatory for all but the most trivial poisonings.
Control of vomiting
Vomiting may be extremely difficult to control even when using high doses of antiemetics. In our experience, ondansetron (8 mg IVI) (or alternative -setrons) appear to be much more effective than even high dose metoclopramide (40-200 mg IVI) and would be first choice as an antiemetic. Patients with vomiting refractory to these measures often have higher theophylline concentrations and may require haemodialyis or haemoperfusion 33. The therapeutic goal is to ensure the majority of doses of activated charcoal are kept down. There have been two cases reported where emesis was controlled only after the addition of ranitidine, the authors postulating that the increase in gastric acid production contributes to nausea .34. At this time, there are no controlled trials which examine the efficacy of this treatment. Inhibition of theophylline metabolism by ranitidine with subsequent theophylline toxicity has been reported 35.
Ipecac induced emesis is not indicated. If the patient is not vomiting then gastric lavage should be performed after the airway is protected with intubation if necessary. A large bore orogastric tube should be used and followed by the administration of activated charcoal. It should be assumed that theophylline is still in the stomach even long after ingestion. We have recently seen a patient who vomited intact tablets 7 hours post poisoning. Tablet bezoars have been documented endoscopically 6,7 and at post mortem 5 and can be responsible for prolonged and sometimes episodic absorption for up to 48 hours.
Activated charcoal binds avidly to theophylline and should be given in a dose of 1 to 2 gm/kg as the first dose regardless of the time of ingestion. Theophylline clearance is enhanced if charcoal is given with a cathartic (providing the cathartic is effective) such as sorbitol 41 but this is no longer routine therapy.
Whole bowel irrigation has also been used successfully to decontaminate patients with both sustained release 42 and conventional preparations 43. It may be used in combination with activated charcoal. There is one case report of its use with activated charcoal during which the theophylline half-life was reduced to the same time as that subsequently achieved using charcoal haemoperfusion 42. The technique is to administer a nonabsorbable iso-osmolar fluid containing polyethylene glycol via nasogastric tube at a rate of 2 L/h in adults (500 mL/h in children) until the rectal effluent is clear. The mean time for this to be achieved is 4 hours 43.
Central nervous system
Control of seizures may be difficult 1,3,17,22 Previous literature reviews suggest that thiopentone is the most efficacious therapy 2 but its use requires intubation and ventilatory support. In clinical practice, diazepam seems to be effective in some patients 1,21 and should be the treatment of first choice followed by phenobarbitone (15 mg/kg) if that fails. Patients whose seizures are refractory to these measures require intubation and thiopentone loading (3-5 mg/kg) and infusion (2-4 mg/kg/hr) 2. Phenytoin as a single agent does not provide good control and may worsen outcomes 1,2. Animal work supports the use of barbiturates. Mice pre-treated with phenobarbital had a significant increase in LD50 and time to seizure compared with controls while pre-treatment with phenytoin reduced the LD50 and time to seizure compared with controls 35.
Case reports and animal work have shown that the mean blood pressure may be improved by the nonselective beta-blocker propranolol without significant change in cardiac output or pulse 10,11,14. Obviously, treatment with a nonselective beta-blocker is potentially hazardous in asthmatics and there has been one report of propranolol induced bronchospasm in the setting of theophylline toxicity 14. The short acting and relatively beta 1-selective blocker, esmolol has been studied in animals and found to control tachycardia while causing a significant increase in systemic vascular resistance without significant change in cardiac output 26. Inotropic agents (dopamine, dobutamine) do not seem to be very effective in case reports and may be inappropriate in view of their potential for further sympathetic stimulation 11. Adequate volume expansion should be assured in any patient in whom the use of inotropes is considered.
Hypokalaemia, hyperglycaemia and acidosis may be partially corrected by propranolol 10,14,39. The hyperglycaemia usually does not require treatment and as the initial hypokalaemia does not represent total body depletion, potassium replacement should be undertaken cautiously if at all. Underlying hypoxia, acid base status and electrolyte abnormalities which may contribute to arrhythmias should be corrected. In theophylline toxicity both intravenous propranolol and verapamil have been used with varying success in the control of supraventricular arrhythmias 11,37,38. The use of a selective beta-blocker may also be useful in the treatment of tachyarrhythmias in some patients. Ventricular arrhythmias have been reported to respond to both lignocaine and propranolol 3,14.
Multiple doses of activated charcoal
Repeated doses of activated charcoal have been shown to enhance significantly the elimination of both parenteral and orally administered theophylline 33,44-46. This is thought to be due to direct dialysis of theophylline across the gut mucosal capillaries. The dose should be 30-40 g every 4 hours however equal benefit has been demonstrated by giving smaller doses more frequently (e.g. 10 g q1h) 47. This method may be better tolerated in a nauseated patient as it has been suggested that large doses of charcoal may increase the episodes of emesis 19. Super activated charcoals with an increase in binding capacity of 3-4 times that of regular charcoal enable smaller amounts of charcoal to be given with equal efficacy 48,49. Activated charcoal has been used successfully in theophylline toxic neonates and infants 44. Both repeated dose activated charcoal and whole bowel lavage are effective treatments but in the clinical setting their use may be limited by protracted vomiting.
These methods may be useful in overcoming this:
Activated charcoal should not be given in the presence of intestinal ileus.
As use of repeated dose activated charcoal is economical, effective and much more readily available than complicated techniques such as charcoal haemoperfusion it should be regarded as the mainstay of treatment of theophylline toxicity. Every effort should be made to ensure this more conservative therapeutic regimen is given an optimum chance of success as failure of this technique to blunt the rise or cause a fall in serum theophylline is an indication to consider more aggressive intervention.
The use of charcoal and resin haemoperfusion represents a significant advance in the treatment of life threatening theophylline toxicity 4. The technique increases theophylline clearance from 2 to 5 fold 2 (average 4 fold) with red cell and plasma clearance being equally increased 51 Theophylline clearances of 112.8 - 350.4 mL/kg/hr have been reported with the use of charcoal haemoperfusion 2. This is a significant increase over the normal values.
The procedure has a low morbidity when used in experienced centers 3,4,51. Coagulopathy and electrolyte imbalance are the most commonly reported complications. Cartridge saturation necessitating replacement occurs at approximately 2 hours 53,54. Although survival of theophylline toxic patients with extremely high concentrations has been reported without intervention such as haemoperfusion their clinical course has been unstable 40,55. In addition, a poor outcome has been documented in patients in whom the major complications of seizures or major cardiac instability have occurred 17,56. Because of this we feel that in patients with clinically severe toxicity the risk benefit ratio favours haemoperfusion even without particularly high concentrations.
Indications for haemoperfusion are:
The data on theophylline concentrations indicate that specific threshold concentrations can be chosen above which an unacceptable risk of life-threatening events exists even if the patient shows only moderate toxicity.
However it has been our and others' 57 experience that some patients can tolerate theophylline concentrations in excess of these treatment thresholds without signs of moderate or severe toxicity. In some patients with high serum concentrations but with signs of mild or even moderate severity and who are tolerating “aggressive conservative” measures it is our opinion that haemoperfusion can be deferred providing that subsequent serum concentrations demonstrate an initial plateau and then fall. These decisions should be made on an individual patient basis. It is important in this context to note the predictive value of hypokalaemia on presentation as a marker of ensuing moderate or severe toxicity.
Haemodialysis increases the clearance of theophylline by 50% 58,63 and is about half as effective as haemoperfusion. Haemodialysis and haemoperfusion have been used in series 53,59, a technique demonstrated to reduce cartridge saturation, maintain normal body temperature and facilitate control of electrolyte imbalance 53 Haemoperfusion provides a higher theophylline clearance rate than haemodialysis. However, haemodialysis appears to have comparable efficacy in reducing the morbidity of severe theophylline intoxication and is associated with a lower rate of procedural complications 64.
Peritoneal dialysis has been used successfully in a few cases 60 and may be of value when haemoperfusion or haemodialysis is not available and the patient cannot be transferred to facilities with access to these techniques.
On theoretical grounds continuous arteriovenous haemofiltration may produce a 24 hour clearance which approaches that of a single two hour charcoal haemoperfusion 61 but there are no data on its use in theophylline poisoning and it cannot therefore be recommended. Continuous venovenous haemodialysis (CVVHD) has been used with success 62.
Patients who have had seizures due to theophylline toxicity should have full neuropsychiatric review. This review should particularly focus on short term memory and areas associated with information processing.