Variability of serum phenytoin levels in critically ill head injured patients in intensive care unit

Variability in serum phenytoin and wide inter patient variability have also been noted in various studies of different population groups.[2-5] Critically ill patients with multiple drug prescriptions and organ dysfunction can have unpredictable drug metabolism. Hence we decided to study the phenytoin levels following intravenous loading dose in head injured patients to study the quantum of variability in the serum phenytoin levels in this group of patients.


Introduction
Phenytoin is a widely prescribed anticonvulsant drug in the intensive care unit.For maximum effectiveness this medication needs to be maintained in steady levels in the body.Therapeutic levels are usually between 10-20 mcg/ml.Intravenous loading, oral loading and starting/restarting oral maintenance-dosing can achieve a serum phenytoin level >10 mcg/ml.Stable phenytoin levels at steady state are obtained in most patients seven to ten days after initiation of oral therapy with recommended doses of 300 mg/day.Following rapid intravenous (I.V) infusion of phenytoin, therapeutic serum levels were achieved within 5 to 30 min with a half-life of 31.2 ± 8.4 h. [1] maintenance intravenous dose of 300 mg per day in divided doses.Serum phenytoin was measured on the fi rst day and fi fth day by (Clia Immunolite Fully Automated Immuno Assay System, Dpc, Usa).

Results
In all, 34 patients serum phenytoin levels, were measured on the fi rst and the fi fth day following loading and maintenance dose of phenytoin.In addition, among these 12 patients phenytoin level were done on day 10 th also.There were 29 males and 5 females all of whom had moderate to severe head injuries.The median serum phenytoin level was 11.45 mcg/ml (2.5-44.28mcg/ml) on the fi rst day and 10.85 mcg/ml (2.4-41 mcg/ml) on the 5 th day.There was no statistically signifi cant difference in the median serum phenytoin level from day 1 to day 5 (P value = 0.9240).Out of the 34 patients only 21% patients remained in the therapeutic range during the monitoring period.
When individual serum phenytoin levels were studied, [Table 1] 41.18% had sub therapeutic, 38.24% had therapeutic and 20.59% had toxic levels on day one.Five days later 47.06% had sub therapeutic, 29.41% had therapeutic and 23.53% had toxic levels.Table 2 shows the serum phenytoin levels on day five.The pharmacokinetics in twelve patients on day ten is shown in Table 3 and Table 4.
Patients with toxic levels on day one and day fi ve had a median age of 55 years (SD ± 18.39) and 49 (SD ± 21.60) years respectively There were three patients who had toxic level on day one and therapeutic or sub-therapeutic by day 5.They had median age of 55 years, while patients with normal or sub therapeutic levels on day one and day fi ve had median age of 38 (SD ± 21.03) years and 39 (SD ± 20.25) years respectively.Liver function tests were available for 20 of these patients and were abnormal in ten patients.But there was no correlation between the toxic levels and raised transaminases nor did low albumin levels relate to sub therapeutic level.In patients with serum phenytoin levels in toxic range on day one, the average Serum Glutamic Oxaloacetic Transaminase (SGOT) and Serum Glutamic Pyruvic Transaminase (SGPT) was 65.4 IU/L and 76.2 IU/L as against 108 IU/L and 55.5 IU/L for those with subtherapeutic and normal levels.On day fi ve, the average SGOT and SGPT was 100.8 IU/L and 52.8 IU/L for patients with toxic levels vs. 49 IU/L and 65.4 IU/L for normal and sub therapeutic levels.Albumin level was 3.75 gm/dl for those with sub therapeutic level vs. 3.63 gm/dl for those with normal or toxic level.

Discussion
This study shows the wide variability of serum phenytoin levels after recommended parental loading and maintenance dose of phenytoin in the critically ill head injured patients in the intensive care unit raising concern about the safety of the drug.
Unbound fraction of phenytoin is the therapeutically active drug, but routinely the total plasma concentration is monitored and used as surrogate in dose adjustment regimes and hence total plasma concentration was used in our study.Phenytoin is highly protein bound but many factors other than serum albumin levels infl uence the drug levels of phenytoin.The liver being the chief site of biotransformation of phenytoin, seriously ill patients with impaired liver function and elderly patients, may show early signs of toxicity.Although toxic levels in our study, both on the fi rst and on the fi fth day occurred in patients older than those with normal or sub therapeutic level, it was not statistically signifi cant probably because numbers in individual groups were small.Markowesky et al., in their study concluded that phenytoin protein binding signifi cantly correlated with albumin and was more variable in ICU and convalescent patients with brain injuries than in healthy volunteers. [6]In our study, neither albumin nor transaminse levels appeared to correlate with raised or sub therapeutic phenytoin levels.In their study the phenytoin requirement decreased during convalescent.Our study did not include convalescent patients but even when studied up to 10 days there were unpredictable fl uctuations in the levels.
Theodore and colleagues reported a pseudo steady state like phenomenon lasting for 5 to 10 days with plasma phenytoin levels remaining stable for 2 to 12 days after a dosage change.Subsequently, the levels fl uctuated by more than 25% in the next 5 to 22 days and a fi nal steady-state level was reached between 13 to 31 days after the fi rst dosage change.In our study steady state therapeutic level was seen to occur only in 21% of cases while in the rest, there was a wide fl uctuation in the levels with a difference of more than 100% between the highest and lowest phenytoin level in individual cases (in four patients the difference exceeds 500%).[8] Boucher et al., correlated fall in phenytoin levels to increase in urinary clearance, which they postulated to be a consequence of changes in protein binding, induction of metabolism or the infl uence of stress on hepatic metabolic capacity.Martinnelli et al., [8] have suggested the necessity to individualize dosing after the fi fth day following rapid intravenous loading to maintain safe and therapeutic levels.Bayesian regression programme has been used to forecast an estimate of each subject's individual pharmacokinetics [9] in various studies.

Conclusion
Only 38.24% patients, had therapeutics phenytoin level on day one, while 20% were in toxic range.On day fi ve, 23% patients were in toxic and 29.41% were in therapeutic range.This study shows that there is a wide variability of phenytoin levels in the ICU patients with a difference of more than 100% between the highest and lowest phenytoin level in individual cases (In four patients the difference exceeds 500%).Potential mechanisms for this include the infl uences of poly pharmacy, altered drug metabolism, acid/base status and other factors beyond serum albumin levels.These results raise concern about the safety of phenytoin doses.Hence it is recommended that ICU patients receiving phenytoin therapy should have periodic serum phenytoin obtained even in absence of seizures or classic signs of phenytoin toxicity.

"HANSRAJ NAYYAR MEMORIAL AWARD"
A prestigious award, called the "HANSRAJ NAYYAR MEMORIAL AWARD" will be awarded annually to the best paper presented by a young scientist, below 40 years of age, at the National Conference in Critical Care organized by ISCCM AT Agra.
The winner will present this paper in a special session Called the HANSRAJ NAYYAR MEMORIAL SESSION to be held during the plenary Session.The award consists of a trophy, certifi cate, and a cash prize, which will be presented to the recipient of the award, at the Annual National Conference.
Last date for submission -15 th December 2008.
The submission should be made to The Organizing Secretary of Conference Dr. Vijaykumar Singhal Advertisement

Table 4 : S. Phenytoin level day 1, day 5 and day 10 [12 patients]
Source of Support: Nil, Confl ict of Interest: None declared.Indian J Crit Care Med January-March 2008 Vol 12 Issue 1