Gail D. Anderson PhD
Professor, Department of Pharmacy, University of Washington, Seattle, Washington
CHEMISTRY AND METABOLIC SCHEME
Phenobarbital is 5-ethyl-5-phenylbarbituric acid, a substituted barbituric acid with a molecular weight of 232.23 (Figure 51.1). The free acid of phenobarbital is a white crystalline material with a melting point of 176°C. Phenobarbital is only sparingly soluble in water (1 g in 1000 mL). In addition to the low aqueous solubility, phenobarbital also has a relatively low lipid solubility. It is soluble in organic solvents such as chloroform (1 g in 40 mL), diethylether (1 g in 15 mL), and ethanol (1 g in 10 mL). The sodium salt is freely soluble in water and is used in the formulations for intravenous or intramuscular administration (1).
Phenobarbital is a weak acid with a negative log of dissociation constant (pKa) of 7.3, approximately the same as physiologic pH. Therefore, changes in the ratio of ionized to nonionized phenobarbital that occur within the normal range of physiologic pH (7.35 to 7.45) can alter both the distribution and the excretion of the drug (Table 51.1).
After oral and intramuscular administration, the absorption of phenobarbital is essentially complete, with greater than 95% bioavailability (1, 2, 3). The time to peak plasma concentrations after oral administration ranges from 0.5 to 4 hours (1, 2, 3, 4, 5). Peak concentrations appear 2 to 8 hours after an intramuscular injection (1,3). Rectal administration of phenobarbital sodium parenteral solution resulted in a mean bioavailability of 90% and a time to peak concentration of 4.4 hours (6).
In a group of newborn infants given a single dose of phenobarbital, the time to peak concentrations ranged from 1.5 to 6 hours (7). Absorption of phenobarbital was delayed in a group of malnourished children compared with healthy children (5.6 hours versus 1.0 hour); however, the maximum concentration obtained was not different, a finding suggesting no differences in bioavailability (8).
Administration of activated charcoal can be used to decrease the equilibrium of phenobarbital across the gastrointestinal tract in cases of overdose. Several studies have demonstrated an increase in total body clearance and reduction of the elimination half-life (t½) of phenobarbital by a factor of two-to 10-fold with the administration of charcoal (9,10, 11, 12, 13).
Formulations and Routes of Administration.
Phenobarbital is available in tablets, capsules, extentabs, elixir, and a parenteral formulation for oral, rectal, intramuscular, and intravenous routes of administration.
After intravenous administration, phenobarbital distributes into the body in two phases, which can be characterized by a two-compartmental mathematical model (1,2,14). Studies in animals have demonstrated that the early distribution includes liver, kidney, and heart. During the late phase, phenobarbital distributes to brain, muscle, and intestine. The average volume of distribution of phenobarbital ranges from 0.54 to 0.73 L/kg in adults (1,15,16). Newborns and young infants have a larger average volume of distribution, ranging from 0.8 to 1 L/kg (17, 18, 19, 20, 21, 22, 23). This corresponds to a relatively larger extracellular fluid volume in neonates (24). In older infants and children, the average volume of distribution is similar to that in adults, ranging from 0.57 to 0.70 L/kg (7,25,26). Because of the low lipid solubility, phenobarbital does not distribute into fat, and loading doses of phenobarbital are usually calculated using ideal body weight or lean body mass (27). However, a case report suggested that using ideal body weight to calculate the loading
dose in a patient whose total body weight was greater than 300% ideal body weight significantly underestimated the loading dose needed to provide therapeutic plasma concentrations of phenobarbital (28).
FIGURE 51.1. The main biotransformation pathways of phenobarbital.
Plasma Protein Binding.
Phenobarbital is approximately 55% bound to albumin in adults (4,29, 30, 31) and children (30,32). In neonates, phenobarbital binding is reduced, presumably because of relatively low albumin concentrations at birth; binding is 57% to 64% in neonates without hyperbilirubinemia and 70% to 72% in neonates with hyperbilirubinemia (30,33). No data are available regarding protein binding in the elderly patient population, in whom an age-related decline in albumin often occurs (34).
TABLE 51.1. PHENOBARBITAL PHARMACOKINETICSa
Cerebrospinal Fluid, Brain, and Other Tissues.
Cerebrospinal fluid (CSF) concentrations of phenobarbital are similar in infants and adults. In infants, CSF levels are 44% to 57% (7), and in adults, they are 43% to 60% (29,35) of plasma concentrations. Thus, CSF concentrations correlate with unbound phenobarbital serum levels (29). Brain penetration of phenobarbital is relatively slow. Animal studies have demonstrated that maximal entry takes 12 to 60 minutes (36, 37, 38). During prolonged status epilepticus, brain uptake is increased, presumably by disruption of the blood-brain barrier (39). The reported concentrations of phenobarbital in surgically removed human epileptic brain tissues relative to plasma concentrations vary from 0.35 to 1.13 (40, 41, 42,43). The brain:plasma phenobarbital concentration ratio in a group of newborns who died while receiving phenobarbital was 0.71±0.21 (19). The ratio of phenobarbital in gray to white matter ranged from 0.86 to 1.11 in this group of newborns, similar to values found in adults (40,42,43). In another group of 11 premature and full-term babies, infants, and children (44), the brain:serum phenobarbital concentration ratio was 0.82±0.22. Phenobarbital concentrations in brain were significantly less than other organs including liver, kidney, spleen, pancreas, and lung (44). A study that evaluated postmortem concentrations of phenobarbital in different regions of the brains from 39 patients with epilepsy determined that the concentrations of phenobarbital in all regions of the cortex and cerebellum were largely comparable (45). Phenobarbital concentrations in the frontal cortex were 1.4 times higher and 2.1 times higher than the total and unbound phenobarbital concentrations in serum, respectively.
Unbound phenobarbital distributes into saliva, with saliva:total phenobarbital serum concentration ratios ranging from 0.21 to 0.52 (32,35,46, 47, 48, 49, 50). Two studies have demonstrated that the distribution of phenobarbital into saliva depends on salivary pH (4,51); however, other studies have not found an effect of pH (32,50). In spite of the contradictory results, when equations incorporate the relative ratio of phenobarbital pKa to the salivary pH, there is an excellent correlation between salivary and unbound serum concentrations (4,51).
Phenobarbital distributes across the placenta and accumulates in fetal liver and brain (52, 53, 54). At the time of delivery, infants of mothers who have received phenobarbital have equivalent serum concentrations (55, 56, 57, 58, 59, 60, 61, 62, 63). In one of the studies, the placental passage of phenobarbital was evaluated in three groups of mothers and infants based on duration of phenobarbital treatment (63). Infants of mothers with epilepsy treated for 229±57 days, mothers treated for gestational hypertension and preeclampsia for 10±16 days, and mothers given prophylaxis of intraventricular hemorrhage in premature birth treated for 3.3±21 days were compared. Phenobarbital arterial cord concentrations were 100±2.8%, 89±21%, and 77±16% compared with the maternal concentrations for the three groups, respectively. In addition, a lower percentage of phenobarbital was found in infants with lower cord arterial pH values at birth.
Phenobarbital is secreted into breast milk, with concentrations in breast milk 30% to 40% of maternal serum concentrations (58,64). Phenobarbital is slowly eliminated by the infants, with resulting infant serum concentrations that may reach or exceed maternal concentrations (58,60,65). Excess sedation, poor suckling and bodyweight
gain, and a high incidence of vomiting have resulted in a recommendation by the American Academy of Pediatrics that phenobarbital be used with caution during lactation (66).
ROUTES OF ELIMINATION
Phenobarbital is eliminated from the body by hepatic metabolism and renal excretion of unchanged drug.
In the liver, phenobarbital is metabolized to two major metabolites, p-hydroxyphenobarbital (PBOH), which partially undergoes sequential metabolism to a glucuronic acid conjugate, and 9-D-glucopyra-nosylphenobarbital, an N-glucoside conjugate (PNG). Minor metabolism includes a dihyrodiol, catechol, and p-methylcatechol (Figure 51.1).
Butler (67) first reported the formation of PBOH from phenobarbital in the dog, and it was later confirmed in humans (11,19,68,69) that PBOH was a major metabolite. A substantial fraction of the PBOH is then conjugated with glucuronic acid to form PBOH glucuronide. In studies in patients treated with phenobarbital, an average of 55% (range, 30% to 87%) of the PBOH found was excreted as the O-glucuronide (29,70,71).
There is a large intersubject variability in the fraction of the phenobarbital dose that is metabolized by aromatic hydroxylation to PBOH. Several single-dose studies have reported a range of 8% to 34% of the dose recovered in urine as total PBOH (70,72, 73, 74). Steady-state experiments found that 6% to 40% of the phenobarbital daily dose was total PBOH (70,71,75,76).
Kadar et al. (69) found a wide range of total PBOH excreted in four children (aged 4 to 15 years). Total PBOH accounted for a range of 4.4% to 32.1%, with 38% to 61% conjugated with glucuronic acid. There was no apparent correlation between the age of the child and the percentage of either total or conjugated PBOH. Boreus (75) found that in four neonates, 15% of the phenobarbital dose was total PBOH, with 33% conjugated with glucuronic acid. These authors concluded that the neonates had a decreased ability to conjugate PBOH. However, because of the lack of pharmacologic activity of PBOH, (67,77) this difference would not be clinically significant.
Glucosidation is an uncommon detoxication pathway in mammals; however, 9-D-glucopyra-nosylphenobarbital is a quantitatively significant metabolite of phenobarbital (PNG). Tang et al. (73) administered a mixture of 14C-labeled and 15N-labeled phenobarbital to two normal volunteers. Thin-layer chromatography of 16-day urine identified the N-glucoside conjugate of phenobarbital after comparison with a synthetic standard. In five volunteers, phenobarbital N-glucoside accounted for 26% (range, 24% to 30%) of the phenobarbital dose (39,40,73,74). Bhargava et al. (78) examined randomly collected urine samples from eight patients treated with phenobarbital monotherapy. PNG was detected in all but one patient. Bernus et al. (76) found 14% (range, 0% to 34%) excreted as PNG in 14 patients treated with phenobarbital.
In one study in four children, PNG accounted for 6% to 22.4% of the dose (69). Bhargava and Garrettson (79) also studied the development of phenobarbital metabolism in four neonates by analyzing serial single daily voided urines. The N-glucosidation pathway was not active at birth, and onset had not occurred until after 2 weeks of age. In one infant, by day 20, PNG accounted for 50% of the drug and metabolites in the urine sample. In the other infants, PNG was still not apparent when urine collections ended at days 14 and 16. Subsequent studies have demonstrated that PNG is subject to decomposition under physiologic conditions of temperature and pH. Unless the urine collected for analysis is acidified immediately before storage, decomposition may occur, and the percentage of PNG will be underestimated (80).
The phenobarbital epoxide could also spontaneously or enzymatically yield the corresponding dihydrodiol (Figure 51.1). Harvey et al.(81) were able to obtain gas chromatography-mass spectrometry evidence that the dihydrodiol was present in rat, guinea pig, and human urine in small amounts. Theoretically, oxidation of the dihydrodiol could yield the corresponding catechol. This substance has been tentatively identified in rat and human urine by gas chromatography-mass spectrometry analysis (82). The 4-hydroxy 3-methoxy derivative of phenobarbital (O-methylcatechol) has been isolated from human urine, and its structure was confirmed by comparison with a synthetic standard. In six normal volunteers who received a single dose of phenobarbital, approximately 1% of the dose was recovered as this metabolite (83).
Cytochrome P450 2C9 (CYP2C9) plays a major role in the metabolism of phenobarbital to PBOH with minor metabolism by CYP2C19 and CYP2E1 (84). There are no data on the identity of the uridine diphosphate glucuronosyltransferase enzyme that is responsible for the formation of PNG.
The effect of CYP2C19 polymorphism on the plasma clearance of phenobarbital was studied in a group of Japanese patients receiving phenobarbital treatment (85). Patients who were homozygous for the CYP2C19*3 or CYP2C19*2 genotype (the CYP2C19 poor metabolizers), had a 19% decreased clearance compared with the wild-type homozygous CYP2C19*1. There was also a trend of decreased plasma clearance in heterozygous extensive metabolizers compared with homozygous extensive metabolizers.
A set of twins exhibiting a genetic deficiency in the formation of the metabolite now known as amobarbital N-D-glucopyranoside has been reported (86), a finding suggesting that a deficiency in the enzymes responsible for the formation of PNG may also occur. There is also limited evidence that some patients do not form PNG. Of the three series of patients evaluated for urinary excretion of PNG, two identified patients without PNG. Bhargava et al. (78) identified one of seven patients and Bernus et al. (76) found two of 14 without detectable urinary excretion of PNG.
Biliary and Renal Excretion.
There is considerable intersubject and intrasubject variability in the amount of phenobarbital excreted unchanged in the urine. Single-dose studies (30,31,72, 73, 74,87) in which urine was collected for a minimum of 15 days reported a range of 9% to 33% (average, 23%) of the phenobarbital dose excreted as unchanged drug. Steady-state studies found that the fraction of the dose excreted unchanged in urine averaged 22% (range, 7% to 55%) in four different studies (70,71,76,88).
Boreus et al. (75) showed that the 8-day urinary excretion of unchanged phenobarbital after a single dose in four newborn infants (17%) was similar to that in two adult volunteers (16%). Similarly, in four critically injured children (ages 5, 7, 10, and 15 years), Kadar et al. (69) found a range of 17.8% to 23.1% excreted as unchanged drug.
The renal clearance of phenobarbital depends on both urine flow (29,70,71,89) and urine pH (89), resulting from the lipophilicity and pKa (7.3) of this drug. This phenomenon may explain some of the intersubject variability found in the fraction of dose excreted unchanged in urine. After a drug is filtered by the glomerulus and possibly is actively secreted into the tubule, it may be subject to passive reabsorption. Drug reabsorption takes place primarily in the distal tubule, where the tubule membranes favor the movement of lipid-soluble and un-ionized compounds. The efficient reabsorption of water from the proximal tubule and loop of Henle results in a large concentration gradient between drug in the distal tubule and drug in the plasma. Increasing urine flow decreases this concentration gradient, resulting in a decrease in passive reabsorption. Small changes in urine pH can cause large increases or decreases in the percentage of an un-ionized weak acid (pK1 3.0 to 7.5) such as phenobarbital subject to passive reabsorption.
Waddell and Butler (89) first demonstrated the effect of urine flow and urine pH on phenobarbital renal clearance in an anesthetized dog model in which diuresis was induced. Kapetanovic et al. (71) studied three epileptic patients receiving prolonged phenobarbital therapy. Twenty-four-hour serial urine samples were collected (n = 26). These investigators found a direct linear correlation between urine flow and urinary excretion of unchanged phenobarbital (r = .913) over a fourfold range of urine flow. In a group of 20 epileptic patients, Lous (29) demonstrated the same linear correlation between urine flow and phenobarbital renal clearance but also observed that the phenobarbital renal clearance was independent of phenobarbital concentration in the therapeutic range. The dependence of phenobarbital renal clearance on urine flow and urine pH has provided the basis for the use of urine alkalization and diuresis in patients who have had a phenobarbital overdose (89,90).
CLEARANCE AND HALF-LIFE
The elimination t½ of phenobarbital has been determined in numerous studies and ranges from 75 to 126 hours in healthy subjects and in patients receiving monotherapy (1, 2, 3, 4,5,14,29,91). Total plasma clearance ranges from 2.1 to 4.9 mL/hr/kg (1,2,14,91).
Comedicated Epileptic Patients.
Because phenobarbital is eliminated by multiple pathways (CYP450, N-glucosidation, and renal elimination of unchanged drug), the effect of enzyme inducers should be minimal. In a study of three patients receiving concurrent carbamazepine, the t½ of phenobarbital was 110, 122, and 128 hours, compared with 77, 83, and 98 hours in another three patients receiving concurrent phenytoin. However, neither was significantly different from the t½ in six normal subjects (75 to 126 hours) (1). Valproate significantly decreases the plasma clearance of phenobarbital by inhibiting the formation of both PNG and PBOH (71,88,92). A population study in South African children reported mean phenobarbital clearances of 7.6, 5.0, and 6.8 mL/hr/kg in children receiving phenobarbital monotherapy, polytherapy with valproate, or polytherapy with carbamazepine or phenytoin, respectively (93).
Newborns receiving phenobarbital for the treatment of neonatal seizures have a t½ ranging from 43 to 217 hours (7,21,24,25,94, 95, 96) and a total clearance of 2.7 to 10.7 mL/hr/kg (21,94). Phenobarbital clearance in asphyxia in neonates (22) was significantly lower (4.1 mL/hr/kg) compared with nonasphyxiated neonates (8.7 mL/h/kg). In children receiving phenobarbital, the t½ ranged from 37 to 198 hours (7,25). In a study of nine pediatric patients, the clearance in the children, ages 8 months to 4 years, was significantly greater than that reported in adults and ranged from 5.3 to 14.1 mL/hr/kg (97).
There is no information of the absorption, volume of distribution, or t½ of phenobarbital in elderly patients. Total clearance of phenobarbital was significantly reduced in a group of patients >40 years old (2.5 ml/hr/kg) compared with patients 15 to 40 years old (4.9 mL/hr/kg) (98).
The t½ of phenobarbital was prolonged in a group of patients with liver cirrhosis (130±15 hours) compared with a group of healthy subjects (86±3
hours). In patients with acute hepatitis, the t½ tended to be longer (60 to 127 hours) but was not significantly different from that of controls (99).
RELATIONSHIP BETWEEN SERUM CONCENTRATION AND DOSE
Large population studies have identified age and cotherapy as the primary determinants of the serum concentration:dose ratio. In an evaluation of data obtained from 536 patients receiving phenobarbital monotherapy, Duran et al. (100) found a decrease of the concentration:dose ratio with increasing age from <3 years, 3 to 6 years, 7 to 9 years, 10 to 14 years, and 15 to 18 years, with corresponding increasing concentration to dose ratios of 3.87±0.15, 4.65±0.16, 6.13±0.29, 7.54±0.73, and 10±1.1, respectively. A similar relationship with age was found in studies by Eadie et al. (98) in a group of 121 patients and by Suzuki et al. in a group of 438 patients (101).
Suzuki et al. (101) found that the concentration:dose ratio was higher in patients receiving phenobarbital in combination with other drugs mainly including phenytoin, valproate, and carbamazepine. Similar results were noted by Yukawa et al. (102) in a population study of 349 pediatric and adult patients with epilepsy in whom polytherapy with valproate or carbamazepine increased the concentration:dose ratio. Botha et al. reported that polytherapy with valproate, phenytoin, or carbamazepine resulted in an increase in the concentration:dose ratio (93).
Eadie et al. (98) demonstrated that sex was a determinant of the concentration:dose ratio only for children <5 years of age; boys required a higher dose (mg/kg) than girls. Sex was not found to be a factor in the other studies (100,101).
RELATIONSHIP BETWEEN SERUM CONCENTRATION AND EFFECT
The recommended therapeutic range for phenobarbital is 10 to 40 µg/mL; however, neither the concentration effect nor the concentration-toxicity relationship is well defined (103). The range is based on studies largely in adults (104, 105, 106). Schmidt et al. (106) showed that the therapeutic range varied according to seizure type. Significantly higher phenobarbital plasma concentrations were necessary to control simple or partial complex seizures with or without secondarily generalized seizures (38±6 µg/mL.) than generalized tonic-clonic seizures only (18±10 µg/mL). Tolerability to the sedative adverse effects of phenobarbital alters the relationship of concentration with toxicity. Initial doses at very low phenobarbital concentration can cause excessive sedation and ataxia; however, over several weeks, tolerability to much higher phenobarbital concentrations can occur (107).