Pharmacokinetics of dapsone in children

We studied dapsone pharmacokinetics in eight children with compromised immune function who were receiving three different preparations. Peak serum concentration was less than 0.25 microgram/ml after doses of an extemporaneous liquid preparation but ranged from 0.72 to 1.33 micrograms/ml after initial tablet or proprietary liquid doses and 1.48 to 2.48 microgram/ml during long-term proprietary liquid administration. Elimination followed first-order kinetics; the mean elimination half-life was 15.1 hours.

Simultaneous determination of dapsone, monoacetyldapsone and pyrimethamine in whole blood and plasma by high-performance liquid chromatography

A sensitive, selective and rapid reversed-phase high-performance liquid chromatographic method was developed for the simultaneous analysis of dapsone, monoacetyldapsone and pyrimethamine in human whole blood and plasma. The procedure involved extraction of the compounds and the internal standard, monopropionyldapsone, with tert.-butylmethyl ether under alkaline conditions. A newly marketed column, Supelcosil LC-ABZ (Supelco, 15 cm x 4.6 mm I.D.), was employed. The mobile phase, consisting of acetonitrile-methanol-phosphate buffer (2:1:7, v/v/v), was delivered at a flow-rate of 1.2 ml/min, and ultraviolet absorbance was monitored at 286 nm. The limit of determination using a 150-microliters sample was 10 ng/ml (40 nM) for dapsone and pyrimethamine and 8 ng/ml (28 nM) for monoacetyldapsone. Given that only a small amount of blood is required in this method, it could now be applied in studies involving blood level monitoring and pharmacokinetics in children on Maloprim (dapsone-pyrimethamine) prophylaxis in malaria endemic areas.

Acetylation of dapsone by human whole blood

We studied the acetylation of dapsone (DDS) in vitro by whole blood taken from subjects with known acetylator phenotype. The acetylation of DDS by whole blood was both incubation time- and DDS concentration-dependent. Thus, it is highly recommended to separate plasma immediately after blood withdrawal during acetylation phenotyping using DDS. para-aminobenzoic acid (PABA) substantially inhibited the acetylation of DDS by whole blood taken from both slow and rapid acetylators, while procainamide (PAD) significantly inhibited DDS acetylation by whole blood taken from slow acetylators. At the highest PAD concentration used (208 microM), DDS acetylation by whole blood taken from rapid acetylators was also inhibited. In contrast, sulphanilamide (SAD) failed to produce any significant inhibition of the acetylation of DDS by whole blood taken from either slow or rapid acetylators. Furthermore, there was no correlation between DDS acetylation by whole blood in vitro and the acetylator status of the subject. It is therefore not possible to predict the acetylator phenotype by studying DDS acetylation by human whole blood. These results indicate that the DDS N-acetyltransferase of human whole blood is most probably of the monomorphic type.

The possible risk of bilirubin encephalopathy as predicted by plasma parameters in neonates with previous severe asphyxia

The study group consisted of nine mature newborn infants with a previous history of severe asphyxia and a control group of 18 mature, healthy newborns with the same postnatal age and sex. The object of the investigation was to compare the possible risk of development of bilirubin encephalopathy between the two groups as estimated by plasma parameters. The asphyxia group had a significantly lower reserve albumin concentrations for binding of monoacetyldiaminodiphenyl sulphone (P = 0.008), a measure of binding of unconjugated bilirubin, and significantly lower total albumin concentrations (P = 0.02). No significant difference was observed in unconjugated bilirubin concentration. It is suggested that mature newborns with previous severe asphyxia are at a slightly increased risk of developing bilirubin encephalopathy over and above the well-known risk associated with increased permeability of the blood brain barrier.

Reduced albumin binding of MADDS--a measure of bilirubin binding--during pregnancy and delivery

It was the purpose of the present study to investigate whether the decrease in the ability of plasma albumin to bind bilirubin, noted in pregnancy and during delivery, takes place in one step, during pregnancy alone, or in two steps, during both pregnancy and delivery. Furthermore, it was investigated whether a possible decrease during labor was related to the method of delivery itself. The material comprised (a) 17 pregnant women who delivered vaginally; (b) 25 women who were delivered by Cesarean Section; and (c) a group of 25 non-pregnant women. The reserve albumin concentration for binding of MADDS (a measure of the binding of unconjugated bilirubin), the total albumin concentration and the ratio between them were constant in the 6-week period up to birth and were significantly lower than the corresponding values in the non-pregnant group. During labor a significant decrease in reserve albumin and ratio of reserve albumin to total albumin was observed, while no change in the total albumin concentration was noted. No significant difference in reserve albumin concentration, total albumin concentration and ratio between them was found when the group of mothers who delivered vaginally was compared to the Cesarean Section group. It is concluded that the decrease in the ability of plasma albumin to bind unconjugated bilirubin takes place stepwise. The first step is a significant reduction during pregnancy followed by a further decrease during labor. The method of delivery is of no significance.

The disposition of dapsone in cirrhosis

Acetylation and N-hydroxylation of dapsone were evaluated in drug-free, non-smoking, normal subjects and subjects with cirrhosis (n = 7 for each group) after oral administration of 100 mg dapsone. Acetylation was not correlated with oral dapsone clearance or reduced in cirrhosis (0.37 +/- 0.43 versus 0.52 +/- 0.32). Fractional metabolic clearance of dapsone to its hydroxylamine was associated with dapsone oral clearance (r = 0.96, p less than 0.001, n = 14). In patients with cirrhosis, liver disease was associated with a trend to reduction in oral clearance (22%) and metabolic clearance of dapsone (48%). Protein binding was minimally reduced by cirrhosis (73% +/- 1% versus 69% +/- 3% in patients with cirrhosis (p less than 0.02). The dapsone recovery ratio was validated as a phenotypic index of the metabolic clearance of dapsone (r = 0.74, p less than 0.05). In an extended comparison of 14 patients with cirrhosis to 70 control subjects, cirrhosis was associated with reductions of 28% in dapsone recovery ratio (p less than 0.001), and 37% in acetylation ratio (p less than 0.01). Neither dapsone recovery ratio nor acetylation ratio correlated with Pugh Score, conventional liver function tests, indocyanine green clearance, or phenotypic measures of S-mephenytoin hydroxylase or debrisoquin hydroxylase activity. We conclude that cirrhosis is associated with minor changes in dapsone disposition and that dosage modification is not required. In addition, there is evidence that cirrhosis has a selective influence on activity of individual isozymes of cytochrome P450.

N-acetylation phenotyping using dapsone in a Jordanian population

1. The N-acetylation of dapsone (DDS) was studied in 160 unrelated healthy Jordanian volunteers. 2. The frequency of slow acetylators determined using the plasma monoacetyldapsone (MADDS) to DDS ratio (MADDS/DDS), was 67.5% with a 95% confidence interval of 59 to 76%. Slow acetylators had an acetylation ratio of less than 0.42. 3. Applying the Hardy-Weinberg Law, the frequency of the recessive allele controlling slow acetylation was found to be 0.82 +/- 0.02. 4. The frequency distribution histogram of the plasma MADDS/DDS ratio showed an apparent trimodal pattern. The number of homozygous (n = 16) and heterozygous (n = 36) rapid acetylators derived from the observed data did not agree with those predicted for the respective rapid acetylators (n = 5 and n = 47) according to the Hardy-Weinberg Law. The suggested antimode used to discriminate the two groups was 0.82. 5. The mean plasma concentration of MADDS and the mean plasma acetylation ratio were about three times lower in slow than in rapid acetylators. However, there was no difference in mean plasma DDS concentration between slow and rapid acetylators. 6. There was a significant correlation (r = 0.853, P less than 0.001) between plasma MADDS concentration and the acetylation ratio. For DDS such a correlation was absent (r = 0.059, P = 0.23).

Simultaneous high performance liquid chromatographic determination of dapsone and monoacetyldapsone in human plasma and urine

A rapid, specific and a one-stage protein precipitation method for simultaneous estimation of dapsone (DDS) and monoacetyldapsone (MAD) concentration in plasma and urine using high performance liquid chromatography (HPLC) is described. The applicability of the method for monitoring DDS and MAD blood levels in two different acetylator phenotype volunteers following the administration of 100-mg oral dose of DDS was shown. Cumulative urinary excretion of DDS and MAD were studied in the same volunteers.

Comparison of a urine spot test and blood tests as indicators of patient compliance

Irregular drug intake has been a concern of leprosy control programmes for many years and various methods have been used to monitor and encourage patient compliance. This study compares the results of a urine spot test for dapsone as proposed by Huikeshoven, with blood levels measured in the same patients by the modified Bratton Marshall method and by high performance liquid chromatography. Two hundred-sixty urine and blood specimens were obtained from subjects who were taking supervised and unsupervised medications as well as from controls who were taking no medications. The results indicate that the urine spot test is simple and easily performed, and for monitoring patient compliance under routine clinical conditions (hospital or field work) it compares favourably with blood levels of dapsone estimated by the Bratton Marshall method or by high performance liquid chromatography. The study also shows that dapsone level is not a good indicator of compliance in patients who are also taking daily rifampicin but the urine spot test remains useful in such patients.

Multiple-dose pharmacokinetics and in vitro antimalarial activity of dapsone plus pyrimethamine (Maloprim) in man

1. The multiple-dose kinetics of dapsone (DDS), its major metabolite monoacetyldapsone (MADDS) and pyrimethamine (PYR) were studied in six healthy adult male volunteers following weekly administration of Maloprim (100 mg DDS plus 12.5 mg PYR). 2. After the last maintenance dose of Maloprim, the following kinetic parameters (mean values) were determined for DDS and PYR, respectively: maximum plasma concentration (Cmax) = 1,134 and 116 ng ml-1; elimination half-life (t1/2) = 23 and 105 h; plasma clearance (CL) = 37.6 and 15.9 ml h-1 kg-1 and apparent volume of distribution (Vss) = 1.20 and 2.29 l kg-1. The mean t1/2 of MADDS was 22 h. 3. The mean whole blood to plasma (B/P) and erythrocyte to plasma (E/P) concentration ratios for DDS were 1.04 and 1.09, respectively. MADDS had a B/P ratio of 0.69 and an E/P ratio of 0.33. The B/P and E/P ratios for PYR were 0.98 and 0.54, respectively. 4. The drug combination was assessed in vitro by measuring inhibition of re-invasion of two Plasmodium falciparum isolates grown in the presence of volunteers' sera. The chloroquine (CQ)- and PYR-sensitive FC-27 isolate was completely inhibited by the sera but the drug combination was ineffective against the CQ- and PYR-resistant K1 strain. The in vitro findings suggest that Maloprim may not be effective against strains of P. falciparum with a high level of resistance to pyrimethamine.

Multiple-dose kinetics in healthy volunteers and in vitro antimalarial activity of proguanil plus dapsone

The multiple-dose kinetics of a daily dose of proguanil (200 mg) coadministered with dapsone (10 mg) was investigated in 6 healthy adult male volunteers. The kinetics of dapsone (DDS), monoacetyldapsone (MADDS), proguanil (PROG) and its active metabolite cycloguanil (CYCLO) were derived from plasma drug concentrations after the last maintenance dose. The following kinetic parameters (mean values) were estimated for DDS and PROG, respectively: maximum concentration (Cmax) = 285 and 151 ng/ml, minimum concentration (Cmin) = 125 and 31 ng/ml, elimination half-life (t1/2) = 23.3 and 18.3 h, plasma clearance (Cl) = 0.032 and 1.27 l/h/kg and apparent volume of distribution (Vss) = 1.05 and 33.32 l/kg. The Cmax, Cmin and t1/2 of CYCLO were 56 ng/ml, 17 ng/ml and 15.0 h, respectively. The antimalarial activity of the proguanil/dapsone combination was assessed in vitro by measuring the inhibition of re-invasion of two Plasmodium falciparum isolates grown in the presence of volunteers' sera. Both FC-27 [chloroquine (CQ)- and pyrimethamine (PYR)-sensitive] and K1 (CQ- and PYR-resistant) isolates were completely inhibited by the drug combination at steady-state concentrations. These findings suggest that the drug regimen may be effective against drug-resistant falciparum malaria.

Genetically determined N-acetylation and oxidation capacities in Japanese patients with non-occupational urinary bladder cancer

Genetically determined polymorphisms of N-acetylation and oxidative capacity have been studied using dapsone and metoprolol in 51 Japanese patients with spontaneous bladder cancer and 203 healthy control subjects. The results for N-acetylation pharmacogenetics were against the initial expectation that there would be a preponderance of slow acetylators in the cancer group, as 3 such patients (5.9%) were found as compared to 13 (6.4%) in the healthy group. There was no poor metabolizer (PM) of metoprolol in the cancer group, whereas in the healthy group one (0.5%) was a PM. There were no significant differences between the groups in the frequency of slow acetylator and poor oxidiser phenotypes, or in the frequency distribution profiles of acetylation (monoacetyldapsone/dapsone) and oxidative metabolic ratio (log metoprolol/alpha-hydroxymetoprolol). The results indicate that neither N-acetylation nor the debrisoquine/sparteine-type oxidative phenotype and/or capacity represent a genetic predisposition to spontaneous bladder carcinogenesis in Japanese patients. In the normal Japanese population there is a great predominance of rapid acetylators and extensive oxidisers.

Effect of cancer chemotherapy on dapsone N-acetylation in man

The effect of cancer chemotherapy on dapsone (DDS) N-acetylation was explored in 28 patients with various malignancies. There was a significant (P = 0.01) increase in plasma monoacetydapsone/dapsone (MADDS/DDS) ratios within 24 h of the start of chemotherapy (CT), indicating an acceleration of N-acetylation.

Drug binding properties of neonatal albumin

Neonatal and adult albumin was isolated by gel chromatography on Sephacryl S-300, from adult and umbilical cord serum, respectively. Binding of monoacetyl-diamino-diphenyl sulfone, warfarin, sulfamethizole, and diazepam was studied by means of equilibrium dialysis and the binding data were analyzed by the method of several acceptable fitted curves. It was found that the binding affinity to neonatal albumin is less than to adult albumin for monoacetyl-diamino-diphenyl sulfone and warfarin. Sulfamethizole binding to the neonatal protein is similarly reduced when more than one molecule of the drug is bound per albumin molecule, and binding of the first sulfamethizole molecule is possibly reduced as well. Diazepam binds with equal affinity to the fetal and adult proteins. Among the two main albumin drug-binding functions, for warfarin and diazepam, the former is thus compromised in the newborn infant while the diazepam binding function is at the adult level.

Dapsone, trimethoprim, and sulfamethoxazole plasma levels during treatment of Pneumocystis pneumonia in patients with the acquired immunodeficiency syndrome (AIDS). Evidence of drug interactions

STUDY OBJECTIVE

To examine the interaction between dapsone and trimethoprim in patients with the acquired immunodeficiency syndrome (AIDS).

DESIGN

Measurement of drug levels as part of an open study of dapsone alone and randomized, double-blind comparison of trimethoprim-dapsone with trimethoprim-sulfamethoxazole in treating Pneumocystis carinii pneumonia in patients with AIDS.

SETTING

County hospital and AIDS clinic.

PATIENTS

Eighteen patients treated with dapsone alone, 30 with trimethoprim-dapsone, and 30 with trimethoprim-sulfamethoxazole.

INTERVENTION

Dapsone, 100 mg/d; trimethoprim, 20 mg/kg body weight per day, and sulfamethoxazole, 100 mg/kg.d; administered for 21 days.

MEASUREMENTS AND MAIN RESULTS

Concentrations of dapsone were 40% higher in patients treated with trimethoprim-dapsone than in those treated with dapsone alone (2.1 compared with 1.5 micrograms/mL; P less than 0.05). Trimethoprimdapsone-treated patients had fewer treatment failures but more side effects and treatment terminations due to toxicity than those treated with dapsone alone. The concentration of trimethoprim was 48.4% higher in patients treated with trimethoprim-dapsone than in those treated with trimethoprim-sulfamethoxazole, (18.4 compared with 12.4 micrograms/mL; P less than 0.05). Discontinuation of therapy due to toxicity was commoner in the trimethoprim-sulfamethoxazole group (57% compared with 30%).

CONCLUSIONS

A bidirectional drug interaction exists between dapsone and trimethoprim, resulting in higher concentrations of each in the presence of the other.

Humoral immune responses in dapsone treated M. leprae infected mice

Humoral responses in M. leprae infected mice were studied through 52 weeks and were found to be directly related to the bacterial load. However, treatment with dapsone (DDS) in the last 12 weeks of infection resulted in an initial enhancement of the humoral responses followed by a gradual decrease, though they were still significantly high at the end of the study.

Plasma dapsone and its metabolite monoacetyldapsone levels in leprotic patients

Dapsone (DDS) is a drug of choice in the treatment of leprosy. The DDS and MADDS levels and different pharmacokinetic parameters after a single dose of dapsone and at steady state in leprotic patients have been studied. At steady state, all the patients showed plasma DDS and MADDS levels above 0.5 micrograms/ml throughout the 24-h duration. There was no significant difference in the elimination half-lives of DDS and MADDS after a single dose as compared to at steady state, but AUC0-alpha for both DDS and MADDS were significantly increased at steady state. From these results, it could be concluded that 100 mg daily dose is sufficient to maintain plasma therapeutic concentration in leprotic patients in Indian population.

Influence of acetylator phenotype of the leprosy patient on the emergence of dapsone resistant leprosy

The half time of disappearance of dapsone and monoacetyl dapsone and the acetylator phenotype of the leprosy patients who harboured dapsone sensitive and dapsone resistant M. leprae was assessed in 27 subjects. Sixteen patients were rapid acetylators, five were slow and six were intermediate acetylators. The mean T 1 1/2 lives of dapsone (30.26 +/- 11.0) and monoacetyl dapsone (31.11 +/- 12.0) were also studied in the above patients. The percentage of different acetylators in both resistant and sensitive groups were similar showing no correlation between the emergence of drug resistance and the phenotype of the patient. The mean time of disappearance of DDS and MAD in the different acetylators did not show significant difference. The ratios of MAD/DDS in an individual at 3, 6 or 24 hours after the dose were similar. The mean T 1 1/2 lives of DDS and MAD in resistant and sensitive patients also showed no difference. Neither T 1 1/2 lives of DDS or MAD nor the acetylator phenotype seem to influence the emergence of dapsone resistance.

N-acetylation polymorphism of dapsone in a Japanese population

1. The N-acetylation of dapsone (DDS) was studied in 182 unrelated healthy Japanese subjects. The frequency of slow acetylators determined using the plasma monoacetyldapsone (MADDS) to DDS ratio (MADDS/DDS, slow acetylators less than 0.30 and rapid acetylators greater than 0.35) at 3 h after an oral dose of DDS (100 mg) was 6.6% (12 of the 182 subjects) with a 95% confidence interval of 3.8 to 11.2%. 2. The frequency distribution histogram of the plasma MADDS/DDS ratio showed an apparent trimodal pattern. However, the numbers of heterozygous (n = 105) and homozygous rapid acetylators (n = 65) derived from the observed data did not agree with those predicted for the respective rapid acetylators (n = 70, and n = 100) by applying the Hardy-Weinberg Law, when the suggested antimode of 0.85 discriminating these two rapid acetylators was employed. 3. The incidence of slow acetylators was unexpectedly lower in the males (1.4%, 1 of the 69 subjects, with a 95% confidence interval of 0.2 to 7.7%) compared with the incidence in the females (9.7%, 11 of the 113 subjects, with a 95% confidence interval of 5.5 to 16.6%). The difference reached a marginally significant level (Fisher's exact probability test, P = 0.02). 4. The mean plasma concentration of MADDS was significantly (P less than 0.001) lower in the slow compared to the rapid acetylators and there was a highly significant correlation (rs = 0.757, P less than 0.001) between plasma MADDS levels and MADDS/DDS ratios. 5. Slow acetylators showed a significantly (P less than 0.001) lower urinary MADDS/DDS ratio and excreted less (P less than 0.001) MADDS than rapid acetylators.(ABSTRACT TRUNCATED AT 250 WORDS)

Metabolism of dapsone to a hydroxylamine by human neutrophils and mononuclear cells

Dapsone is an effective anti-inflammatory agent in conditions in which inflammation is mediated by neutrophils. Dapsone also has been associated with agranulocytosis. We found that neutrophils, which had been activated by a phorbol ester or opsonized zymosan, oxidized dapsone to its nitroderivative. It appears as if this is due to oxidation of dapsone by myeloperoxidase to the hydroxylamine, followed by nonenzymatic oxidation of the hydroxylamine to the nitroderivative. The hydroxylamine can be isolated if ascorbic acid is added to the incubations. Monocytes also contain myeloperoxidase and activated mononuclear leukocytes also metabolize dapsone to the hydroxylamine. Dapsone also causes a mononucleosis-like syndrome. The reactive hydroxylamine could be responsible for both the pharmacologic and toxic properties of dapsone.

Influence of once-monthly rifampicin and daily clofazimine on the pharmacokinetics of dapsone in leprosy patients in Nigeria

In leprosy patients in Nigeria the influence of daily clofazimine and of once-monthly rifampicin on the pharmacokinetics of dapsone has been investigated. Three days after rifampicin the elimination half-life of dapsone was reduced from 40.4 to 25.3 h (n = 23). Correspondingly, the plasma dapsone 24 h after the last dose had fallen significantly from 2.63 to 2.02 mg/l. Clofazimine did not cause change in the pharmacokinetics of dapsone. It was concluded that, although rifampicin had a considerable influence on the pharmacokinetics of dapsone, there is no reason to adjust the dose of dapsone during multidrug therapy of leprosy.

Role of dapsone hydroxylamine in dapsone-induced hemolytic anemia

The hemolytic anemia which frequently accompanies treatment of individuals with dapsone and other arylamine drugs is believed to be caused not by the parent drugs per se, but rather by metabolites which are formed during the clearance of the drugs in vivo. To determine whether the N-hydroxyarylamine metabolites of dapsone could be responsible for dapsone-induced hemolysis, dapsone, dapsone hydroxylamine (DDS-NOH) and monoacetyldapsone hydroxylamine were administered to rats which had previously received 51Cr-labeled red blood cells. All three compounds caused an increase in the rate of disappearance of radioactivity from the blood as compared with saline-treated controls. In parallel in vitro studies, incubation of 51Cr-labeled red blood cells with DDS-NOH, but not dapsone or monoacetyldapsone, induced a decrease in survival time of the radiolabeled cells when they were reintroduced into isologous rats. The disappearance of radioactivity from the blood was matched by its selective uptake into the spleen. The amount of damage (as measured by decreased red cell survival in vivo) was proportional to both concentration and time of exposure to DDS-NOH. The area under the blood concentration vs. time curve for total arylhydroxylamines (DDS-NOH + monacetyldapsone hydroxylamine) in rats given a hemotoxic dose of dapsone was similar to that of rats given an equitoxic dose of DDS-NOH. Collectively, these data indicate that the hydroxylamine metabolites of dapsone are direct acting hemolytic agents that are formed from dapsone in sufficient amounts to account for their being the sole mediators of dapsone-induced hemolytic anemia in the rat.

Albumin binding of MADDS--a measure of bilirubin binding--in women during pregnancy and after delivery and in their infants

The purpose of this study was to investigate the binding potential of MADDS (monoacetyldiaminodiphenyl sulphone) to albumin, a measure for binding of unconjugated bilirubin, in healthy women during pregnancy, during and after delivery, and in their infants. The serum concentrations of unconjugated bilirubin, reserve albumin for binding of MADDS and total albumin were measured in: (a) 21 non-pregnant women; (b) 16 pregnant women in the 16th-24th, 28th-32nd, and 36th-38th gestational weeks, and at the time of delivery from both mother and infant; and (c) 15 women at the time of delivery, and 24 and 72 hours after delivery. The bilirubin concentrations did not change during pregnancy or at delivery and were very small compared with the concentrations of reserve albumin for binding of MADDS and total albumin. Therefore, the ratio of reserve albumin to total albumin was an expression of the binding potential of the albumin for MADDS in women. During pregnancy, the reserve albumin decreased equal to total albumin, so that the ratio was not significantly changed (p greater than 0.05). In contrast, at delivery the reserve albumin was significantly lower (p less than 0.01) than in the 36th-38th gestational weeks, without any significant difference in total albumin, i.e. the ratio was significantly lower than during pregnancy (p less than 0.01). During the first 3 days after delivery the reserve albumin increased significantly (p less than 0.05), also without any significant change in total albumin, so that the ratio increased significantly (p less than 0.05) and was normalized. The binding potential for MADDS to albumin in newborn infants was even lower than that of their mothers. Since the binding-potential is reduced in both mother and infant at delivery, a relation is suggested.

Ceftriaxone effect on bilirubin-albumin binding

The effect of ceftriaxone on bilirubin-albumin binding was measured in vitro using the peroxidase method with human serum albumin and a dialysis rate method with adult and newborn serum. Ceftriaxone competes with bilirubin for binding to human serum albumin; the displacement constant is 1.5 X 10(4) L/mol. Therapeutic levels of ceftriaxone decrease the reserve albumin concentration in newborn serum by 39%. These results indicate that ceftriaxone may increase the risk of bilirubin encephalopathy in jaundiced premature infants.