- Email: [email protected]
The effect of paraben preservatives on albumin binding of bilirubin
PEDIATRIC PHARMACOLOGY AND THERAPEUTICS Paul S. Lietman,Editor
The effect of paraben preservatives on albumin binding of bilirubin The interaction of methylparaben and propylparaben with bilirubin-albumin complexes was studied using difference speetra, Sephadex gel filtration, red blood cell uptake of bilirubin, and the peroxidase assay. Methylparaben was found to be a weak eompetitor with bilirubin for binding to primary albumin-binding sites but a strong binding competitor (similar to sulfisoxazole) at secondary sites. The displacing effects of methylparaben and sulfisoxazole were additive. Propylparaben bound to albumin but did not displace bilirubin. Drugs and injectable saline and water preparations which contain methylparaben should be avoided in jaundieed newborn infants when the high-affinity albumin-binding sites approach saturation.
L. Fraser Rasmussen, M.S., Charles E. Ahlfors, M.D., and Richard P. W e n n b e r g , M.D.,* S a c r a m e n t o , Calif.
D R U G P R E S E R V A T I V E S methylparaben a n d propylparaben have recently been reported to produce a marked displacement of bilirubin from albumin/ Parabens are present in many multidose preparations including gentamicin sulfate and bacteriostatic water and saline and therefore could potentiate kernicterus in jaundiced infants receiving intravenous drugs and saline infusions. This study analyzes the interaction of methyl- and propylparaben with bilirubin-albumin complexes and compares their effect with the well-known binding competitor sulfisoxazole. THE
MATERIALS AND METHODS Methyl- and propylparaben (methyl and propyl esters of p-hydroxybenzoic acid) were obtained from the University of Washington Drug Service Department, Seattle, Washington. Human serum albumin (Fraction V) and crystalline bilirubin were purchased from Sigma Chemical Co., St. Louis, Missouri. The bilirubin had a millimolar extinction coefficient (1 cm pathlength) of 60.0 in chloroform, contained approximately 10% isomers, and From the Department of Pediatrics, Sacramento Medical Center, University of California, Davis. Supported by a grant from the United Cerebral Palsy Foundation. *Reprint address: Sacramento Medieal Center, Department of Pediatrics, 4301 X St. Sacramento, Calif 9581Z
was used without further purification. Bilirubin was dissolved in 0.1N NaOH, diluted to the desired concentration with distilled water, titrated to pH 9.0 with dilute HC1, and immediately added to purified human serum albumin (model serum) or adult human plasma dissolved or diluted in distilled water. The icteric serum or plasma was buffered with an equal volume of 0.11M phosphate buffer, pH 7.4, giving a final ionic strength of 0.15. All procedures were performed in subdued light to prevent photooxidation.
See related articles, pp. 479 and 483. Shifts in the absorption spectrum of bilirubin will occur when bilirubin is displaced from albumin. 2 The absorption spectrum of "free" bilirubin in aqueous solution at pH 7.4 has a maximum at 436-440 nm whereas albuminbound bilirubin produces a maximum at 458-462 nm. We performed difference spectroscopy to detect small spectral changes induced by drugs. The difference spectrum of unbound bilirubin (sample) versus bound bilirubin (reference) produces a maximum at 410 nm and a minimum at 485 nm. Experimentally, model serum was placed in sample and reference cuvettes; parabens were then added to the sample cuvette and an equal quantity of water added to the reference so that the concentration of bilirubin was identical in both cuvettes. Difference
The Journal o f P E D I A T R I C S Vol. 89, No. 3, pp. 475-478
475
476
Rasmussen,A hlfors, and Wennberg
B
+ .04
+
The Journal of Pediatrics September 1976 Table 1. Peroxidase assay
E
Drug
Unbound biBrubin (nmol/1 • SE)
Control Sulfisoxazole (500 ffM) Methylparaben (500/xM) Propylparaben (500/zM)
9.9 + 0.4 47.5 • 4. l 25.2 _+ 8.5 6.2• 4.1
Molar ratio bilirubin /albumin
03
+ .02 -
0.8
g
+.01
o '~ -.01
Table II. Sephadex G-25 gel filtration
t Bilirubin concentration
-.02
- .03-
- 04
Initial* I (mg/dl)
Drug (3/) I~
340
I
410
I
485
I
550
Wave Length (nm) Fig. 1. Perturbations of the bilirubin-albuminabsorption spectrum as measured by difference spectra. A, Baseline; conditions are identical in sample and reference. B, The spectrum of unbound vs. bound bilirubin; biiirubin 4.0 ffmol/1 in sample and reference; sample has no albumin but reference contains 8.0 ffmol/l albumin. C, Methylparaben 41.9 ffg/ml in sample; both cuvettes contain 20 ffmol/1 bilirubin and 30 pmol/l albumin (molar ratio 0.67). D, Propylparaben 6.45 fig/rot in the sample (molar ratio 0.67). E, Propylparaben 6.45 ~g/mI in sample, albumin 15 ~mot/1 (molar ratio 1.33). Difference spectra induced by methyl- and propylparabens are ahnost mirror images, with methylparaben producing a shift consistent with increased dissociation of bilirubin from albumin.
absorption spectra were performed at 27~ using a Cary model 16 recording spectrophotometer. Unbound bilirubin concentrations were determined using the peroxidase assay described by Jacobsen and Wennberg? The relative affinity of albumin and alternative substrates for bilirubin binding was evaluated by measuring red blood cell uptake 4 (using bilirubinenriched human plasma) and by Sephadex G-25 gel filtration~ (using both model serum and plasma). Slight modifications of these methods were used as previously described. 6 RESULTS Methylparaben and sulfisoxazole produced difference spectra, indicative of displacement at bilirubin-albumin ratios of 0.67 (Fig. 1) and 1.3. When compared to the difference spectrum of "free" versus bound bilirubin, methylparaben produced a decrease in absorbance a~ 485 nm which was disproportionately greater than the
Control Methylparaben (250 filM) Propylparaben
%
Adsorbed (p~mol/l)
Change from control
(4) (3)
11.4 11.4
2.00 • 0.04 3.51 • 0.02
+75.5
(3)
11.4
1.90 -+ 0.06
--5.0
(3)
11.4
2.47 _+ 0.04
+ 23.5
(3)
11.4
3.83 • 0.04
+91.7
(250 ~M) Sulfisoxazole (250 ffM) Methylparaben (250 ffM) + sulfisoxazole (250 IzM) *Molar ratio = 1.6. increase at 410 nm, suggesting noncompetitive as well as competitive binding to the albumin.1 Propylparaben produced an increase rather than a decrease in absorbance at 485 nm with a decrease at 420 nm. This spectrum is compatible with noncompetitive binding to albumin and/or atigmentation of bilirubin binding. When both methyl- and propylparaben were added to the model serum, the difference spectrum represented the algebraic sum of their respective influences. The peroxidase assay demonstrated a marked acceleration in bilirubin oxidation in the presence of both methylparaben and propylparaben. Further invesugation i,dicated that much of the increased oxidation reaction induced by high concentrations of parabens was due to accelerated oxidation of bound bilirubin. Normally the peroxidase assay requires only a small correction for oxidation of bound bilirubin3; the presence of parabens required a large correction factor, whereas sulfisoxazole increased the oxidation of bound bilirubin only slightly. Corrected values for unbound bilirubin indicate greater displacement at the primary binding site by sulfisoxazole than by methylparaben, and no displacement by propylparaben (Table I).
Volume 89 Number 3
Secondary site binding was evaluated by Sephadex G-25 gel filtration. Methylparaben displaced more bilirubin than did sulfisoxazole and the effects of the drugs were additive (Table II). Propylparaben and control samples produced similar results. Red blood cell uptake of bilirubin was increased by both methylparaben and sulfisoxazole at a bilirubin/ albumin molar ratio of 1.6 (Table III). Again, the displacing effects of sulfisoxazole and methylparaben were additive. Displacement of bilirubin at a 0.8 molar ratio was observed only with sulfisoxazole. Propylparaben produced no increase in red blood cell uptake of bilirubin at either molar ratio. Because of the marked difference in disp]acing properties of methyl- and propylparabens, several benzene derivatives were comPared (Table IV). Methylparaben was the most potent binding competitor, whereas ethylparaben and benzoic acid were moderately effective displacers of bilirubin. In three of four repeated experiments (1.6 molar ratio) propylparaben appeared to augment binding. Nonesterified p-hydroxyben: - acid produced no bilirubin displacement. DISCUSSION These data indicate that both methyl- and propylparaben bind to albumin, but that only methylparaben displaces bilirubin from albumin. Methylparaben appears to be a weak primary site competitor, but a strong competitor (similar to sulfisoxazole) when bilirubin is bound to secondary binding sites. Propylparaben has a unique effect on albumin binding of bilirubin; our data suggest a slight enhancement of bilirubin binding. Methylparaben is present in a wide variety of multidose pharmaceutical preparations frequently administered t o sick premature infants including some preparations of injectable saline and bacteriostatic water. It is common practice to flush umbilical artery catheters with I to 3 ml saline fc!iowing blood sampling, and a rapid infusion of 5 or even 10 ml is not uncommon. A single intra-artcrial flush of 2.0 ml saline (containing 2.4 mg methylparaben) in a one kilogram infant could acutely raise the plasma paraben concentration to 342 /*mol/1 (assuming perfect mixing in a 50 ml/kg plasma volume). Methylparaben at this concentration would be expected to compete with bilirubin for binding when the high-affinity binding sites on albumin approach saturation (see Table III). Since the equilibrium of bilirubin with tissue and albumin occurs very rapidly, the administration of methylparaben via catheter flushes is potentially dangerous when the highaffinity binding sites on albumin approach saturation. In such patients it would be prudent to use single dose vials
Effect o f paraben on bilirubin-albumin binding
477
Table IlL Erythrocyte uptake of bilirubin RBC uptake of bilirubin (l~mol/l ++-SE) o.8 molar ratio*
l. 6 molar ratio*
% Change from control
5.4 _+ 0.12 5.6 + 0.04 5.4 _+ 0.43 6.7 _+ 0.03 --
17.7 _+ 0.06 17.3 _+ 0.03 26.7 _ 0.33 26.8 +__0.23 38.0 _+ 0.30
-2 +5t +51 + 114
28.3 _+ 0.09
+ 59
Drug (~tmol/l) Control Propylparaben (250) Methylparaben (250) Sulfisoxazole (250) Sulfisoxazole (250) + methylparaben (250) Sulfisoxazole (250) + propylparaben (250)
*Albuminconcentration= 1.7 gm/dl. Table IV. Effect of benzene compounds on erythrocyte uptake of bilirubin*
Drug (Ixmol/l) Control Propylparaben (250) p-OH benzoic acid (250) Ethylparaben (250) Benzoic acid (250) Methylparaben (250)
RBC uptake of bilirubin (l~mol/l + SE) 17.8 _~ 0.2 15.5 + 0.1 16.5 _+ 0.1 21.3 _+ 0.8 22.1 _+ 0.2 25.6 + 0.2
% Change from control --13 --7 +20 + 24 +43
*Molar ratio bilirubin/albumin1.6;duplicatesamples. of injectable saline and water or to select multidose preparations containing other bacteriostatic agents. CONCLUSION Methylparaben is a potent displacer of bilirnbin from albumin when bilirubin is bound to secondary binding sites and a weak competitor with bilirubin for binding to primary binding sites. When high-affinity binding sites become saturated, the displacing potential of methylparaben is similar to that of sulfisoxazole and greater than that of sodium benzoate. The contribution of various competitors is cumulative and must be taken into consideration when administering multiple drugs to sick jaundiced infants. The authors are grateful for the skilled technical assistance of Ms. Nadja Kronlund. REFERENCES
1. Brodersen R: Competitive binding of bilirubin and drugs to human serum studied by enzymatic oxidation, J Clin Invest 54:1353, 1974.
478
2.
Rasmussen, Ahlfors, and Wennberg
Odell GB, Cohen SN, and Kelly PC: Studies in kernicterus: II. The determination of the saturation of serum albumin with bilirubin, J PEDIATR74:214, 1969. 3. Jacobsen J, and Wennberg RP: Determination of unbound bilirubin in the serum of newborns, Clin Chem 20:782, 1974. 4. Bratlid D: Bilirubin binding by human erythrocytes, Scand J Clin Invest 29:91, 1972.
The Journal of Pediatrics September 1976
5. Shift D, Chan G, and Stern L: Sephadex G-25 quantitative estimation of free bilirubin potential in jaundiced newborn infants' sera: A guide to the prevention of kernicterus, J Lab Clin Med 80:455, 1972. 6. Wennberg RP, and Rasmussen LF: Effects of gentamicin on albumin binding of bilirubin, J PEDIATR 86:611, 1975.
The effect of paraben preservatives on albumin binding of bilirubin The interaction of methylparaben and propylparaben with bilirubin-albumin complexes was studied using difference speetra, Sephadex gel filtration, red blood cell uptake of bilirubin, and the peroxidase assay. Methylparaben was found to be a weak eompetitor with bilirubin for binding to primary albumin-binding sites but a strong binding competitor (similar to sulfisoxazole) at secondary sites. The displacing effects of methylparaben and sulfisoxazole were additive. Propylparaben bound to albumin but did not displace bilirubin. Drugs and injectable saline and water preparations which contain methylparaben should be avoided in jaundieed newborn infants when the high-affinity albumin-binding sites approach saturation.
L. Fraser Rasmussen, M.S., Charles E. Ahlfors, M.D., and Richard P. W e n n b e r g , M.D.,* S a c r a m e n t o , Calif.
D R U G P R E S E R V A T I V E S methylparaben a n d propylparaben have recently been reported to produce a marked displacement of bilirubin from albumin/ Parabens are present in many multidose preparations including gentamicin sulfate and bacteriostatic water and saline and therefore could potentiate kernicterus in jaundiced infants receiving intravenous drugs and saline infusions. This study analyzes the interaction of methyl- and propylparaben with bilirubin-albumin complexes and compares their effect with the well-known binding competitor sulfisoxazole. THE
MATERIALS AND METHODS Methyl- and propylparaben (methyl and propyl esters of p-hydroxybenzoic acid) were obtained from the University of Washington Drug Service Department, Seattle, Washington. Human serum albumin (Fraction V) and crystalline bilirubin were purchased from Sigma Chemical Co., St. Louis, Missouri. The bilirubin had a millimolar extinction coefficient (1 cm pathlength) of 60.0 in chloroform, contained approximately 10% isomers, and From the Department of Pediatrics, Sacramento Medical Center, University of California, Davis. Supported by a grant from the United Cerebral Palsy Foundation. *Reprint address: Sacramento Medieal Center, Department of Pediatrics, 4301 X St. Sacramento, Calif 9581Z
was used without further purification. Bilirubin was dissolved in 0.1N NaOH, diluted to the desired concentration with distilled water, titrated to pH 9.0 with dilute HC1, and immediately added to purified human serum albumin (model serum) or adult human plasma dissolved or diluted in distilled water. The icteric serum or plasma was buffered with an equal volume of 0.11M phosphate buffer, pH 7.4, giving a final ionic strength of 0.15. All procedures were performed in subdued light to prevent photooxidation.
See related articles, pp. 479 and 483. Shifts in the absorption spectrum of bilirubin will occur when bilirubin is displaced from albumin. 2 The absorption spectrum of "free" bilirubin in aqueous solution at pH 7.4 has a maximum at 436-440 nm whereas albuminbound bilirubin produces a maximum at 458-462 nm. We performed difference spectroscopy to detect small spectral changes induced by drugs. The difference spectrum of unbound bilirubin (sample) versus bound bilirubin (reference) produces a maximum at 410 nm and a minimum at 485 nm. Experimentally, model serum was placed in sample and reference cuvettes; parabens were then added to the sample cuvette and an equal quantity of water added to the reference so that the concentration of bilirubin was identical in both cuvettes. Difference
The Journal o f P E D I A T R I C S Vol. 89, No. 3, pp. 475-478
475
476
Rasmussen,A hlfors, and Wennberg
B
+ .04
+
The Journal of Pediatrics September 1976 Table 1. Peroxidase assay
E
Drug
Unbound biBrubin (nmol/1 • SE)
Control Sulfisoxazole (500 ffM) Methylparaben (500/xM) Propylparaben (500/zM)
9.9 + 0.4 47.5 • 4. l 25.2 _+ 8.5 6.2• 4.1
Molar ratio bilirubin /albumin
03
+ .02 -
0.8
g
+.01
o '~ -.01
Table II. Sephadex G-25 gel filtration
t Bilirubin concentration
-.02
- .03-
- 04
Initial* I (mg/dl)
Drug (3/) I~
340
I
410
I
485
I
550
Wave Length (nm) Fig. 1. Perturbations of the bilirubin-albuminabsorption spectrum as measured by difference spectra. A, Baseline; conditions are identical in sample and reference. B, The spectrum of unbound vs. bound bilirubin; biiirubin 4.0 ffmol/1 in sample and reference; sample has no albumin but reference contains 8.0 ffmol/l albumin. C, Methylparaben 41.9 ffg/ml in sample; both cuvettes contain 20 ffmol/1 bilirubin and 30 pmol/l albumin (molar ratio 0.67). D, Propylparaben 6.45 fig/rot in the sample (molar ratio 0.67). E, Propylparaben 6.45 ~g/mI in sample, albumin 15 ~mot/1 (molar ratio 1.33). Difference spectra induced by methyl- and propylparabens are ahnost mirror images, with methylparaben producing a shift consistent with increased dissociation of bilirubin from albumin.
absorption spectra were performed at 27~ using a Cary model 16 recording spectrophotometer. Unbound bilirubin concentrations were determined using the peroxidase assay described by Jacobsen and Wennberg? The relative affinity of albumin and alternative substrates for bilirubin binding was evaluated by measuring red blood cell uptake 4 (using bilirubinenriched human plasma) and by Sephadex G-25 gel filtration~ (using both model serum and plasma). Slight modifications of these methods were used as previously described. 6 RESULTS Methylparaben and sulfisoxazole produced difference spectra, indicative of displacement at bilirubin-albumin ratios of 0.67 (Fig. 1) and 1.3. When compared to the difference spectrum of "free" versus bound bilirubin, methylparaben produced a decrease in absorbance a~ 485 nm which was disproportionately greater than the
Control Methylparaben (250 filM) Propylparaben
%
Adsorbed (p~mol/l)
Change from control
(4) (3)
11.4 11.4
2.00 • 0.04 3.51 • 0.02
+75.5
(3)
11.4
1.90 -+ 0.06
--5.0
(3)
11.4
2.47 _+ 0.04
+ 23.5
(3)
11.4
3.83 • 0.04
+91.7
(250 ~M) Sulfisoxazole (250 ffM) Methylparaben (250 ffM) + sulfisoxazole (250 IzM) *Molar ratio = 1.6. increase at 410 nm, suggesting noncompetitive as well as competitive binding to the albumin.1 Propylparaben produced an increase rather than a decrease in absorbance at 485 nm with a decrease at 420 nm. This spectrum is compatible with noncompetitive binding to albumin and/or atigmentation of bilirubin binding. When both methyl- and propylparaben were added to the model serum, the difference spectrum represented the algebraic sum of their respective influences. The peroxidase assay demonstrated a marked acceleration in bilirubin oxidation in the presence of both methylparaben and propylparaben. Further invesugation i,dicated that much of the increased oxidation reaction induced by high concentrations of parabens was due to accelerated oxidation of bound bilirubin. Normally the peroxidase assay requires only a small correction for oxidation of bound bilirubin3; the presence of parabens required a large correction factor, whereas sulfisoxazole increased the oxidation of bound bilirubin only slightly. Corrected values for unbound bilirubin indicate greater displacement at the primary binding site by sulfisoxazole than by methylparaben, and no displacement by propylparaben (Table I).
Volume 89 Number 3
Secondary site binding was evaluated by Sephadex G-25 gel filtration. Methylparaben displaced more bilirubin than did sulfisoxazole and the effects of the drugs were additive (Table II). Propylparaben and control samples produced similar results. Red blood cell uptake of bilirubin was increased by both methylparaben and sulfisoxazole at a bilirubin/ albumin molar ratio of 1.6 (Table III). Again, the displacing effects of sulfisoxazole and methylparaben were additive. Displacement of bilirubin at a 0.8 molar ratio was observed only with sulfisoxazole. Propylparaben produced no increase in red blood cell uptake of bilirubin at either molar ratio. Because of the marked difference in disp]acing properties of methyl- and propylparabens, several benzene derivatives were comPared (Table IV). Methylparaben was the most potent binding competitor, whereas ethylparaben and benzoic acid were moderately effective displacers of bilirubin. In three of four repeated experiments (1.6 molar ratio) propylparaben appeared to augment binding. Nonesterified p-hydroxyben: - acid produced no bilirubin displacement. DISCUSSION These data indicate that both methyl- and propylparaben bind to albumin, but that only methylparaben displaces bilirubin from albumin. Methylparaben appears to be a weak primary site competitor, but a strong competitor (similar to sulfisoxazole) when bilirubin is bound to secondary binding sites. Propylparaben has a unique effect on albumin binding of bilirubin; our data suggest a slight enhancement of bilirubin binding. Methylparaben is present in a wide variety of multidose pharmaceutical preparations frequently administered t o sick premature infants including some preparations of injectable saline and bacteriostatic water. It is common practice to flush umbilical artery catheters with I to 3 ml saline fc!iowing blood sampling, and a rapid infusion of 5 or even 10 ml is not uncommon. A single intra-artcrial flush of 2.0 ml saline (containing 2.4 mg methylparaben) in a one kilogram infant could acutely raise the plasma paraben concentration to 342 /*mol/1 (assuming perfect mixing in a 50 ml/kg plasma volume). Methylparaben at this concentration would be expected to compete with bilirubin for binding when the high-affinity binding sites on albumin approach saturation (see Table III). Since the equilibrium of bilirubin with tissue and albumin occurs very rapidly, the administration of methylparaben via catheter flushes is potentially dangerous when the highaffinity binding sites on albumin approach saturation. In such patients it would be prudent to use single dose vials
Effect o f paraben on bilirubin-albumin binding
477
Table IlL Erythrocyte uptake of bilirubin RBC uptake of bilirubin (l~mol/l ++-SE) o.8 molar ratio*
l. 6 molar ratio*
% Change from control
5.4 _+ 0.12 5.6 + 0.04 5.4 _+ 0.43 6.7 _+ 0.03 --
17.7 _+ 0.06 17.3 _+ 0.03 26.7 _ 0.33 26.8 +__0.23 38.0 _+ 0.30
-2 +5t +51 + 114
28.3 _+ 0.09
+ 59
Drug (~tmol/l) Control Propylparaben (250) Methylparaben (250) Sulfisoxazole (250) Sulfisoxazole (250) + methylparaben (250) Sulfisoxazole (250) + propylparaben (250)
*Albuminconcentration= 1.7 gm/dl. Table IV. Effect of benzene compounds on erythrocyte uptake of bilirubin*
Drug (Ixmol/l) Control Propylparaben (250) p-OH benzoic acid (250) Ethylparaben (250) Benzoic acid (250) Methylparaben (250)
RBC uptake of bilirubin (l~mol/l + SE) 17.8 _~ 0.2 15.5 + 0.1 16.5 _+ 0.1 21.3 _+ 0.8 22.1 _+ 0.2 25.6 + 0.2
% Change from control --13 --7 +20 + 24 +43
*Molar ratio bilirubin/albumin1.6;duplicatesamples. of injectable saline and water or to select multidose preparations containing other bacteriostatic agents. CONCLUSION Methylparaben is a potent displacer of bilirnbin from albumin when bilirubin is bound to secondary binding sites and a weak competitor with bilirubin for binding to primary binding sites. When high-affinity binding sites become saturated, the displacing potential of methylparaben is similar to that of sulfisoxazole and greater than that of sodium benzoate. The contribution of various competitors is cumulative and must be taken into consideration when administering multiple drugs to sick jaundiced infants. The authors are grateful for the skilled technical assistance of Ms. Nadja Kronlund. REFERENCES
1. Brodersen R: Competitive binding of bilirubin and drugs to human serum studied by enzymatic oxidation, J Clin Invest 54:1353, 1974.
478
2.
Rasmussen, Ahlfors, and Wennberg
Odell GB, Cohen SN, and Kelly PC: Studies in kernicterus: II. The determination of the saturation of serum albumin with bilirubin, J PEDIATR74:214, 1969. 3. Jacobsen J, and Wennberg RP: Determination of unbound bilirubin in the serum of newborns, Clin Chem 20:782, 1974. 4. Bratlid D: Bilirubin binding by human erythrocytes, Scand J Clin Invest 29:91, 1972.
The Journal of Pediatrics September 1976
5. Shift D, Chan G, and Stern L: Sephadex G-25 quantitative estimation of free bilirubin potential in jaundiced newborn infants' sera: A guide to the prevention of kernicterus, J Lab Clin Med 80:455, 1972. 6. Wennberg RP, and Rasmussen LF: Effects of gentamicin on albumin binding of bilirubin, J PEDIATR 86:611, 1975.