and metocurine in mother and neonate, male and nonpregnant female subjects, and women n = number of subjects. *S~gnificantly different from mother.
Changes in plasma drug binding and a,-acid glycoprotein in mother and newborn infant
Margaret Wood, M.B., and Alastair J. J. Wood, M.B.* Nushoillc, Tenn. Dtycrrtrncwt c?f'A~~esth~~siology crnd Dil'isiot~of' Clinic~ulPhurmuc~)logy,Vunderbilt Uniwrsity Motlicul School
Plasma binding of drugs has important implications for drug disposition and drug action. Drug effect depends on free drug in plasma because it is this fraction which is available for binding to receptor sites and exerting pharSupported by U.S. Public Health Service grant GM 15431. Received for publication Dec. 8, 1980. Accepted for publication Dec. 9, 1980. Reprint requests to: Margaret Wood, M.B., Department of Anesthesiology, Vanderbilt University Medical School, Nashville, TN 37232. *Recipient of a Faculty Development Award from The Pharmaceutical Manufacturers Association Foundation.
522
macologic action. Drug disposition is also influenced by alterations in plasma protein binding because an increase in the free fraction will result in more drug being available for distribution outside the plasma space with resultant decrease in total concentration and increase in volume of distribution. A specific example of the importance of these changes is in the pregnant patient where placental drug transfer is predicated by the free fraction of drug in plasma. This has been shown in studies of the effect of plasma binding on the placental trans-
0009-9236/81/040522+05$00.50/0 0 1981 The C. V. Mosby Co
Volume 29
Drug binding and a,-acid gl~coproteinin mother and fetus
N ~ r m h c r4
523
and metocurine in mother and neonate, male and nonpregnant female subjects, and women receiving oral contraceptives Slrhjcc,~
Neonate Mother Adult male Adult nonpregnant female Adult female on oral contraceptives
d-Tubocurarine (%free)
68.7* (23.5) n = I0 58.9 ( 2 2 . 6 ) n = I0 57.4 (+ 1.2) n = 10 54.6 ( t l .6) n = I0 54.7 ( 2 1.7) n = I0
Metocurine ( % ,fi-ee)
Propranolol (%free)
75.6* (+2.3) 43.3" ( 2 3 . 1 ) n = 13 n = 15 67.1 (+1.2) 19.9: (21.3) n = 13 n = 15 68.2 ( 2 2 . 1 ) 14.9 ( t 0 . 4 ) n = I0 n = 17 64.9 (+ 1.5) 14.6 ( 2 0 . 8 ) n = 10 n = 16 64.3 ( t 1.8) 15.8 (k0.7) n = 16 n = I0
Lidomine (5% free)
Diazepclt?l (%free)
69. I* (ir 1.5) n = 5 47.7: ( k 4 . 7 ) n = 5 33.3 (ir 1.2) n = I0 32.0 ( t 1.4) n = 16 36.8t ( t 1.6) n = 16
3.6* ( 2 0 . 1 9 ) n = ll 4.5t (t0.28) n = Il 2.3 (20.07) n = 16 2.4 ( t 0 . 0 7 ) n = 16 2.6t ( t 0 . 0 7 ) n = 16
n = number of subjects. *S~gnificantlydifferent from mother tSignificantly d~fferentfrom nonpregnant women
fer of bupivacaine. '" As predicted, there was no difference between the free concentrations of bupivacaine in maternal venous and umbilical venous blood, indicating rapid equilibration of free drug across the placenta. It was found, however, that drug binding was higher in mother than in neonate, resulting in higher total drug level in mother than in neonate. Others have confirmed that plasma drug binding is decreased in the newborn infant and it has been shown that the free fraction of lidocaine,"' ampi~illin;~ phen~barbital;~phenytoin,". '-' imipramine,I3 and diazoxideI4 is higher in umbilical cord plasma than adult plasma. Differences in plasma drug binding in the neonate may explain, at least in part, their altered drug sensitivity. Although it is recognized that the clinical response of the newborn infant to the nondepolarizing muscle relaxants decreases with age,', the plasma binding of d-tubocurarine and metocurine has not been defined in the newborn infant. A number of factors including disease states,'. I;' aging,:' and drug interactions2%ave been shown to affect drug binding in man and consideration of these variables is of critical importance to the determination of drug action. It has recently been demonstrated that a number of basic drugs, such as propranolol and quinidine, are bound to a,-acid glycoprotein." I " Several local anesthetics (e.g., lidocaine and bupivacaine) have also been shown to bind to
a,-acid glycoprotein.'" I " The purpose of our study was therefore to evaluate age-related changes in a,-acid glycoprotein and relate these changes to plasma drug binding in the newborn infant. The effects of sex, pregnancy, and the oral contraceptive pill on drug binding were also determined. Methods
Our subjects were 15 pregnant women (aged 19 to 35 yr) at 39 to 41 wk gestation and their newborn infants at delivery; in addition, 16 adult nonpregnant women (20 to 31 yr), 17 adult men (20 to 32 yr), and 16 nonpregnant women (20 to 26 yr) who were taking the oral contraceptive pill were included as part of the study. None of the subjects were receiving any medication apart from the oral contraceptive pill and, in the case of the pregnant patients, prenatal vitamin tablets. Blood samples (40 ml) were drawn from the nonpregnant subjects by venipuncture from the antecubital vein into heparinized glass stoppered tubes. Venous blood (20 to 40 ml) from the maternal end of the umbilical cord was collected immediately after the baby was delivered but before the placenta was delivered, and a simultaneous venous sample (40 ml) was drawn from the antecubital vein of the mother. Drug binding in plasma was determined by equilibrium dialysis by dialyzing 1.0 ml plasma obtained, as described above, against 1.0 ml
Clrn Phurmac 01. Ther. Aprd 1981
524 Wood and Wood
i Fetus
0 er Males
immunodiffusion was 89% (-+ 1.7) of the concentration by weight. The results were analyzed by Student's t test for paired or unpaired values as appropriate and p < 0.05 was taken as the minimal level of significance. Results
i
les Oral Contra
Fig. 1. Mean a,-acid glycoprotein concentrations (+-SEM) in mother and fetus, male and nonpregnant female subjects, and women receiving the oral contraceptive pill.
phosphate buffer, pH 7.4, in a dialysis cell (Spectrum Medical Industries). The buffer contained one of the following: 2.2 ng propranolol (specific activity 81 mCiImg), 6.1 ng "H-diazepam (specific activity 294 mCi/ mg) , 1.6 p g IT-lidocaine (specific activity 0.1 mcilmg), 7.7 ng :'H-tubocurarine (specific activity 23.5 mcilmg), or 3.4 p g IT-metocurine (specific activity 0.104 mCi1mg). The cells were rotated in a water bath maintained at 37' for 4 hr, after it was shown that for all the drugs studied, equilibrium is achieved within this time. The concentration of drug in the plasma and buffer was determined and the free fraction in plasma calculated as the concentration in the buffer divided by the concentration in the plasma. The concentration of a,-acid glycoprotein was determined by radial immunodiff~sion.'~' To estimate the contribution of a,-acid glycoprotein to the binding of these drugs, drug binding as described above was determined in solutions of a,-acid glycoprotein ranging in concentration from 14 to 700 mgI100 ml by weight. The concentration of a,-acid glycoprotein in these solutions as determined by radial
"-
The free fraction of propranolol, lidocaine, diazepam, d-tubocurarine, and metocurine in mother and neonate and in male and nonpregnant female subjects is shown in Table I. The free fraction of d-tubocurarine (p < 0.05), metocurine (p < 0.01), propranolol (p < 0.001), and lidocaine (p < 0.02) was higher in the neonatal plasma than in the maternal plasma at delivery, indicating decreased binding in the newborn infant. The free fraction of diazepam, however, was less in the neonatal plasma than in that of the mother (p < 0.02), but the free fraction of diazepam in neonatal plasma was still significantly higher than in that of female nonpregnant adults (p < 0.001). The free fractions of diazepam, propranolol, and lidocaine were higher (p < 0.001) in pregnant women than in nonpregnant women. Diazepam and lidocaine free fractions were higher (p < 0.05) in nonpregnant women on the contraceptive pill than in age-matched women not using the pill. The mean a,-acid glycoprotein concentration was lower in the cord plasma (15.3 -+ 4.7 mgIlO0 ml) than in maternal plasma (49.6 t 6.5 mg1100 ml) (p < 0.002) (Fig. l ) , more than a threefold difference. There was no difference in the a,-acid glycoprotein concentrations in the male and nonpregnant female subjects, whether receiving the oral contraceptive or not (Fig. 1). There was a positive correlation between plasma &,-acid glycoprotein concentrations and the binding ratio (boundlfree concentrations) of lidocaine (p < 0.001; r = 0.623; Fig. 2) and propranolol (p < 0.001; r = 0.652, Fig. 3). The in vitro studies using &,-acid glycoprotein solutions showed excellent correlation between binding ratio (B/F) and a,-acid glycoprotein concentration ([aGP]) for tubocurarine (BIF = 0.13 0.18 [aGP], r = 0.981, p < 0.001), lidocaine (BIF = -.08 0.39 [aGP], r = 0.966, p < 0.001), diazepam (BIF = -0.01 0.32 [aGP], r = 0.985, p < 0.001),
+
+
+
Drug binding and a,-acid glycoprotein in mother and fetus
0
I
0
1
I
I
4
12.5 25 37.5 50 62.5 a , ACIDGLYCOPROTEIN (rng/IOOrni I
525
I
75
0
2 5
25
37.5
50
62.5
75
a l ACIDGLYCOPROTEIN (rng/IOOrnl 1
Fig. 2. The relationship between plasma a,-acid glycoprotein concentrations and the binding ratio (boundlfree) of lidocaine in the subjects studied.
Fig. 3. The relationship between plasma a,-acid glycoprotein concentrations and the binding ratio (boundlfree) of propranolol in the subjects studied.
propranolol (BIF = 0.56 + 1.41 [aGP], r = 0.981, p < 0.001), and metocurine (BIF = 0.20 0.11 [aGP], r = 0 . 9 3 9 , ~< 0.001).
curine is higher in the neonate than in the mother and that the concentration of a,-acid glycoprotein is three times as high in the mother as in the neonate. For lidocaine and propranolol, which are known to bind to a,-acid glycoprotein, we found, as others have reported, a correlation between binding ratio and plasma a,-acid glycoprotein concentration, and for all of the drugs studied there was excellent correlation between in vitro a,-acid glycoprotein concentrations and binding ratio. We therefore conclude that it is likely that the elevation of the free fraction of certain drugs in neonatal plasma is due to lower levels of a,-acid glycoprotein, although there are many other factors that are also important in determining the amount of binding to plasma in neonates. Newborn infants and pregnant women have reduced levels of plasma albumin and the albumin concentration has been reported in one study to be decreased by 26% in pregnant women.' The serum binding of salicylic acid, sulfisoxazole, phenytoin, diazepam, and dexamethasone has been shown to be reduced during pregnancy4 and with the exception of dexamethasone the free fraction of all these drugs correlated negatively with their serum albumin concentration^.^ Although low plasma albumin levels may contribute to reduced binding of some drugs in neonates, there is also evidence of decreased binding capacity, which is inde-
+
Discussion
Although many basic drugs are highly bound to plasma proteins, their binding to albumin often accounts for only a small proportion of their total plasma binding," suggesting that some other plasma component accounts for a large portion of their total plasma binding. Propranolol and lidocaine, both basic drugs, have recently been shown to bind to a,-acid glycoprotein, an acute-phase protein that is elevated in certain situations such as stress, hepatic damage, malignant tumors, and acute and chronic inflammatory disease.Viafsky et al. demonstrated that the binding of propranolol and chlorpromazine is increased in patients with Crohn's disease and rheumatoid arthritis, and that there is an inverse relationship between the free fraction of drug in plasma and plasma a , acid glycoprotein concentrations, indicating that the increase in plasma binding in these patients is mediated by an increase in a,-acid glycoprotein." Others have shown that lidocaine is bound to &,-acid glycoprotein and that the free fraction of lidocaine in plasma is inversely related to plasma a,-acid glycoprotein concentrations. ''L I" We have shown that the free fraction of propranolol, lidocaine, d-tubocurarine, and meto-
526
Wood and Wood
pendent of albumin concentration. 2 1 Krasner et a1 reported that neonatal albumin, although having a reduced affinity for many drugs, has greater ability than adult albumin to bind to bilirubin.Vt is therefore interesting that we demonstrated that binding of diazepam in the mother and neonate was lower than in adult men and women. Diazepam is known to be principally bound to plasma albumin and it is therefore likely that these changes reflect the reported" reduction in plasma albumin in mothers at term. This suggestion is reinforced by the finding that the free fraction of diazepam was 73% higher in the mothers than in the nonpregnant women. It may therefore be possible from a knowledge of the principle binding protein of a drug to predict the effects of pregnancy, oral contraceptive use, and age on the plasma binding of that drug.
'
We wish to thank Dr. K. M. Piafsky for his generous donation of the al-acid glycoprotein and Dr. R. A. P. Burt, Lilly Research Laboratories, Indianapolis, Ind., for his support.
References Bennett EJ, Ignacio A, Patel K, Grundy EM, Salem MR: Tubocurarine and the neonate. Br J Anaesth 48:687-688, 1976. Craig WA, Evenson MA, Ramgopal V: The effect of uremia, cardiopulmonary bypass and bacterial infection on serum protein binding, it1 Benet LZ, editor: The effect of disease states on drug pharmacokinetics. Washington, D. C . , 1976, American Pharmaceutical Association, pp. 125-136. Davis D, Grossman SH, Kitchell BB, Shand DG, Routledge PA: Age related changes in the plasma protein binding of lidocaine and diazepam. Clin Res 28:234A, 1980. (Abst.) Dean M, Stock B, Patterson RJ, Levy G: Serum protein binding of drugs during and after pregTHER nancy in humans. C L I N PHARMACOL 28:253-161, 1980. Ehrnebo M , Agurell S, Jalling B, Boreus LO: Age differences in drug binding by plasma proteins: Studies on human foetuses, neonates and adults. Eur J Clin Pharmacol 3: 189- 193, 1971. Fremsted D, Bergerud K, Haffner JFW, Lunde PKM: Increased plasma binding of quinidine after surgery: A preliminary report. Eur J Clin Pharmacol 10:44 1-444, 1976.
Clin. Pharmacol. Thrr April 1981
7. Ganrot PO: Variation of the concentrations of some plasma proteins in normal adults, in pregnant women and in newborns. Scand J Clin Lab Invest 29(suppl. 124):83-88, 1972. 8 . Kawai T: Clinical aspects of the plasma proteins. Philadelphia, 1973, J. B. Lippincott Co., p. 20. 9 . Krasner JK. Giacoia GP, Yaffe SJ: Drug-protein binding in the newborn infant. Ann NY Acad Sci 226:lOl-114, 1973. 10. Mancini G , Carbonara AO, Heremans JF: Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2: 235-254, 1965. 1 1 . Piafsky KM: Disease-induced changes in the plasma binding of basic drugs. Clin Pharmacokinet 5246-262, 1980. 12. Piafsky KM, Borga 0 , Odar-Cederlof I, Johansson C, Sjoqvist F: Increased plasma protein binding of propranolol and chlorpromazine mediated by disease-induced elevations of plasma alpha, acid glycoprotein. N Engl J Med 299:1435-1439, 1978. 13. Piafsky KM, Knoppert D: Binding of local anesthetics to alpha, acid glycoprotein. Clin Res 263836.4, 1978. 14. Pruitt AW, Dayton PG: A comparison of the binding of drugs to adult and cord plasma. Eur J Clin Pharmacol 459-62, 1971. 15. Reidenberg MM, Affrime M: Influence of disease on binding of drugs to ulasma moteins. Ann NY Acad Sci 226: 115- 125: 1973.' 16. Routledge PA, Barchowsky A, Bjornsson TD, Kitchell BB, Shand DG: Lidocaine plasma proT H E R27:347tein binding. CLINPHARMACOL 351. 1980. 17. Sager G, Nilsen OG, Jacobsen S: Variable binding of propranolol in human serum. Biochem Pharmacol 28:905-911, 1979. 18. Stead AL: The response of the newborn infant to muscle relaxants. Br J Anaesth 27: 124- 130, 1955. 19. Thomas J , Long G , Moore G , Morgan D: Plasma protein binding and placental transfer of THER19:426bupivacaine. CLIN PHARMACOL 434, 1976. 20. Tucker GT, Boyes RN, Bridenbaugh PO, Moore DC: Binding of anilide-type local anesthetics in human plasma. Anesthesiology 33:304-314, 1970. 21. Wallace S: Factors affecting drug-protein binding in the plasma of newborn infants. Br J Clin Pharmacol 3:510-5 12, 1976. 22. Wood M, Shand DG, Wood AJJ: Altered drug binding due to the use of indwelling heparinized cannulas (heparin lock) for sampling. CLIN PHARMACOL THER25103-107, 1979.
