brain bilirubin transfer equilibrium

Ear!? Human

Deoelopmenr.

Elsevier Biomedical

6 (1982)

341-355

341

Press

Risk of bilirubin acid precipitation in preterm infants with respiratory distress syndrome: considerations of blood/brain bilirubin transfer equilibrium F. Ebbesen ’ and R. Brodersen’ ’Depurfmwr

qf Neonatclogv,

Rigshospitalet,

Copenhagen.

and ’ lnstlrute of~.Medical Bwchenlistrr.

Unioersity of Aarhu.r. Aorhus.

Accepted

for pubhcation

Dmmurh-

23 March

19X2

Summap Twienty-six preterm infants with respiratory distress syndrome (RDS). were examined daily during the first 6 days of life. Twenty-six equally preterm but clinically well infants served as controls. In the RDS infants, plasma albumin concentration was low, hyperbilirubinemia was prolonged, plasma pH was decreased during the first two days, and the concentration of reserve albumin for binding of monoacetyldiaminodiphenylsulfone (MADDS), a deputy ligand for bilirubin. was decreased on the second throughout the sixth day, when compared with the controls. These factors concur in increasing the likelihood of bilirubin acid precipitation in RDS above the increased risk present in preterm infants. The plasma of the preterm controls was supersaturated with respect to crystalline bilirubin acid by an average factor 5 (index of plasma bilirubin toxicity = 0.7) on the first day of life, peaking at a factor 10 (index 1.0) on the third and fourth days while these factors were 10 and 20 (index 1.O and 1.3), respectively, in the RDS infants. Two of the latter surpassed a level of 60 times supersaturation (index 1.8) where acute precipitation of amorphous bilirubin acid becomes possible. hyperbilirubinemia; syndrome; preterm

reserve albumin infants

for bilirubin

binding:

respiratory

distress

introduction Transfer of bilirubin from blood plasma to certain observed, especially in preterm infants. In animal 0378/3782,~82,~0000-0000/.$02.75

‘n 1982 Elsevier Biomedical

areas of the brain is sometimes and tissue culture work such

Press

342

deposition of bilirubin causes toxic damage to the nerve cells [9,10,30,31] and considerable efforts have been made accordingly in order to prevent this. Exchange transfusion and phototherapy, have been accepted for this purpose. Indications for these measures have been much discussed and are usually based upon plasma concentration of unconjugated bilirubin in combination with other criteria such as acidosis, prematurity, etc. Several laboratory methods are also used for estimation of the degree of binding of bilirubin to albumin since binding to albumin is considered to counteract transfer of bilirubin into the nerve cells. It is very difficult in a clinical setting to evaluate the preventive value of such methods because overt cases of bilirubin encephalopathy are rare and cases of minor damage, if they occur, can only be demonstrated by long-term follow-up and have to be distinguished from sequelae of anoxia, hemorrhage and other possible causes of brain damage. The latter distinction is probably a nearly impossible task in a population of preterm infants who often suffer from multiple abnormalities during their first few days of postnatal life. Under these conditions it appears reasonable to consider the physical chemistry of bilirubin deposition in the brain from an experimental and theoretical point of view. Indications for preventive measures and the choice of laboratory methods should be directed by the best available theoretical and experimental evidence until their value can be finally assessed by controlled clinical studies. Physical chemistry of bilirubin deposition in the brain must deal with three problems, (1) equilibrium conditions for transfer of bilirubin from plasma to nerve cells; (transport is impossible if equilibrium conditions favor a transport in the opposite direction), (2) the velocity of the transport process (transfer will not take place if prevented by an effective blood-brain barrier, even if favored by the condition of equilibrium), and (3) possible disposition of bilirubin after its entry into the target cell. The present paper presents some theoretical considerations of the equilibrium conditions. The thermodynamical basis for the theory has been presented previously [3] and is here reformulated for possible application in the clinic, It is shown that, under certain presumptions, the energetic tendency to precipitation of bilirubin in the cells can be described by an Index of Plasma Bilirubin Toxicity, containing three measured parameters, plasma concentration of tinconjugated bilirubin, plasma pH, and the concentration of reserve albumin for binding of bilirubin. The index is calculated in a group of preterm infants with respiratory distress syndrome (RDS) and in a group of equally preterm patients without this syndrome and it is proposed that a high value of the index is a necessary but not a sufficient condition for development of bilirubin encephalopathy. Other conditions are an open blood-brain barrier and failing oxidation of bilirubin in the cells. Physical chemistry of bilirubin acid precipitation The overall process

Deposition of bilirubin on erythrocytes and in tissues has often been described in newborn infants, particularly with emphasis on kernicterus, yellow staining of basal

343

ganglia of the brain. The bilirubin stems from blood plasma where it is bound to circulating albumin as the bilirubin dianion [2]. In the target tissues. bilirubin is presumably present as the acid, aggregated on lipoid membranes [18]. The overall process of precipitation thus involves uptake of two hydrogen ions and release of one albumin for each precipitated molecule of the pigment. Brb’- albumin

+ 2 H ’ = BrbH,(solid)

+ albumin

(1)

According to this reaction scheme, the tendency to precipitation of bilirubin should increase with high concentration of bilirubin-albumin, low pH, and low concentration of reserve albumin, all in good agreement with clinical experience and results from animal and tissue culture work (see review [4]). It should be noted that the toxic agent here is a combination of hydrogen ions and the bilirubin-albumin complex, while albumin. capable of binding bilirubin. plays an antitoxic role. This is a more realistic concept than talking of unbound bilirubin as the toxic substance since we do not know whether this, or the complex with albumin or some other form of bilirubin. is transported from plasma into the cells. Equilibrium

of’precipitation

Irrespective process is

of the mechanism

involved.

the condition

of equilibrium

of the above

or log;-ZpH-logK=O where B is the concentration of the bilirubin-albumin complex (measured as unconjugated plasma bilirubin), p is the reserve albumin concentration for binding of bilirubin, and K is the equilibrium constant. The numerical value of K can be found by studying the solubility of bilirubin in buffered solutions of albumin. The measurement must be carried out by indirect means since dissolution of solid bilirubin in an albumin solution results in formation of colloid bilirubin acid-albumin coaggregates, a product which is different from the bilirubin dianion-albumin complex found in blood plasma. The process can be carried out in two steps, dissolution of bilirubin in buffer, BrbH,(solid) and binding Brb’-

= Brb’-

+2 Ht

of the dissolved

+ albumin

(4) bilirubin

dianion

to albumin,

= Brb2- albumin

At 37°C and physiological buffer is described by log !i = 2 pH - 23.24

salt concentration,

(5) the equilibrium

of dissolution

in

344

or S. [H+]’ = 5.75 . 1O-24 (mol/$

(6)

where S is the solubility of bilirubin [2,3]. The equilibrium constant of the second step, i.e. the binding constant of bilirubin dianion to albumin, is under these circumstances [3]: -

B

P.S

= 5.9 . 10’ l/m01

Combination S.

of Eqns. 6 and 7 gives

[H+]‘.+&. = 5.75.

[H+12

lo-24 .5.9. 10’ = 3.39.

10-‘h (mol/l)’

or log;-2pH+

15.5=0

(8)

K in Eqns. 2 and 3 is thus 3.39. lo-‘” (mol/l)2, or logK= - 15.5. This result is obtained if bilirubin crystallized from methanol/chloroform is used for the solubility determinations. A plasma sample at 37’C, to which some crystalline bilirubin is added, will therefore neither dissolve more bilirubin nor deposit any if the above numerical relation is fulfilled. At pH 7.75 (2 pH = 15.5) plasma is thus saturated with crystalline bilirubin when one bilirubin molecule is present for every two of binding albumin (B/p = 1). At pH 7.4, plasma is saturated with a molar ratio of bilirubin/binding albumin = 1 : 6, and at pH 7.25 already when one molecule of binding albumin among eleven has combined with bilirubin. At lower molar ratios, the plasma will dissolve more crystalline bilirubin and, vice versa, will deposit crystals of the pigment at higher molar ratios of bilirubin/binding albumin. Plasma samples with higher bilirubin contents than the above limits of saturation are often encountered in newborns. Under these conditions, plasma is supersaturated with respect to crystalline bilirubin. The degree of supersaturation can be calculated from the equilibrium condition as (B/p). [H’12/(3.4. 10p16) and may be as high as 10 or even 100. Rate of the process - precipitation

of amorphous

bilirubin

Crystallization of bilirubin in plasma is a very slow process and may for practical purposes never take place. However, if a phospholipid membrane is present, deposition of bilirubin acid is enhanced, within a period of seconds to 1 h forming large aggregates of the pigment, connected to the membrane [2,18]. It has recently been found that such aggregates are about 60-fold more soluble than crystalline bilirubin [ 181. This may possibly be explained by the fact that bilirubin is precipitated as an amorphous hydrate of bilirubin acid when an alkaline solution of the pigment is neutralized [27]. This amorphous hydrate is more soluble than crystalline bilirubin

345

and, once formed, is only slowly transformed into the crystalline substance. Formation of such aggregates is possible when a plasma sample is more than 60-fold supersaturated with respect to crystalline bilirubin and is in touch with a suitable phospholipid membrane. If conditions favor precipitation of bilirubin acid in one form or the other in plasma, actual formation of aggregates may take place if local tissue factors allow. In the brain, the blood-brain barrier may have to be opened before bilirubin can get access to the target lipoid membranes in mitochondria and other organelles [25] and other f:lctors may determine the rate of aggregation in kidneys. skin, erythrocytes. etc. .A hilirubin oxidase present in brain mitochondria and in many tissues [6] may also prebcnt precipitation and thus preclude damage to the cells even when plasma conditions are strongly in favor of deposition of bilirubin. The distribution of yellow staining in the brain. as seen by autopsy. could be determined by differences in the blood-brain barrier and in the cellular content of bilirubin oxidase. On the other hand, deposition is impossible if plasma is not supersaturated with bilirubin. Index of plasma hiliruhin toxici? The factor of supersaturation. i.e. the plasma bilirubin toxicity, may serve as a me,asure of one factor in development of toxic damage, as outlined above. It has been suggested [4] to rate plasma bilirubin toxicity by an Index of Plusma Biliruhin Toxicity. which is equal to the logarithm of the supersaturation factor and is I=logE--2pH+ P

15.5

(9)

If I is negative, precipitation of bilirubin acid is impossible: if I has a value between 0 and 1.8. bilirubin may precipitate as crystals but probably does so very slowly: if I is above 1.8 (= log 60). a faster process, precipitation of amorphous bilirubin acid. becomes thermodynamically possible and will take place if favored by local tissue factors such as an open blood-brain barrier and low intracellular bilirubin oxidase activity. Influence

of intrucellular pH

One might suppose that pH inside the cell where bilirubin precipitation takes place would be more decisive for the process than pH in blood plasma. This, however, is not so. provided a condition of equilibrium exists across the cell mernbrane, as seen from the following. If the process takes place by transport of bilirubin dianions and hydrogen ions through the cell membrane and if an equilibrium has been established in which the hydrogen ion concentration is higher inside the cell, a membrane potential must be present which again determines the equilibrium distribution of bilirubin dianion on both sides of the membrane, TIC

7T _ 2

RT F

ln

[H+]I _ RT -[H+ ]z -

-2F

I@ II In [B’ I2

346

from which we obtain [H+]: . [B*-1,

= [H+];

.[B2-12

The product of the squared hydrogen ion concentration and the concentration of bilirubin dianion is thus equal on both sides of the cellular membrane. This product determines the possibility of bilirubin acid precipitation, as seen from Eqn. 4, and thus is a measure of the bilirubin toxicity. Conditions for precipitation of bilirubin acid inside the cell may therefore equally well be evaluated by measuring plasma parameters as intracellular concentrations. Throughout the above reasoning it has been presumed that active transport of bilirubin and hydrogen ions across the cell membrane does not take place. This presumption seems reasonable in itself and will be further supported by the clinical observations reported below. Determination

of reserve albumin for binding of bilirubin

The reserve albumin concentration for binding of bilirubin must be determined at 37”C, using undiluted serum or plasma, and cannot be found by titration with bilirubin since this would displace any loosely bound competitors. The only possible way seems to be addition of a trace amount of a deputy ligand, binding to albumin in the same fashion as bilirubin, and determination of the binding ratio. The reserve albumin concentration is then found from a series of standard determinations with pure albumin. MADDS (monoacetyldiaminodiphenyl sulfone) seems to function as a satisfactory deputy ligand for bilirubin [ 11.

Materials and Methods 52 preterm newborn infants (gestational age (259 days) with low birth weight (G 2500 g) were studied during the first 6 days of life. They were single-born, and their birth weight was normal for the gestational age according to the growth chart of Gairdner and Pearson [19]. Their direct Coombs’ test was negative. One group consisted of 26 infants suffering from RDS, which required treatment w’ith intermittent positive pressure ventilation (IPPV) and/or continuous positive airway pressure (CPAP) and/or nasal CPAP (N-CPAP). The disease was classified as RDS if the following conditions were fulfilled: At least 24 h of respiratory insufficiency with tachypnea, retractions, expiratory grunting, hypoxemia, hypercapnea, and characteristic X-ray with diffuse reticulogranular pattern and air bronchogramme. 22 of the infants were treated with IPPV or with IPPV plus CPAP during an average of 91 h, range 9- 129. The remaining 4 infants were treated with N-CPAP during an average of 39 h, range 15-50. One infant died on the second day of life. The control group consisted of 26 infants, who were clinically asymptomatic. The two groups were matched so that the gestational age and birth weight of the infants were equal, as shown in Table I. By ordinary phototherapy the patients were exposed from above with a mixture

347 TABLE Patient

I and control

material

Gestational age (days) median (range) Birth weight (g) median (range) Ratio. female/male Number of infants with Apgar scores<7 5 mln after birth Onset of phototherapy. (h of life) median (range). P>O.O5 Duration of ordinary phototherapy, (h) median (range). P>O.O5 Number of infants receiving simple blood transfusion Number of infants in whom a blood glucose concentration< 1.7 mmol/l was measured

Preterm infants wth respiratory distress syndrome ( PI= 26)

Preterm, clinically well infants (n = 26)

230 (206-255) 1830 (1200-2430) 16/10

231 (206-254) 1870 (1200-2290) 11/15

4

0

67 (25-X4)

53 (15-78)

59 (O- 100)

59 (O-85)

7

3

I

2

of ‘daylight’ and blue light [ 121. The radiance at infant level was 5.5 W/m’ (B), as measured with a photometer [26]. By double-light treatment the infants were exposed simultaneously from above and below to blue light [ 171. The radiance from above at infant level was 9.5 W/m2 (B) and from below at mattress level 10.5 W/m’. Phototherapy was given on equal indications to both groups of infants. Ordinary phototherapy was initiated when the plasma bilirubin concentration increased above the following limits: 120 pmol/l in infants of birth weight less than 1250 g, 150 pmol/l in infants of 1250-2000 g, and 180 pmol/l in those weighing 2000-2500 g. Double light treatment was given when the bilirubin concentration approached the exchange transfusion limit, i.e. 90 pmol/l higher than the limit for ordinary phototherapy. The time of beginning phototherapy and the duration of ordinary phototherapy did not differ significantly in the two patient groups, as seen in Table I. Three RDS and four control infants received double-light treatment during an average of 17 and 21 h, respectively. None of the infants had exchange transfusion or albumin infusion; simple blood transfusion was given to seven RDS patients and to three controls. 14 infants in each group were treated with gentamicin (Garamycin@‘, containing methyl- and propylparaoxybenzoate) combined with ampicillin, either in relation to the treatment of RDS or due to rupture of the maternal membranes more than 24 h prior to delivery (almost exclusively the clinically well infants). No bacterial growth in blood cultures was seen. Capillary blood was drawn by heel prick daily at 8 a.m. for the first 6 days of life. The serum unconjugated bilirubin concentration was determined by a modification [8] of the method of Mertz and West [29]. as previously described [ 121. This method was chosen on the assumption that it is more specific than the diazo method

348

[8]. Purified bilirubin [28] dissolved in cyanide-formamide [ 1 l] and added to human serum was used for standardization.The coefficient of variation of the analysis was 2% by duplicate determinations. The albumin concentration was determined in serum by the electrophoretic method of Laurel1 [24]. Human serum albumin from AB Kabi, Stockholm, Sweden, was used for standardization. The coefficient of variation of the analysis was 3% by duplicate determinations. The reserve albumin concentration for binding of MADDS was determined in serum by the [14C]MADDS method [5]. Human serum albumin (Kabi) was used for standardization. By triplicate determinations the coefficient of variation of the analysis was 5-6%, the intraday coefficient of variation being 4-5s. The pH of plasma was measured on an Acid-Base Laboratory 2 from Radiometer, Copenhagen, Denmark. Statistical analyses were performed using Mann-Whitney’s test for independent observations. Informed consent was obtained from mothers of all infants included in the study.

Results Concentrations of unconjugated bilirubin, albumin, and of reserve albumin for binding of MADDS are shown in Fig. 1, and plasma pH values in Table II. The preterm infants with RDS as well as the preterm controls were hyperbilirubinemic. Levels of unconjugated bilirubin were not significantly different in the two groups until a median peak value of 150 pmol/l was reached on the 3rd day of life. Thereafter the plasma bilirubin of the controls declined slowly while that of the RDS infants remained high. Median plasma albumin concentrations ranged about 520-530 pmol/l in the control group. The RDS patients had significantly lower albumin concentrations throughout the period of observation, on the 1st day of life by a median factor 0.83, on the 6th day by 0.92. Reserve albumin concentrations for binding of MADDS were low in all infants of this study, median values ranging from 45 to 90 vmol/l. This should be compared with albumin concentrations of the order of 500 pmol/l, of which typically 150 pmol/l was occupied by unconjugated bilirubin. The albumin of all the infants thus showed poor binding of MADDS, and presumably of bilirubin. The RDS patients had lower concentrations of reserve albumin than the controls, the difference being highly significant on the 3rd to 5th day of life. Plasma pH was slightly decreased in the RDS infants on the 1st day, median 7.33, and on the 2nd day, median 7.35. Thereafter it remained normal and not significantly different from that of the controls, 7.36-7.40, median values. The three plasma parameters which determine the tendency to precipitation of bilirubin acid, i.e. unconjugated bilirubin. reserve albumin, and pH, were changed significantly in favor of precipitation when taken together in the index of plasma bilirubin toxicity (Eqn. 9) (Fig. 2). The preterm control infants had indices in the

349

600

100

50

0 1

2

3

I

5 6 days of Ilfe

Fig. I. Serum concentrations of albumin (0 .- .-. 0. ?? ‘-.-.O). unconjugated bilirubin (0 -0. 0). and reserve albumin for binding of MADDS (O------O, O------e). Preterm mfanta wth o---respiratory distress syndrome, 0: and clinically well preterm infants. 0; medians (25-75 percentiles).

range from 0.7 on the 1st day to about 1.0 peak median levels on the 3rd to 4th day, indicating that their plasmas were supersaturated by factors 5-10 with respect to crystalline bilirubin. The RDS patients had indices approximat.ely 0.3 higher than the controls, corresponding to supersaturation factors 10-20. Index levels above 1.8, where acute precipitation of amorphous bilirubin becomes possible, were seen in two infants with RDS. Individual findings in these patients are shown in Table III. Even though the index of one infant was as high as 1.99 (lOO-fold supersaturation of plasma with respect to crystalline bilirubin) on the 3rd da:y of life while an index of 1.87- 1.90 was seen for more than 24 h (2nd and 3rd day) in the other patient, no abnormal neurological signs were seen, particularly the muscle tone and the Moro reflex were normal, and no oculogyric movements, muscle twi tchings or convulsions occurred.

350

TABLE II Plasma pH in preterm

infants

with respiratory

Day of life

Median

(25-75

distress

syndrome

well preterm

infants

percentile) Well infants (n=26)

RDS infants (n =26)

7.39 (7.36-7.41)

7.33 (7.25-7.40)

1st

and clinically

P
7.41 (7.38-7.44)

7.35 (7.30-7.43) PSO.05

7.40 (7.36-7.42)

7.37 (7.34-7.45)

3rd

P10.05 7.37 (7.35-7.41)

7.37 (7.34-7.42)

4th

P>O.O5 7.38 (7.35-7.43)

7.39 (7.36-7.44)

5th

P10.05 7.36 (7.33-7.40)

7.39 (7.32-7.43)

6th

PlO.05

Index of plosmo bllirubin toxicity

O.Oi

6 days of life

Fig. 2. Index of plasma bilirubin toxicity in preterm infants with respiratory clinically well preterm infants, (0); medians (25-75 percentiles).

distress

syndrome

(0)

and

Discussion Kernicterus concentrations concentrations

has been observed in preterm infants in spite of low plasma bilirubin [20]. It has accordingly been generally accepted that plasma bilirubin alone cannot serve as a sufficient indicator for therapy in threatening

III

A B

Patients

216 237

t days)

age

Gestational

Data of RDS patients

TABLE

1200 1625

(g)

F F

3 2 3 1.99 1X7 1.90

Index

Sex

Day of life

Birth weight

hilirubin

with an index of plasma

above

1.80

205 141 202

Unconjugated bilirubin 25 22 20

447

Reserve albumin

( pmol /I)

483 434

Albumin

Serum concentrations

toxicity Plasma

7.21 7.22 7.30

PH

Ordmary None Ordinary

Phorotherapy

IPPV I PPV IPPV

lPPV/ CPAP, N-CPAP

None None None

Neurologic symptoms

352

bilirubin encephalopathy and considerable efforts have been directed towards establishment of better criteria. A few theoretical considerations may be helpful. Bilirubin is present in blood plasma as the dianion, reversibly bound to albumin, and is probably deposited in the brain as the nearly insoluble bilirubin acid. Transport of bilirubin from plasma to brain must occur at the expense of free energy. Provided that active transport processes are not involved, deposition of bilirubin is thermodynamically possible if the concentration of dissolved bilirubin acid in plasma is above the level of saturation, i.e. if the plasma is supersaturated with bilirubin acid and the likelihood of deposition increases with the degree of supersaturation. As shown in the first part of the present paper, the degree of supersaturation can be assessed by measuring the concentration of unconjugated bilirubin, the concentration of available albumin for binding of bilirubin, and pH of the plasma. These three parameters are entered in a simple formula for calculation of an index of plasma bilirubin toxicity. The index is the logarithm of the factor of supersaturation of plasma with respect to crystalline bilirubin. If the index is negative, deposition of bilirubin acid is thermodynamically impossible, positive index values show that crystalline bilirubin acid can be precipitated, and at indices higher than 1.8, deposition of a more soluble, amorphous form of bilirubin acid becomes possible. The latter process, precipitation of amorphous bilirubin, is catalyzed by phospholipid membranes and results in formation of large bilirubin aggregates on polar lipid structures inside the cells. From a theoretical and experimental standpoint it would accordingly be expected that values of index of plasma bilirubin toxicity above 1.8 would indicate a danger of acute cellular damage. In the present paper it is found that the index is increased in preterm infants with RDS above values in a group of equally preterm, but otherwise healthy infants, i.e. the index values in preterm infants with RDS fall in the expected range, classifying these patients as a high risk group. This finding tends to strengthen the validity of the presumption made on theoretical grounds, namely that precipitation of bilirubin in the brain is furthered by a high concentration of unconjugated bilirubin (bound to albumin) and by a low plasma pH and is counteracted by a high concentration of albumin capable of binding bilirubin. The result is in good keeping with clinical expectations and agrees with the usual preventive praxis. Novel is only the specific way of putting these measured parameters together in one index, based on simple equilibrium considerations. On the other hand, we have not been in a position to investigate any possible correlation of the index with clinical findings. Two infants in the present material of patients with RDS had index values in the range of 1.8-2.0, i.e. the plasma of these patients was supersaturated with respect to crystalline bilirubin by a factor 60-100, and l-2.5fold with respect to amorphous bilirubin. Under these circumstances, deposition of bilirubin acid takes place on phospholipid liposomes in vitro within a period of a few minutes. The two infants showed no neurologically abnormal signs. Detailed neurological studies or follow-up of these patients were not attempted and would hardly be rewarding in this small material. One may speculate, in accordance with the theory, that these patients were protected, at least from severe damage, by local tissue factors. An intact blood-brain barrier, and the presence of bilirubin oxidase in nerve cell mitochondria, may

353

prevent damage to the cells even when this would be strongly favored by a high de,gree of supersaturation of blood plasma with bilirubin acid. It is interesting to note that concentrations of unconjugated bilirubin were not especially high, mostly within the range 100-200 ~mol/l, and were only slightly higher in the RDS patients than in the controls. Plasma pH values were usually normal. The factor contributing most to the high plasma bilirubin toxicity was the low reserve albumin concentrations for binding of bilirubin measured as reserve for binding of MADDS. A large proportion of albumin was non-binding, i.e. not occupied by bilirubin and also not available for binding of MADDS. The proportion of non-binding albumin averaged about 60% and was equal in RDS patients and controls. The RDS infants had lower plasma albumin concentrations. Treatment of the infants with gentamicin. given as an injectable preparation containing parabens. has not contributed to the lowering of reserve albumin [7]. Ampicillin has a slight bilirubin displacing effect, in vitro as well as in vivo [7], and ma:y have caused a minimal decrease of reserve albumin, equally in RDS patients and controls, since the amount of ampicillin given to the two groups was equal. The group of RDS patients had a higher fraction of girls than the control group. Plasma bilirubin concentration tends to be equal in both sexes or slightly higher in boys than in girls 1211. Binding capacity of albumin, measured by the HBABA method, is sex-independent [22] as is the reserve albumin concentration for binding of MADDS (Ebbesen, F., unpublished data). The higher bilirubin concentration and lowler reserve albumin and thereby the higher index in the RDS patients compared with the controls, thus cannot be related to the different ratio of sexes in the two groups. Up till now, clinical determinations of the index of plasma bilirubin toxicity have shown the following. Indices are negative in healthy adults [23]. Positive index values usually in the range O-l are found in healthy term newborns [ 12,161, showing that precipitation of crystalline bilirubin, but not of amorphous bilirubin, is thermodynam.ically possible. Higher indices are present in healthy preterm infants and even higher in preterm infants with RDS as shown in the present paper. Levels above 1.X are reached in very few sick preterm newborns, with RDS or with rhesus hemolytic disease [14], in whom the faster process of amorphous bilirubin deposition thus is theoretically possible. Indices of plasma bilirubin toxicity are further decreased by exchange transfusion [ 141, by phototherapy [12,13] and by giving albumin [15,16]. Intravenous dosage of a weakly bilirubin displacing drug. ampicillin, results in a slight increase of the index [7].

Acknowledgements The authors wish to thank Professor B. Friis-Hansen for guidance during the study. [14C]MADDS was prepared by C. Jacobsen, Ph.D.; Inger Bonnevie is thanked for technical assistance, and the staffs of the Departments of Neonatology and Clinical Chemistry for assistance. This study was supported by the Danish Medical Research Council (Grant Nos.

354

512-10626, Izumsfond,

512-10767, and and by Mmand

512-15538), Nordisk Gjenforsikrings Selskabs i Odense Johann og Hanne Weimanns Legat.

Jubi-

References of the vacant amount of high-affinity bilirubin binding site on 1 Brodersen, R. (1978): Determination serum albumin. Acta Pharmacol. Toxicol., 42, 153- 158. with albumin and phospholipid. J. Biol. 2 Brodersen, R. (1979): Bilirubin solubility and interaction Chem.. 254. 2364-2369. 3 Brodersen, R. (1979): Binding of bilirubin to albumin. CRC Crit. Rev. Clin. Lab. Sci., 11, 305-399. 4 Brodersen. R. (1980): Bilirubin transport in the newborn infant. reviewed with relation to kernicterus. J. Pediatr.. 96. 349-356. 5 Brodersen, R., Andersen. S., Jacobsen, C., S&derskov. 0.. Ebbesen. F.. Cashore, W.J. and Larsen, S. (1982): Determination of reserve albumin for ligand-binding, probing two distinct binding functions of the protein. Anal. Biochem., 121, 395-408. 6 Brodersen. R. and Bartels. P. (1969): Enzymatic oxidation of bilirubin. Eur. J. Biochem.. 10. 468-473. 7 Brodersen, R. and Ebbesen, F. (1982): Bihnrbin displacing effect of ampicillin, indomethacin, chlorpromazine, gentamicin and antibacterial additives in vitro and in newborn infants. J. Pharmacol. Sci.. in press. 8 Brodersen, R. and Jacobsen. J. (1969): Separation and determination of bile pigments. Methods Biochem. Anal., 17, 31-54. 9 Cowger. M.L. (1971): Mechanism of bilirubin toxicity on tissue culture cells: Factors that affect toxicity. reversibility by albumin, and comparison with other respiratory poisons and surfactants. Biochem. Med., 5, 1-16. 10 Diamond, I. and Schmid, R. (1966): Experimental bilirubin encephalopathy. The mode of entry of bilirubin-14C into the central nervous system. J. Clin. Invest., 45, 678-689. 11 Dybkzr, R. and Hertz, H. (1970): A reference for determination of bilirubin concentration in serum using bilirubin in cyanide-formamide for enrichment of serum. Stand. J. Clin. Lab. Invest., 25, 151-160. 12 Ebbesen. F. (1981): Bilirubin, reserve albumin for binding of bilirubin and pH in plasma during phototherapy (ordinary and double light) of term newborn infants. Acta Paediatr. Stand., 70. 223-228. 13 Ebbesen. F. (1981): The relationship between serum bilirubin and reserve albumin for binding of bilirubin during phototherapy of preterm infants. Acta Paediatr. Stand., 70. 405-407. 14 Ebbesen, F. (1981): Effect of exchange transfusion on serum reserve albumin for binding of bilirubin and index of serum bilirubin toxicity. Acta Paediatr. Stand., 70, 643-648. 15 Ebbesen, F. and Brodersen, R. (1981): Albumin administration combined with phototherapy in treatment in low-birth-weight infants. Acta Paediatr. Stand., 70, 649-653. 16 Ebbesen, F. and Brodersen, R. (1982): Comparison between two preparations of human serum albumin in treatment of neonatal hyperbilirubinaemia. Acta Paediatr. Stand., 71, 85-90. 17 Ebbesen. F. and Moller. J. (1976): Blue double light. Improved method of phototherapy. Arch. Dis. Child.. 51. 476-478. 18 Eriksen. E.F., Danielsen, H. and Brodersen, R. (198 I): Bilirubin-liposome interaction: Formation of a bilirubin dianion-phospholipid complex, followed by protonization and bilirubin acid aggregation. J. Biol. Chem., 256, 4269-4274. 19 Gairdner. D. and Pearson, J. (1971): A growth chart for Child., 46. 783-787. 20 Gartner, L.M., Snyder, R.N., Chabon, R.S. and Bernstein, premature infants with low serum bilirubin concentrations. 21 Hugh-Jones, K., Slack, J., Simpson, K., Grossman, A. and hyperbilirubinemia in premature infants. N. Engl. J. Med..

premature

and other

infants.

Arch.

Dis.

J. (1970): Kernicterus: High incidence Pediatrics, 45, 906-917. Hsia, D.Y.-Y. (1960): Clinical course 263, 1223-1229.

in of

355

22 Johnson, L. and Boggs. T.R. (1974): Bihrubin-dependent brain damage: Incidence and indications for treatment. In: Phototherapy in the Newborn: An overview. pp. 122-149. Editors: G.B. Odell. R. Schaffer and A.P. Simopoulos. National Academy of Sciences. Washington. DC. 23 Jrarnerot. G.. Andersen. S., Esbjorner. E.. Sandstrom, B. and Brodersen. R. (1981): Albumin reserve for binding of bilirubin in maternal and cord serum under treatment with sulphasalazine. Eur. J. Gastroenterol.. 16. 1049-1055. 24 LaurelI. C.-B. (1966): Quantitative estimation of proteins by electrophorests tn agarose gel containing antibodies. Anal. Biochem.. 15. 45-51. 25 Levine, R.L. (1979): Bilirubin: Worked out years ago? Pediatrics. 64. 380-385. 26 Lund. H.T., Gudum, J. and Kop. V. (1974): The “blue” meter, a photometer designed to measure light emission during phototherapy of hyperbilirubinaemia. Acta Paedtatr. Stand.. 63. 89-91. 27 h4cDonagh. T. (1981): Inernational Workshop on Pigment Gallstone Disease. Philadephia. PA. 28 McDonagh. A.F. and Assisi, F. (1972): The ready isomerization of bilirubin 1X-o in aqueous solution. Biochem. J.. 129. 797-800. 29 Mertz, J.E. and West. C.D. (1956): A rapid micromethod for the determinatton of indirect bilirubin. Am. J. Dis. Child.. 91. 19-22. 30 Slberberg. D.H.. Johnson, L. and Ritter, L. (1970): Factors influencing toxxity of hilirubin in cerebellum tissue culture. J. Pediatr.. 77. 386-396. 31 Wennberg. R.. Law M. and Rasmussen. L.F. (1976): Clinical significance of unbound bilirubin. Pediatr. Res.. IO. 434.