Plasma hypoxanthine in neonatal hypoxia: a comparison of two methods

EUROP. J. OBSTET. GYNEC. REPROD. BIOL., 1978,8/2,89-94 o Elsevier/North-Holland Biomedical Press

Plasma hypoxanthine in neonatal hypoxia: a comparison of two methods P. Guicheney I*, J.-R. Zorn I, E. Rey l, C. Sureau 2 and G. Olive l 1Laboratoire de Biochimie, H6pital Saint-Vincent-de-Paul,74 Avenue Denfert-Rochereau, and 2 Clinique Universitaire Baudelocque, 123 Boulevard de Port-Royal, 75014 Pan’s,France

GUICHENEY, P., ZORN, J.-R., REY, E., SUREAU, C. and OLIVE, G. (1978): Plasma hypoxanthine in neonatal hypoxia: a comparison of two methods.Europ. J. Obstet. Gynec. reprod. Biol., 8/Z, 89-94. Hypoxanthine levels were determined in both venous and arterial cord blood of 42 neonates. Two methods were compared, a Po2 electrode determination and an HPLC (high-pressure liquid chromatography) method. A good correlation was found between the two methods. However, the HPLC method was more sensitive, more reproducible and easier to perform. Hypoxanthine levels in the umbilical artery were found to be higher than in the vein. A significant negative correlation between pH and hypoxanthine level was established. The studies showed that plasma hypoxanthine levels by themselves did not provide an absolute diagnosis of intrauterine hypoxia. hypoxia; fetus; oxypurines; liquid chromatography

intrauterine hypoxia during or after labor. The level of plasma HX has been determined by two methods, the first one described by Saugstad (1975b) using a PO* electrode and the second a direct method using high-pressure liquid chromatography (HPLC). The results obtained with these two methods are compared in the present study. In addition, the use of HPLC permitted a rapid determination on microquantities of blood. The micromethod is of potential interest for determining hypoxanthine in small quantities of scalp vein blood samples during labor. The object of the present study was to discover if there was any correlation between plasma HX levels and clinical and biochemical parameters of hypoxia.

Introduction

Fetal hypoxia is usually assessed either by clinical symptoms, for example fetal heart rate deceleration, or by biochemical changes such as a decrease of pH. Estimation of rapidly reversible parameters on blood samples gives an inadequate idea of the degree of hypoxia since they may be normal at the time of sampling between protracted periods of hypoxia. A more useful index of hypoxia would be a value which remains altered for some time after restoration of normal PO,. Hypoxanthine (HX), a degradation product of red cell adenosine triphosphate (ATP), accumulates in plasma during hypoxia and, following reoxygenation, only returns gradually to basal levels. Plasma HX could therefore be a useful index of

Material and methods * Present address and address for reprints: INSERM U7, Physiologie et Pharmacologic, Hdpital Necker, 161 Rue de S&es, 75015 Paris, France.

A simple method for determination of the hypoxanthine concentration in plasma was described by 89

P. Guicheney et al.: Plasmahypoxanthine in neonatal hypoxia

90 X&the

oxidm

Xanthina

Hypox.

axidass

Theoph.

Mi$fT)T

$& n,o*ol

HVPOXANTHINE

!4102 ~wv202 XANTHINE

Fig. 1. Degradation of hypoxanthine

fJ& n2o*ri2

w202 L W~tYzn2 URIC ACIO

to uric acid.

Saugstad (1975b). The method is based on the principle that .the oxygen is consumed quantitatively when hypoxanthine is oxidized to urate in the presence of xanthine oxidase. Figure 1 shows that when one molecule of hypoxanthine is metabolized to uric acid, one molecule of oxygen is utilized. The decrease in plasma partial pressure of oxygen (PO,) is measured with Clark’s electrode (E/5046) connected to a Radiometer pH-meter (BMS 3) and to a recording device (Servograph REC 61, Radiometer). Hypoxanthine concentration can be calculated using Henry’s law which states that dPoz = d(Oz) X K (HX), where K is a constant depending on the medium, and dPr+ is the change in partial pressure of oxygen in a solution where its concentration is changed d(0,). The constant K (determined experimentally) = 0.65 mm Hg/pmol/l. The level of plasma hypoxanthine was also measured by HPLC using a modification of a method used for the determination of purine bases (Whatman, 1976). A high-pressure liquid chromatograph (Varian 8510) was connected to an A2S-1 pen-recorder. The chromatography column was composed of silica cationexchange resin Partisil 10 SC X (4.5 mm X 25 cm), the mobile phase was 0.1 M NH4H2P04, pH 3.5, and the flow rate was 80 ml/h. An ultraviolet Varian detector was used and measurements were made at a wavelength of 254 nm. Column efficiency expressed in height equivalent to a theoretical plate was 0.11 mm. Under the conditions described, hypoxanthine retention time was 2 min 57 set and that of the internal standard, theophylline, was 3 min 30 sec. A standard curve (Fig. 2) was constructed by adding hypoxanthine (lo-70 pmol/l) to Hyland lyophilized serum, and theophylline (96 ~ol/l) to all samples. There was a linear relationship between concentrations of hypoxanthine and peak height hypoxanthine : theophylline ratio. Plasma levels were measured after

Fig. 2. Standard curve using high-pressure liquid chromatography.

a 1 : 2 dilution with the mobile phase containing the internal standard. 5 pmol/l of this mixture was injected. into the column. Specificity of the method was tested by the addition of xanthine oxidase to plasma containing hypoxanthine. The addition of the enzyme resulted in the disappearance of the peak. 42 neonates were studied. Arterial and venous blood samples were taken from the umbilical cord after clamping, immediately after delivery. The plasma was quickly separated by centrifugation (3600 rpm for 5 min) to avoid spurious results (Saugstad, 1975b; Jorgensen and Engelund-Poulsen, 1955). Plasma was stored temporarily at 4°C until hypoxanthine measurement. Blood samples were also taken and analysed by pH-meter (Radiometer), and uric acid was determined using the method of Henry, Sobel and Kim (1957). Clinical profiles were established for the mother and the child at each delivery and the following information was grouped: - signs of intrauterine hypoxia, i.e. fetal heart rate deceleration whatever the type, staining of the amniotic fluid during labor, Apgar score, cord pH; - fetal presentation and details of labor (Table III); - drugs taken during the third trimester of pregnancy and during labor.

Results

The hypoxanthine measurements obtained by the two methods were analysed by linear regression. Re-

P. Guicheney

et al.: Plasma hypoxanthine

in neonatal

hJspoxia

pmoles/t

Art.

“en. hypoxanthlne (H.F! L.C.)

cone

y = 0.65x + 2.99 ” = 29 rE0.861 p c 0.00,

1

cont. Art. hydoxanthine (~02 electrode method)

Fig. 3. Correlation of arterial hypoxanthine obtained by the two methods.

ZOpmoles/l

VW. hypoxonthine cont. (~0~ electrode method)

concentrations

sults obtained on arterial or venous blood were positively correlated. The regression coefficients and the lines of linear regression are shown in Figures 3 and 4. Comparison of the arterial levels by Student’s paired t-test showed that arterial concentrations measured with either technique did not differ significantly (t = 0.456, P = NS); the same results were found for venous concentrations (t = 1 S78, P= NS). The HPLC method is 5 times more sensitive and 4-5 times more precise (measured as the ratios of the coefficients of variation) than the PO* electrode method (Table I). The difference in sensitivity was confirmed by Fisher’s F-test by which the probability that the variance of the two methods was the same was less than 0.001. Arterial levels of hypoxanthine were found to be higher than venous levels. An arterio-venous difference of 4.6 + 1.5 pmol/l (n = 26) was found using Saugstad’s method and of 3.5 + 1.1 pmol/l (n = 15)

Fig. 4. Correlation of venous hypoxanthine by the two methods.

levels obtained

by HPLC. A decrease in blood pH accompanied the increase of hypoxanthine plasma level. The regression coefficients are given in Table II. The results are grouped according to signs of intrauterine hypoxia and other criteria (Table III). The mean hypoxanthine levels calculated in each group were compared to the normal level, 8.5 +_1.6 pmol/l (n = 15), by the use of Student’s unpaired r-test (the normal value is the mean of arterial hypoxanthine levels found in neonates showing no sign of intrauterine hypoxia, having a blood pH superior to 7.20 and an Apgar score at 1 min greater than 7). Where the Apgar score was less than 7, where pH was less than 7.20 and when forceps extraction was required, the mean hypoxanthine values were twice normal, respectively 17.6, 17.5 and 16.3 pmol/l. The lack of statistical significance, however, reflects the small sample size of each group and their wide range. It is noteworthy

TABLE I Comparison of the precision and the sensitivity of the two methods _____..__~_ Method used

Sensitivity (fimol/l)

Precision n

--

.-.- ~ ~... ...~~~

~~~~ _

Zi(fimol/l)

SD(pmol/l)

cv

P

-__

PO* electrode 5 20 18 3.1 0.17
P. Guicheney et al.: PIasma hypoxanthine in neonatal hypoxia

92

TABLE II Correlation between pH and hypoxanthine cord plasma levels _____. __ _.... ..__ --_. --. --. ------- ------ --.---Method used No. of cases (n) Regression coefficient (r) _.-________ .._-. --._.-__pH arterial hypox. Pea electrode 32 -0.35 pH arterial hypox. HPLC 17 -0.65 pH venous hypox. pH venous hypox.

TABLE III

Pea electrode HPLC

37 30

___- ._.- -.-

GO.01 GO.05

levels and clinical features

-~

.- -.

Criteria for intrauterine hypoxia: Apgar score <7 (at 1 min) pH <7.20 Meconium staining Siiitly stained fluid Fetal heart rate deceleration

-

HX mean concentration 0 knoV1)

17.6 17.5 11.4 10.0 1.8

Factors which may influence hypoxanthine levels: ROP (right occipito-posterior) 8.5 LOP (left occipito-posterior) 34.4 Cord around the neck or shoulder 8.7 Forceps 16.3 Cesarean section before labor 9.3 Artificial induction of labor 11.4 Episiotomy 14.4 Oxytocin 11.4 Pethidine 12.8 Dipropyline 16.0 ~---_ ._---.-___n = number of cases; t = Student’s coefficient.

GO.05
-0.42 -0.38

Correlation between plasma arterial hypoxanthine

__..~------

P

___

that the mean value in cases of fetal heart rate deceleration is the lowest found. Drugs currently used during labor, such as oxytotin, pethidine chlorhydrate and dipropyline, did not modify plasma hypoxanthine level. No correlation was found between neonatal hypoxanthine and uric acid blood concentrations.

Discussion

The HPLC method gave more reproducible results, even at low levels, than the PO, electrode technique. HPLC thus appeared to be the best method. It was rapid, sensitive, and used a small volume of blood.

n

Standard error

Range of values

OlmoVI)

Olmolll)

---

r

P

NS

-_--

5 6 3 3 6

5.1 6.5 3.3 3.0 3.4

6.1-34.4 0 -38.6 6.1-17.4 4.5-15.0 0 -22.0

1.3458 0.9163 0.7080 0.3681 0.1292

11 1 4 5 3 19 19 23 17 I

2.1

0 -20.5

0.0097

NS

1.9 5.8 4.6 2.4 4.0 2.3 2.9 5.1

6.1-14.4 6.1-38.6 1.5-17.4 0 -22.0 0 -38.6 0 -42.4 0 -42.4 0 -42.4

0.0525 1.0215 0.1709 0.2413 0.3024 0.2113 0.3275 0.8083

NS NS NS

_-_.---~-

NS NS NS

NS

NS

NS NS

NS NS

HPLC results correlated well with the PO* electrode technique. The marked arterio-venous difference observed with both methods suggests a significant degree of placental hypoxanthine clearance. Since the control of such clearance is not yet well defined, it seems to be more appropriate to determine hypoxanthine levels in arterial cord blood. This is likely to be more representative of values in the circulating blood of the fetus. A negative correlation between pH and hypoxanthine blood concentrations was found, but when the children were separated into three groups (I’able IV), I (nonhypoxic), II (slightly hypoxic) and III (severely hypoxic), according to Lipp, Tuchschmid, Silberschmidt and Due (1976), no significant increase

P. Guicheney et al.: Plasma hypoxanthine in neonatal hypoxia TABLE IV

Cord arterial plasma hypoxanthine levels according to the classification of Lipp et al. (1976)

No.

Group

of cases

Hypox. mean (fimol/ 1)

Standard error (pmol/ 1)

I

Neonates without detected hypoxia

15

8.55

t1.47

II

Neonates with pH <7.18, meconium staining or bradycardia

10

9.15

+2.22

III

Neonates with pH <7.18, Apgar score (4, meconium staining and/or bradycardia

7

15.4

k5.84

-.--_

in hypoxanthine levels induced by hypoxia was apparent. Table V summarizes results obtained by different methods and authors and shows a broad range of results. High levels were found in some neonates showing no sign of intrauterine hypoxia and, conversely, low levels were found in hypoxic neonates. The high levels could be explained if hypoxanthine is a more sensitive parameter of intrauterine hypoxia than all other recognized criteria. There may, how-

TABLE V Authors

Range of concentrations of cord plasma hypoxanthine ~____ -- _-__ No. of cases Method used

_-___~ Neonates with a normal Saugstad (1975a) Lipp et al. (1976) Present study

delivery: 29 15 15 17 10 17

PO, electrode Paz electrode PO? electrode PO* electrode HPLC HPLC

Neonates showing signs of intrauterine hypoxia: Saugstad (1975a) 12 PO, electrode Lipp et al. (1976) PO* electrode Present study 17 PO? electrode 19 Paz electrode 9 HPLC 13 HPLC --___-__ ..-- ~_

93

ever, be another explanation for the low hypoxanthine levels found. During hypoxia, red cell ATPis degraded and plasma hypoxanthine levels increase. The release of hypoxanthine diminishes as red cell ATP is catabolized. When hypoxia is long-lasting, plasma hypoxanthine could be progressively reduced as a result of this diminished release in the face of continued excretion and metabolism by placental transfer or hepatic xanthine oxidase, even though this enzyme is partly inhibited by hypoxia. If this were the case, plasma hypoxanthine concentration would fall and this could explain the low hypoxanthine levels found in certain hypoxic neonates. With reoxygenation, hypoxanthine reenters the red blood cells to form ATP. However, with long-lasting hypoxia the low level of plasma hypoxanthine does not permit the reestablishment of normal ATP level in red blood cells and this could explain why these children are more seriously affected by hypoxia and need more time to recover. In order to regain normal ATP values a de novo synthesis of purine is necessary. The low levels in fetal heart rate deceleration support this hypothesis. This suggests that high hypoxanthine levels may be found in children showing a short but severe hypoxia while low levels may represent either unaffected cases or moderate but long-lasting fetal hypoxia. This could explain the fluctuations found in the different groups of children. Systematic analysis should be continued. At the

Origin of blood

Range of concentration (CtmoVl)

Arterial + venous Arterial Arterial Venous Arterial Venous

0 -11 12 -15 1.5-20.4 0 -15.9 1 -19 3 -17

Arterial + venous Arterial Arterial Venous Arterial Venous _____

6 6 0 0’ 6 2

-61.0 -27 -38.6 -19.7 -28 -15

94

P. Guicheney et al.: Plasmahypoxanthine in neonatal hypoxia

present time, hypoxanthine plasma levels cannot be used as a routine clinical test but the use of HPLC in research can enable determinations to be made of hypoxanthine levels in fetal scalp vein blood, thus permitting a closer examination of the pattern of change of plasma hypoxanthine during the different phases of labor.

of uric acid and comparison with the spectrophotometric uric% method. Amer. J. clin. Path., 28, 152-i60. Jorgensen, S. and Engelund-Poulsen,H. (1955). On accumulation of hypoxanthine and xanthine in human plasma and urine.Acta pharmacol. (Kbh.), II, 287. Lipp, A., Tuchschmid, P., Silberschmidt, M. and Due, G. (1976): Arterial cord blood hypoxanthine and intrauterine hypoxia. fioceedings, V European Congress of PerinatalMedicine, Uppsak Saugstad, O.D. (1975a): Hypoxanthine as a measurement of hypoxia. Pediat. Res., 9, 158-161. Saugstad, O.D. (1975b): The determination of hypoxanthine and xanthine with a PO* electrode. Pediat. Res., 9, 575579. Whatman (1976): Bulletin No. 116.

References Henry, R.J., Sobel, C. and Kim, J. (1957): A modified carbonate-phosphotungstate method for the determination