Main

Several years ago we published in this journal our observation that in the human neonate immediately after birth the plasma creatinine levels in preterm babies are significantly higher than in term infants (1). This difference persisted until approximately 3 wk after birth, much longer than could be accounted for by the maternal passage of creatinine, a small molecular mass substance (133 D, 0.3-nm radius) which is freely filterable at the placental barrier. In mammals urinary creatinine excretion is determined by glomerular filtration and to some extent by tubular secretion (2,3). Due to this tubular secretion of creatinine the creatinine clearance normally overestimates the inulin clearance, the gold standard of GFR. Two studies in neonates (piglets and humans) (4,5) seem to indicate that the creatinine clearance in these subjects underestimates GFR, suggesting that in the newborn there is tubular reabsorption of creatinine. The present studies were undertaken to evaluate the dynamics of the creatinine clearance in the newborn. The newborn rabbit is a good model for the study of renal function of the human premature infant, because in the rabbit nephrogenesis continues until approximately 10-14 d of postnatal life, thus reflecting the renal immaturity present in the premature human infant (6–8)

METHODS

Two comparable groups of 5-10-d-old newborn (groups 1 and 2) and a group of adult (group 3) New Zealand White rabbits were studied.

Preparation. In the newborn animals anesthesia was initiated with an intraperitoneal dose of 25 mg/kg of a 0.5% sodium pentobarbital solution; the adult animals received an i.v. dose of 30 mg/kg. Additional, small i.v. doses were given when indicated. All animals were mechanically ventilated through a tracheotomy with a rodent ventilator (Harvard 683, Millis, MA). The respiratory rate was kept constant at 40 breaths/min, and tidal volume was adjusted for age and weight. With a heating table and an infrared light, the body temperature of the animals was kept constant between 38.0 and 38.5°C, monitored with an intraesophageal thermometer (Hewlett Packard temperature recorder 78231C, Palo Alto, CA). The femoral artery and vein were catheterized with appropriately sized tubing for blood sampling, fluid administration, as well as measurements of mean arterial blood pressure and heart rate. The bladder was catheterized for urine collection. Statham transducers recorded arterial and ventilatory pressures (model 7B polygraph, Grass Instruments, Quincy, MA).

After surgery the animals received an i.v. priming dose of inulin (100 mg/kg), followed by a continuous infusion (10 mL kg-1 h-1) of a modified rabbit Ringer's solution containing per liter 50 g of mannitol, 3 g of inulin (groups 1 and 2), or 20 g of inulin (group 3), 150 mmol of sodium, 105 mmol of chloride, 5 mmol of potassium, and 50 mmol of sodium bicarbonate. The preparation of the animals took approximately 1 h. All rabbits were studied under normoxemic conditions. None of the rabbits was exposed to drugs or chemicals that are known to interfere with tubular creatinine transport. After the surgical procedures, a 90-min equilibration period was allowed, followed by four 30-min (groups 1 and 2) or 20-min (group 3) urine collections. Blood (0.4 mL) was withdrawn at the midpoint of each urine collection. The red blood cells were reconstituted in 0.4 mL of human albumin and reinfused.

Animals. Group 1 (n = 9): the animals in this group of neonatal animals had a weight of 116.8 ± 7.8 g (X ± SEM) and a basic plasma creatinine level of 30-50 µmol/L (0.34-0.45 mg/dL), which was artificially raised to 74.4 ± 7.8 µmol/L (0.85 ± 0.02 mg/dL) by the addition of creatinine (Sigma Chemical Co., Buchs, Switzerland) to the constant infusion (vide supra), both as a loading dose (330 µmol/L) and as a sustaining dose (0.33 µmol/L). Group 2 (n = 11): these newborn animals had the same weight (116.0 ± 18.5 g) and the same basic plasma creatinine levels (30-50 µmol/L, 0.34-0.45 mg/dL) as the animals in group 1, but their plasma creatinine level was raised to 155.0 ± 3.9 µmol/L (1.75 ± 0.04 mg/dL) with the doubling of the exogenous administration of creatinine (loading dose, 660 µmol/L; sustaining dose, 0.66 µmol/L). Group 3 (n = 8): the adult animals in this group weighed 2100 ± 250 g; their baseline plasma creatinine level was 84.1 ± 1.0 µmol/L (0.95 ± 0.01 mg/dL) (see Table 1). In these adult animals basic endogenous creatinine and exogenous inulin clearance were determined; no exogenous creatinine was added.

Table 1 The weight, enhanced plasma creatinine levels and creatinine and inulin clearances (creat-clearance, inul-clearance) in two groups of newborn (group 1 and 2) and a group of adult (group 3) New Zealand White rabbits
Full size table

Laboratory techniques. Creatinine and inulin in blood and urine were measured with the creatinine kinetics (BioMérieux, Lyon, France) and the anthrone methods, respectively (9,10).

Statistics. All data are expressed as mean ± SEM. The significance of the changes were tested with the paired t test. A p < 0.05 was considered significant. The clearance values were corrected for kilograms of body weight (11).

RESULTS

The results are summarized in Table 1. The addition of a relatively small amount of exogenous creatinine raised the plasma creatinine level of the animals of group 1 from a value of 30-50 µmol/L (0.34-0.45 mg/dL) to 74.4 ± 7.8 µmol/L (0.85 ± 0.02 mg/dL) (Table 1). Doubling the amount of creatinine in the infusion of the animals of group 2 more than doubled (×2.1) the plasma creatinine level with only a relatively small rise in the creatinine clearance (+17.7%). This latter value (1.33 ± 0.17 mL min-1 kg-1) was still much lower (a factor of 2.6) than the creatinine clearance observed in the adult animals of group 3 (Table 1). The most interesting findings of these experiments, however, were obtained when the creatinine and inulin clearances in each group of animals were compared. In the adult animals of group 3 the creatinine clearance was 21% greater than the inulin clearance, the universally accepted best measurement of GFR. In these animals the creatinine clearance thus overestimated true GFR. In contrast, the creatinine clearance underestimated GFR in the newborn animals of group 1 (-16%). This is graphically depicted in Figure 1. In the neonatal animals of group 2 the creatinine and inulin clearances were virtually identical (-4%). The mean arterial pressure and the heart rate were stable in all experiments.

Figure 1
figure 1

The relationship between the clearances of creatinine and inulin in newborn and adult rabbits with comparable plasma levels of creatinine.

Full size image

DISCUSSION

GFR in the human neonate and in newborn animals such as the rabbit is very low, even corrected for kidney mass, body surface, and total body water (12). This major limitation of neonatal renal function poses great practical difficulties in measuring GFR at this very young age. Inulin is the gold standard for the measurement of GFR at any age. For matters of convenience, however, the endogenous creatinine clearance is generally used in clinical practice and clinical research (13–15). One of the basic requirements for the validity of the creatinine clearance is that plasma creatinine levels remain constant during a finite period of time. Under steady state conditions the plasma creatinine level is dependent on total body muscle mass, which in older children and healthy adults does not change appreciably. This is not true in the rapidly growing neonates in whom the stable periods of creatinine production are rather short. This is even more pronounced in premature babies whose plasma creatinine levels are much higher than can be expected for their body size (1). Despite these basic limitations, even in the neonate the short time stable plasma creatinine level and the fact that creatinine is completely filtered at the glomerular filtration barrier, come close to fulfil the criteria for the two main conditions for an ideal marker of GFR. The third condition, that creatinine is not handled by the tubule, is less well controlled at all ages. In the adult, the endogenous creatinine clearance overestimates GFR by a factor of 20-40%, although this figure is rather variable (2,3). The overestimation is due to tubular secretion of creatinine. Both Smith (2) and Pitts (3) emphasized, and this is clearly reiterated in a modern textbook of physiology (16), that when a certain substance is only filtered and not reabsorbed nor secreted along the renal tubule, the clearance of that substance equals the clearance of inulin. To quote Guyton and Hall (16): "If the clearance rate of the substance is greater than the inulin clearance, the substance must be secreted by the nephron tubules and if the clearance rate is less than the inulin clearance, the substance must have been reabsorbed."

In the present experiments, the normally occurring overestimation of the creatinine clearance was indeed seen in the adult animals of group 3 (Table 1). In the neonatal animals, however, the opposite phenomenon was observed. At a level of plasma creatinine, which was artificially raised to be in the same range as that of the adult animals, a definite underestimation of the GFR was found (group 1). The only valid explanation for this finding is that in the neonate creatinine is reabsorbed along the tubule. When the already high plasma creatinine level in the newborn animals of group 1 was even raised further (group 2), the creatinine clearance increased relatively little, whereas at this point there was virtually no difference between the clearances of creatinine and inulin. Because plasma creatinine doubled and the urine flow remained unchanged (the amount of exogenous creatinine was doubled, not the rate of the infusion), the urinary excretion of creatinine must have almost doubled as well. In other words, the tubular reabsorption of creatinine had previously already almost reached its maximal capacity, and after the doubling of the plasma creatinine level the remainder of the filtered load of creatinine was excreted unchanged in the final urine sample.

Tubular reabsorption of creatinine is a very unusual physiologic phenomenon. There are, however, previous data which support our experimental findings. In studies on fetal piglets, undertaken to evaluate the existence of glomerulotubular balance in the developing kidney, Alt et al. (4) state that net tubular reabsorption of creatinine was found. In premature and term infants Coulthard et al. (5) also observed that creatinine clearances were lower than inulin clearances; these authors attributed their findings to the techniques used for measuring creatinine. In dehydrated human adult volunteers with a low urinary flow, tubular reabsorption of creatinine was also observed (17).

The present study is the first to unequivocally show, in an animal experimental setting, that immediately after birth the creatinine clearance underestimates "true" GFR. These data, which were obtained by the simultaneous measurement of creatinine and inulin clearances in newborn rabbits, are physiologically applicable to premature human infants. It may well be that tubular reabsorption of creatinine is a physiologic phenomenon in the "immature" developing kidney. We propose that this is due to a slow urinary flow along the tubule at this age and in particular to increased back-leak of creatinine along the immature tubular structures.

It is refreshing to find that in this day and age of ever more sophisticated biologic study methods, an astute clinical observation can still be "answered" by time-proven basic renal physiologic studies. For obvious reasons these studies had to be performed in an animal model and could not be verified in the human neonate. The reasons why and precisely how creatinine is reabsorbed along the "immature" tubules need further study.