INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

EXAGGERATED ARTERIAL pressure fluctuations are associated with morbidity, especially for extremely premature human neonates (15). Attempts to examine the development of baroreflex regulation have focused on heart rate and arterial pressure responses in term and preterm newborns to whole body tilting. Data are inconsistent, suggesting variation in levels of baroreflex control (12, 22, 25). Cardiac output is highly sensitive to changes in central venous filling pressure (11, 20), but the contribution of the developing venous system to these observations on the postural baroreflex has not been investigated.

Deviations in central venous pressure with body reorientation were recognized in the 19th century to relate directly to distance from the hydrostatic indifference point (HIP) (24). The HIP is a unique axial reference within the column of venous blood where pressure is not altered by a given postural reorientation. HIP location is determined by the distribution of venous mural compliance, which, in turn, has many potential controlling mechanisms (6). The venous HIP of adult humans resides just below the cupola of the diaphragm (6), and in the mature dog it is within the right ventricle (8). This proximity of the HIP to the heart ensures relative stability of cardiac filling and stroke volume despite changes in body posture.

Organisms have evolved a degree of postural baroreflex control commensurate with the requirements for survival in their natural habitats. For example, head-up tilt is poorly tolerated in aquatic reptiles taken out of water compared with closely related terrestrial and especially arboreal species (19). In humans, postural reflexes need to be maintained by regular conditioning, as demonstrated by the tendency to orthostatic hypotension after prolonged bed rest or spaceflight (2, 5). The veins of the late-gestation human fetus are buoyed by pressure transmitted by the amniotic fluid and so are not exposed to orthostatic strain (14). We postulated that, should an active venous pressure control system exist in mammals, at birth it will be poorly conditioned to the extrauterine environment. Native maturation of such a system would be evidenced by improved stability of cardiac filling pressure during body tilt, marked by postnatal shifting of the venous HIP toward the heart.

The piglet is a well-established model for the investigation of human neonatal cardiovascular control (7). In this study, changes in the HIP were traced in the piglet by using both head-up and head-down tilt during the first 6 days of life and after lethal injection. As there are no previous data describing HIP location in the living neonate, a range of tilts was employed to determine the effect of tilt magnitude and direction on HIP location. A separate group of piglets was studied on postnatal day 5, before and after autonomic blockade, to establish the influence of adrenoceptor and cholinoceptor mechanisms on venous HIP location.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Eight domestic piglets, opposite-sex pairs from four litters, underwent surgery within 12 h of birth. They were studied after recovery and also on postnatal days 3 and 6. Four other piglets, two of each sex, were studied on day 5 before and after drug-induced total autonomic blockade (TAB). The experimental procedures and care of the piglets were in accord with the Code of Practice adopted by the Animal Care and Ethics Committee of the University of Newcastle.

Tilt-table design. The rigid cradle that supported the piglets in these experiments was fashioned from a polyvinyl chloride pipe (75 mm diameter, 500 mm length) cut longitudinally to produce a U-shaped trough. Four oval-shaped holes were located to accommodate the legs, and multiple 8-mm holes were drilled at 10-mm centers to provide skin traction and allow visualization of a cross marking, on the lateral thorax, the center of the piglet's right atrium. The entire cradle was heat molded and padded so as to provide snug support; urinary balloon catheters were incorporated into the cradle front and back so that, when the catheters were inflated with warm water, either the buttocks or shoulders were supported during tilting. These measures ensured that the piglet shifted <5 mm relative to the plastic cradle over the entire range of head-up and head-down tilt.

Surgery. Piglets were anesthetized with 1% isoflurane in oxygen via a close-fitting mask. An umbilical artery was catheterized by using a 3.5-Fr polyvinyl chloride umbilical vessel catheter (Argyle, St. Louis, MO) inserted 1 cm further than the distance between the umbilicus and shoulder tip. This procedure is the same as that widely performed on sick human neonates, and ultrasound confirmed that it also results in the catheter tip lodging within the midthoracic aorta in the newborn piglet. The right external jugular vein was dissected through a 1-cm incision in the root of the neck and another 3.5-Fr catheter was inserted 4-5 cm. A double-lumen catheter (4-Fr, Arrow Pediatric, Reading, PA) was used for the venous catheterization in the four piglets who underwent the autonomic blockade experiments to allow simultaneous drug infusion and central venous pressure monitoring. The venous catheter was tunneled subcutaneously to emerge between the scapulae. Both catheters were primed and flushed daily with saline containing heparin (2 U/ml). Penicillin (60 mg/kg) and gentamicin (2.5 mg/kg) were given preoperatively, and the surgery was completed within 30 min. Piglets recovered in the neutral thermal environment of a Humidicrib. They were fed milk formula every 4 h (Di-Vetalact, Sharpe Laboratories) via a bottle and nipple.

Determination of location of the right atrium. Immediately after the surgery, the center of the right atrium, as a point of reference, was determined by two-dimensional ultrasound, and its location was marked on the lateral chest in two planes by using an indelible marker pen. The locations of the venous and arterial catheters were checked by ultrasound to confirm their tips were in the right atrium and midthoracic aorta, respectively.

Protocol. Before each experiment, piglets were given intravenous (iv) midazolam (Roche; 0.1 mg/kg body wt); this rendered them less excitable but still able to walk and drink. They were loaded into the polyvinyl chloride cradle, which was then attached to the tilt-table frame by locating bolts. The arterial and venous pressure transducers (DT-NN, Ohmeda, Madison, WI) were calibrated before and after each experiment by using mercury and saline manometers, respectively. These lightweight transducers were chosen because their small domes allowed them to be accurately centered at a desired point relative to the piglet. They have a stable and linear response over the pressure ranges of interest (±20 cmH₂O for venous pressure and 0-100 mmHg for arterial pressure).

Autonomic blockade. In four additional piglets, - and -adrenoceptor and cholinoceptor blockade were employed on day 5 to investigate the role of autonomic control on the HIP. The blocking procedures were those used in the newborn lamb by Jones et al. (13). Restated briefly, -adrenoceptor blockade was achieved with iv propranolol (2 mg, Inderal, ICI) followed by an infusion at 0.17 mg/min through the side port of the double-lumen central venous catheter; the block was confirmed by observing ablation of the tachycardic response to boluses of iv isoproterenol (1 µg, Isuprel, Winthrop). -Adrenoceptor blockade was achieved with iv phentolamine (Regitine, Ciba) by 2-mg bolus injection and 0.32 mg/min infusion; the block ablated the pressor response to iv phenylephrine (125 µg). An iv bolus of the cholinoceptor-antagonist atropine (1 mg, Apex Laboratories) was followed by an infusion at 0.2 mg/min. The drop in arterial pressure due to a test bolus of iv acetylcholine (50 µg, Sigma Chemical, St. Louis, MO) was ablated by the atropine dose. The study protocol was conducted during combined infusion of the three blocking agents; after completion of the tilts, the blockade was again checked by serial bolus injections of the three agonists.

HIP in the deceased piglet. While still in the tilt table on postnatal day 6, piglets were given iv heparin (5,000 U) and killed by rapid iv injection of 1 ml/kg of pentobarbitone sodium (Nembutal, Abbott). The HIP was determined by repeating the tilt protocol, commencing immediately after asystole. During this period of ~20 min, the venous pressure deviations with tilts remained stable and the HIP estimates were highly reproducible.

Signal acquisition and handling. Amplified arterial pressure, central venous pressure, and tilt-angle signals were digitized at 500 Hz and displayed along with heart rate by using the AMLAB computer system (Associative Measurement, Sydney, Australia). Heart rate was derived in real time from the arterial pressure trace by computerized timing of the diastolic minima. Signals were saved to magnetic tape (PCM 4000, A R Vetter, Rebersburg, PA) for further analysis.

HIP determination. The change in venous pressure due to a defined tilt was calculated by deducting the average venous pressure for the 30 s immediately before tilt from that for the 30 s immediately after achieving venous pressure stability in the new tilt angle (Fig. 1).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Central venous pressure (CVP) recordings from a single piglet comparing postnatal days 1 and 6 during 45° (head-up) tilt. Top: inclination of tilt table relative to horizontal as recorded by inclinometer. Middle: 4 CVP traces for each day arranged in order of transducer position relative to right atrium (RA; 10, 5, 0, and 5 cm). Traces with higher control pressures (dotted lines) were recorded on day 6. Bottom: same CVP traces as in middle but offset so that 30-s average immediately before tilt is 0. Superimposed is a plot of change in CVP (ordinate) vs. distance from RA (abscissa). , Day 1; , day 6. Points of hydrostatic indifference (HIP) can be inferred as intersection of regression lines with abscissa. Over first 6 days, HIP migrated in cephalad direction, toward RA.

Data analysis. Data were not included in the analysis if, because of piglet arousal and movement, the venous pressure did not quickly settle to a stable level after tilt. Overall, some 15% of tilt events were excluded from the analysis, but this varied among piglets and on different days; in particular, older piglets were more likely to arouse. Because the HIP can be calculated from the change in venous pressure recorded at a single transducer location, the HIP for each tilt angle on each day is represented in the analysis even though data from some transducer positions are missing. Data are expressed as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Piglets weighed 1.62 ± 0.04 (SE) kg, with body length (snout to rump) 36.1 ± 1.8 cm on day 1; on day 6, the corresponding measurements were 1.72 ± 0.07 kg and 37.1 ± 0.9 cm, respectively. On day 1, the ratio of the distances from the snout to the center of the right atrium and right atrium to rump was 0.47 ± 0.02, whereas in the transverse plane across the thorax, the ratio of the distance from dorsal thoracic spinous process to right atrium and right atrium to sternum was 0.73 ± 0.03.

A 30-s epoch of baseline data was collected while piglets were in quiet sleep before each tilt event; these data segments were averaged on each day to provide resting data (Table 1). Within-piglet comparison revealed a rise in the central venous pressure between days 1 and 3 but no further rise between days 3 and 6. The rising trend in resting arterial pressure evident after day 3 failed to achieve statistical significance (P = 0.07). The resting heart rate was lower on day 3 than on either of the other days.

Analysis of the +45° tilt data from all piglets revealed a cephalad shifting of the HIP toward the heart (Fig. 2). In addition, statistically significant but smaller shifts were evident for the other (+30 and +15°) positive tilts (Fig. 3). In all but one piglet, this shifting of the HIP for head-up tilt (HIP_up) occurred after day 3. With head-down tilt (HIP_down; 30 and 15°), HIP was also caudal to the heart but was not influenced by postnatal age (Fig. 3).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Pooled CVP changes during +45° (head-up) tilt. Values are means ± SE. Piglet outline was traced postmortem from 1 of the piglets, and diaphragm and heart were drawn from a radiograph of same animal. Axes have their origin at center of RA, the mean coordinates of which were determined by ultrasound (see METHODS). Abscissa, drawn to same scale as piglet outline, is distance of transducer from RA, whereas ordinate is change in CVP with 45° tilting. , , and : data for same piglets at 1, 3, and 6 days of age, respectively; arrow, migration of HIP over first 6 days of life.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   HIP distance from RA vs. body tilt from horizontal. Values are means ± SE. Solid lines are postnatal days 1, 3, and 6; dotted line connects measurements immediately after death on day 6. Same-day across-tilt comparisons of HIP estimates are the following: day 1: all different (P < 0.05) except 30 vs. 15 and 30 vs. 15°; day 3: all different except 45 vs. 30, 30 vs. 15, and 15 vs. 15°; day 6: no difference except 45 vs. 15, 15, and 30, as well as 30 vs. 15 and 30°; and deceased day 6: no differences. * HIP significantly different (P < 0.05) from postnatal days 1, 3, and 6 (deceased).  HIP measured after death significantly different from days 1, 3, and 6 (alive).

On all days there were differences in the measured HIP, depending on grade and direction of tilt and the extremes of tilt tested; i.e., +45 and 30° had the most disparate HIP values (Fig. 3). These differences due to the nature of the tilt became smaller with increasing age, reflecting improving HIP stability over the range of tilt stimuli. For instance, the separation of the HIP values for +45° (HIP_up) and 30° (HIP_down) tilt was 3.9 ± 0.3, 3.6 ± 0.7, and 1.3 ± 0.3 cm on days 1, 3, and 6, respectively.

In the deceased 6-day-old piglet, the HIP was the same regardless of the direction or grade of tilt and was more caudal to the heart than on the same day while the piglet was alive by 2.1 ± 0.3 cm (Fig. 3).

Resting data from the separate group of four piglets who underwent TAB with phentolamine, propranolol, and atropine on day 5 are shown (Table 2). TAB caused no significant change in resting heart rate, arterial pressure, or central venous pressure. HIP location was similar to that seen in the other 6-day-old piglets and was not affected by TAB (Fig. 4).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   In 4 piglets (5 days old), HIP for all tilt angles were not significantly different before () and after total chemical autonomic blockade (TAB; ).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Using an appropriate tilt table, we have shown that HIP location can be tracked in a small neonatal animal model without the need for deep anesthesia, which has the potential to ablate cardiovascular responses of interest. HIP_up migrated cranially, converging on both the heart and HIP_down. This appears to be a useful postnatal adaptation as a HIP close to the heart implies minimal changes in cardiac filling during alterations in posture and hence improved cardiac output homeostasis.

Midazolam has been reported to cause a small reduction in arterial pressure and vascular resistance (9) but has no significant effect on central venous pressure (16), nor does the rate of midazolam metabolism appear to be influenced by early postnatal age (3). There was no trend in HIP_up during experiments (which took ~1.5-2 h to complete), and so any age-related change in the piglet's midazolam metabolism does not explain the differences in HIP_up between days.

In 6-day-old piglets, both HIP_up and HIP_down reside some 3 cm caudal to the right atrium. These results are similar to those of previously published studies in adult humans and dogs (6, 8). We thus do not expect that there would be appreciable HIP_up migration after day 6 in the piglet. Because Guyton and Greganti (8) were unable to shift the venous HIP of the dog by spinal blockade, hemorrhagic shock, or epinephrine infusion, they postulated that the juxtaposition of the HIP and heart is a consequence of a direct mechanical relationship between right ventricular filling pressure and stroke volume. For instance, a transient reduction in filling pressure due to head-up tilt would be reversed by a reduction in right ventricular stroke volume and cardiac output. Similarly, any rise in venous pressure with head-down tilt might be attenuated by an increase in stroke volume, thus maintaining the observed HIP in proximity to the heart. In the application of this postulate to the piglet, it might be argued that the increased separation of the piglet's heart and HIP_up in the first 3 days of life might be due to the relative noncompliance of the ventricles at this age (12). Reduced right ventricular compliance would imply that stroke volume adjustments are relatively insensitive to changes in filling pressure, resulting in a caudal HIP_up and HIP_down. HIP_down is not more caudally located in the first 3 days, and so our data would support the proposition that the venous HIP is determined by characteristics of the veins rather than the heart.

Clark et al. (4) described a passive model for venous column hydrostatics. In any inert, fluid-filled container, the HIP marks the arithmetic center of the compliances distributed in the walls of that container. For instance in the simplest model, that of a fluid-filled rigid tube sealed at either end by compliant caps, the HIP will be closest to the end with the most compliant cap and is the same no matter how the tube is reoriented relative to gravity (4). The deceased 6-day-old piglet is analogous to such an inert system in that the same HIP was recorded regardless of tilt direction or grade. The situation in the living animal is more complicated in that the distribution of compliances among the veins may change in response to body tilt and also the pressurization of the veins may be altered by mechanisms that adjust total venous capacity. Notably, in the living animal it cannot be assumed that the measured HIP will be independent of the direction and magnitude of tilt.

In life, constriction of the hindlimb, pelvic, and splanchnic veins during head-up tilt would effect a cephalad shift in the HIP_up for two reasons. First, it would significantly reduce total venous capacity and so enhance average venous pressure in the head-up position. Second, the caudal location of these veins would mean that the reduction in their compliance that accompanies their constriction would result in a cranial shift in the HIP analogous to the effect of stiffening one of the caps in the fluid-filled-cylinder model (4). Thus the caudal veins may well be important in HIP control because they have a large capacity and they are relatively distant from the HIP.

Our observations of the piglet, confirming those previously reported of the dog, demonstrate that the venous HIP is not materially shifted by autonomic blockade (8). Thus unlike the arterial HIP (21), the venous HIP is not maintained by centrally integrated baroreflexes but by the local properties of the veins. For instance, changes in pelvic and splanchnic venous tone might be due to their myogenicity, that is, their ability to respond to the changing transmural hydrostatic pressures that accompany whole body tilt. A significant myogenic response has recently been documented in human and canine saphenous veins to physiological changes in distending pressure (1). Vascular myogenic responses are controlled by voltage-operated calcium channels, which are in turn modulated by many factors including nitric oxide released from endothelium, as well as prostaglandins and epinephrine. Alternatively, venous constriction during head-up tilt might be due to the release of a nonadrenergic transmitter through a barosensitive axonal reflex analogous to the dermal venivasomotor reflex activated during orthostatic stress (10, 23). The local stretch set point and pressure sensitivity about which these myogenic or axonal responses operate would need to relate to distance from the HIP.

Whatever the mechanism that controls intrathoracic venous pressure during postural changes, in the newborn the process appears to require conditioning or resetting. From the moment of birth, the veins have to adapt to the loss of the supra-atmospheric pressure and buoyancy provided by the amniotic fluid. The rise in resting central venous pressure observed during the first 3 days of life is perhaps a manifestation of this process; the shift in the HIP_up occurs after postnatal day 3 and is therefore not simply a result of this correction in resting venous pressure. The extremely premature human neonate has limited cerebral autoregulation in the face of fluctuations in arterial pressure and is prone to intraventricular hemorrhage (15). The time course of the maturation in human HIP control and the influence of premature birth on this process may have important clinical relevance. In addition, adaptation in venous HIP regulation may occur in adults undergoing prolonged bed rest, repeated acceleration, or zero-gravity spaceflight (2, 5).

On days 1 and 3, there are significant differences in the measured HIP, depending on tilt direction and grade. The HIP_down but not the HIP_up values at these ages are different from those in the deceased piglet, suggesting that there is active control of venous pressure at birth but only in the direction that can diminish venous pressurization and so produce an appropriate response to head-down tilt. If the data on the 6-day-old piglets were studied in isolation, the juxtaposition of heart, HIP_up, and HIP_down might be interpreted simply that the heart resides at the passive center of venous compliance. The day 6 HIP_up and HIP_down values, however, are both different from those measured in the deceased animal and thus appear to require active regulation.

In summary, this study details a convenient method for characterizing the neonatal venous system by determining its HIP. In the mature animal, HIP control is precise and not susceptible to chemical autonomic blockade and thus is difficult to unravel by conventional means. In this context, a newborn model is useful because the control system during the first days of life is not as effective as in the adult. These data support the hypothesis that there is an active mechanism that maintains the HIP close to the heart and therefore attenuates changes in cardiac filling pressure during alterations in posture. In the newborn, this control is achieved for head-down tilt before it is achieved for head-up tilt. The implications for the development of human cardiovascular control, especially in the preterm infant, need further study.