Journal of Applied Physiology
Vol. 82, No. 2, pp. 607-612, February 1997
CONTROL OF BREATHING, CIRCULATION, AND TEMPERATURE

Periodic breathing induced on demand in awake newborn lamb

Emmanuel Canet, Jean-Paul Praud, and Michel A. Bureau

Unité de Recherche Pulmonaire, Département de Pédiatrie, Université de Sherbrooke, Sherbrooke, Quebec, Canada J1H 5N4

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Canet, Emmanuel, Jean-Paul Praud, and Michel A. Bureau. Periodic breathing induced on demand in awake newborn lamb. J. Appl. Physiol. 82(2): 607-612, 1997.Spontaneous periodic breathing, although a common feature in fullterm and preterm human infants, is scarce in other newborn mammals. The aim of this study was to induce periodic breathing in lambs. Four 10-day-old and two <48-h-old awake lambs were instrumented with jugular catheters connected to an extracorporeal membrane lung aimed at controlling arterial PCO₂ (Pa_CO2). Arterial PO₂ (Pa_O2) was set and maintained at the desired level by changing inspired O₂ fraction and providing O₂ through a small catheter into the "apneic" lung. At a critical Pa_O2/Pa_CO2 combination, the four 10-day-old lambs exhibited periodic breathing that could be initiated, terminated, and reinitiated on demand. In the 2-day-old lambs with low chemoreceptor gain, periodic breathing was hardly seen, regardless of the trials done to find the critical PO₂/PCO₂ combination. We conclude that periodic breathing can be induced in lambs and depends on critical Pa_O2/Pa_CO2 combinations and maturity of the chemoreceptors.

control of breathing; extracorporeal CO₂ removal; apnea

INTRODUCTION

IN THE LAST DECADE, as our knowledge has grown in regard to the control of breathing, a clear understanding has emerged as to the mechanism by which adult humans and other mammals develop periodic breathing (2, 8, 14, 16, 20), a condition in which periods of central apnea, or reduced respiratory efforts, alternate at reasonably regular intervals with periods of normal or increased efforts (12, 22). In contrast, the mechanisms leading to periodic breathing in newborn mammals have remained unclear (1, 6, 13, 17, 21). This study addresses the conditions under which newborn lambs develop and maintain periodic breathing.

In adults, the chemoreflex drive to breathe appears to be key to the induction, maintenance, and termination of periodic breathing. Periodic breathing occurs when the "oxystatic" compound of the chemical drive to breathe has become disproportionnally increased relative to its "capnostatic" component (8). In other words, periodic breathing requires two essential conditions. The first is a low arterial PO₂ (Pa_O2) that increases the peripheral reflex gain, thereby increasing the overall gain of the respiratory control system (16). Conversely, conditions of low chemoreceptor responsiveness to hypoxia predictively diminish the capacity to develop periodic breathing (16). The second condition is a sufficiently low level of arterial PCO₂ (Pa_CO2), i.e., a resting Pa_CO2 set close to the apnea point to induce apnea (10). Indeed, the understanding of adult periodic breathing is such that Lahiri et al. (16), in high-altitude experiments, could predict and find the variable capacity for developing periodic breathing in sherpas compared with low-altitude natives during Mount Everest climbing expeditions.

Periodic breathing in the neonate is not so well understood. First, there is a major discrepancy among the different species of neonates as to the occurrence of periodic breathing. Apparently, all human neonates, preterm or fullterm, have bouts of periodic breathing, whereas spontaneous periodic breathing is still unreported in other mammals. Furthermore, although investigators have reported increased periodic breathing in human neonates with hypoxia (19), there have been no such reports of periodic breathing in newborn mammals, despite the hundreds of experiments performed on ventilatory response to hypoxia in a few species of newborn mammals. There are, however, two reports of interest: one in monkeys (21), in which hypoxia induced periodic breathing in one animal; and one in lambs, in which periodic breathing was observed in the posthypoxia period (6). Interestingly enough, in our laboratory, we have induced hypoxia in >100 lambs (4-7) and in piglets (9) of various ages but have not yet observed periodic breathing during hypoxia, regardless of the hypoxic challenge. In the present study, we tested the hypothesis that, should we find the critical Pa_O2/Pa_CO2 combination, we could induce and maintain periodic breathing in lambs, provided they have an adequate chemoreflex maturation to respond to hypoxia and hypocapnia. We succeeded in producing periodic breathing, on demand, in awake newborn lambs.

METHODS

Experimental subjects. Six Rambouillet-Suffolk lambs, born by spontaneous vaginal delivery at sea level and housed in our animal quarters, were used in the study. Two lambs were between 12 and 48 h of age and four were between 10 and 12 days of age. The study protocol was approved by the ethics committee for animal research at our institution.

Design of study. The experimental design allowed us to set the Pa_CO2 and Pa_O2 at levels expected to induce sustained periodic breathing by using two gas exchangers connected in series (7, 15). The first gas exchanger, an extracorporeal membrane lung connected to the lamb by a venovenous shunt, was used for adjustment of Pa_CO2; the second gas-exchanger was the lamb's lung, used for adjustment of Pa_O2. At constant Pa_O2, Pa_CO2 could be altered by changing gas flow or CO₂ fraction through the extracorporeal membrane. Conversely, at constant Pa_CO2, Pa_O2 could be altered by changing the inspired O₂ fraction (FI_O2) of the lamb's lung or by supplying humidified and heated O₂ in small quantities (25- 100 ml/min) by using a polyvinyl 5-Fr catheter introduced in the endotracheal tube beyond 1 cm of its tip.

After surgical preparation (see below), the lamb was placed comfortably in a prone position on an animal cart under gentle restraint that did not interfere with chest or abdominal movement. The first baseline ventilatory measurements and arterial blood-gas samplings were performed during a 2-min period during normal room-air breathing. The animal was then connected to the extracorporeal circuit (progressively set to obtain a constant blood flow) while no gas flowed through the membrane (no gas exchange). A minimum of 30 min was allowed for the lamb to become accustomed to the venovenous shunt and the blood flowing through the membrane. The gas flow through the membrane was then opened, and the critical Pa_O2/Pa_CO2 values for periodic breathing were searched as follows.

First, based on our previous results (6), we tested whether a decrease in Pa_CO2 (approximately to the 25-35 Torr range) and Pa_O2 (approximately to the 30-40 Torr range) would lead to the development of periodic breathing. Gas flow to the membrane was set to a value expected to lead to the aimed Pa_CO2 range, and FI_O2 delivered to the lamb's lung was decreased to 10%. If necessary, these parameters were then slightly modified in a stepwise fashion (most often empirically) until periodic breathing developed. At least 10 min were allowed for stabilization between each modification. Critical Pa_O2/Pa_CO2 values were measured in arterial blood systematically drawn at the onset of periodic breathing.

Second, because one of our objectives was to evaluate the gap between Pa_CO2 at the onset of periodic breathing and Pa_CO2 at the point of apnea, we further determined the apneic threshold in each lamb by decreasing Pa_CO2 to the point of apnea as previously described (7). Pa_O2 was kept constant during this procedure by supplying small quantities of O₂ via the endotracheal tube catheter on the basis of blood gas analysis obtained from frequent arterial blood withdrawal. By anticipating occurrence of apnea, we were able to precisely determine the Pa_O2/Pa_CO2 combination that resulted in apnea. After a few minutes of apnea, gas flow through the membrane was discontinued, thereby allowing for spontaneous reinitiation of breathing.

Finally, we tested the general hypothesis that periodic breathing is the result of disturbing an unstable respiratory chemical system. Because significant association between periodic breathing and sighs has been reported in human infants, we tried to induce periodic breathing by mimicking a sigh. After adjusting Pa_CO2 and Pa_O2 near critical Pa_O2/Pa_CO2 values for periodic breathing, one sigh breath [the volume of which was at least twice the tidal volume (VT)] was induced by bagging the lamb with a hand bag connected to the endotracheal tube.

Surgical preparation. Each lamb was prepared surgically, as reported in detail elsewhere (7). Briefly, 3 h before the experiment, after applying a 2% lidocaine spray to the oropharynx and larynx, a peroral intubation was performed. An appropriately sized (4.5-6.0 mm) endotracheal tube was inserted, with cuff pressure maintained at 25 cmH₂O; then, under sterile conditions and local anesthesia (2% lidocaine), the catheters were put in place. An arterial catheter was inserted into the innominate artery for arterial blood-gas sampling and for monitoring systemic blood pressure and heart rate. After administrating a loading dose of heparin (300 U/kg), the animal's coagulation status was monitored throughout the experiment via the Hemochron apparatus (International Technidyne) so that an activating clotting time could be maintained in the range of 180-300 s. Venous cannulations were performed for connection to the extracorporeal circuit through the jugular veins by using thin-walled spring-wire-reinforced polyurethane catheters (Bypass Systems). One cannula was inserted through the left jugular vein (12 Fr) and advanced to the superior vena cava for use as the venous return line. A second cannula (12 Fr) was inserted through the right jugular vein and advanced centrally to 1 cm above the diaphragm to allow drainage of both the inferior vena cava and the atrium through lateral side holes. Cannulas (12 Fr) were inserted into each jugular vein in a cephalad direction to allow for drainage of the head on each side. Positioning of the catheters was controlled by X ray of the neck and chest. Due to the easy accessibility of the jugular veins in the lamb, all catheters could be introduced under local anesthesia (2% lidocaine) without any undue discomfort to the animal.

Extracorporeal circuit. As described previously (7), a venous-to-venous extracorporeal bypass circuit was used. The venous blood was drawn by gravity from the atrium, inferior vena cava, and both head-drainage jugular cannulas into a 50-ml reservoir bag that, in turn, activated a servo-controlled centrifugal pump (Biomedicus). From the centrifugal pump, the blood passed through one membrane lung (4.5 m² surface area, Bentley) and then returned to the lamb through the left venous catheter to the superior vena cava. A heat exchanger affixed to the membrane lung was adjusted to enable the lamb's rectal temperature to be maintained at its baseline value (39.5°C) during the entire experimental procedure. Gas flow to the membrane lung was supplied from a cylinder and regulated with precision rotameters. The gas was heated to 37°C and humidified before flowing through the membrane lung.

On the morning of the experiments, the extracorporeal perfusion circuit was primed with Ringer-lactate solution and subsequently displaced with homologous sheep blood (800 ml) containing 5 units of heparin/ml. The blood, recirculated through the circuit for 1 h before any connection was made to the lamb, was adjusted to approximate normal venous pH as well as venous PO₂ and PCO₂. Aseptic procedures were followed throughout the experiment. Once the circulated blood was conditioned, the lamb was connected to the circuit and the blood flow through the pump was slowly increased to ~150 ml ^· min^{1 ·} kg¹ and maintained at this level throughout the entire experiment.

Ventilatory measurements. The endotracheal tube was connected to a size 0 pneumotachograph (Hewlett-Packard 21070B, Pomona, CA). The resulting flow signal, integrated with respect to time to give volume (model 8815A Hewlett-Packard respiratory integrator), was recorded on a strip-chart recorder. End-tidal PO₂ and PCO₂ were sampled from the endotracheal tube and analyzed by a mass spectrometer (model MGA-1100, Perkin-Elmer). Simultaneously, the pneumotachograph flow signal and the mass spectrometer tidal PO₂ and PCO₂ signals were sent to a PDP-11/23 computer (Digital Equipment) for a breath-by-breath computer analysis. The flow wave, recorded by an analog-to-digital converter, was analyzed at a sampling rate of 40 Hz. The collected signals were stored on disk for further analysis. The respiratory variables derived by the computer for each breath were minute ventilation; VT; respiratory rate (f ); inspiratory time (TI), expiratory time, and total time of each cycle; mean inspiratory flow (VT/TI), duty cycle (TI/total time), end-tidal PO₂, and end-tidal PCO₂. For blood-gas analysis, duplicate 1-ml samples were taken anaerobically into 3-ml syringes and analyzed immediately for Pa_CO2, Pa_O2, and pH determinations (Micro 13, Instrument Laboratory).

Definition of periodic breathing. For this study, periodic breathing was defined as cyclic fluctuations of respiratory efforts interrupted by periods of apnea (3 s) or hypopnea in a crescendo-decrescendo pattern.

RESULTS

The fact that the lambs were connected to an extracorporeal circuit set at a constant blood flow (while the membrane was closed to gas flow) had no effect on baseline minute ventilation or on arterial blood gases compared with baseline values (Table 1). The absence of cyclic fluctuation in ventilation confirmed that this procedure, by itself, did not affect the stability of the ventilatory controller.

Table 1. Baseline ventilatory and blood gas data in instrumented animals with and without extracorporeal circulation VI, ml · min1 · kg1 pH PaCO2, Torr PaO2, Torr 10-Day-Old Lambs Lamb 1    Off 486 7.35 36.7 105   On 443 7.34 37.7 103 Lamb 2    Off 405 7.30 38.1 91   On 459 7.29 40.9 92 Lamb 3    Off 383 7.32 40.2 89   On 411 7.39 38.1 83 Lamb 4    Off 336 7.41 43.9 83   On 391 7.42 41.5 75 2-Day-Old Lambs Lamb 5    Off 455 7.39 48.3 84.3   On 454 7.39 44.1 109 Lamb 6    Off 460 7.34 53.9 72.7   On 468 7.35 50.2 90 VI, minute ventilation; PaCO2, arterial PCO2, PaO2, arterial PO2; Off, without extracorporeal circulation; On, with extracorporal circulation.

Table 2. Arterial blood gas values during baseline recordings, at onset of PB, and at onset of apnea Baseline PaCO2, Torr PB PaCO2/PaO2, Torr/Torr Apnea PaCO2, Torr Lamb 1  37 29.2/40.2 27.8 Lamb 2  40.9 23.6/30.8 22.8 Lamb 3  38.1 22.6/28 19.5 Lamb 4  41.5 38/29.5 34.3 Baseline values were taken in normoxia. PB, periodic breathing.

DISCUSSION

This study shows that respiratory controllers in awake lambs can produce periodic breathing. Periodic breathing requires the same two conditions as in other mammals, one being the critical Pa_O2/Pa_CO2 combination, the second being a sufficient state of maturity of the chemoreceptors to allow for high-chemoreceptor gain when the critical blood-gas level is reached.

The data obtained in the present study are interesting with regard to one point. They establish without a doubt that the periodic breathing of awake lambs is a true chemoreceptor event. Given the critical combination of Pa_O2/Pa_CO2 in the lambs, we were able to produce periodic breathing on demand, maintain periodic breathing, stop periodic breathing, even reinitiate periodic breathing. In addition, at critical Pa_O2/Pa_CO2 values, by maneuvers known to change blood-gas tensions and, consequently, chemoreceptor output (such as a sigh, a sharp decrease or increase in FI_O2, or a burst of increased inspired CO₂ fraction), predictable results with respect to enhancement or stoppage of periodic breathing could be elicited. These data clearly establish that periodic breathing can be induced in awake newborn lambs and that it critically depends on the chemical "drive to breathe." We have not, as yet, tested the potentially modulatory effect of sleep state on periodic breathing.

The difficulty of producing periodic breathing in the less-than-2-day-old lambs deserves comments. At this age, prior studies from our laboratory (5) and others (3, 21) have shown that the chemoreceptors are still set to respond to very low levels of Pa_O2 (3) and high levels of Pa_CO2 (18). In other words, immediately after birth, as witnessed by the 2-day-old lambs, the chemoreceptors are not capable of eliciting a sufficiently brisk chemoreflex response to blood-gas changes to produce sustained periodic breathing. After birth, the chemoreceptors have to adapt to new blood-gas values; they must shift to new response thresholds to these blood gases. In the process, they acquire sufficient oxystatic and capnostatic chemoreceptor gains to produce sustained periodic breathing when critical blood gas tensions are reached. In lambs, the maturation process of the chemoreceptors is progressive over the first 2 wk of life (3, 5). It is therefore not surprising that sustained periodic breathing could not be induced in our younger lambs inasmuch as they had not developed their capacity of high chemoreceptor gain. One could argue that the window of critical Pa_O2/Pa_CO2 levels needed to produce this periodic breathing does exist at this age, but that we were unable to find it in our numerous attempts to identify the critical blood-gas combinations. In our view, however, this supposition is unlikely. We believe that newborns in the first days of life, as well as fetuses in prenatal life, simply do not have a sufficiently high chemoreflex gain to produce sustained periodic breathing. This is in agreement with our previous study of posthyperventilation breathing oscillations in lambs (6), showing that 2-day-old lambs, because of lower chemoreceptor responsiveness, did not experience as much periodic breathing as did 10-day-old lambs.

The relationship between the critical Pa_CO2 that produces periodic breathing and the critical Pa_CO2 that produces apnea is interesting. Those two critical points are separated by only 1-3 Torr. This means that when Pa_CO2 is at 2-3 Torr from the apnea point, large variations in breathing, as a sigh, could reach the point of apnea and induce periodic breathing, provided that Pa_O2 is also low enough to increase the chemoreceptor gain.

Despite differences between species, extrapolation of our findings regarding periodic breathing in lambs to periodic breathing in human neonates is warranted. The difficulty of producing periodic breathing in lambs during the first 2 days of life is in agreement with the current literature dealing with human infants. We have observed that at that age, periodic breathing simply does not exist in human infants (4); it occurs later, at a time when chemoreceptors have been reset for activation at a higher level of responsiveness. Similarly, Barrington and Finer (1) demonstrated that the propensity of human infants to develop periodic breathing is both age related and associated with chemoreceptor maturation. Additionally, in our lambs, when Pa_CO2/Pa_O2 was set at the appropriate critical level, a simple sigh or passive ventilation produced periodic breathing. Along the same lines, Bureau et al. (4) have reported that spontaneous periodic breathing in human neonates is often preceded by a brief period of hyperpnea believed to lower Pa_CO2 toward the critical level for periodic breathing; a simple sigh could then initiate periodic breathing in infants (11). Moreover, in both species, inhalation of O₂ ends periodic breathing.

Hence, periodic breathing of both newborn lambs and human neonates seems to be a chemoreceptor-driven event, although quantitative differences between these species are still remarkable. For example, in human infants, a few large breaths or a sigh are enough to bring infants to the critical Pa_O2/Pa_CO2 window needed to induce periodic breathing. This phenomenon implies that the breathing of human infants is naturally set at a blood-gas level close to the level needed for periodic breathing. In contrast, in awake lambs, much like adults, the critical Pa_O2/Pa_CO2 window required to induce periodic breathing is much lower than the baseline Pa_O2/Pa_CO2, and drastic experimental conditions are needed to reproduce periodic breathing. It is therefore not surprising that lambs do not have spontaneous periodic breathing, whereas most healthy human infants do, even though their mechanisms leading to periodic breathing are alike. The fact that our studies were conducted in awake animals might also have contributed to this discrepancy. The effect of sleep on ventilatory responses to chemical stimuli is well established, and recurring changes in sleep state could be considered as a contributing factor that may promote ventilatory instability.

In conclusion, this study shows that it is possible to produce periodic breathing on demand in awake lambs when appropriate settings of blood gases are reached to drive the chemoreceptors. It also strongly suggests that, in newborn mammals, periodic breathing is a chemoreceptor-mediated event.

ACKNOWLEDGEMENTS

The authors acknowledge the expert secretarial assistance of Marguerite Cloutier and the technical assistance of D. Edgell.

FOOTNOTES

Address for reprint requests: M. Bureau, Unité de Recherche Pulmonaire, Département de Pédiatrie, Université de Sherbrooke, Québec, Canada J1H 5N4.

Received 1 March 1995; accepted in final form 3 October 1996.

REFERENCES