- Email: [email protected]
Measuring Functional Residual Capacity in Normal and Oleic Acid-Injured Lungs
JOURNAL OF SURGICAL RESEARCH ARTICLE NO.
63, 204–208 (1996)
0248
Measuring Functional Residual Capacity in Normal and Oleic Acid-Injured Lungs1 PAUL G. GAUGER, M.D., MICHAEL C. OVERBECK, B.S., SEAN D. CHAMBERS, M.S.E., ERIK D. WEBER, AND RONALD B. HIRSCHL, M.S., M.D.2 Department of Surgery, University of Michigan, Ann Arbor, Michigan 48109 Presented at the Annual Meeting of the Association for Academic Surgery, Dearborn, Michigan, November 8–11, 1995
Functional residual capacity (FRC) is an important oxygen reserve that is often depleted in acute respiratory failure. Recent interest in the mechanisms of liquid ventilation and limited experience in measuring FRC in paralyzed, mechanically ventilated, normal and lung-injured animal models have mandated development of accurate laboratory techniques. Eight sheep, from 17 to 27 kg, were anesthetized and instrumented to provide a tracheostomy, a pulmonary artery catheter, and carotid arterial line. They were randomized to two groups, one of which received 0.07 ml/kg of intravenous oleic acid to induce lung injury. Gas ventilation of both groups was identical except for respiratory rate, which was adjusted to normalize PaCO2 . FRC was measured in duplicate by both helium dilution (HD) and body plethysmography (BP). When measurements were completed, the animals were euthanized and their endotracheal tubes clamped at end expiration. The lungs were then removed and their water displacement (WD) FRC values were measured. FRC was the difference between WD and tissue weight assuming 1 ml Å 1 g. Pearson’s correlation coefficient (R 2) was calculated. During in vitro measurement of test lungs, HD had an R 2 value of 0.99 and BP had an R 2 value of 0.98. When compared to WD, in vivo measurement of FRC by HD had an R 2 value of 0.94 while the value for BP was 0.97. In conclusion, both HD and BP are accurate methods of determining FRC in an uninjured and injured lung model when compared to postmortem WD. Documenting changes in FRC will aid in elucidating the mechanisms of alternative ventilatory techniques. q 1996 Academic Press, Inc.
INTRODUCTION
Functional residual capacity (FRC) is the volume of gas remaining in the lungs at the end of a tidal expiration. FRC is a crucial oxygen reservoir, second in impor1
Funded in part by NIH grant no. RO1HD15434. To whom correspondence and reprint requests should be addressed at Section of Pediatric Surgery, F3970 Mott, Box 0245, 1500 East Medical Center Drive, Ann Arbor, MI 48109–0245. Phone: (313) 764-6846. Fax: (313) 936-9784. 2
0022-4804/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AID
JSR 4799
/
6n0e$$$521
tance only to the hemoglobin pool, and is often depleted in acute respiratory failure. Functional residual capacity is the sum of the expiratory reserve volume and the residual volume. As it includes the residual volume, the absolute value of FRC is not amenable to direct measurement techniques and must rely on indirect techniques such as indicator gas dilution, indicator gas washout, or body plethysmography [1]. FRC has an important effect on gas exchange. It has been demonstrated that lungs of patients with the acute respiratory distress syndrome (ARDS) have a decreased volume of pulmonary gas and that physiologic shunt fraction increases as the FRC decreases. Supportive therapy has targeted increasing the FRC, thus improving ventilation-perfusion matching and subsequently improving gas exchange. This is a well known mechanism of positive end expiratory pressure (PEEP) [2, 3]. Developing therapies to augment FRC obviously requires accurate techniques of measurement applicable to normal and lung-injured, paralyzed and mechanically ventilated large animal models. For this purpose, our goal was to combine a closed-circuit helium dilution system and a body plethysmography system to measure FRC by two different, but complimentary methods. Additionally, we wished the methods to be applicable to both normal and lung-injured models. MATERIALS AND METHODS Eight Suffolk sheep, from 17 to 27 kg, were anesthetized with ketamine HCl (10 mg/kg IV) and guaifenesin (2.2 ml/kg of a 5% solution) after being fasted for 24 hr. Anesthesia was maintained with a standardized mixture of guiafenesin (50 g/l) and ketamine (1.5 g/l). The animals were instrumented to provide a 9-mm endotracheal tube (Mallinckrodt Medical, St. Louis, MO) via a surgical tracheostomy, a Swan-Ganz pulmonary artery catheter, and carotid arterial pressure monitoring. The tracheostomy was tightly sealed and leak tested. Pancuronium was then administered (0.1 mg/kg IV) and was redosed as necessary. All animals were ventilated in volume control mode with a Siemens Servo 900 E (Siemens Elema AB, Sweden). The initial ventilator settings were similar for all animals and included a tidal volume (TV) of 15 ml/kg, inspired oxygen fraction (FiO2) of 1.0, PEEP of 5–7 cm H2O, inspiratory to expiratory ratio of 1:1, and respiratory rate of 10–16 breaths per minute to normalize PaCO2 to 35–45 torr. The animals were then suctioned of their gastric rumen. The sheep were placed in a specially designed full body plethysmo-
204
05-13-96 07:38:39
srga
AP: Surg Res
GAUGER ET AL.: MEASURING FRC IN NORMAL AND INJURED LUNGS
205
FIG. 1. Diagram of the large animal plethysmograph. Water trough served to detect leaks across the lid seal during box pressurization (see text).
graph constructed of 41-inch Plexiglas. Empty, the box had a volume of approximately 130 l. The box is illustrated in Fig. 1. The lid sealed tightly to the box, creating an airtight chamber. A water trough around the lid allowed for detection of any potential air leak across the lid seal. All IV and monitoring lines were connected via sealed luer locks through the side of the box. Likewise, the ventilator was connected to the tracheostomy tube by a sealed connector through the lid of the box. Above this connector was a pinch valve that was tightly closed during plethysmographic measurements (see below). Through a side port connected to the airway, a differential transducer (PX170–28DV, Omega Engineering, Inc., Stamford, CT), protected by a sputum trap, measured the differential between the airway pressure and the ambient box pressure. A strain gauge (CDX III, Cobe, Lakewood CO) connected through a sealed adapter in the box’s side directly measured the box pressure. A schematic of the setup is illustrated in Fig. 2. Both transducer signals were processed and recorded in real time by Lab View software (National Instruments, Inc., Austin, TX). The closed-circuit helium dilution device was constructed to rest atop the box lid. It was constructed of noncompliant tubing and incorporated one way valves to maintain circular gas flow. The circuit included a CO2 scrubbing chamber filled with soda lime (Mallinckrodt Specialty Chemicals, Paris, KY), an in-line O2 analyzer (Instrumentation Laboratories, Lexington MA), a mixing motor (Warren E. Collins, Inc., Braintree, MA), and a ‘‘bag in box’’ type slave bellows. Circuit gas was driven by the motor through a continuous helium analyzer (Warren E. Collins, Inc). Total device volume was 3120 ml. A diagram of the circuit is shown in Fig. 3. The circuit was connected to the airway above a pinch valve by a two way valve (Hans-Rudolph, Kansas City, MO). By this construction, the ventilator was coupled to the airway through the open valves across the ‘‘bridge’’ during most of the experimental period. When helium dilution measurements were performed (see below), the ventilator was coupled to drive the slave bellows by pressurizing the surrounding space, and all ventilating gas then passed into and out of the closed circuit.
FIG. 2. Schematic of setup required for measurement of FRC by null-point body plethysmography. Differential transducer measured the relationship of tracheal pressure and ambient box pressure.
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
FIG. 3. Construction of our closed-circuit device for measurement of FRC by helium dilution. Important features included the PEEP valve, the CO2 scrubber, and the slave bellows, which were driven by the ventilator during measurement.
PEEP was maintained in the airway by a PEEP valve on the expiratory limb of the circuit calibrated to 6 cm H2O. Airway pressures during closed-circuit measurement were monitored via the differential transducer (while the box communicated with atmosphere) and were nearly identical to those generated by the ventilator when the closed circuit was bypassed through the majority of the experiment. To ensure similarity of tidal volumes used during closed-circuit ventilation and ventilation across the bridge, an in-line spirometer (Boeringer, Inc., Ridgefield, CT) was placed directly above the pinch valve and a constant tidal volume of 15–16 ml/kg was maintained. The sheep were randomized to two groups, one of which received intravenous oleic acid to generate a model of acute respiratory failure characterized by hemorrhagic pulmonary edema, increased transpulmonary shunt, decreased compliance, and decreased FRC (n Å 4). The oleic acid was administered, after emulsification with 10 ml blood, in a dose of 0.07 ml/kg into the right atrial port over 20 min. The other group received no lung injury to maintain a presumably normal FRC (n Å 4). Both groups underwent the same anesthesia, instrumentation, and ventilation as described above. Helium dilution (HD) measurement of FRC was performed in duplicate as was nullpoint body plethysmography (BP). Both techniques are explained fully below. Helium dilution. As the animal was ventilated as usual across the bridge, the helium dilution circuit was flushed with 100% oxygen until an oxygen fraction (FO2) of 90–100% was obtained. The circuit was then evacuated of 300 ml of gas to prevent pressurization as 300 ml of pure helium was added. The circulating motor was used to mix the gasses until a stable helium concentration between 7 and 11% was observed on the helium meter. A small diameter bridge was used to bypass the PEEP valve only during mixing to promote homogeneity of the gas mixture. While the circuit was mixing, three inspired gas samples and three mixed expired gas samples were aspirated from the ventilator circuit and analyzed for oxygen percentage on an ABL 30 (Radiometer A/S, Denmark). These percentages within 0.5% were averaged to derive the FiO2 and the FeO2 (expired oxygen fraction). The minute volume ventilation (MVV) was measured (in ml) and the rate of oxygen uptake (O2 consumption) in ml/min was calculated as O2 consumption Å (FiO2 0 FeO2) 1 MVV [4]. The initial equilibrated helium percentage (He%) and the FO2 were noted. The measurement of circuit FO2 was necessary as O2 concentrations above 21% cause an artifactual elevation of the He% due to competing thermal conductance. This overestimation occurs in a linear fashion and can be corrected by the formula Hec Å Hed 0 (FO2 0 0.2093)/37.7 where Hec is the corrected He% and Hed is the He% displayed on the meter [5]. The ventilator was then paused at end expiration (lungs at FRC) and the valve was opened to the closed circuit as the ventilator was switched to drive the slave bellows. A stopwatch was started as the animal was ventilated through the
srga
AP: Surg Res
206
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 1, JUNE 1996
helium dilution circuit. Constancy of tidal volume was assured through monitoring of the in-line spirometer. Helium is an inert indicator gas that is minimally absorbed across the respiratory membrane, and therefore as the circuit gas equilibrated with the FRC gas volume, a new stable helium concentration was reached. At this point, the stopwatch was stopped and the final Hed and FO2 were recorded and the Hec was calculated for the initial and final readings. Again, three samples of inspired gas and three samples of mixed expired gas were measured, and the oxygen uptake rate was calculated as above. This rate was averaged with that determined prehelium dilution and this value was multiplied by the time (in minutes) that the animal was breathing through the closed circuit to more accurately estimate the volume decrease in the circuit attributed to oxygen consumption (O2 consumption). As the animal consumed oxygen at his basal rate, the circuit volume would decrease, as evidenced by the decreasing excursion of the slave bellows. FRC was then calculated using the formula (see appendix for complete derivation): FRC (ml) Å [(He%initial/He%final) 1 V1] 0 V1 / O2 consumption (ml) The previously measured dead space of this circuit configuration included the two way valve, the pinch valve, the sputum trap, and the endotracheal tube (90 ml total) and was subtracted to derive the FRC by HD. This was mechanical dead space and did not include the anatomic and functional dead space of the individual sheep. The value was corrected to BTPS, and an average of both helium dilution measurements was used as the final FRC by HD value. Null point body plethysmography. Duplicate measurements of FRC by body plethysmography were then performed. The technique was based on that originally described for measurement of FRC in dogs by Laver et al. in 1964 and in cats by Colebatch et al. in 1974 [6, 7]. As with all plethysmographic methods, the null-point variation is based on Boyle’s law which is a permutation of the Ideal Gas Law that states that in a closed system at constant temperature, the pressure of a gas multiplied by the volume of a gas is a constant, or P1 V1 Å P2 V2 . Standard plethysmographic techniques were not appropriate for this model as they require spontaneously breathing subjects and the sheep were paralyzed and mechanically ventilated. For each measurement, the ventilator was stopped at end expiration (FRC), and the pinch valve at the lid of the plethysmograph was tightly closed. Twenty to thirty milliliters of room air were injected into the trachea through the sputum trap access from a calibrated gas syringe (DV). This caused a deflection in the signal from the differential transducer (Pd). High flow, compressed air was then injected into the plethysmograph until the value of Pd had been restored to the level recorded prior to injection of DV. The difference in the initial and final pressure of the plethysmograph chamber was termed DPG . The box was vented to atmosphere and ventilation was resumed. End expiratory thoracic gas volume, or FRC, was calculated by the equation derived from Boyle’s law [7]: FRC (ml) Å DV 1 PB 0 PH2O/DPG (see appendix). The dead space of the endotracheal tube and the sputum trap was subtracted (65 ml). The calculated FRC was then corrected to BTPS and the two values were averaged for the FRC by BP determination. After both helium dilution and plethysmographic measurements had been completed, the sheep were removed from the box while continuing the identical ventilatory pattern. They were then euthanized with a barbiturate overdose and the lungs and trachea were rapidly removed en bloc after the endotracheal tube was clamped at end expiration (FRC). Their water displacement volume was then measured in triplicate. The lungs and trachea were then weighed, and allowing for a tissue density of 1 g/ml, their water displacement value (WD) was calculated. Again, this was the mean of the three measurements. The functional residual capacity values measured by helium dilution, null-point plethysmography, and water displacement were then compared. Descriptive statistics (expressed as the mean { SEM) and Pearson’s correlation coefficient (expressed as R 2) were all determined using Systat software (Systat, Inc., Evanston, IL). All FRC values discussed have been normalized for the animal’s body weight and are expressed in ml/kg. Both in vitro and in vivo evaluations of the techniques are reported.
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
FIG. 4. Correlation of measured and known test lung volumes during in vitro development of the techniques. Pearson’s correlation coefficient was 0.99 for helium dilution and 0.98 for body plethysmography.
RESULTS
Before animal models were used, bench top testing of both systems was performed to validate both methods. The helium dilution circuit was used to measure multiple test lungs ranging in volume from 140 to 2900 ml. Comparison with the known volumes of the test lungs revealed excellent correlation with a R 2 value of 0.99. Similar in vitro validation was performed for the null-point plethysmography method using test lungs (volume range 250–1850 ml) with a compliant and noncompliant component as described by Laver et al. [6]. Pearson’s correlation coefficient was 0.98. The in vitro correlation analysis is graphically displayed in Fig. 4 for both techniques. The mean FRC of uninjured animals measured by helium dilution was 34.0 { 3.3 ml/kg, while the mean FRC was 34.8 { 2.8 ml/kg when measured by plethysmography. In comparison, the mean FRC by water displacement was 35.0 { 4.4 ml/kg in uninjured animals. When oleic acid lung injury was induced, the mean FRC by helium dilution was 21.8 { 4.2 ml/kg and the mean FRC by plethysmography was 22.8 { 4.0 ml/kg. When water displacement measurement was used, the mean FRC was 22.3 { 4.9 ml/kg for lung-injured animals. The helium dilution technique was compared to water displacement. The results of all animals measured are displayed in Fig. 5, demonstrating a Pearson’s correlation coefficient of 0.94. Similarly, body plethysmography was compared to water displacement in all animals which revealed an R 2 value of 0.97. This is shown in Fig. 6. DISCUSSION
Helium dilution and null-point body plethysmography are two distinct methods of measuring the functional residual capacity in experimental models and in clinical investigation. Although different, they are complimentary. Each technique has its own advantages and shortcomings. With helium dilution, one of
srga
AP: Surg Res
GAUGER ET AL.: MEASURING FRC IN NORMAL AND INJURED LUNGS
FIG. 5. Correlation of helium dilution FRC values and postmortem water displacement FRC values. Pearson’s correlation coefficient was 0.94 with a linear regression model described by y Å 125 / 0.78x.
the potential sources of inaccuracy is a lack of proper mixing of the gas between the lungs and the circuit. This would leave a higher helium concentration at the end of the measurement period and would thus underestimate FRC. Similarly, another potential source of inaccuracy is the presence of occlusive airway lesions such as excessive secretions or profound pulmonary edema that may prevent the helium mixture from reaching the most peripheral alveolar units in a timely fashion. This would also leave a higher helium reading at the end of the measurement and underestimate FRC. Helium dilution can overestimate FRC if there is even a slight leak in the closed rebreathing circuit, airway, or endotracheal tube cuff. Regarding our specific application of the technique, another possible problem could be the estimation of oxygen uptake to quantify the decrease in circuit volume. Most closed circuits described in the literature include a spirometer as an oxygen or air reservoir [5, 8]. We abandoned this approach for two reasons. First, the spirometer increased the volume of the circuit to a point that we felt the method was losing sensitivity for measuring animal, not adult human lung volumes. Second, with the decreased compliance and increased airway pressures seen with the oleic acid-injured animals, we had trouble controlling the emptying of the spirometer during the varying pressures of the respiratory cycle. Our method avoided both of these confounding situations and kept the circuit volume as small as possible. A possible cause for inaccuracy with the null-point plethysmography method is related to the time needed to complete the measurement. If DV was injected without compressed air subsequently entering the box, the signal would deteriorate back to baseline within 15 sec. Although not as profound, this signal (Pd) would decrease even if DV were not injected. As Colebatch explained, this is therefore not due to simple stress relaxation of the lungs, but is instead due to a decrease in
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
207
intrathoracic volume caused by displacement of thoracic blood volume [7]. Due to the rapid flow capability of our compressed air lines, we were able to consistently complete our measurements in less than 4 sec, and thus before the deterioration in Pd became noticeable. To avoid unintentional forced deflation of the lungs, it was important to vent the box pressure back to atmospheric pressure levels before opening the pinch valve and resuming ventilation. If this was not done, the subsequent FRC measurements could be significantly decreased as exhalation was forced by the pressurized box environment, thus diminishing FRC. Since this method relies on the pressure-volume relationships of the pulmonary gas volume and the plethysmograph during a short period of time, the problem of including accumulated gastric and esophageal gas in the measurement is minimized [6]. Likewise, the problem of changing temperature inside the plethysmograph during the course of the experiment is negligible as the temperature is assumed to be unchanged during the 4 sec needed for the measurement. Thus, the validity of Boyle’s law is maintained. In contrast to helium dilution, plethysmography measures the entire thoracic gas volume, whether or not it communicates freely with the airways. Thus, obstructive airway disease with peripheral air trapping would tend to increase the FRC when compared to a simultaneous measurement by helium dilution. It is interesting to note that in both subsets, the mean FRC was slightly less with helium dilution than with plethysmography. This is likely due to the fact that any occlusive, small airway secretions would tend to decrease the measurement of FRC by HD compared to BP, as already discussed. Also, if Pd deteriorated more significantly than appreciated, then less pressurization of the plethysmograph would be necessary to nullify the change, thus decreasing the value of DPG and increasing the measured FRC by BP. Regardless, the differences in mean FRC values measured by helium dilution and FRC are neither clinically nor statistically significant. In uninjured lungs, the mean FRC by HD was 34.0 { 3.3 ml/kg and the mean FRC by BP
FIG. 6. Correlation of null-point body plethysmography FRC values and postmortem water displacement FRC values. Pearson’s correlation coefficient was 0.97 with a linear regression equation of y Å 137 / 0.81x.
srga
AP: Surg Res
208
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 1, JUNE 1996
of oxygen consumed during the measurement as explained above. Rearranging, V1 C1/C2 Å FRC / (V1 0 O2cons). Rearranging again, FRC Å [(C1/C2) V1] 0 (V1 0 O2cons). Substituting, the final equation becomes FRCHD (ml) Å [(He%initial/He%final) 1 V1] 0 V1 / O2cons (ml). All volumes are corrected to BTPS. Derivation of Null-Point Body Plethysmography Equation
FIG. 7. Correlation of helium dilution FRC and body plethysmography FRC values. Pearson’s correlation coefficient was 0.96 with a linear regression equation of y Å 44 / 0.98x.
was 34.8 { 2.8 ml/kg (P Å 0.74 by Student’s t-test). Similarly, in injured lungs, the mean FRC was 21.8 { 4.2 ml/kg with HD and was 22.75 { 4.0 ml/kg with BP (P Å 0.25, t-test). When directly compared, helium dilution and body plethysmography had an R 2 value of 0.96 (see Fig. 7). CONCLUSION
In summary, we have shown that helium dilution and null-point body plethysmography can be used successfully in combination to measure functional residual capacity in a large animal model. In addition, these methods can be used with success in animals with normal lung physiology and in those with a lung injury produced by intravenous injection of oleic acid. Both methods are accurate when compared to each other and also when compared to postmortem water displacement. These methods may be useful in elucidating the mechanisms of novel ventilatory techniques.
Begin with Boyle’s Law, P1 V1 Å P2 V2 . Then, (PB 0 PH2O) 1 (FRC / DV) Å (PB 0 PH2O / DPPG) 1 FRC, where PB is the barometric pressure (mm Hg) and PH2O is the vapor pressure of water (47 mm Hg), DV is the volume increment injected into the airway at end expiration, and DPG is the pressure increment above atmospheric required to reduce the volume FRC / DV back to FRC. Solving for FRC, the equation becomes FRC Å DV 1 PB 0 PH2O/DPPG . All volumes are corrected to BTPS. ACKNOWLEDGMENTS The authors thank Robert H. Bartlett, M.D., Constance Wise, R.R.T., Ronald E. Dechert, M.S., R.R.T., and Rekha Threja for their assistance in the completion of this work.
REFERENCES
Start with V1 C1 Å V2 C2 , where V1 is the initial volume of the closed circuit (3120 ml), C1 is the corrected initial helium concentration, V2 is the final volume of the closed circuit plus the FRC, and C2 is the corrected final helium concentration. Substituting for V2 , V1 C1 Å [(V1 0 O2cons) / FRC] C2 , where O2cons is the volume
1. Freedman, S. Lung volumes. Br. J. Clin. Pharmacol. 8: 99, 1979. 2. Dueck, R., Wagner, P. D., and West, J. B. Effects of positive end expiratory pressure on gas exchange in dogs with normal and edematous lungs. Anesthesiology 47: 359, 1977. 3. Dantzker, D. R., Brook, C. J., Dehart, P., Lynch, J. P., and Weg, J. G. Ventilation-perfusion distributions in the adult respiratory distress syndrome. Am. Rev. Respir. Dis. 120: 1039, 1979. 4. Branson, R. D. The measurement of energy expenditure: instrumentation, practical considerations, and clinical application. Respir. Care 35: 103, 1990. 5. Cooper, K. T., and Boswell, P. A. Accurate measurement of functional residual capacity and oxygen consumption of patients on mechanical ventilation. Anaesth. Intensive Care 11: 151, 1983. 6. Laver, M. B., Morgan, J., Bendixen, H. H., and Radford, E. P. Lung volume, compliance, and arterial oxygen tensions during controlled ventilation. J. Appl. Physiol. 19: 725, 1964. 7. Colebatch, H. J., and Engel, L. A. Measurement of lung volume in paralyzed cats. J. Appl. Physiol. 36: 614, 1974. 8. Suter, P. M., and Schlobohm, R. M. Determination of functional residual capacity during mechanical ventilation. Anesthesiology 41: 605, 1974.
AID
srga
APPENDIX
Derivation of Helium Dilution Equation
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
AP: Surg Res
63, 204–208 (1996)
0248
Measuring Functional Residual Capacity in Normal and Oleic Acid-Injured Lungs1 PAUL G. GAUGER, M.D., MICHAEL C. OVERBECK, B.S., SEAN D. CHAMBERS, M.S.E., ERIK D. WEBER, AND RONALD B. HIRSCHL, M.S., M.D.2 Department of Surgery, University of Michigan, Ann Arbor, Michigan 48109 Presented at the Annual Meeting of the Association for Academic Surgery, Dearborn, Michigan, November 8–11, 1995
Functional residual capacity (FRC) is an important oxygen reserve that is often depleted in acute respiratory failure. Recent interest in the mechanisms of liquid ventilation and limited experience in measuring FRC in paralyzed, mechanically ventilated, normal and lung-injured animal models have mandated development of accurate laboratory techniques. Eight sheep, from 17 to 27 kg, were anesthetized and instrumented to provide a tracheostomy, a pulmonary artery catheter, and carotid arterial line. They were randomized to two groups, one of which received 0.07 ml/kg of intravenous oleic acid to induce lung injury. Gas ventilation of both groups was identical except for respiratory rate, which was adjusted to normalize PaCO2 . FRC was measured in duplicate by both helium dilution (HD) and body plethysmography (BP). When measurements were completed, the animals were euthanized and their endotracheal tubes clamped at end expiration. The lungs were then removed and their water displacement (WD) FRC values were measured. FRC was the difference between WD and tissue weight assuming 1 ml Å 1 g. Pearson’s correlation coefficient (R 2) was calculated. During in vitro measurement of test lungs, HD had an R 2 value of 0.99 and BP had an R 2 value of 0.98. When compared to WD, in vivo measurement of FRC by HD had an R 2 value of 0.94 while the value for BP was 0.97. In conclusion, both HD and BP are accurate methods of determining FRC in an uninjured and injured lung model when compared to postmortem WD. Documenting changes in FRC will aid in elucidating the mechanisms of alternative ventilatory techniques. q 1996 Academic Press, Inc.
INTRODUCTION
Functional residual capacity (FRC) is the volume of gas remaining in the lungs at the end of a tidal expiration. FRC is a crucial oxygen reservoir, second in impor1
Funded in part by NIH grant no. RO1HD15434. To whom correspondence and reprint requests should be addressed at Section of Pediatric Surgery, F3970 Mott, Box 0245, 1500 East Medical Center Drive, Ann Arbor, MI 48109–0245. Phone: (313) 764-6846. Fax: (313) 936-9784. 2
0022-4804/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AID
JSR 4799
/
6n0e$$$521
tance only to the hemoglobin pool, and is often depleted in acute respiratory failure. Functional residual capacity is the sum of the expiratory reserve volume and the residual volume. As it includes the residual volume, the absolute value of FRC is not amenable to direct measurement techniques and must rely on indirect techniques such as indicator gas dilution, indicator gas washout, or body plethysmography [1]. FRC has an important effect on gas exchange. It has been demonstrated that lungs of patients with the acute respiratory distress syndrome (ARDS) have a decreased volume of pulmonary gas and that physiologic shunt fraction increases as the FRC decreases. Supportive therapy has targeted increasing the FRC, thus improving ventilation-perfusion matching and subsequently improving gas exchange. This is a well known mechanism of positive end expiratory pressure (PEEP) [2, 3]. Developing therapies to augment FRC obviously requires accurate techniques of measurement applicable to normal and lung-injured, paralyzed and mechanically ventilated large animal models. For this purpose, our goal was to combine a closed-circuit helium dilution system and a body plethysmography system to measure FRC by two different, but complimentary methods. Additionally, we wished the methods to be applicable to both normal and lung-injured models. MATERIALS AND METHODS Eight Suffolk sheep, from 17 to 27 kg, were anesthetized with ketamine HCl (10 mg/kg IV) and guaifenesin (2.2 ml/kg of a 5% solution) after being fasted for 24 hr. Anesthesia was maintained with a standardized mixture of guiafenesin (50 g/l) and ketamine (1.5 g/l). The animals were instrumented to provide a 9-mm endotracheal tube (Mallinckrodt Medical, St. Louis, MO) via a surgical tracheostomy, a Swan-Ganz pulmonary artery catheter, and carotid arterial pressure monitoring. The tracheostomy was tightly sealed and leak tested. Pancuronium was then administered (0.1 mg/kg IV) and was redosed as necessary. All animals were ventilated in volume control mode with a Siemens Servo 900 E (Siemens Elema AB, Sweden). The initial ventilator settings were similar for all animals and included a tidal volume (TV) of 15 ml/kg, inspired oxygen fraction (FiO2) of 1.0, PEEP of 5–7 cm H2O, inspiratory to expiratory ratio of 1:1, and respiratory rate of 10–16 breaths per minute to normalize PaCO2 to 35–45 torr. The animals were then suctioned of their gastric rumen. The sheep were placed in a specially designed full body plethysmo-
204
05-13-96 07:38:39
srga
AP: Surg Res
GAUGER ET AL.: MEASURING FRC IN NORMAL AND INJURED LUNGS
205
FIG. 1. Diagram of the large animal plethysmograph. Water trough served to detect leaks across the lid seal during box pressurization (see text).
graph constructed of 41-inch Plexiglas. Empty, the box had a volume of approximately 130 l. The box is illustrated in Fig. 1. The lid sealed tightly to the box, creating an airtight chamber. A water trough around the lid allowed for detection of any potential air leak across the lid seal. All IV and monitoring lines were connected via sealed luer locks through the side of the box. Likewise, the ventilator was connected to the tracheostomy tube by a sealed connector through the lid of the box. Above this connector was a pinch valve that was tightly closed during plethysmographic measurements (see below). Through a side port connected to the airway, a differential transducer (PX170–28DV, Omega Engineering, Inc., Stamford, CT), protected by a sputum trap, measured the differential between the airway pressure and the ambient box pressure. A strain gauge (CDX III, Cobe, Lakewood CO) connected through a sealed adapter in the box’s side directly measured the box pressure. A schematic of the setup is illustrated in Fig. 2. Both transducer signals were processed and recorded in real time by Lab View software (National Instruments, Inc., Austin, TX). The closed-circuit helium dilution device was constructed to rest atop the box lid. It was constructed of noncompliant tubing and incorporated one way valves to maintain circular gas flow. The circuit included a CO2 scrubbing chamber filled with soda lime (Mallinckrodt Specialty Chemicals, Paris, KY), an in-line O2 analyzer (Instrumentation Laboratories, Lexington MA), a mixing motor (Warren E. Collins, Inc., Braintree, MA), and a ‘‘bag in box’’ type slave bellows. Circuit gas was driven by the motor through a continuous helium analyzer (Warren E. Collins, Inc). Total device volume was 3120 ml. A diagram of the circuit is shown in Fig. 3. The circuit was connected to the airway above a pinch valve by a two way valve (Hans-Rudolph, Kansas City, MO). By this construction, the ventilator was coupled to the airway through the open valves across the ‘‘bridge’’ during most of the experimental period. When helium dilution measurements were performed (see below), the ventilator was coupled to drive the slave bellows by pressurizing the surrounding space, and all ventilating gas then passed into and out of the closed circuit.
FIG. 2. Schematic of setup required for measurement of FRC by null-point body plethysmography. Differential transducer measured the relationship of tracheal pressure and ambient box pressure.
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
FIG. 3. Construction of our closed-circuit device for measurement of FRC by helium dilution. Important features included the PEEP valve, the CO2 scrubber, and the slave bellows, which were driven by the ventilator during measurement.
PEEP was maintained in the airway by a PEEP valve on the expiratory limb of the circuit calibrated to 6 cm H2O. Airway pressures during closed-circuit measurement were monitored via the differential transducer (while the box communicated with atmosphere) and were nearly identical to those generated by the ventilator when the closed circuit was bypassed through the majority of the experiment. To ensure similarity of tidal volumes used during closed-circuit ventilation and ventilation across the bridge, an in-line spirometer (Boeringer, Inc., Ridgefield, CT) was placed directly above the pinch valve and a constant tidal volume of 15–16 ml/kg was maintained. The sheep were randomized to two groups, one of which received intravenous oleic acid to generate a model of acute respiratory failure characterized by hemorrhagic pulmonary edema, increased transpulmonary shunt, decreased compliance, and decreased FRC (n Å 4). The oleic acid was administered, after emulsification with 10 ml blood, in a dose of 0.07 ml/kg into the right atrial port over 20 min. The other group received no lung injury to maintain a presumably normal FRC (n Å 4). Both groups underwent the same anesthesia, instrumentation, and ventilation as described above. Helium dilution (HD) measurement of FRC was performed in duplicate as was nullpoint body plethysmography (BP). Both techniques are explained fully below. Helium dilution. As the animal was ventilated as usual across the bridge, the helium dilution circuit was flushed with 100% oxygen until an oxygen fraction (FO2) of 90–100% was obtained. The circuit was then evacuated of 300 ml of gas to prevent pressurization as 300 ml of pure helium was added. The circulating motor was used to mix the gasses until a stable helium concentration between 7 and 11% was observed on the helium meter. A small diameter bridge was used to bypass the PEEP valve only during mixing to promote homogeneity of the gas mixture. While the circuit was mixing, three inspired gas samples and three mixed expired gas samples were aspirated from the ventilator circuit and analyzed for oxygen percentage on an ABL 30 (Radiometer A/S, Denmark). These percentages within 0.5% were averaged to derive the FiO2 and the FeO2 (expired oxygen fraction). The minute volume ventilation (MVV) was measured (in ml) and the rate of oxygen uptake (O2 consumption) in ml/min was calculated as O2 consumption Å (FiO2 0 FeO2) 1 MVV [4]. The initial equilibrated helium percentage (He%) and the FO2 were noted. The measurement of circuit FO2 was necessary as O2 concentrations above 21% cause an artifactual elevation of the He% due to competing thermal conductance. This overestimation occurs in a linear fashion and can be corrected by the formula Hec Å Hed 0 (FO2 0 0.2093)/37.7 where Hec is the corrected He% and Hed is the He% displayed on the meter [5]. The ventilator was then paused at end expiration (lungs at FRC) and the valve was opened to the closed circuit as the ventilator was switched to drive the slave bellows. A stopwatch was started as the animal was ventilated through the
srga
AP: Surg Res
206
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 1, JUNE 1996
helium dilution circuit. Constancy of tidal volume was assured through monitoring of the in-line spirometer. Helium is an inert indicator gas that is minimally absorbed across the respiratory membrane, and therefore as the circuit gas equilibrated with the FRC gas volume, a new stable helium concentration was reached. At this point, the stopwatch was stopped and the final Hed and FO2 were recorded and the Hec was calculated for the initial and final readings. Again, three samples of inspired gas and three samples of mixed expired gas were measured, and the oxygen uptake rate was calculated as above. This rate was averaged with that determined prehelium dilution and this value was multiplied by the time (in minutes) that the animal was breathing through the closed circuit to more accurately estimate the volume decrease in the circuit attributed to oxygen consumption (O2 consumption). As the animal consumed oxygen at his basal rate, the circuit volume would decrease, as evidenced by the decreasing excursion of the slave bellows. FRC was then calculated using the formula (see appendix for complete derivation): FRC (ml) Å [(He%initial/He%final) 1 V1] 0 V1 / O2 consumption (ml) The previously measured dead space of this circuit configuration included the two way valve, the pinch valve, the sputum trap, and the endotracheal tube (90 ml total) and was subtracted to derive the FRC by HD. This was mechanical dead space and did not include the anatomic and functional dead space of the individual sheep. The value was corrected to BTPS, and an average of both helium dilution measurements was used as the final FRC by HD value. Null point body plethysmography. Duplicate measurements of FRC by body plethysmography were then performed. The technique was based on that originally described for measurement of FRC in dogs by Laver et al. in 1964 and in cats by Colebatch et al. in 1974 [6, 7]. As with all plethysmographic methods, the null-point variation is based on Boyle’s law which is a permutation of the Ideal Gas Law that states that in a closed system at constant temperature, the pressure of a gas multiplied by the volume of a gas is a constant, or P1 V1 Å P2 V2 . Standard plethysmographic techniques were not appropriate for this model as they require spontaneously breathing subjects and the sheep were paralyzed and mechanically ventilated. For each measurement, the ventilator was stopped at end expiration (FRC), and the pinch valve at the lid of the plethysmograph was tightly closed. Twenty to thirty milliliters of room air were injected into the trachea through the sputum trap access from a calibrated gas syringe (DV). This caused a deflection in the signal from the differential transducer (Pd). High flow, compressed air was then injected into the plethysmograph until the value of Pd had been restored to the level recorded prior to injection of DV. The difference in the initial and final pressure of the plethysmograph chamber was termed DPG . The box was vented to atmosphere and ventilation was resumed. End expiratory thoracic gas volume, or FRC, was calculated by the equation derived from Boyle’s law [7]: FRC (ml) Å DV 1 PB 0 PH2O/DPG (see appendix). The dead space of the endotracheal tube and the sputum trap was subtracted (65 ml). The calculated FRC was then corrected to BTPS and the two values were averaged for the FRC by BP determination. After both helium dilution and plethysmographic measurements had been completed, the sheep were removed from the box while continuing the identical ventilatory pattern. They were then euthanized with a barbiturate overdose and the lungs and trachea were rapidly removed en bloc after the endotracheal tube was clamped at end expiration (FRC). Their water displacement volume was then measured in triplicate. The lungs and trachea were then weighed, and allowing for a tissue density of 1 g/ml, their water displacement value (WD) was calculated. Again, this was the mean of the three measurements. The functional residual capacity values measured by helium dilution, null-point plethysmography, and water displacement were then compared. Descriptive statistics (expressed as the mean { SEM) and Pearson’s correlation coefficient (expressed as R 2) were all determined using Systat software (Systat, Inc., Evanston, IL). All FRC values discussed have been normalized for the animal’s body weight and are expressed in ml/kg. Both in vitro and in vivo evaluations of the techniques are reported.
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
FIG. 4. Correlation of measured and known test lung volumes during in vitro development of the techniques. Pearson’s correlation coefficient was 0.99 for helium dilution and 0.98 for body plethysmography.
RESULTS
Before animal models were used, bench top testing of both systems was performed to validate both methods. The helium dilution circuit was used to measure multiple test lungs ranging in volume from 140 to 2900 ml. Comparison with the known volumes of the test lungs revealed excellent correlation with a R 2 value of 0.99. Similar in vitro validation was performed for the null-point plethysmography method using test lungs (volume range 250–1850 ml) with a compliant and noncompliant component as described by Laver et al. [6]. Pearson’s correlation coefficient was 0.98. The in vitro correlation analysis is graphically displayed in Fig. 4 for both techniques. The mean FRC of uninjured animals measured by helium dilution was 34.0 { 3.3 ml/kg, while the mean FRC was 34.8 { 2.8 ml/kg when measured by plethysmography. In comparison, the mean FRC by water displacement was 35.0 { 4.4 ml/kg in uninjured animals. When oleic acid lung injury was induced, the mean FRC by helium dilution was 21.8 { 4.2 ml/kg and the mean FRC by plethysmography was 22.8 { 4.0 ml/kg. When water displacement measurement was used, the mean FRC was 22.3 { 4.9 ml/kg for lung-injured animals. The helium dilution technique was compared to water displacement. The results of all animals measured are displayed in Fig. 5, demonstrating a Pearson’s correlation coefficient of 0.94. Similarly, body plethysmography was compared to water displacement in all animals which revealed an R 2 value of 0.97. This is shown in Fig. 6. DISCUSSION
Helium dilution and null-point body plethysmography are two distinct methods of measuring the functional residual capacity in experimental models and in clinical investigation. Although different, they are complimentary. Each technique has its own advantages and shortcomings. With helium dilution, one of
srga
AP: Surg Res
GAUGER ET AL.: MEASURING FRC IN NORMAL AND INJURED LUNGS
FIG. 5. Correlation of helium dilution FRC values and postmortem water displacement FRC values. Pearson’s correlation coefficient was 0.94 with a linear regression model described by y Å 125 / 0.78x.
the potential sources of inaccuracy is a lack of proper mixing of the gas between the lungs and the circuit. This would leave a higher helium concentration at the end of the measurement period and would thus underestimate FRC. Similarly, another potential source of inaccuracy is the presence of occlusive airway lesions such as excessive secretions or profound pulmonary edema that may prevent the helium mixture from reaching the most peripheral alveolar units in a timely fashion. This would also leave a higher helium reading at the end of the measurement and underestimate FRC. Helium dilution can overestimate FRC if there is even a slight leak in the closed rebreathing circuit, airway, or endotracheal tube cuff. Regarding our specific application of the technique, another possible problem could be the estimation of oxygen uptake to quantify the decrease in circuit volume. Most closed circuits described in the literature include a spirometer as an oxygen or air reservoir [5, 8]. We abandoned this approach for two reasons. First, the spirometer increased the volume of the circuit to a point that we felt the method was losing sensitivity for measuring animal, not adult human lung volumes. Second, with the decreased compliance and increased airway pressures seen with the oleic acid-injured animals, we had trouble controlling the emptying of the spirometer during the varying pressures of the respiratory cycle. Our method avoided both of these confounding situations and kept the circuit volume as small as possible. A possible cause for inaccuracy with the null-point plethysmography method is related to the time needed to complete the measurement. If DV was injected without compressed air subsequently entering the box, the signal would deteriorate back to baseline within 15 sec. Although not as profound, this signal (Pd) would decrease even if DV were not injected. As Colebatch explained, this is therefore not due to simple stress relaxation of the lungs, but is instead due to a decrease in
AID
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
207
intrathoracic volume caused by displacement of thoracic blood volume [7]. Due to the rapid flow capability of our compressed air lines, we were able to consistently complete our measurements in less than 4 sec, and thus before the deterioration in Pd became noticeable. To avoid unintentional forced deflation of the lungs, it was important to vent the box pressure back to atmospheric pressure levels before opening the pinch valve and resuming ventilation. If this was not done, the subsequent FRC measurements could be significantly decreased as exhalation was forced by the pressurized box environment, thus diminishing FRC. Since this method relies on the pressure-volume relationships of the pulmonary gas volume and the plethysmograph during a short period of time, the problem of including accumulated gastric and esophageal gas in the measurement is minimized [6]. Likewise, the problem of changing temperature inside the plethysmograph during the course of the experiment is negligible as the temperature is assumed to be unchanged during the 4 sec needed for the measurement. Thus, the validity of Boyle’s law is maintained. In contrast to helium dilution, plethysmography measures the entire thoracic gas volume, whether or not it communicates freely with the airways. Thus, obstructive airway disease with peripheral air trapping would tend to increase the FRC when compared to a simultaneous measurement by helium dilution. It is interesting to note that in both subsets, the mean FRC was slightly less with helium dilution than with plethysmography. This is likely due to the fact that any occlusive, small airway secretions would tend to decrease the measurement of FRC by HD compared to BP, as already discussed. Also, if Pd deteriorated more significantly than appreciated, then less pressurization of the plethysmograph would be necessary to nullify the change, thus decreasing the value of DPG and increasing the measured FRC by BP. Regardless, the differences in mean FRC values measured by helium dilution and FRC are neither clinically nor statistically significant. In uninjured lungs, the mean FRC by HD was 34.0 { 3.3 ml/kg and the mean FRC by BP
FIG. 6. Correlation of null-point body plethysmography FRC values and postmortem water displacement FRC values. Pearson’s correlation coefficient was 0.97 with a linear regression equation of y Å 137 / 0.81x.
srga
AP: Surg Res
208
JOURNAL OF SURGICAL RESEARCH: VOL. 63, NO. 1, JUNE 1996
of oxygen consumed during the measurement as explained above. Rearranging, V1 C1/C2 Å FRC / (V1 0 O2cons). Rearranging again, FRC Å [(C1/C2) V1] 0 (V1 0 O2cons). Substituting, the final equation becomes FRCHD (ml) Å [(He%initial/He%final) 1 V1] 0 V1 / O2cons (ml). All volumes are corrected to BTPS. Derivation of Null-Point Body Plethysmography Equation
FIG. 7. Correlation of helium dilution FRC and body plethysmography FRC values. Pearson’s correlation coefficient was 0.96 with a linear regression equation of y Å 44 / 0.98x.
was 34.8 { 2.8 ml/kg (P Å 0.74 by Student’s t-test). Similarly, in injured lungs, the mean FRC was 21.8 { 4.2 ml/kg with HD and was 22.75 { 4.0 ml/kg with BP (P Å 0.25, t-test). When directly compared, helium dilution and body plethysmography had an R 2 value of 0.96 (see Fig. 7). CONCLUSION
In summary, we have shown that helium dilution and null-point body plethysmography can be used successfully in combination to measure functional residual capacity in a large animal model. In addition, these methods can be used with success in animals with normal lung physiology and in those with a lung injury produced by intravenous injection of oleic acid. Both methods are accurate when compared to each other and also when compared to postmortem water displacement. These methods may be useful in elucidating the mechanisms of novel ventilatory techniques.
Begin with Boyle’s Law, P1 V1 Å P2 V2 . Then, (PB 0 PH2O) 1 (FRC / DV) Å (PB 0 PH2O / DPPG) 1 FRC, where PB is the barometric pressure (mm Hg) and PH2O is the vapor pressure of water (47 mm Hg), DV is the volume increment injected into the airway at end expiration, and DPG is the pressure increment above atmospheric required to reduce the volume FRC / DV back to FRC. Solving for FRC, the equation becomes FRC Å DV 1 PB 0 PH2O/DPPG . All volumes are corrected to BTPS. ACKNOWLEDGMENTS The authors thank Robert H. Bartlett, M.D., Constance Wise, R.R.T., Ronald E. Dechert, M.S., R.R.T., and Rekha Threja for their assistance in the completion of this work.
REFERENCES
Start with V1 C1 Å V2 C2 , where V1 is the initial volume of the closed circuit (3120 ml), C1 is the corrected initial helium concentration, V2 is the final volume of the closed circuit plus the FRC, and C2 is the corrected final helium concentration. Substituting for V2 , V1 C1 Å [(V1 0 O2cons) / FRC] C2 , where O2cons is the volume
1. Freedman, S. Lung volumes. Br. J. Clin. Pharmacol. 8: 99, 1979. 2. Dueck, R., Wagner, P. D., and West, J. B. Effects of positive end expiratory pressure on gas exchange in dogs with normal and edematous lungs. Anesthesiology 47: 359, 1977. 3. Dantzker, D. R., Brook, C. J., Dehart, P., Lynch, J. P., and Weg, J. G. Ventilation-perfusion distributions in the adult respiratory distress syndrome. Am. Rev. Respir. Dis. 120: 1039, 1979. 4. Branson, R. D. The measurement of energy expenditure: instrumentation, practical considerations, and clinical application. Respir. Care 35: 103, 1990. 5. Cooper, K. T., and Boswell, P. A. Accurate measurement of functional residual capacity and oxygen consumption of patients on mechanical ventilation. Anaesth. Intensive Care 11: 151, 1983. 6. Laver, M. B., Morgan, J., Bendixen, H. H., and Radford, E. P. Lung volume, compliance, and arterial oxygen tensions during controlled ventilation. J. Appl. Physiol. 19: 725, 1964. 7. Colebatch, H. J., and Engel, L. A. Measurement of lung volume in paralyzed cats. J. Appl. Physiol. 36: 614, 1974. 8. Suter, P. M., and Schlobohm, R. M. Determination of functional residual capacity during mechanical ventilation. Anesthesiology 41: 605, 1974.
AID
srga
APPENDIX
Derivation of Helium Dilution Equation
JSR 4799
/
6n0e$$$521
05-13-96 07:38:39
AP: Surg Res