Changes in Functional Residual Capacity and Regional Diaphragm Lengths After Upper Abdominal Surgery in Anesthetized Dogs Juraj Sprung, MD, PhD, George M. Barnas, PhD, Eugene Y. Cheng, Joseph R. Rodarte, MD
MD,
and
Department of Anesthesiology, University of Maryland, Baltimore, Maryland; Department of Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin; Pulmonary Section, Department of Medicine and Physiology, Baylor College of Medicine, Houston, Texas; and Thoracic Disease Research Unit, Mayo Clinic, Rochester, Minnesota
The respiratory performance of the diaphragm may be altered by changes in mechanical or neural factors, or both, induced by upper abdominal surgery. We conducted this study to examine the effects of upper abdominal surgery on postoperative respiratory function. We studied resting lengths of four diaphragm regions, three in the costal and one in the crural diaphragm, with biplane videoroentgenography in six dogs immediately after upper abdominal surgery and up to 30 days postoperatively. Functional residual capacity was 16.7% smaller immediately after surgery compared with values obtained in the same animals after 30 days. Simultaneously measured resting lengths of each of the diaphragm regions immediately after surgery were longer, on average by
P
ostoperative respiratory dysfunction occurs after upper abdominal surgery (UAS) (1-6) and is caused by inhibition of central respiratory drive, not mechanical dysfunction of the diaphragm (2,7,8). Afferent traffic from the diaphragm, if it is stretched, inhibits breathing (9). Because functional residual capacity (FRC) is decreased in patients after UAS (3), and because we have found that diaphragm length varies inversely with lung volume in dogs (lo), we postulated that part of the respiratory dysfunction may be attributed to the inhibition of breathing secondary to an increase in diaphragm stretch. However, results from the two previous studies that have measured diaphragm resting length using sonomi-
This research was supported by Grant HL21584 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. Accepted for publication July 2, 1992. Address correspondence to Dr. Sprung, University of Maryland, Department of Anesthesiology, 22 South Greene Street, Baltimore, MD 21201.
8.3%, than 30 days postoperatively. During the postoperative course, resting diaphragm lengths gradually and uniformly decreased as functional residual capacity increased. Phrenic nerve stimulation in four other dogs immediately after identical surgery resulted in large diaphragm shortening (from 42% to 55%), indicating that neither the diaphragm nor phrenic nerves were injured by the surgical manipulation. We hypothesize that respiratory dysfunction after upper abdominal surgery may be, at least in part, attributed to a decreased central drive for breathing caused by activation of the afferent limb of an inhibitory reflex owing to stretching of the diaphragm. (Anesth Analg 1992;75:977-82)
crometry during recovery from UAS seem to contradict this. In awake, standing sheep, there was no change in resting length of the crural diaphragm, and there was a shortening of the costal region after UAS compared with recovery (8). In awake dogs, there were no significant changes of the crural or costal diaphragm compared with 22 days after surgery (11). Because FRC was not measured in these studies, it is difficult to evaluate the applicability of these results to humans. Presently, there is no study that simultaneously correlates FRC and resting diaphragm length after UAS. In this study, we quantified resting lengths of four regions of the diaphragm by using computerized biplane-videoroentgenography (12) immediately after UAS and during the 30-day recovery period. We were able to correlate postoperative regional length changes in FRC. We conclude that postlaparotomy respiratory dysfunction is, at least in part, due to decreased central respiratory drive caused by reflex inhibition originating from afferents in the stretched diaphragm.
01992 by the International Anesthesia Research Society 0003-2999/92/$5.00
Anesth Analg 1992;75:977432
977
978
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
Figure 1. Excised canine diaphragm (abdominal side) showing schematically the typical location of markers sutured along muscle bundles in the ventral (CoV), middle (CoM), and dorsal (COD) regons of the costal diaphragm and in the crural (Cr) diaphragm. The actual markers were lead beads, 2 mm in diameter, relatively smaller than the white dots shown here. The regional diaphragm length was calculated as a cumulative sum of all intermarker distances. White bor = 8 cm.
Methods Study approval was obtained from the institutional Animal Investigation Committee. Ten beagle dogs (10-13 kg) were anesthetized with pentobarbital (30 mgkg, supplemented as needed), tracheally intubated with a cuffed endotracheal tube (inner diameter 9 mm) without neurornuscular blocking agents, and mechanically ventilated. After induction of anesthesia, a midline laparotomy was performed under sterile conditions. Silicone-coated lead spheres (2-mm diameter) with central holes were stitched with 5-0 Prolene to the peritoneal serosa of the left hemidiaphragm. The rows of spheres, which served as radiographic markers, were placed along muscle bundles from the insertion of the diaphragm on the rib cage to their origin on the central tendon. The spheres were spaced at 1.5-2.0-cm intervals so that the sum of the intermarker distances would closely approximate the arc length along the diaphragm region (12). Figure 1 shows the four regions in which the markers were located: the ventral, middle, and dorsal portion of the costal diaphragm and the crural diaphragm region. All measurements of resting diaphragm lengths and FRC were made with the dogs anesthetized and spontaneously breathing. Anesthesia was induced with pentobarbital (30 mg/kg); sufficient supplemental doses (10%-20% induction dose) were administered to abolish spontaneous movements. Regional
ANESTH ANALG 1992;75:977-82
diaphragm resting length was measured in six dogs on the same day of marker placement, 4-8 h after surgery (OD) and on postoperative day (POD) 30. In addition, measurements were made on POD 3 and 15 in four of the dogs. The dogs were placed supine in a radiolucent volume-displacement body plethysmograph in the center of an orthogonal biplane videoroentgenographic system. The lower rib cage and upper abdomen, with the diaphragm markers, were scanned during consecutive breaths. If fluoroscopy revealed signs of residual postlaparotomy pneumoperitoneum, the air was evacuated with needle aspiration before studies of the diaphragm were attempted. The animals breathed through a Fleisch pneumotachograph that measured flow, and thoracic gas volume was measured at FRC by Boyle’s law during inspiratory efforts against an occluded airway. The fluoroscopic images were encoded on a stopaction videotape. We also measured regional diaphragm shortening during 15 spontaneous breaths and during mechanical ventilation (Harvard pump). For the latter measurements, the same tidal volume as spontaneous breathing was used. Initially, a respiratory frequency more rapid than spontaneous breathing was used to ensure inhibition of breathing efforts. Then, the frequency was reduced to that observed during spontaneous ventilation, and 15 mechanical breaths were recorded. During these breaths, there was no evidence of spontaneous respiratory efforts in any of the dogs. After all data were recorded, each of the six dogs was killed with an overdose of barbiturate. Macroscopic and microscopic (histologic slides) examination of the diaphragm did not show any signs of significant inflammation or other abnormalities. In four additional dogs, the phrenic nerves were electrically stimulated to determine the extent of maximal diaphragm shortening shortly after surgical manipulation identical to that used in the other six dogs. The cervical roots comprising the phrenic nerve (C5-7) were identified and isolated on both sides of the neck. Insulated hook electrodes were placed under the nerve roots and covered with mineral oil. The phrenic nerves were stimulated with 50 Hz using a Grass S88 nerve stimulator. Diaphragmatic shortening was measured during three consecutive phrenic nerve stimulations. The projection of each marker in the two orthogonal fluoroscopic images was identified by an operatorinteractive computer program. A cursor was moved in regular numerical order along the rows of markers on both views of the same video frame. Once the geometric coordinates of all the parenchymal markers in the first frame had been entered, at FRC, the same sequence was repeated at the end of inspiration. The distance between individual markers was computed
ANESTH ANALG 1992;75:977-82
from the three-dimensional spatial coordinates of each marker using solid geometry. The regional diaphragm length was calculated as the sum of individual intermarker lengths. Regional diaphragm shortening was expressed as relative changes in length at endinspiration from the length at FRC (resting length). "Average" diaphragm shortening was calculated by finding the mean shortening of the four regions in each dog and then finding the average of this mean in the six dogs. The biplane videoroentgenography technique does not allow preoperative control values to be established, because the diaphragm markers require surgical placement. However, in a previous longterm study in dogs (12), diaphragm regional length and shortening 2 mo postoperatively were similar to that 30 days after surgery, indicating that 30 days after UAS, diaphragm function has recovered to its preoperative baseline level. Because there were no significant differences among regional diaphragm shortening or lengths on POD 15 and 30 in this study, we assume that recovery was achieved by POD 15. All data are presented as mean and SEM. One-way analysis of variance for repeated measures was used for all paired comparisons for each variable. The Fisher Least Square Difference test was used to compare the statistical significance between repeated measurements. A level of P < 0.05 was considered statistically significant.
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AlTER LAPAROTOMY
OCr
0 55
Figure 2. Resting diaphragm lengths at functional residual capacity (FRC) and corresponding FRC after upper abdominal surgery, linear regressions are indicated by dashed lines. The FRC was significantly lower on the day of surgery compared with postoperative days (POD) 3,15, and 30 POD 0 and 30, n = 6; POD 3 and 15, n = 4. *Difference from POD 0 (P < 0.05). SEM indicated by verticai and horizontal bars. Other abbreviahons as in Figure 1.
0
20
Cr
0
CoV
ACoM
A COD
Spontaneousventilation h
0 15-
e E J
10-
0
c .S
L
0
Mean respiratory frequency immediately postoperatively was 7.5 ? 1.5 breathslmin, less frequent (P < 0.05) than on POD 30 (16.9 ? 5.8 breathdmin). Tidal volumes gradually increased from 135 ? 31 mL on OD to 160 k 33,235 ? 43, and 240 ? 46 mL on POD 3, 15, and 30, respectively. Significant differences ( P < 0.05) in tidal volume were found between OD and all other days and among POD 3, 15, and 30. The regional diaphragm lengths and FRC after UAS are shown in Figure 2. The FRC was lower on OD compared with POD 3, 15, and 30 ( P < 0.05). Although there was a trend for FRC to further increase after POD 3, this increase was not statistically significant. Immediately postoperatively, all diaphragm regions were longer than on POD 30 ( P < 0.05). The average length of the four diaphragm regions on OD was 8.3% _t 0.8% longer than on POD 30 (range 7%-11%). Diaphragm resting lengths in the postoperative period were inversely correlated with FRC. The correlation coefficients between FRC and the diaphragm lengths were as follows: Y = -0.98 for the ventral and middle, -0.92 for the dorsal, and -0.97 for the crural portions.
0 50
FRC (liter)
Q) c
Results
ACoD
--
0 45
0 40
ACoM
eCoV
979
r v)
-
,
5-
1
01
MechanicalVentilation
m
C
I
I
0
3
15
30
Postoperative days Figure 3. Regional diaphragmatic shortening, expressed as percent resting length at functional residual capacity (%LFRC),during spontaneous and mechanical ventilation after upper abdominal surgery. SEM indicated by vertical bars. Postoperative days 0 and 30, n = 6; postoperative days 3 and 15, n = 4. 'Difference in average tidal breath shortening from postoperative days 3, 15, and 30 (P < 0.05). **Difference in average tidal breath shortening from postoperative days 15 and 30 (P < 0.05). Other abbreviations as in Figure 1.
Figure 3 shows regional diaphragmatic shortening during spontaneous and mechanical ventilation. "Average" diaphragmatic shortening during spontaneous breathing was decreased ( P < 0.05) in the immediate postoperative period compared with
980
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
Cr
cov
Diaphragm
CoM
COD
regions
Figure 4. Regonal diaphragmatic shortening during bilateral phrenic nerve stimulation in four dogs, expressed as percent resting length at functional residual capacity (%LFRC). SEM indicated by vertical burs. Other abbreviations as in Figure 1.
PODs 15 and 30. Over the first 15 PODs, the diaphragmatic shortening gradually increased, and by POD 15 the diaphragmatic shortening was comparable to POD 30 (P > 0.05). Diaphragmatic shortening during mechanical ventilation was decreased on OD compared with PODs 3, 15, and 30 (P < 0.05). The ratio of the average shortening in the four diaphragm regions to tidal volume increased during spontaneous ventilation from 31% ? 3% per liter on OD to 49% ? 2% per liter on POD 30. During mechanical ventilation, the ratio increased from 23% ? 3% per liter on OD to 32% 2 3% per liter on POD 30. Figure 4 shows regional diaphragmatic shortening after phrenic nerve stimulation in four dogs. All regions shortened at least 42% after the stimulation.
Discussion Biplane videoroentgenography necessitates the use of anesthetized animals because any gross body movement may result in erroneous measurement. However, barbiturate anesthesia does not affect FRC in supine dogs (13), and the changes in FRC displayed by the anesthetized dogs in our study were similar to those seen in most patients after UAS (3). In contrast, in awake animals FRC does not seem to change after UAS, as inferred from only small, inconsistent changes in resting diaphragm lengths (8,ll). Thus, the anesthetized dog seems a fitting model for the clinical situation. It is also probable that the prolonged period of anesthesia and surgical trauma on the day of marker implantation in the dogs is roughly comparable to those encountered in many abdominal surgeries, because one could expect wide variabilities in patient response to anesthesia and in difficulty of surgery. It is unlikely that the presence of the marker beads on the diaphragm exerted effects on
ANESTH ANALG 1992;75:977-82
afferent stimulation independent of the surgery, because extreme care was taken during surgical implantation, and the markers were sutured to the serosa of the diaphragm undersurface, thus sparing muscle bundles. In summary, the protocol used in the present study offers some advantages over other animal studies, and the results should be reasonably applicable to the clinical situation. According to Craig (3), during POD 1 after UAS, the FRC decreases up to 30% and gradually returns toward normal over 7-10 days. This change in FRC is probably not immediate, because Dureuil et al. (2) did not find significantly reduced FRC 4 h after cholecystectomy. We found a distinct decrease in FRC 4-8 h after UAS that gradually recovered by POD 15. In our study, the regional diaphragm length at FRC was inversely related to FRC and decreased relatively uniformly throughout the recovery period. In contrast, Torres et al. (8) found that after laparotomy in awake sheep, costal resting length was decreased 17% ( P < 0.01) on POD 1 compared with POD 7, whereas the crural resting length remained unchanged throughout the recovery. Easton et al. (11) found different trends in awake, resting dogs: costal resting length at FRC increased 1.56 mm, whereas crural length decreased 1.93 mm, both approximately 11% compared with recovery. However, these length changes were not significant. The videoroentgenography technique allows us to measure length change of a relatively large diaphragm segment ( 4 6 cm) compared with the small segment (1-1.5 cm) measured with sonomicrometry. In our previous study (12), as well as in this one, we found that there were significant differences in intraregional fractional shortenings and that no single location was representative of the shortening of the entire diaphragm. This is in contrast to the report of uniform shortening across the diaphragm segment (14). Videoroentgenography, which measures the length of the entire diaphragm region by summing all intraregional separations between markers, should give a more accurate estimate of global regional diaphragm behavior than sonomicrometry. However, because the studies using sonomicrometry (8,ll) did not measure FRC, it is not possible to decide whether discrepancies among resting lengths in their and our experiments were due solely to limitations of sonomicrometry technique or additionally due to species differences and protocol. The decreased FRC and increased diaphragm length after upper abdominal surgery may be at least partly due to activation of expiratory muscles. Duggan and Drummon (15) reported increased expiratory activation of the abdominal external oblique muscle in humans after cholecystectomy. Farkas and De Troyer (16) found that expiratory activation of the
ANESTH ANALG 1992;75:977-82
triangularis sterni muscle was maintained or increased after laparotomy in dogs. Activation of expiratory muscles may cause cephalad displacement of the diaphragm. The cephalad diaphragm displacement will result in decreased FRC and elongated crural and costal diaphragm parts (17). Whether other factors also contribute to the decrease in FRC after UAS is not clear. Although some workers have suggested that mechanical diaphragm dysfunction due to local irritation, inflammation, or trauma secondary to UAS may be responsible for alteration in contractile properties of the diaphragm (1,5), recent evidence suggests that decreased respiratory drive due to afferent inhibitory traffic is primarily involved (7,8,18). Immediately postoperatively, the dogs in our study displayed much lower minute ventilation (owing to reduction of both tidal volume and respiratory frequency) than on POD 30, which suggests a centrally mediated mechanism of respiratory dysfunction. In addition, evidence from our results is consistent with the conclusion that respiratory dysfunction is not primarily due to mechanical impairment of the diaphragm. First, regional shortening to phrenic nerve stimulation on OD was at least 42% in each region of the diaphragm. The shortening that we found is in a range similar to that described in dogs or sheep during phrenic stimulation after recovery from surgery (8,ll). This indicates that our method, which uses implantation of up to 18 markers on the diaphragm undersurface, did not physically impair diaphragmatic ability to contract. Second, during both spontaneous and mechanical ventilation, the ratio of average regional shortening to tidal volume was less on OD than on POD 30. One might expect that the ratio would be increased if a given amount of diaphragmatic shortening was less effective in increasing lung volume. It is possible, but unlikely, that an extremely abnormal pattern of respiratory muscle activation could cause such a change during spontaneous breathing; however, this might not be the case during mechanical ventilation. Rather, passive changes in chest wall mechanics, such as the observed decrease in FRC, and the likely decrease in abdominal compliance (15) probably affected the ratio. In any case, the results indicate no gross mechanical dysfunction of the diaphragm. There is evidence that increased diaphragmatic stretch, as observed in the present study, could lead to decreased respiratory drive. Cheeseman and Revelette (9) demonstrated that an increase in diaphragmatic operating length decreases phrenic nerve efferent drive, an effect abolished after bilateral cervical dorsal root transection. The increase in diaphragm length tested was approximately 17%,measured with sonomicrometry. It is difficult to directly compare localized sonomicrometric measurements with our
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
981
more global measurements. However, the range of immediate postoperative stretch in our study was only slightly less, between 7% and 11%,and it is likely that increased diaphragmatic stretch activated inhibitory diaphragmatic aff erents. Such a conclusion is consistent with the report by Jammes et al. (19) that muscle spindle and Golgi tendon organ receptors in the diaphragm reflexly inhibited phrenic motor output through phrenic afferents. It should be mentioned that in addition to afferent inhibition from the diaphragm, afferents from other sources could be involved. For example, it has been demonstrated that intercostal and abdominal muscle proprioceptors (20), afferents from the gallbladder (21), and cutaneous afferents in an area of the chest wall where the diaphragm inserts on the ribs (22) could decrease respiratory drive. However, epidural block with local anesthetic up to the T4 level only partially reversed respiratory dysfunction after UAS (17), whereas epidural analgesia had no effect (5'23). Because suppression of vagal inputs and spinal afferents below C8 does not prevent the inhibition of phrenic motoneuron activity due to stimulation of phrenic afferents (19), it is probable that a phrenic-phrenic reflex is involved with inhibition of breathing after UAS. In conclusion, decreased central respiratory drive after UAS could be due in part to reflex inhibition originating in receptors in the stretched diaphragm, such as Golgi tendon organs. The authors are indebted to Dr. Claude Deschamps, who performed the surgery in the dogs.
References 1. Ford GT, Whitelaw WA, Rosenal TW, Cruse PJ, Guenter CA. Diaphragm function after upper abdominal surgery in humans. Am Rev Respir Dis 1983;127431-6. 2. Dureuil B, Viires N, Cantineau JP, Aubier M, Desmonts JM. Diaphragmatic contractility after upper abdominal surgery. J Appl Physiol 1986;61:1775-80. 3. Craig DB. Postoperative recovery of pulmonary function. Anesth Analg 1981;60:46-52. 4. Engberg G. Factors influencing the respiratory capacity after upper abdominal surgery. Acta Anaesthesiol Scand 1985;29: 434-55. 5. Simmoneau G, Vivien A, Sartene R, et al. Diaphragm dysfunction induced by upper abdominal surgery. Am Rev Respir Dis 1983;128:899-903. 6. Chutter TAM, Weissman C, Mathews DM, Starker PM. Diaphragmatic breathing maneuvers and movement of diaphragm after cholecystectomy. Chest 1990;97:1110-4. 7. Road JD, Burgess KR, Whitelaw WA, Ford GT. Diaphragm function and respiratory response after upper abdominal surgery in dogs. J Appl Physiol 1984;57:576-82. 8. Torres A, Kimball WR, Qvist J, et al. Sonomicrometric regional diaphragm shortening in awake sheep after thoracic surgery. J Appl Physiol 1989;67:2357-68. 9. Cheeseman M, Revelette R. Phrenic efferent contribution to
982
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOW
reflexes elicited by changes in diaphragm length. J Appl Physiol 1990;69:64&7. 10. Sprung J, Deschamps C, M a r p h e s SS, Hubmayer RD, Rodarte JR. Effect of body position on regional diaphragm function in dogs. J Appl Physiol 1990;69:2296302. 11. Easton PA, Fitting JW, Amoux R, Guerraty A, Grassino AE. Recovery of diaphragm function after laparotomy and chronic sonomicrometer implantation. J Appl Physiol 1989;66:613-21. 12. Sprung J, Deschamps C, Hubmayr RD, Walters BJ, Rodarte JR. In vivo regional diaphragm function in dogs. J Appl Physiol 1989;67:655-62. 13. Rehder K. Anesthesia and the mechanics of respiration. In: Covino BG, Fozzard HA, Rehder K, Strichartz G, eds. Effects of anesthesia. Bethesda, Md: American Physiological Society, 1985:91-106. 14. Newman S, Road J, Bellemare F, Clozel J, Lavigne CM, Grassino A. Respiratory muscle length measured by sonomicrometry. J Appl Physiol 1984;56:753-&. 15. Duggan JE, Drummon GB. Activity of lower intercostal and abdominal muscle after upper abdominal surgery. Anesth Analg 1987;66:852-5. 16. Farkas GA, De Troyer A. Effects of midline laparotomy on
ANESTH ANALG 1992;75:977-82
17. 18.
19. 20. 21. 22. 23.
expiratory muscle activation in anesthetized dogs. J Appl Physiol 1989;67599405. Duggan JE, Drummon GB. Abdominal muscle activity and intraabdominal pressure after upper abdominal surgery. Anesth Analg 1989;69:59-03. Mankikian B, Cantineau JP, Bertrand M, Kieffer E, Sartene R, Viars P. Improvement of diaphragmatic function by a thoracic extradural block after upper abdominal surgery. Anesthesiology 1988;68:379-86. Jammes Y, Buchler 8, Delpierre S, Rasidakis A, Grimaud C, Roussos C. Phrenic afferents and their role in inspiratory control. J Appl Physioll986;60854-60. Shannon R. Intercostal and abdominal muscle afferent influence on medullary dorsal respiratory group neurons. Respir Physiol 1980;39:73-94. Ford GT, Rideout KS, Bozdech LK, Whitelaw WA, Davidson JS. Inhibition of breathing arising from the gallbladder in dogs (abstract). Physiologist 1983;26:A38. Bellemare F, Garzaniti N. Inhibition of the diaphragm by cutaneous afferents (abstract). Physiologist 1985;28:A338. Clergue F, Montembault C, Despierre 0, Ghesquiere F, Harari A, Viars P. Respiratory effects of intrathecal morphine after upper abdominal surgery. Anesthesiology 19866k677-85.
MD,
and
Department of Anesthesiology, University of Maryland, Baltimore, Maryland; Department of Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin; Pulmonary Section, Department of Medicine and Physiology, Baylor College of Medicine, Houston, Texas; and Thoracic Disease Research Unit, Mayo Clinic, Rochester, Minnesota
The respiratory performance of the diaphragm may be altered by changes in mechanical or neural factors, or both, induced by upper abdominal surgery. We conducted this study to examine the effects of upper abdominal surgery on postoperative respiratory function. We studied resting lengths of four diaphragm regions, three in the costal and one in the crural diaphragm, with biplane videoroentgenography in six dogs immediately after upper abdominal surgery and up to 30 days postoperatively. Functional residual capacity was 16.7% smaller immediately after surgery compared with values obtained in the same animals after 30 days. Simultaneously measured resting lengths of each of the diaphragm regions immediately after surgery were longer, on average by
P
ostoperative respiratory dysfunction occurs after upper abdominal surgery (UAS) (1-6) and is caused by inhibition of central respiratory drive, not mechanical dysfunction of the diaphragm (2,7,8). Afferent traffic from the diaphragm, if it is stretched, inhibits breathing (9). Because functional residual capacity (FRC) is decreased in patients after UAS (3), and because we have found that diaphragm length varies inversely with lung volume in dogs (lo), we postulated that part of the respiratory dysfunction may be attributed to the inhibition of breathing secondary to an increase in diaphragm stretch. However, results from the two previous studies that have measured diaphragm resting length using sonomi-
This research was supported by Grant HL21584 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. Accepted for publication July 2, 1992. Address correspondence to Dr. Sprung, University of Maryland, Department of Anesthesiology, 22 South Greene Street, Baltimore, MD 21201.
8.3%, than 30 days postoperatively. During the postoperative course, resting diaphragm lengths gradually and uniformly decreased as functional residual capacity increased. Phrenic nerve stimulation in four other dogs immediately after identical surgery resulted in large diaphragm shortening (from 42% to 55%), indicating that neither the diaphragm nor phrenic nerves were injured by the surgical manipulation. We hypothesize that respiratory dysfunction after upper abdominal surgery may be, at least in part, attributed to a decreased central drive for breathing caused by activation of the afferent limb of an inhibitory reflex owing to stretching of the diaphragm. (Anesth Analg 1992;75:977-82)
crometry during recovery from UAS seem to contradict this. In awake, standing sheep, there was no change in resting length of the crural diaphragm, and there was a shortening of the costal region after UAS compared with recovery (8). In awake dogs, there were no significant changes of the crural or costal diaphragm compared with 22 days after surgery (11). Because FRC was not measured in these studies, it is difficult to evaluate the applicability of these results to humans. Presently, there is no study that simultaneously correlates FRC and resting diaphragm length after UAS. In this study, we quantified resting lengths of four regions of the diaphragm by using computerized biplane-videoroentgenography (12) immediately after UAS and during the 30-day recovery period. We were able to correlate postoperative regional length changes in FRC. We conclude that postlaparotomy respiratory dysfunction is, at least in part, due to decreased central respiratory drive caused by reflex inhibition originating from afferents in the stretched diaphragm.
01992 by the International Anesthesia Research Society 0003-2999/92/$5.00
Anesth Analg 1992;75:977432
977
978
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
Figure 1. Excised canine diaphragm (abdominal side) showing schematically the typical location of markers sutured along muscle bundles in the ventral (CoV), middle (CoM), and dorsal (COD) regons of the costal diaphragm and in the crural (Cr) diaphragm. The actual markers were lead beads, 2 mm in diameter, relatively smaller than the white dots shown here. The regional diaphragm length was calculated as a cumulative sum of all intermarker distances. White bor = 8 cm.
Methods Study approval was obtained from the institutional Animal Investigation Committee. Ten beagle dogs (10-13 kg) were anesthetized with pentobarbital (30 mgkg, supplemented as needed), tracheally intubated with a cuffed endotracheal tube (inner diameter 9 mm) without neurornuscular blocking agents, and mechanically ventilated. After induction of anesthesia, a midline laparotomy was performed under sterile conditions. Silicone-coated lead spheres (2-mm diameter) with central holes were stitched with 5-0 Prolene to the peritoneal serosa of the left hemidiaphragm. The rows of spheres, which served as radiographic markers, were placed along muscle bundles from the insertion of the diaphragm on the rib cage to their origin on the central tendon. The spheres were spaced at 1.5-2.0-cm intervals so that the sum of the intermarker distances would closely approximate the arc length along the diaphragm region (12). Figure 1 shows the four regions in which the markers were located: the ventral, middle, and dorsal portion of the costal diaphragm and the crural diaphragm region. All measurements of resting diaphragm lengths and FRC were made with the dogs anesthetized and spontaneously breathing. Anesthesia was induced with pentobarbital (30 mg/kg); sufficient supplemental doses (10%-20% induction dose) were administered to abolish spontaneous movements. Regional
ANESTH ANALG 1992;75:977-82
diaphragm resting length was measured in six dogs on the same day of marker placement, 4-8 h after surgery (OD) and on postoperative day (POD) 30. In addition, measurements were made on POD 3 and 15 in four of the dogs. The dogs were placed supine in a radiolucent volume-displacement body plethysmograph in the center of an orthogonal biplane videoroentgenographic system. The lower rib cage and upper abdomen, with the diaphragm markers, were scanned during consecutive breaths. If fluoroscopy revealed signs of residual postlaparotomy pneumoperitoneum, the air was evacuated with needle aspiration before studies of the diaphragm were attempted. The animals breathed through a Fleisch pneumotachograph that measured flow, and thoracic gas volume was measured at FRC by Boyle’s law during inspiratory efforts against an occluded airway. The fluoroscopic images were encoded on a stopaction videotape. We also measured regional diaphragm shortening during 15 spontaneous breaths and during mechanical ventilation (Harvard pump). For the latter measurements, the same tidal volume as spontaneous breathing was used. Initially, a respiratory frequency more rapid than spontaneous breathing was used to ensure inhibition of breathing efforts. Then, the frequency was reduced to that observed during spontaneous ventilation, and 15 mechanical breaths were recorded. During these breaths, there was no evidence of spontaneous respiratory efforts in any of the dogs. After all data were recorded, each of the six dogs was killed with an overdose of barbiturate. Macroscopic and microscopic (histologic slides) examination of the diaphragm did not show any signs of significant inflammation or other abnormalities. In four additional dogs, the phrenic nerves were electrically stimulated to determine the extent of maximal diaphragm shortening shortly after surgical manipulation identical to that used in the other six dogs. The cervical roots comprising the phrenic nerve (C5-7) were identified and isolated on both sides of the neck. Insulated hook electrodes were placed under the nerve roots and covered with mineral oil. The phrenic nerves were stimulated with 50 Hz using a Grass S88 nerve stimulator. Diaphragmatic shortening was measured during three consecutive phrenic nerve stimulations. The projection of each marker in the two orthogonal fluoroscopic images was identified by an operatorinteractive computer program. A cursor was moved in regular numerical order along the rows of markers on both views of the same video frame. Once the geometric coordinates of all the parenchymal markers in the first frame had been entered, at FRC, the same sequence was repeated at the end of inspiration. The distance between individual markers was computed
ANESTH ANALG 1992;75:977-82
from the three-dimensional spatial coordinates of each marker using solid geometry. The regional diaphragm length was calculated as the sum of individual intermarker lengths. Regional diaphragm shortening was expressed as relative changes in length at endinspiration from the length at FRC (resting length). "Average" diaphragm shortening was calculated by finding the mean shortening of the four regions in each dog and then finding the average of this mean in the six dogs. The biplane videoroentgenography technique does not allow preoperative control values to be established, because the diaphragm markers require surgical placement. However, in a previous longterm study in dogs (12), diaphragm regional length and shortening 2 mo postoperatively were similar to that 30 days after surgery, indicating that 30 days after UAS, diaphragm function has recovered to its preoperative baseline level. Because there were no significant differences among regional diaphragm shortening or lengths on POD 15 and 30 in this study, we assume that recovery was achieved by POD 15. All data are presented as mean and SEM. One-way analysis of variance for repeated measures was used for all paired comparisons for each variable. The Fisher Least Square Difference test was used to compare the statistical significance between repeated measurements. A level of P < 0.05 was considered statistically significant.
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AlTER LAPAROTOMY
OCr
0 55
Figure 2. Resting diaphragm lengths at functional residual capacity (FRC) and corresponding FRC after upper abdominal surgery, linear regressions are indicated by dashed lines. The FRC was significantly lower on the day of surgery compared with postoperative days (POD) 3,15, and 30 POD 0 and 30, n = 6; POD 3 and 15, n = 4. *Difference from POD 0 (P < 0.05). SEM indicated by verticai and horizontal bars. Other abbreviahons as in Figure 1.
0
20
Cr
0
CoV
ACoM
A COD
Spontaneousventilation h
0 15-
e E J
10-
0
c .S
L
0
Mean respiratory frequency immediately postoperatively was 7.5 ? 1.5 breathslmin, less frequent (P < 0.05) than on POD 30 (16.9 ? 5.8 breathdmin). Tidal volumes gradually increased from 135 ? 31 mL on OD to 160 k 33,235 ? 43, and 240 ? 46 mL on POD 3, 15, and 30, respectively. Significant differences ( P < 0.05) in tidal volume were found between OD and all other days and among POD 3, 15, and 30. The regional diaphragm lengths and FRC after UAS are shown in Figure 2. The FRC was lower on OD compared with POD 3, 15, and 30 ( P < 0.05). Although there was a trend for FRC to further increase after POD 3, this increase was not statistically significant. Immediately postoperatively, all diaphragm regions were longer than on POD 30 ( P < 0.05). The average length of the four diaphragm regions on OD was 8.3% _t 0.8% longer than on POD 30 (range 7%-11%). Diaphragm resting lengths in the postoperative period were inversely correlated with FRC. The correlation coefficients between FRC and the diaphragm lengths were as follows: Y = -0.98 for the ventral and middle, -0.92 for the dorsal, and -0.97 for the crural portions.
0 50
FRC (liter)
Q) c
Results
ACoD
--
0 45
0 40
ACoM
eCoV
979
r v)
-
,
5-
1
01
MechanicalVentilation
m
C
I
I
0
3
15
30
Postoperative days Figure 3. Regional diaphragmatic shortening, expressed as percent resting length at functional residual capacity (%LFRC),during spontaneous and mechanical ventilation after upper abdominal surgery. SEM indicated by vertical bars. Postoperative days 0 and 30, n = 6; postoperative days 3 and 15, n = 4. 'Difference in average tidal breath shortening from postoperative days 3, 15, and 30 (P < 0.05). **Difference in average tidal breath shortening from postoperative days 15 and 30 (P < 0.05). Other abbreviations as in Figure 1.
Figure 3 shows regional diaphragmatic shortening during spontaneous and mechanical ventilation. "Average" diaphragmatic shortening during spontaneous breathing was decreased ( P < 0.05) in the immediate postoperative period compared with
980
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
Cr
cov
Diaphragm
CoM
COD
regions
Figure 4. Regonal diaphragmatic shortening during bilateral phrenic nerve stimulation in four dogs, expressed as percent resting length at functional residual capacity (%LFRC). SEM indicated by vertical burs. Other abbreviations as in Figure 1.
PODs 15 and 30. Over the first 15 PODs, the diaphragmatic shortening gradually increased, and by POD 15 the diaphragmatic shortening was comparable to POD 30 (P > 0.05). Diaphragmatic shortening during mechanical ventilation was decreased on OD compared with PODs 3, 15, and 30 (P < 0.05). The ratio of the average shortening in the four diaphragm regions to tidal volume increased during spontaneous ventilation from 31% ? 3% per liter on OD to 49% ? 2% per liter on POD 30. During mechanical ventilation, the ratio increased from 23% ? 3% per liter on OD to 32% 2 3% per liter on POD 30. Figure 4 shows regional diaphragmatic shortening after phrenic nerve stimulation in four dogs. All regions shortened at least 42% after the stimulation.
Discussion Biplane videoroentgenography necessitates the use of anesthetized animals because any gross body movement may result in erroneous measurement. However, barbiturate anesthesia does not affect FRC in supine dogs (13), and the changes in FRC displayed by the anesthetized dogs in our study were similar to those seen in most patients after UAS (3). In contrast, in awake animals FRC does not seem to change after UAS, as inferred from only small, inconsistent changes in resting diaphragm lengths (8,ll). Thus, the anesthetized dog seems a fitting model for the clinical situation. It is also probable that the prolonged period of anesthesia and surgical trauma on the day of marker implantation in the dogs is roughly comparable to those encountered in many abdominal surgeries, because one could expect wide variabilities in patient response to anesthesia and in difficulty of surgery. It is unlikely that the presence of the marker beads on the diaphragm exerted effects on
ANESTH ANALG 1992;75:977-82
afferent stimulation independent of the surgery, because extreme care was taken during surgical implantation, and the markers were sutured to the serosa of the diaphragm undersurface, thus sparing muscle bundles. In summary, the protocol used in the present study offers some advantages over other animal studies, and the results should be reasonably applicable to the clinical situation. According to Craig (3), during POD 1 after UAS, the FRC decreases up to 30% and gradually returns toward normal over 7-10 days. This change in FRC is probably not immediate, because Dureuil et al. (2) did not find significantly reduced FRC 4 h after cholecystectomy. We found a distinct decrease in FRC 4-8 h after UAS that gradually recovered by POD 15. In our study, the regional diaphragm length at FRC was inversely related to FRC and decreased relatively uniformly throughout the recovery period. In contrast, Torres et al. (8) found that after laparotomy in awake sheep, costal resting length was decreased 17% ( P < 0.01) on POD 1 compared with POD 7, whereas the crural resting length remained unchanged throughout the recovery. Easton et al. (11) found different trends in awake, resting dogs: costal resting length at FRC increased 1.56 mm, whereas crural length decreased 1.93 mm, both approximately 11% compared with recovery. However, these length changes were not significant. The videoroentgenography technique allows us to measure length change of a relatively large diaphragm segment ( 4 6 cm) compared with the small segment (1-1.5 cm) measured with sonomicrometry. In our previous study (12), as well as in this one, we found that there were significant differences in intraregional fractional shortenings and that no single location was representative of the shortening of the entire diaphragm. This is in contrast to the report of uniform shortening across the diaphragm segment (14). Videoroentgenography, which measures the length of the entire diaphragm region by summing all intraregional separations between markers, should give a more accurate estimate of global regional diaphragm behavior than sonomicrometry. However, because the studies using sonomicrometry (8,ll) did not measure FRC, it is not possible to decide whether discrepancies among resting lengths in their and our experiments were due solely to limitations of sonomicrometry technique or additionally due to species differences and protocol. The decreased FRC and increased diaphragm length after upper abdominal surgery may be at least partly due to activation of expiratory muscles. Duggan and Drummon (15) reported increased expiratory activation of the abdominal external oblique muscle in humans after cholecystectomy. Farkas and De Troyer (16) found that expiratory activation of the
ANESTH ANALG 1992;75:977-82
triangularis sterni muscle was maintained or increased after laparotomy in dogs. Activation of expiratory muscles may cause cephalad displacement of the diaphragm. The cephalad diaphragm displacement will result in decreased FRC and elongated crural and costal diaphragm parts (17). Whether other factors also contribute to the decrease in FRC after UAS is not clear. Although some workers have suggested that mechanical diaphragm dysfunction due to local irritation, inflammation, or trauma secondary to UAS may be responsible for alteration in contractile properties of the diaphragm (1,5), recent evidence suggests that decreased respiratory drive due to afferent inhibitory traffic is primarily involved (7,8,18). Immediately postoperatively, the dogs in our study displayed much lower minute ventilation (owing to reduction of both tidal volume and respiratory frequency) than on POD 30, which suggests a centrally mediated mechanism of respiratory dysfunction. In addition, evidence from our results is consistent with the conclusion that respiratory dysfunction is not primarily due to mechanical impairment of the diaphragm. First, regional shortening to phrenic nerve stimulation on OD was at least 42% in each region of the diaphragm. The shortening that we found is in a range similar to that described in dogs or sheep during phrenic stimulation after recovery from surgery (8,ll). This indicates that our method, which uses implantation of up to 18 markers on the diaphragm undersurface, did not physically impair diaphragmatic ability to contract. Second, during both spontaneous and mechanical ventilation, the ratio of average regional shortening to tidal volume was less on OD than on POD 30. One might expect that the ratio would be increased if a given amount of diaphragmatic shortening was less effective in increasing lung volume. It is possible, but unlikely, that an extremely abnormal pattern of respiratory muscle activation could cause such a change during spontaneous breathing; however, this might not be the case during mechanical ventilation. Rather, passive changes in chest wall mechanics, such as the observed decrease in FRC, and the likely decrease in abdominal compliance (15) probably affected the ratio. In any case, the results indicate no gross mechanical dysfunction of the diaphragm. There is evidence that increased diaphragmatic stretch, as observed in the present study, could lead to decreased respiratory drive. Cheeseman and Revelette (9) demonstrated that an increase in diaphragmatic operating length decreases phrenic nerve efferent drive, an effect abolished after bilateral cervical dorsal root transection. The increase in diaphragm length tested was approximately 17%,measured with sonomicrometry. It is difficult to directly compare localized sonomicrometric measurements with our
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOMY
981
more global measurements. However, the range of immediate postoperative stretch in our study was only slightly less, between 7% and 11%,and it is likely that increased diaphragmatic stretch activated inhibitory diaphragmatic aff erents. Such a conclusion is consistent with the report by Jammes et al. (19) that muscle spindle and Golgi tendon organ receptors in the diaphragm reflexly inhibited phrenic motor output through phrenic afferents. It should be mentioned that in addition to afferent inhibition from the diaphragm, afferents from other sources could be involved. For example, it has been demonstrated that intercostal and abdominal muscle proprioceptors (20), afferents from the gallbladder (21), and cutaneous afferents in an area of the chest wall where the diaphragm inserts on the ribs (22) could decrease respiratory drive. However, epidural block with local anesthetic up to the T4 level only partially reversed respiratory dysfunction after UAS (17), whereas epidural analgesia had no effect (5'23). Because suppression of vagal inputs and spinal afferents below C8 does not prevent the inhibition of phrenic motoneuron activity due to stimulation of phrenic afferents (19), it is probable that a phrenic-phrenic reflex is involved with inhibition of breathing after UAS. In conclusion, decreased central respiratory drive after UAS could be due in part to reflex inhibition originating in receptors in the stretched diaphragm, such as Golgi tendon organs. The authors are indebted to Dr. Claude Deschamps, who performed the surgery in the dogs.
References 1. Ford GT, Whitelaw WA, Rosenal TW, Cruse PJ, Guenter CA. Diaphragm function after upper abdominal surgery in humans. Am Rev Respir Dis 1983;127431-6. 2. Dureuil B, Viires N, Cantineau JP, Aubier M, Desmonts JM. Diaphragmatic contractility after upper abdominal surgery. J Appl Physiol 1986;61:1775-80. 3. Craig DB. Postoperative recovery of pulmonary function. Anesth Analg 1981;60:46-52. 4. Engberg G. Factors influencing the respiratory capacity after upper abdominal surgery. Acta Anaesthesiol Scand 1985;29: 434-55. 5. Simmoneau G, Vivien A, Sartene R, et al. Diaphragm dysfunction induced by upper abdominal surgery. Am Rev Respir Dis 1983;128:899-903. 6. Chutter TAM, Weissman C, Mathews DM, Starker PM. Diaphragmatic breathing maneuvers and movement of diaphragm after cholecystectomy. Chest 1990;97:1110-4. 7. Road JD, Burgess KR, Whitelaw WA, Ford GT. Diaphragm function and respiratory response after upper abdominal surgery in dogs. J Appl Physiol 1984;57:576-82. 8. Torres A, Kimball WR, Qvist J, et al. Sonomicrometric regional diaphragm shortening in awake sheep after thoracic surgery. J Appl Physiol 1989;67:2357-68. 9. Cheeseman M, Revelette R. Phrenic efferent contribution to
982
SPRUNG ET AL. FRC AND DIAPHRAGM LENGTH AFTER LAPAROTOW
reflexes elicited by changes in diaphragm length. J Appl Physiol 1990;69:64&7. 10. Sprung J, Deschamps C, M a r p h e s SS, Hubmayer RD, Rodarte JR. Effect of body position on regional diaphragm function in dogs. J Appl Physiol 1990;69:2296302. 11. Easton PA, Fitting JW, Amoux R, Guerraty A, Grassino AE. Recovery of diaphragm function after laparotomy and chronic sonomicrometer implantation. J Appl Physiol 1989;66:613-21. 12. Sprung J, Deschamps C, Hubmayr RD, Walters BJ, Rodarte JR. In vivo regional diaphragm function in dogs. J Appl Physiol 1989;67:655-62. 13. Rehder K. Anesthesia and the mechanics of respiration. In: Covino BG, Fozzard HA, Rehder K, Strichartz G, eds. Effects of anesthesia. Bethesda, Md: American Physiological Society, 1985:91-106. 14. Newman S, Road J, Bellemare F, Clozel J, Lavigne CM, Grassino A. Respiratory muscle length measured by sonomicrometry. J Appl Physiol 1984;56:753-&. 15. Duggan JE, Drummon GB. Activity of lower intercostal and abdominal muscle after upper abdominal surgery. Anesth Analg 1987;66:852-5. 16. Farkas GA, De Troyer A. Effects of midline laparotomy on
ANESTH ANALG 1992;75:977-82
17. 18.
19. 20. 21. 22. 23.
expiratory muscle activation in anesthetized dogs. J Appl Physiol 1989;67599405. Duggan JE, Drummon GB. Abdominal muscle activity and intraabdominal pressure after upper abdominal surgery. Anesth Analg 1989;69:59-03. Mankikian B, Cantineau JP, Bertrand M, Kieffer E, Sartene R, Viars P. Improvement of diaphragmatic function by a thoracic extradural block after upper abdominal surgery. Anesthesiology 1988;68:379-86. Jammes Y, Buchler 8, Delpierre S, Rasidakis A, Grimaud C, Roussos C. Phrenic afferents and their role in inspiratory control. J Appl Physioll986;60854-60. Shannon R. Intercostal and abdominal muscle afferent influence on medullary dorsal respiratory group neurons. Respir Physiol 1980;39:73-94. Ford GT, Rideout KS, Bozdech LK, Whitelaw WA, Davidson JS. Inhibition of breathing arising from the gallbladder in dogs (abstract). Physiologist 1983;26:A38. Bellemare F, Garzaniti N. Inhibition of the diaphragm by cutaneous afferents (abstract). Physiologist 1985;28:A338. Clergue F, Montembault C, Despierre 0, Ghesquiere F, Harari A, Viars P. Respiratory effects of intrathecal morphine after upper abdominal surgery. Anesthesiology 19866k677-85.
