Eur Respir J 199, 6, 19- Effet of pleural effusion on respiratory mehanis, and the influene of deep inflation, in dogs G. Dehman, J. Sato, J.H.T. Bates Effet of pleural effusion on respiratory mehanis, and the influene of deep inflation, in dogs. G. Dehman, J. Sato, J.H.T. Bates. ABSTRACT: We wished to study how pleural effusion affets dynami mehanis of the lung and the hest wall. We also determined if these hanges ould be reversed by deep lung inflations. Pleural effusion was produed by saline infusion into the pleural spae. During the infusion and over the following hours, dynami elastane and resistane of the lungs, the hest wall and the whole respiratory system were reorded. Dynami elastane and resistane of the lung inreased ignifianl.ly during fluid loading and were partially, and only transiently, reversed by deep inflations. Dynami elastane and resistane of the hest wall were little affeted by these l>roedures. Thus, pleural effu sion an have signifiant l'f'ets on d. nami el.nstane and resistane ur the respiratory. yslem CEts, RRs). The transient nature of the h;lnge in lung pammeters afl'r deep inflat:ion suggests tllllf the1 apies ba d on periodj lung intlations may be of little benefit to patients with this ondition. Eur Respir J., 199, 6, 19-. Meakins-Christie Laboratories, MGill University, Montreal, Quebe, Canada. Correspondene: G. Dehman Meakins-Christie Laboratories 66 St. Urbain St Montreal Quebe Canada HX P Keywords: Elastane resistane tissue and airway properties volume hange Reeived: Marh I 0 199 Aepted after revision July 9 199 Pleural effusions are a very ommon linial entity. To date, investigations of the effets of pleural effusions on lung funtion have onentrated on quantifying the hanges whih our in gas exhange, lung volume [1-], or fored expiratory flow rates [, ]. A few studies have examined the alterations in stati elastane of the respiratory system whih our as a result of pleural effusion [-]. However, none have assessed hanges in dynami elastane (E) and resistane (R) that aompany pleural effusion. These dynami properties of the respiratory system, whih may differ markedly from its stati properties, are obviously more relevant during breathing. The purpose of the present study was to investigate the hanges in the dynami mehanial properties of the anine respiratory system and its omponents, the lung and hest wall, produed by the progressive development of pleural effusion. We also sought to determine the extent to whih suh hanges ould be reversed by deep lung inflations, in order to disern the therapeuti effetiveness of suh a manoeuvre. Methods Six mongrel dogs, weighing 18-5 kg were anaesthetized with pentobarbital sodium (5 mg kg- 1 i.v.) and maintained with an hourly dose of 10--15% of the initial dose. The dogs were positioned supine, traheostomized and a annula was inserted in the airway (ID 0 mm). Traheal pressure (Ptr) was measured by a piezoresistive pressure transduer (ICS 1 00 g 8051610, SPR Control Systems Ltd, Rexdale, Ontario, Canada) at a lateral tap in the annula. A heated Fleish No. pneumotahograph was onneted o the annula for the measurement of traheal flow (V). The pressure drop aross the pneumotahograph was measured by a piezoresistive pressure transduer (MiroSwith 16PCOID6, Honeywell, Sarborough, Ontario, Canada). A three-way valve was onneted to the pneumotahograph to allow olusion of the airway during the olusion test (see below). Pleural pressure (Poes) was measured with a thin latex balloon, 5.5 m long and sealed over one end of a thin polyethylene atheter (88 m long, 1.7 m ID). The other end of the atheter was onneted to another ICS pressure transduer. The balloon was filled with 0.7 m! of air, whih plaed it on the flat portion of its volume pressure urve. All signals were passed through 8 pole Bessel filters (90LPF, Frequeny Devies, Haverhill, MA, USA) with their orner frequenies set at 0 Hz. They were then sampled at 100 Hz with a 1-bit analogueto-digital onverter (DT801 A, Data Translation, Marlborough, MA, USA) and stored on omputer. All data were olleted using LABDAT software (RHTlnfoData In., Montreal, Quebe, Canada).
0 G. DECHMAN, J. SATO, J.H.T. BATES The oesophageal balloon was positioned 10 m above the oesophageal sphinter and the olusion test was performed at funtional residual apaity (FRC) and FRC+500 ml [5]. The slope of Poes vs Ptr, obtained during a spontaneous breathing effort, did not vary more than 5% from unity in any dog at either lung volume. We have previously shown that the hanges in Poes and Ptr during an olusion test are even loser after paralysis than during spontaneous breathing efforts [6]. However, the ability of the oesophageal balloon to aurately measure hanges in pleural pressure with an effusion in plae is ontroversial [, 7, 8]. Therefore, we tested the auray of Poes as a measure of pleural pressure in a similar manner to that desribed in our previous investigations, in two paralysed dogs with unilateral pleural effusion. Speifially, the dogs were traheostomized and an oesophageal balloon plaed 10 m above the gastro-oesophageal juntion. Eah dog was plaed in a plethysmograph, paralysed with panuronium bromide, and ventilated. The olusion test was performed by oluding the airway and measuring Ptr and Poes, while the pressure around the animal was osillated by injeting and withdrawing I of air from the plethysmograph. This aused Ptr and Poes to osillate quasi-sinusoidally, with an amplitude of 1 kpa and a frequeny of 0.08-0.8 Hz. We then infused saline into the pleural spae, until we obtained the same maximum volume per kg used in the dogs reported in this paper. The osillation test was then repeated. The slope of Ptr vs Poes did not vary more than % between the ontrol and effusion states. Therefore, we are onfident that Poes aurately refleted mean pleural pressure in the investigations reported here. Dogs were paralysed with a bolus of panuronium bromide ( mg) and paralysis maintained by mg hourly doses thereafter. The three-way valve was removed from the pneumotahograph and a Harvard ventilator (model 618, Harvard Apparatus, Southnatik, MA, USA) onneted in its plae. The dogs were mehanially ventilated with a tidal volume of 15 ml kg- 1 at a frequeny of 0 breaths min- 1 A polyethylene atheter was inserted into either the right or left pleural spae, at the level of the seventh or eighth interostal spae. The atheter was sutured in plae and air was evauated from the pleural spae using an underwater sealed sution apparatus. Just prior to saline infusion, three deep inflations were given. A deep inflation was aomplished by oluding the expiratory port of the ventilator for three onseutive breaths, whih raised transpulmonary pressure to approximately kpa. Saline was then infused into the pleural spae in 60 m! inrements, given eah minute, until an effusion of 60 ml kg- 1 body weight had been administered. Ptr, Poes and V were measured ontinuously for min prior to, during, and 10 min following loading of the effusion. They were also measured for 0 s every 5 min over the next h. Following loading of the effusion, three deep inflations were given every 0 min, immediately after a data olletion period. At the termination of the experiment we opened the hests of two of the dogs and removed the pleural effusion using a 60 ml syringe. In both ases we reovered at least 90% of the effusate. Data Analysis Forty seond segments of Ptr, Poes and V following eah inrement in effusate were isolated from the ontinuous data reord. Volume (V) was alulated by numerial integration of V. A small onstant was added to V prior to integration so that the resulting V had no baseline drift. The segments were divided into individual breaths, the breaths superimposed and the data ensemble averaged [9]. Similar ensemble averaging was performed on the data olleted in disrete 0 s samples in the h period following fluid loading. We then fitted the equation: P = ExV + RxV + K to eah ensemble averaged data set by multiple linear regression in order to alulate elastane (E) and resistane (R) of the respiratory system (Rs), lung (L) and hest wall (w), where RS = L + w.. K (onstant) is the value of pressure when V and V are both zero. Of ourse, these model parameters do not haraterize the respiratory system perfetly, but they do embody the great majority of its elasti and dissipative properties during regular mehanial ventilation [ 10]. P was represented by Ptr, Poes or Ptr-Poes (Ppt), yielding ERs and RRs, Ew and Rw and EL and RL, respetvely. In all ases, the equation P = ExV + RxV + K aounted for at least 98% of the variane in the dependent variable P as indiated by the oeffiient of variation. All data were analysed using ANADAT data analysis software (RHT-InfoDat In., Montreal, Quebe, Canada). Results Figure 1 shows hanges in ERs, Ew and EL during baseline until the end of effusate loading for eah dog studied. Figure illustrates the hanges in R for the same time period in the same dogs. In all ases, there are steady inreases in ERs, EL, RRs and RL. There are muh smaller dereases in Ew and Rw. Changes in ERs and RRs are very similar in shape and amplitude to those observed for the lung. Table 1 presents the perentage hanges in E and R from baseline values produed by effusate loading for eah dog. In all ases the hanges in E and R for the RS, L, and w were signifiant at p<0.05. Figures 1 and demonstrate an abrupt derease in E and R in dog no. at 15 min after initiation of loading. Examination of the orresponding Ptr signal revealed a sudden inrease in the baseline and a derease in the amplitude of the pressure swing.
RESPIRATORY MECHANICS WITH PLEURAL EFFUSION 1 5 Dog no. 1 " -10 0 5 10 15 0 Dog no. 1 0. 0. 0.0-10 0 5 10 15 0 0. Dog no. 0. 0 0 5 10 15 0 5 0.0 0 5 10 15 0 5..._ eo a... u ro u; ro [jj 6 Dog no. 5 1 " 0 5 10 15 0 0. <p ";'"..._ eo a... 0.0 u 0.8 Dog no. u; 0. Dog no. 0. 0 5 10 15 0 0-5 0 5 10 15 0 5 0.0-5 0 5 10 15 0 5 Dog no. 5 0. 0. Dog no. 5 eeeee eee e eeeeeeeee eee e 0 5 10 15 0 0 5 10 15 0 0. 0 5 10 15 0 5 0 Fig. l. - ERs (e), EL (D) and Ew () plotted against time. "0" time indiates the initiation of effusate loading, whih is omplete at the end of eah plot. Points before initiation of loading are baseline data. (In dogs no. 1, and 6 the break on the absissa is used beause baseline data were measured 8-10 min prior to initiation of loading whih would unneessarily extend graph). ERs, EL and Ew: elastane of the respiratory system, lung and hest wall. respetively. 0.0 L.,_-T--'--'---'--,L_---'---' 0 5 10 15 0 5 0 Fig.. - RRs (e), RL (D) and Rw () plotted against time. "0" time indiates the initiation of effusate loading, whih is omplete at the end of eah plot. Points before initiation of loading are baseline data. (In dogs no. I, and 6 the break on the absissa is used beause baseline data were measured 8-10 min prior to initiation of loading whih would unneessarily extend graph). RRs, RL and Rw: resistane of the respiratory system, lung and hest wall, respetively.
G. DECHMAN, J. SATO, J.H.T. BATES Table 1. Dog no. 1 5 6 Mean so Changes in E and R during effusate loading ERS EL Ew RRs RL Rw 6-5 57 9-6 17-8 66 77 19 61-1 91 15-5 79-5 11-5 7 56-5 0 88-75 - 18-5 1 7 6 60-7 107-9 1 0 19 9 9 Data are expressed as a perentage of baseline values. All hanges are signifiant at p<0.05. ERs, EL and Ew: elastane of the respiratory system, lung and hest wall, respetively; RRs, RL and Rw: resistane of the respiratory system, lung and hest wall respetively. patterns of hange in E and R were similar to those of the other dogs examined. Figure illustrates the hanges in E and R whih ourred 5-10 min following loading of the effusion and inludes the effets of three deep inflations for a representative dog. These data illustrate the immediate, marked dereases in ERs and EL and the small dereases in Ew resulting from deep inflations. These dereases were followed by returns towards original values. RRs and RL also dereased following deep inflations, whilst Rw was little affeted. Again, the resistanes inreased toward original values following the deep inflations. Table presents the perentage hanges in E and R produed by deep inflations in all dogs studied. The lung rather than the hest wall was primarily affeted by deep inflations in all ases..._ ((j Cl....>tt:. (.) re [jj!!! '!' "';- ":"- re Cl....::<: (.) u; a: 0.!!! 0. 0 -'--r-----r---,-----.---...----,-- 5 60 75 90 105 10 o -'--.-- -,---r---.-- -.-----.- 5 60 75 90 105 10 Fig.. - Elastane and resistane for dog no. 5 against time from the initiation of loading of effusate. l : administration of a total lung apaity (TLC) manoeuvre; e: respiratory system data; D: lung data; V : hest wall data. Table. - Changes in E and R with deep inflations Dog no. ERS EL Ew RRS RL Rw 1 9 (1) 5 (1 ) 15 (0.1) 17 (6) 1 (7.5).5 (.6) 6 (1.7) 9 (.1) 1 (l) 6 (.1) 0 (.1 ) 17 () 8 (8.5) 1 (9.9) 1 (.) 6 (6.) 7 (6.) 9 (.5) 8 () 9 (.8) 6 () 50 (.8) 71 (.6) (.7) 5 (1) 9 ( 1.5) (0.58) (.6) 5 (5.) (l) 6 17 (.1) (.5) ( 1.8) 6 (5.5) 1(6.1) 5 (.6) Mean 7 8 0 0 7 so 7 8.7 5.6 15.6 0. 5. Data are expressed as a perentage of end-loading values and are the means of the data points reorded immediately following three deep inflations given between 15 and 10 min post-loading. Numbers in parentheses are the standard deviations of the points. For abbreviations see legend to table l. This apparent abnormality in pressure measurement resolved spontaneously at the end of the reording period and was not observed at any other time. We suspet that this may have been due to the formation of a fluid bubble at the opening to the traheal annula side port. Despite this anomaly in Ptr the Disussion Effusate loading of the pleural spae produed progressive hanges in the dynami E and R of the lung and hest wall in the animals studied (figs 1 and, table 1). Following the ompletion of loading, the
RESPIRATORY MECHANICS WITH PLEURAL EFFUSION ourses of these hanges were stable, apart from transient reversals indued by deep inflations (fig. ). This indiates that the hanges in E and R during fluid loading are not due solely to the arual of ateletasis as a result of regular ventilation over the reording period. These results, therefore, demonstrate that pleural effusion an reate signifiant hanges in the dynami mehanial properties of the respiratory system, and that the extent of the hanges is losely related to the volume of the effusion. EL inreased steadily throughout effusate loading (fig 1). This was probably due to a ombination of parenhyma! distortion, whih ours as the lung rotates around its long and transverse axes during effusate loading [, 11, 1], and to a derease in FRC as dependent lung regions are displaed by the fluid [1]. The derease in FRC an our as a result of airspae losure, or a uniform derease in volume throughout the lungs. For example, several investigators using animal models have reported an inrease in stati EL at very low lung volumes, whilst noting that airspaes remained open, even at negative transpulmonary pressures [1-16]. On the other hand, researh on humans has demonstrated impaired oxygen exhange in the presene of pleural effusion, whih is indiative of some degree of airspae losure [ 17, 18, ]. In reality, there is probably a ombination of the above phenomena, whih ould potentially result in signifiant regional inhomogeneities of ventilation, and so ause an inrease in dynami elastane via the mehanism desribed by Ons et al. [19]. MEAD and CoLLIER [0] desribed a deline in ompliane assoiated with airspae losure in dogs ventilated for several hours without periodi deep inflations. In the present study, we found sudden marked dereases in EL, followed by slower returns toward original values in response to deep inflations (fig. ). Similar results have been desribed for quasi-stati elastane following a period of low volume breathing [15] and for dynami elastane determined from normal FRC or above [1, ]. The dereases in EL may be explained either by stress adaptation of the lung tissues, reruitment of previously losed airspaes, or a ombination of the two phenomena [11]. Although previous work has demonstrated that the lung and hest wall exhibit a similar degree of stress adaptation in response to moderate hanges in lung volume at normal FRC [], we found deep inflations produed muh smaller hanges in Ew than EL (fig. ). This suggests that airspae reruitment may have been responsible for the marked transient dereases in EL that ourred in response to deep inflations in the present study. On the other hand, LuowiG et al. [] and LoRING et al. [1] observed large hanges in EL in response to deep inflations from normal FRC, where pre-existing airspae losure is presumed absent. Therefore, it is still possible that our results are a refletion of a severe volume dependene of stress adaptation of lung tissue. In any ase, our results (fig. ) learly show that the effets of deep inflations on mehanis following fluid loading of the pleural spae are only transitory and that the hanges indued by the loading itself are stable over h. Figure demonstrates a steady inrease in RL as mean lung volume dereases below FRC during effusate loading. This is in ontrast to the work of LuowiG et al. [] who demonstrated a positive dependene between mean lung volume and RL. Previous work has established that at low ventilation frequenies, suh as the one used in this study, tissue and not airway properties are the prinipal determinants of RL [, ]. However, all of these previous measurements were made at mean lung volumes at or above normal FRC. It is possible that airways ontribute signifiantly more to RL at volumes below FRC, thus explaining the ontradition between our work and that of LuowiG et al. []. However, the derease in lung volume indued by the effusion in the present study probably affeted the peripheral more than the entral airways and, thus, would not have signifiantly inreased airway resistane. We also observed a derease in RL in response to deep inflations whih neither LoRING et al. [1] nor LuowiG et al. [], operating at mean lung volumes at or above normal FRC, demonstrated. This supports the notion that the development of airspae losure below normal FRC, temporarily removed by deep inflations, may have ontributed signifiantly to the inrease in RL that we observed with effusate loading. Figure 1 and table 1 show a derease in Ew during effusate loading. KRELL and RooARTE [] also reported dereases in Ew assoiated with an inrease in hest wall volume during fluid loading of the pleural spae. Their findings are onsistent with those of Bamas (personal ommuniation), who demonstrated a derease in Ew with a derease in mean lung volume. Suh evidene supports our opinion that hest wall volume inreased during effusate loading in the present study. It is interesting to note that in dogs, unlike humans, the pleural spae is inomplete and ommuniates bilaterally [5]. Therefore, it is possible some of the effusate ould have rossed the mediastinum, resulting in a bilateral, not unilateral, effusion in the dogs that we studied. The fat that we were able to reover at least 90% of the effusate by evauation via the hest tube, suggests that muh of the effusion remained unilaterally distributed. In any ase, we are interested in the behaviour of the respiratory system with pleural effusion and the exat distribution of the effusion is not pivotal to our findings. In summary, the results of our study suggest that hanges in respiratory system, lung and hest wall volume during effusate loading of the pleural spae alter the dynami E and R of these strutures. It also appears that breathing at mean lung volumes below normal FRC may hange the relative extents to whih airways and tissues ontribute to EL and RL. In partiular, our data suggest that airspae losure may be an important determinant of the hanges in lung mehanis that we observed. Deep inflations, whih may at to dissipate airspae losure, were only able to
G. DECHMAN, J. SATO, J.H.T. BATES transiently reverse the observed hanges in EL and RL. Furthermore, it suggests that a measure, suh as mask ontinuous positive airway pressure (CPAP), whih raises mean lung volume over prolonged periods of time, may be a more effetive treatment for this patient population. Aknowledgements: This work was supported by the Medial Researh Counil of Canada (MRC), the J.T. Costello Memorial Researh Fund and the Respiratory Health Network of Centres of Exellene. G.D. is supported by the Physiotherapy Soiety of the Canadian Lung Assoiation. J.S. was supported by Chiba University Hospital, Chiba, Japan. J.H.T.B. is a Cherheur-Boursier of the Fonds de la Reherhe en Sante du Quebe. Referenes 1. Anthonisen NR, Martin RR. - Regional lung funtion in pleural effusion. Am Rev Respir Dis 1977; 116: 01-07.. Gilmartin JJ, Wright AJ, Gibson GJ. - Effets of pneumothorax or pleural effusion on pulmonary funtion. Thorax 1985; 0: 60-65.. Krell WS, Rodarte JR. - Effets of aute pleural effusion on respiratory system mehanis in dogs. 1 Appl Physiol 1985; 59: 158-16.. Yoo HO, Ting EY. - The effets of pleural effusion on pulmonary funtion. Am Rev Respir Dis 196; 89: 55-6. 5. Baydur A, Behrakis PK, Zin WA, Jaeger M, Mili Emili J. - A simple method for assessing the validity of the esophageal balloon tehnique. Am Rev Respir Dis 198; 16: 788-791. 6. Dehman G, Sato J, Bates JHT. - Pleural pressure measurement in dogs using esophageal balloon: effet of posture, lung volume, balloon position and paralysis. 1 Appl Physiol 199; 7: 8-88. 7. Estenne M, Yernault JC, De Troyer A. - Mehanism of relief of dyspnea after thoraoentesis in patients with large pleural effusions. Am J Med 198; 7: 81-819. 8. Wohl ME, Turner BJ, Mead J. - Stati volume pressure urves of dog lungs in vivo and in vitro. 1 Appl Physiol 1968; : 8-5. 9. Bates JHT, Shardonofsky F, Stewart DE. The low-frequeny dependene of respiratory system resistane and elastane in normal dogs. Respir Physiol 1989; 78: 69-8. IO. Bates JHT, Lauzon AM. - A nonstatistial approah to estimating onfidene intervals about model parameters: appliation to respiratory mehanis. IEEE Trans Biomed Eng 199; 9: 9-100. 11. Hoppin FG Jr, Lee GC, Dawson SV. - Properties of lung parenhyma in distortion. J Appl Physiol 1915; 9: 7-751. 1. Vawtere DL, Fung YC, West FB. Elastiity of exised dog lung parenhyma. J Appl Physiol: Respirat Environ Exerise Physiol 1978; 5: 61-69. 1. Cavagna GA, Stemmler EJ, DuBois AB. - Alveolar resistane to ateletasis. J Appl Physio/ 1967; : 1-5. 1. Forrest JB. - The effet of hyperventilation on the size and shape of alveoli. Br 1 Anaesth 1910; : 810-817. 15. Williams JV, Tierney DF, Parker HR. Surfae fores in the lung, ateletasis, and transpulmonary pressure. J Appl Physiol 1966; 1 : 819-87. 16. Young SL, Tierney DF, Clements JA. - Mehanism of ompliane hange in exised rat lungs at low transpulmonary pressure. J Appl Physiol 1970; 9: 780-785. 17. Chang SC, Shiao GM, Perng RP. - Postural effet on gas exhange in patients with unilateral pleural effusions. Chest 1989; 96: 60-6. 18. Sonnenblik M, Melzer E, Rosin AJ. - Body positional effet on gas exhange in unilateral pleural effusion. Chest 198; 8: 78-786. 19. Otis AB, MKerrow CB, Bartlett RA, et al. - Mehanial properties of lungs and hest wall during spontaneous breathing. J Appl Physiol 1956; 8: 7-. 0. Mead J, Collier C. - Relation of volume history of lungs to respiratory mehanis in anesthetized dogs. 1 Appl Physiol 1959; 1: 669-678. 1. Loring SH, Ingram RH Jr, Drazen JM. - Effets of lung inflation on airway and tissue responses to aerosol histamine. J Appl Physio/: Respirat Environ Exerise Physiol 1981; 51: 806-811.. Ludwig MS, Dreshaj I, Solway J, Munoz A, Ingram RH Jr. - Partitioning of pulmonary resistane during onstrition in the dog: effets of volume history. 1 Appl Physiol 1987; 6: 807-815.. Bates JHT, Brown KA, Kohi T. - Respiratory mehanis in the normal dog determined by expiratory flow interruption. J Appl Physiol 1989; 67: 76-85.. Bahofen H. - Lung tissue resistane and pulmonary hysteresis. J Appl Physiol 1968; : 96-01. 5. Hare WCD. - Respiratory System. In: Sisson and Grissmans. The anatomy of the domesti animals. Vol. 1, Philadelphia, WD Saunders Co., 1975, Chp 8, p. 11.