MSU RETURNING MATERIALS: P1ace in book drop to remove this checkout from LIBRARIES —:-a--. your record. FINES wfl] be charged if book is returned after the date stamped be10w. £7" :4; ”*9 '-' 3 an?" ‘45:; 3' if” g,; a? 2‘? a; THE EFFECT OF INHOMOGENEOUS INFLATION OF A SUBLOBAR LUNG SEGMENT 0N COLLATERAL CHANNEL RESISTANCE By Steven Douglas Fuller AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1984 ABSTRACT THE EFFECT OF INHOMOGENEOUS INFLATION OF A SUBLOBAR LUNG SEGMENT ON COLLATERAL CHANNEL RESISTANCE By Steven Douglas Fuller Inhomogeneous inflation of a sublobar lung segment occurs when the segment is inflated to a different distending pressure than the remainder of the lobe. The effect of inhomogeneity on collateral channel resistance (Rcoll) was studied in left cranial, left caudal, and right caudal lobes of excised dogs' lungs. A double lumen catheter was advanced through the trachea and wedged in a small bronchus supply— ing a sublobar segment. Helium (He), air, or sulfurhexafluoride (SF6) flowed into the outer lumen of the catheter (Vcoll) while segment pressure (Pct) was measured at the tip of the inner lumen. Trans- pulmonary pressure (Pao) was measured either at the trachea or lobar bronchus as the lobe was inflated with air. At constant Pao, raising Rcoll inflated the segment by increasing Pct-Pao, thereby creating inhomogeneity. Collateral channel resistance was calculated as Rcoll = (Pct-Pao)/Vcoll. Lobar inflation (i.e., raising Pao at constant Pct-Pao) decreased Rcoll. In contrast, segment inflation (i.e., raising Pct-Pao at constant Pao) increased Rcoll. Similar results occurred in the lungs of closed chest anesthetized dogs. In excised lungs, the increase in Rcoll during segment inflation was accentuated when the segment gas flow regime was turbulent and Steven Douglas Fuller eliminated when the flow regime was laminar. The fact that Rcoll failed to decrease during segment inflation, even when flow was laminar, sug- gests that segment inflation and lobar inflation may have different effects on segment airway geometry. In additional studies designed to determine the arrangement of airways in the segment-lobar interface, corrosion casts of excised dogs' lungs showed that the interface con- tains bronchi coursing within and parallel to the interface and pro- viding branches to both the segment and lobe. Also, examination of casts of parenchymal tissue from the interface revealed interdigitation of segment and lobar acini. This airway arrangement suggests three possible routes followed by gas flowing out of an obstructed segment, and each route is likely influenced by tissue distortion at the segment-lobar interface occurring during measurements of Rcoll. Further studies in excised dogs' lungs suggests that most of the gas leaving a segment enters the bronchi in the interface and flows directly out of the lobe without entering the remainder of the lobe. ACKNOWLEDGMENTS A most significant and special person in my life during the past eight years has been Dr. Norman Edward Robinson. As my major advisor, he has successfully guided me through M.S. and Ph.D. degrees and has thus played a critical role in my formative academic years. He uniquely blends discipline, intellect, and compassion which allows him to consistently demonstrate academic and research excellence as well as kindness and understanding. He demands high performance but is always supportive and caring during times of frustration and dis- appointment. It is my hope to emulate these and his many other posi- tive qualities, and I look forward to an enduring friendship and frequent reunions. I also express my gratitude to Dr. S. Richard Heisey whose thorough critique of this dissertation resulted in its significant improvement. We have also shared a friendship these past eight years which has been very meaningful to me, and I'll miss our long conver- sations and his frequent attempts at humor. To this day, I still feel just a little guilty for what I did to him on the racquetball court! The other members of my guidance committee--Dr. Jack Hoffert, Dr. John Chimosky, and Dr. Robert Echt--have also had important duties in guiding my academic progress and critiquing this dissertation. To them, I extend my sincere appreciation, and I wish them well in all future endeavors. ii Last, but not at all least, is Roberta Milar who has aided me, directly and indirectly, in every experiment I have ever performed during my graduate school career. For me she has been technician extraordinaire, graphic artist, occasional taxi, and always an understanding and loyal friend. I will miss her greatly. iii TABLE OF CONTENTS LIST OF TABLES ......................... LIST OF FIGURES ........................ LIST OF ABBREVIATIONS ..................... LIST OF DEFINITIONS ...................... CHAPTER I. GENERAL INTRODUCTION AND LITERATURE REVIEW ....... General Introduction . . . . . . . .......... Introduction to Literature Review .......... Anatomy of Collateral Pathways ............ Interalveolar Pores ............... Bronchiolealveolar Canals ............ Interbronchiolar Respiratory Bronchioles ..... Interacinar Ducts ................ Methods Used to Study Collateral Ventilation ..... Pressure-Flow Relationships ............. Basic Concepts of Fluid Mechanics ........ Pressure-Flow Relationships in a Branched System ..................... Problems of analysis ............. Rohrer's equation .............. Modification of Rohrer's equations ...... Moody diagram ................ Pressure-Flow Relationships in a Sublobar Lung Segment .................. Segment-Lobar Inhomogeneity ............. Purpose of the Present Studies ............ List of References .................. II. THE EFFECT OF REGIONAL INHOMOGENEITY ON COLLATERAL AIRWAY RESISTANCE .................. Introduction ..................... Methods ....................... Results ....................... iv Page vi vii CHAPTER Page Discussion ....................... 73 List of References ................... 81 III. MECHANISM OF INCREASED COLLATERAL AIRWAY RESISTANCE DURING NONHOMOGENEOUS SEGMENT INFLATION ........ 83 Introduction ...................... 83 Methods ........................ 84 Results ........................ 91 Discussion ....................... 99 List of References ................... 116 IV. PATHWAYS CONNECTING OBSTRUCTED AND NONOBSTRUCTED SUBLOBAR REGIONS IN THE DOG LUNG ................ 118 Introduction ...................... 118 Methods ........................ 119 Series One ..................... 119 Series Two . . . . . . . . . . ........... 120 Series Three: Distribution of Gas Leaving an Obstructed Segment ................ 121 Results ........................ 125 Series One ..................... 125 Series Two ..................... 128 Series Three: Distribution of Gas Leaving an Obstructed Segment ................ 133 Discussion ....................... 139 List of References ................... 145 V. DISCUSSION AND SUMMARY .................. 146 List of References ................... 151 VI. CONCLUSIONS ....................... 152 APPENDICES A. MEAN (i SEM) VALUES OF COLLATERAL RESISTANCE FOR CHAPTER II DATA .................... 155 B. RELATIONSHIP OF SEGMENT AIRWAY DIAMETER TO LOBAR VOLUME DURING LOBAR INFLATION FOR CHAPTER III DATA . . . 158 Table 1-1. 2-1. 2-2. 3-2. A-1. B-1. LIST OF TABLES Synopsis of literature describing anatomy of collateral channels ........................ Protocols for series one, series two, and series three experiments ....................... Values of Reynolds' number (Re) for each series of experiments ....................... Physical properties of ases [helium (He), air, and sulfurhexafluoride (SF5TJ used in calculation of collateral resistance .................. Intrasegmental airway resistance (Rs) expressed as a fraction of collateral resistance (Rcoll) (Rs/Rcoll, Y': SEM) as the segment was inflated to a segment- lobar pressure difference (Pct-Pao) = 1-7 cm H20 at transpulmonary pressure (Pao) = 2-6 cm H20 in the left caudal lobes of excised dogs' lungs ........ Data points used in Figures 2-1, 2-2, and 2-3 in Chapter II showing collateral resistance [Rcoll; cm HZO/(mllsec); 2': SEM] in left cranial and right caudal lobes in series one (excised lungs with the segment inflated relative to the lobe), series two (lungs of closed chest anesthetized dogs), and series three (excised lungs with the segment deflated relative to the lobe) ............. Relationship of segment airway diameter to lobar volume when the curves in each panel of Figure 3-5 are forced to approximate a single curve ........ vi Page 16 58 75 89 96 Figure 1-1. 1-2. 1-3. 1-5. 1-6. 2-1A. LIST OF FIGURES Page Photograph of excised lung showing inhomogeneously inflated sublobar segment (i.e., a segment which is inflated to a relatively greater volume than the remaining lobe) at extreme left ............. 3 Scanning electron micrograph of a portion of the segment-lobar interface from an excised left caudal lobe of dog lung ..................... 5 Schematic diagram illustrating three different types of collateral channels (interbronchiolar respiratory bronchiole, bronchiole-alveolar canal, and inter- alveolar pore) connecting two adjacent sublobar lung segments ...................... ll Schematic diagram illustrating Hilpert's technique (22) for calculating collateral channel resistance (Rcoll) ......................... 20 Pressure-flow characteristics in a straight smooth circular tube ...................... 27 Schematic illustration of Moody diagram where the logarithm of the normalized pressure drop (Log Pn; ordinate) is plotted against the logarithm of Reynold's number (Log Re; abscissa) ................ 38 Effect of transpulmonary pressure (Pao; cm H20); abscissa) on collateral resistance [Rcoll; cm HZO/ (ml/sec); ordinate] in left cranial lobes (left panel) and right caudal lobes (right panel) at different levels of segment pressure (Pct; cm H20) in five excised dogs' lungs ventilated with air ......... 61 . Same data as in Figure 2-1A showing effect of segment- lobar pressure difference (Pct-Pao; cm H O; abscissa) on collateral resistance [Rcoll; cm H20/Tml/sec); ordinate] at different values of transpulmonary pressure (pao; cm H20) in five excised dogs' lungs. . . . 63 vii Figure 2-2A. 2-28. 2-3A. 2-38. 3-1. 3-2. Effect of transpulmonary pressure (Ptp; cm H20; abscissa) on collateral resistance [Rcoll; cm H20/ (ml/sec); ordinate] in seven left cranial lobes (left panel) and six right caudal lobes (right panel) at different levels of segment pressure (Pct; cm H20) in lungs from closed chest dogs ventilated with air . . . Same data as Figure 2-2A showing effect of segment— 1obar pressure difference (Pct-Pao; cm H O; abscissa) on collateral resistance [Rcoll; cm HZO/Tml/sec); ordinate] at different values of transpulmonary pressure (Ptp; cm H20) in lungs from closed chest dogs .......................... Effect of transpulmonary pressure (Pao; cm H20; abscissa) on collateral resistance [Rcoll; cm HZO/ (ml/sec); ordinate] in five left cranial lobes (left panel) and six right caudal lobes (right panel) at different levels of segment pressure (Pct; cm H20) when the segment was deflated relative to the lobe in excised dogs' lungs ventilated with air ......... Same data as in Figure 2-3A showing effect of segment- lobar pressure difference (Pao-Pct; cm H O; abscissa) on collateral resistance [Rcoll; cm HZO/Tml/sec); ordinate] at different values of transpulmonary pressure (Pao; cm H O) in excised dogs' lungs when the segment was def ated relative to the lobe ...... Illustration of left caudal lobe from an excised dog lung showing double lumen catheter wedged in a brochus supplying a sublobar segment .............. Effect of segment inflation by raising the segment- lobar pressure gradient (Pct-Pao; cm H20; abscissa) on total collateral resistance [Rcoll; cm HZO/(ml/sec); ordinate], top panel; intrasegmental airway resistance [Rs; cm HZO/(ml/sec); ordinate], middle panel, and intersegmental airway resistance [Ri; cm HZO/(ml/sec), ordinate], bottom panel, at five transpulmonary pressures (Pao = 2-6 cm H20) .............. Same data as Figure 3-2 showing the effect of trans- pulmonary pressure (Pao; cm H20; abscissa) on total collateral resistance [Rcoll; cm HZO/(ml/sec), ordinate], top panel; intrasegmental airway resistance [Rs; cm HZO/(ml/sec), ordinate], middle panel; and viii Page 66 68 7O 72 87 93 Figure 3-4. 3-5. 3-6. 3-7. 3-8. 4-1. 4-3A. intersegmental airway resistance [Ri; cm HZO/ (ml/sec), ordinate], bottom panel; as the segment- lobar pressure gradient was held constant at seven values (Pct-Pao = 1-7 cm H20) .............. The logarithm of the normalized pressure drop (Log Pn; ordinate) is shown as a function of the logarithm of Reynolds' number (Log Re; abscissa) and is referred to as a Moody diagram (Figure l-6) ........... The logarithm of the normalized pressure drop (Log Pn; ordinate) is shown as a function of the logarithm of Reynolds' number (Log Re; abscissa) and is referred to as a Moody diagram (Figure 1-6) ........... Effect of scaling segment airway diameter to cube root of lobar volume on data in the Moody diagrams illustrated in Figure 3-5 ................ Effect of scaling segment airway diameter to transpulmonary pressure (Pao) on data in the Moody diagrams illustrated in Figure 3-5 ........... Comparison of the effect of raising lobar volume [plotted as percent vital capacity (VC); abscissa] on the following five resistances (cm HZO/LPS, where LPS = liters per second): homogeneous collateral resistance [Rcoll (homogeneous) obtained by extrap- olating the curves at each Pao in the center panel of Figure 3-2 to the ordinate]; total lung resistance (RL) and central (Rc) and peripheral (Rp) airway resistance in vagally intact dogs as reported by Macklem and Mead (5); and R in vagotomized dogs as reported by Macklem et aI. Schematic diagram of excised left caudal lobe suspended by lobar bronchus in an airtight box ..... Corrosion cast of individual bronchopulmonary segment (Figure 4-2A; red) and subsegmental segment (Figure 4-28; white) .............. Corrosion cast of three adjacent subsegmental segments ........................ ix (6) ............ Page 95 98 109 115 Figure 4-38. 4-4A. 4-4B. 4-4C. 4-4D. 4-5. 4-6. B-1. 8-2. The large arrow points to a large airway, which I term "interface" airway, coursing within and parallel to the red—blue interface and which provides branches into both red and blue segments ........ Corrosion cast of three adjacent bronchopulmonary segments ........................ At the extreme left a blue nonrespiratory bronchiole branches into a blue respiratory bronchiole which further divides into red and blue respiratory bronchioles ....................... A portion of the interface from another corrosion cast .......................... A different view of the same specimen from Figure 4-4C ....................... Distribution of outflow of gas leaving an obstructed bronchopulmonary segment (as a percent of total inflow volume, i’with SEM bars) as a function of inflow segment, inflow rate (low = l liter/minute, high = 2 liters/minute), and transpulmonary pressure (Ptp; cm H20) in six excised left caudal lobes ..... Model illustrating possible course of resin or gas flow through an obstructed segment and interface airway ......................... Attempt to forcibly approximate the curves in each panel of Figure 3-5 into a single smooth curve ..... Attempt to forcibly approximate the curves in each panel of Figure 3-5 into a single smooth curve ..... Page 132 132 135 He Pao Pct Pn Ppl Ps Ptp Rcoll Re Ri Rs SF Vcoll LIST OF ABBREVIATIONS Helium. Gas density (gm/ml). Pressure (cm H20) at airway opening of the trachea or lobar bronchus, equivalent to transpulmonary pressure in the excised lungs of Chapters II and III. Pressure (cm H20) in an obstructed segment measured at the tip of the wedged catheter or bronchoscope. Normalized pressure drop; i.e., the ratio of the static pressure drop to dynamic pressure. Intrapleural pressure (cm H20). Pressure (cm H20) in the segment subpleural alveoli in excised lungs. Transpulmonary pressure (cm H O) in the lungs of closed chest anesthetized dogs (Chapter II). In Chapter IV, in which measurements were taken in excised lungs suspended in an airtight box, Ptp represents the difference between box pressure and atmospheric pressure. Collateral channel resistance [cm HZO/(ml/sec)] (equation l-l). Reynolds' number (equation l-ll). Airflow resistance [cm H O/ (ml/sec)] in airways between the segment subpleural alveoIi and the airway opening. Airflow resistance [cm H 20/(ml/sec)] in airways between the tip of the wedged catheter and the segment subpleural alveoli. Sulfurhexafluoride. Gas viscosity [gm/(sec- cm)]. Flow (ml/sec) of gas through the wedged catheter and into the obstructed segment. xi LIST OF DEFINITIONS Acinus All the respiratory bronchioles, alveolar ducts, and alveoli distal to a terminal bronchiole. Airways Refers to the series of bronchi and bronchioles leading from the trachea to the alveolar ducts. CbZZateraZ Channel A pathway connecting acini originating from two different airways. Flow Regime The predominant pattern of flow in a system (i.e., whether flow is laminar, transitional, or turbulent). Laminar Flow Fluid flow having the following characteristics: (1) flow in all parts of the stream is constant and occurs in parallel laminae which slide over one another; (2) the profile of the leading edge of flow is parabolic, the axial velocity of flow being twice as fast as the average velocity. Transitional Flow Fluid flow having characteristics of both laminar and turbulent flow. Turbulent Flow Fluid flow having the following characteristics: (1) the velocity of flow at any point in the stream fluctuates vigorously and randomly in both magnitude and direction; (2) the profile of the leading edge of flow is only slightly curved, the axial velocity of flow being 1.2 times the average velocity. xii CHAPTER I GENERAL INTRODUCTION AND LITERATURE REVIEW General Introduction Inhomogeneous inflation of a sublobar lung segment occurs when the segment is inflated to a relatively different distending pressure than the remainder of the lobe. Inhomogeneity may be created by flowing gas into a sublobar segment through a catheter obstructing the bronchus supplying the segment (Figure 1-1). The incoming gas inflates the segment to a relatively greater volume than the lobe, then exits the segment via collateral channels to flow out of the lobe. Inhomogeneity creates an area of tissue distortion at the segment-lobar parenchymal interface. This is demonstrated in Figure 1-2 which shows a scanning electron micrograph of a section of the interface from an inhomogeneously inflated air-dried left caudal lobe of dog lung. Parenchyma from the relatively inflated segment is on the right of the figure, and parenchyma from the relatively deflated lobe is on the left. The interface is observed as a thin line of distorted, compressed parenchymal tissue in the center of the figure. Collateral channels serving as connections between the segment and the lobe may also be distorted during inhomogeneity, for they not only pass through the zone of parenchymal tissue distortion at the interface but they also are subjected to abrupt changes in distending pressure as they pass from the segment to the lobe. Distortion of Figure 1—1. Photograph of excised lung showing inhomogeneously inflated sublobar segment (i.e., a segment which is inflated to a relatively greater volume than the remaining lobe) at extreme left. Figure 1-1 .5: o.ooop mcwucmmwgamg mpnum a m? Amgammm mo Eouuonv Ema uppom ugh .udmp mg» co m? mama?» canop .ugmwg use no my mama?» acwEmmm .Awnop wcu cog» waspo> Lmummgu apm>wumpmg m o» amen—Cc? mm: acmsmmm mg» ..m.wv maop as» cmzawz xpmaomcmmoeoscr umumpmcw no: newsmmm mg» mm coagulgpm mm: mnop as» .mcsp new we wnop panama ummp cmmwuxm cm soc» mummcmucw canopuucwsmmm mg» mo covugoa a we samgmogows coguumpm mcwccmum .~-F acsmma a" f‘ v’ .- [— 35h Figure 1-2 collateral channels should affect collateral airflow resistance (Rcoll), but there are few reports describing the effect of inhomogeneity is reported to increase Rcoll, decrease Rcoll, or have no effect on Rcoll. In addition, these studies used several methods to measure Rcoll in three animal species, and measurements were made over a restricted range of inhomogeneities. Thus, these studies are difficult to compare. The major purpose of this dissertation is to provide a systematic examination of the effect of inhomogeneity on Rcoll. , Chapter II documents the effect, and two hypotheses are proposed to explain the mechanism. Chapter III reports testing of the hypotheses. In Chapter IV, the arrangement of the airways in the segment-lobar interface is determined, so that the function of the airways discussed in Chapters II and III may be correlated with anatomy. Chapter I describes the known anatomy of the collateral channels, discusses the various methods used to measure Rcoll, and provides background information needed to appreciate the hypotheses proposed in Chapter II and tested in Chapter III. Introduction to Literature Review The tracheobronchial tree is an irregularly dichotomous branching system of airways, beginning at the trachea, branching into the bronchi, bronchioles, and alveolar ducts, and terminating in the alveolar sacs and alveoli. Prior to 1930 these airways were considered to be a series of dead-end tubes because there were no known inter- connections. This model of the lung explained the common clinical observation that obstruction of a lobar bronchus produced atelectasis, because the gas trapped inside the segment was absorbed by the blood. In 1930, however, using this concept of the lung, VanAllen, Lindskog, and colleagues (78, 80, 82) were unable to explain why atelectasis failed to occur in the dog lung in which a segmental bronchus had been experimentally obstructed. In addition, they were able to flow India ink through the airways of the obstructed segment and into the lobe. They reasoned that the airways, rather than being isolated from one another, were interconnected by collateral channels through which air could enter an obstructed segment from the surrounding lobe and prevent segmental collapse. They termed this phenomenon "collateral respiration" and concluded that the collateral channels must be interalveolar pores since these were the only known potential segment-lobar connections. During the next 35 years these investigators studied collateral respiration (more recently referred to as collateral "ventilation") in a series of simple and direct experiments using the lungs of a variety of species (2, 7, 38-40, 77-84). They cannulated two portions of an excised lung lobe from humans, dogs, cats, and rabbits: one cannula was ligated in a segmental bronchus, and a second cannula was ligated in the lobar bronchus leading to the remaining segments. When air was gently blown into either of the cannulas, it escaped from the other (81, 82). In excised rabbit lungs, air could be passed between can- nulas ligated in the lobar bronchi of adjacent lobes, indicating incomplete separation of the lobes (83). In the lungs of closed- chested anesthetized dogs, segmental bronchi were obstructed by a variety of means. At autopsy performed three hours to one-and-one-half months following the obstruction, none of the obstructed segments were collapsed, indicating they had been ventilated collaterally (80). Collateral ventilation was not found in all species investi- gated, for when a cannula obstructed an airway in calf and pig lungs, air flowing into the cannula only inflated the obstructed segment (83). To explain their results, VanAllen and Lindskog observed that the lobes of pig and calf lungs are divided into many small lobules by connective tissue septae [which extend from the visceral pleura down to the airways (67)]. They concluded that collateral channels do not penetrate the septae, thus preventing collateral ventilation. In contrast, the septae are incomplete in human lungs and are totally absent in dog and cat lungs, accounting for the presence of collateral ventilation in these species. VanAllen, Lindskog, and colleagues also demonstrated the functional importance of collateral ventilation. They sampled gas from an obstructed lung segment of a healthy anesthetized dog and found that the 02 and CO2 tensions closely correlated with those of arterial blood (39), demonstrating that segmental bronchial obstruction does not impair the gas exchange function of the segment. They showed that the mechanical behavior of collateral channels and airways is similar because the pressure required for gas flow in the channels varies inversely with lung volume (82) as does airway resistance (46). They demonstrated that the volume of an obstructed segment's collaterally respired gas is inversely proportional to respiratory frequency (7), and that an atelectatic segment can be inflated through collateral channels (84). In addition, pulmonary artery ligation or the injection of histamine or curare into the pulmonary artery decreases collateral ventilation (2, 7) while serotonin has no effect (7), indicating that collateral channels must have smooth muscle capable of reacting with vasoactive compounds. These experiments provided the framework for future investigators of collateral ventilation and were remarkable in that the majority of the subsequent research, although more sophisticated and detailed, is primarily a confirmation of VanAllen and Lindskog's original findings. Anatomy of Collateral Pathways Collateral channels connect acini originating from different parent airways. The majority of research on collateral channels centers on their function rather than anatomy, and the few investigators who have attempted anatomical studies have succeeded in describing channels which are only theoretically or indirectly linked with a collateral ventilatory function. Interalveolar Pores Interalveolar pores (Figure 1-3) are round or oval openings in alveolar walls. Pore diameter is 2-10 u (47) although the exact dimension depends on the lung fixation technique. The number of pores per alveolus may be as high as 50 (44), and they have been found in every mammalian species investigated including the bat, dog, rabbit, pig, cat, human, monkey, rat, guinea pig, orangutan, chimpanzee, 1O .mucosmwm asap gmnopnsm ucmumnua oz» mcwuumccou Amgoa gmpom>pmsmucp can .chmu Empom>—m-mpow;ucogn .wpowgucogo xgoumgwammg gapopgocogngwucwv mpmccmgu Fmgmumppoo mo mama» acmgmemwu owes» mcvumsumappw Emgmmwu ovumswsum .m-F aczmwa 11 m-F mc=mwu m<._o__.._ozommmw._.z_ $483455. 1 l «.3834 q -ufizozomm / 12 baboon, mouse, mole, hedgehog, horse, ox, sheep, goat, opossum, and manatee (43, 44). Interalveolar pores were first described in 1847 (l) and were well characterized anatomically by the time of the original investiga- tions of collateral ventilation by VanAllen, Lindskog, and colleagues (2, 7, 38-40, 77-84) nearly 100 years later. These investigators reported that obstruction of sublobar airways in humans, dogs, cats, and rabbits did not produce atelectasis, so they reasoned that the obstructed segment must be ventilated by the surrounding normal lung tissue through some type of collateral channel. Since the interalveolar pore was the only known anatomical structure which could theoretically provide such a connection, the pore was assigned the function of a collateral channel. Later reports suggest that pores offer much too high a resistance to participate in collateral ventilation. Martin (48) calculated that the opening pressure of a collateral pore would be 192 cm H20, and Sasaki et al. (73) reported that airflow resistance through pores may exceed 4000 cm H20/1/sec. Both investigators con- clude that larger pathways must participate in collateral ventilation. Bronchiolealveolar Canals Bronchiolealveolar canals (BAC) (Figure 1-3), first described by Lambert (34) and later by others (5, 13, 19, 33, 65), are connections between a bronchiole (nonterminal, terminal, or respiratory) and the immediately surrounding alveoli. They are either a single interruption in a bronchiolar wall which opens into an adjacent alveolar sac or a tubule connecting the bronchiole with the sac. The BAC is 30 u in 13 diameter and has been found in humans of all ages, cats, rabbits, rats, and sheep. They are more numerous in the lungs of patients with emphysema and in certain cystic lung diseases, such as honeycomb lungs (37). The BAC cannot be ascribed a route for collateral flow merely because of its presence within the lung. Indeed, the following argu- ments suggest only a minor role of the BAC in collateral ventilation: (l) the BAC connects alveoli, via recurrent airways, with their parent bronchiole. Thus, there is no evidence demonstrating that the BAC connects acini having different parent airways. (2) Collateral ventilation still occurs when obstruction is found in airways much more proximal than the bronchiole (i.e., bronchi), and this would not occur if the BAC only connects alveoli with their parent bronchiole. The BAC is therefore not required for collateral ventilation. Interbronchiolar Respiratory Bronchioles An interbronchiolar respiratory bronchiole (Figure 1-3) is a respiratory bronchiole which connects two terminal bronchioles. In 1966 Martin (48) cannulated the first two bronchial branches in left upper lobes of excised dogs' lungs. In one series of experiments a suspension of polystyrene spheres having diameters from 60 to 710 u were made to flow through one cannula and were collected from the other cannula. Examination of the outflow revealed that the largest sphere which passed between the two obstructed segments was 120 u in diameter. In a second series of experiments he caused aerosolized India ink to flow into the lobe through one of the cannulas. After fixing the lobe 14 in the inflated state with formaldehyde, he dissected out the segment-lobar parenchymal interface, embedded it in paraffin, and made serial sections. He found India ink deposited on a respiratory bronchiole which connected two terminal bronchioles each lying on opposite sides of the interface. Martin concluded that respiratory bronchioles are the pathways for collateral ventilation. He also concluded that alveolar pores are not collateral channels since India ink was rarely deposited within alveoli in the segment-lobar interface. Martin's conclusions may not be justified, however, because his serial sections simply identify a respiratory bronchiole passing from one bronchiole towards another. At no point does he demonstrate the respiratory bronchiole opening into the two bronchioles which it supposedly connects, and yet this is the crucial evidence needed to consider the respiratory bronchiole as a collateral pathway. Interacinar Ducts In 1968 Henderson et a1. (21) cannulated two segmental bronchi in disease free post mortem human lung lobes. They used a similar perfusion pressure as Martin (48) to flow a suspension of polystyrene spheres into one cannula and collected the effluent from the other, and they found that 90 percent of the spheres had a diameter of less than 64 p. In addition, they made resin casts of the lungs by pouring a different colored resin down each cannula. After the resin cured and tissue was dissolved away, the cast was dissected along the segment-lobar interface, revealing collateral airway connections between alveolar ducts originating from two acini supplied by different 15 airways. Boyden (6) described a similar interacinar pathway in a plastic reconstruction of an acinus from a child. In 1975 Raskin and Herman (66) used a micropuncture technique to inject a radio-opaque silicone rubber into subpleural acini of inflated and fixed human lungs. The flow and distribution of the rubber was studied by cinematography. A frame-by—frame analysis of the motion pictures revealed ducts having a diameter approximately 200 u connecting acini separated by interlobular septae. The much larger size of these ducts in comparison to those described by Henderson et al. (21) may be related to the state of lung inflation: Raskin and Herman fixed the lungs at 25 cm H20, whereas Henderson et al. perfused deflated lungs with a driving pressure of 5 cm H20. Table l-l is a synopsis of the anatomical literature described above. Methods Used to Study Collateral Ventilation Collateral ventilation was first assessed by measuring the pressure required to initiate collateral flow in excised dog lungs. VanAllen et al. (82) placed a freshly excised lung lobe in an air— tight chamber, cannulated the lobar and one segmental bronchus with separate cannulas, and connected the cannulas to the chamber's exterior with tubes. The lobe was inflated by reducing chamber pressure, follow- ing which the tip of one tube was submerged in water while air was blown into the other tube. When pressure in the inflow tube equalled one cm H20, the initiation of collateral airflow was indicated by bubbling in the water at the tip of the submerged tube. When the lobe was 16 Table 1-1. Synopsis of literature describing anatomy of collateral channels Ti Speciesa Collateral Channel Human Dog Others Reference Datesb Interalveolar Yes Yes All pore mammals 57a 1847 Bronchioloalveolar Yes No cats, rats, 5, l3, l9 canal rabbits, 33, 34, 37 sheep 65 1953 Interbronchiolar No Yes No 48 respiratory bronchiole 1966 Interacinar Yes No No 6, 21, 66 1968 duct aMost research on collateral channels uses the lungs of humans and dogs. separately from other mammalian species. b Dates of first investigation. Therefore, for ease of comparison, these species are listed 17 completely deflated, six cm H20 was required to initiate collateral flow and two cm H20 to maintain it. Thus, the pressure required to initiate collateral flow was inversely proportional to lung volume. A similar method was used to measure collateral gas flow in the lungs of closed chested anesthetized dogs. A cannula was passed through the trachea and into a segmental bronchus of the right lower lobe where the cannula tip was dilated to form a leak-proof seal (76, 82). The proximal end of the cannula was attached to glass tubing whose tip was submerged in water. Each time the spontaneously breathing dog exhaled, gas escaped the obstructed segment through the cannula, causing bubbling at the submerged tip of glass tubing. In other experiments, exhaled gas from the segment was collected in a Krogh spirometer attached to the proximal port of the cannula, and the volume of gas collected was directly proportional to lung volume (39). VanAllen and colleagues used these methods to study collateral ventilation in the excised and intact lungs of a variety of species, and they determined how collateral channels respond to changes in lung volume (39), histamine and serotonin (7), isoproterenol and C02 (31), rate and depth of respiration, pulmonary artery ligation, pulmonary embolism, and pulmonary venous ligation (7). These early experiments were landmarks in the history of collateral ventilation because they not only provided much of the current knowledge of the function and behavior of the collateral pathways, but they also provided the basic method used in virtually all subsequent research: namely, obstructing an airway with a catheter or bronchoscope which was used to flow gas into or collect gas from the obstructed segment. 18 More recently, airflow resistance through collateral channels (Rcoll) was measured by Hilpert (22) who used a variation of the methods of VanAllen and colleagues (Figure 1-4). In Hilpert's technique, a double lumen catheter is inserted into the trachea and advanced until it becomes wedged in a small bronchus supplying a sublobar segment. A constant flow of gas (Vcoll), injected into the segment through the outer lumen of the catheter, inflates the segment and exits the segment via collateral channels. Segment pressure (Pct) is measured by the inner lumen of the catheter while transpulmonary pressure (Pao) is measured at the trachea. Collateral channel resistance is calculated as: Rcoll =-EgELLE§9- (1'1) Vcoll This method has been used to confirm many of the original findings of VanAllen and colleagues as well as measure vagal influences on Rcoll (S9) and the effect of ozone on Rcoll (20). Also, segment gas may be withdrawn through the catheter, allowing analysis of segment gas exchange (16, 17, 55). In addition, the resistance through collateral channels could be compared to airway resistance, for the route having the lower resistance is the one through which an unobstructed segment receives the majority of its tidal volume. Subsequent reports have shown that Rcoll in excised dogs' lungs is greater than airway resis- tance (52, 54), demonstrating that a segment in the healthy dog is likely ventilated primarily by airways. Figure 1-4. 19 Schematic diagram illustrating Hilpert's technique (22) for calculating collateral channel resistance (Rcoll). A double lumen catheter is inserted into the trachea and advanced until it becomes wedged in a small bronchus supplying a sublobar segment. A constant flow of gas (Vcoll; ml/sec), made to flow into the outer lumen of the catheter, inflates the segment and exits the segment via collateral channels. Segment pressure (Pct; cm H20) is measured by the inner lumen of the catheter while transpulmonary pressure (Pao; cm H20) is measured at the trachea. Collateral resistance is calculated as Rcoll = (Pct-Pao)/Vcoll. 20 vCOLL .2 E (i x! S. 3} Pct Rco" : M Vcou. Figure 1-4 21 Several investigators have attempted to partition Rcoll into various components. Hogg et a1. (23) cannulated the trachea of excised human lungs. A double lumen catheter was inserted through a side-arm in the tracheal cannula and ligated in the lower lobe bronchus beyond the first branch to the superior segment. Air which flowed into the bronchial catheter entered the basal segments and flowed through col- lateral pathways into the superior segment and out of the lobe. The inner lumen of the catheter measured pressure at the catheter tip. Alveolar pressure in the basal and superior segments was obtained by inserting a catheter through the pleural surface on either side of the superior-basal segment interface and ligating the two catheters to the visceral pleura. The following resistances were calculated: _ Pct- Palv (base) R - . (1-2) base Vcoll Rcoll = Palv(base)- Palv (sup) (1_3) Vcoll Rsup = Palv (sup)- Pao (1-4) Vcoll where Rbase = airway resistance in basal segments, Rcoll = resistance in collateral pathways between basal and superior segments, Rsup = air- way resistance of superior segment airways, Pct = pressure at the tip of the bronchial catheter, Palv (base) = alveolar pressure in the basal segments, Palv (sup) = alveolar pressure in the superior segment, 22 Pao = tracheal pressure, and Vcoll = rate of gas flow into the bronchial catheter. Hogg et a1. (23) found that in health lungs Rcoll greatly ). In exceeded Rbase and that Rsu was zero (Rcoll>Rbase>Rsu P emphysematous lungs, however, there was a marked fall inpRcoll and increase in Rbase and Rsup’ such that Rbase>’Rsup>'RC°1]‘ They suggested that an obstructed segment is ventilated primarily by airways in healthy lungs but may be ventilated preferentially by collateral pathways in emphysematous lungs. Menkes and colleagues (8, 54) used Hilpert's technique (equation 1-1) and observed that when the constant Vcoll into the segment was suddenly interrupted causing Ps-Pao to decay to zero, the initial portion of the decay occurred more rapidly than the remaining portion. They attributed the rapid initial drop in Ps-Pao to pressure equilibration (and thus resistance) in the small airways between the tip of the wedged catheter and collateral channels, and the subsequent slower decline to deflation of the obstructed segment through collateral channels. They used this analysis in an attempt to differentiate the reSponse of small airways and collateral channels to methacholine, C02, and variations in lung volume. Sasaki et a1. (73), realizing that Rcoll measured with Hilpert's technique (72) includes the resistance of small airways and one or more types of collateral channels, attempted to measure airflow resistance through interalveolar pores only. The first measured Rcoll using Hilpert's technique in the excised lower lobes of dogs' lungs. Then they glued 2 small capsules to the segment's 23 pleural surface and a third capsule to the lobar pleural surface, and the pleural surface under each capsule received 30 small punctures. A silicone rubber solution, having an experimentally adjusted viscosity, was poured into the lobar bronchus, and following vulcanization of the rubber, lobar airways were obstructed to the bronchiolar level. Segment airways were prevented by the wedged catheter from receiving the rubber. Sasaki et a1. (73) reasoned that when air flowed into one of the segment capsules and was collected from the lobar capsule, the only collateral pathways available would be interalveolar pores. The resistance through the pores was often greater than 100 times higher than Rcoll measured by Hilpert's technique, so they concluded that collateral ventilation does not normally occur through interalveolar pores. The time constant for collateral ventilation (Tcoll) provided another assessment of the mechanical behavior of collateral channels. When a constant Vcoll into a segment was suddenly interrupted, Hilpert found that Ps-Pao decayed exponentially to zero, and Tcoll was calcu- lated as the time required for Ps-Pao to decrease to 37 percent of its initial value. The shorter the Tcoll, the more readily collateral ventilation occurs. Woolcock and Macklem used two other methods to measure Tcoll (87). In the first method, a small volume of air was injected into the outer lumen of a double lumen catheter which obstructed an airway. The pressure in the obstructed segment, measured by the inner lumen of the catheter, initially rose, and Tcoll was estimated as the pressure fell back to baseline. In the second method, an excised lung was oscillated sinusoidally at different frequencies (f) through a tracheal cannula attached via a loud speaker 24 powered by a variable frequency wave generator. Simultaneously, pressure in an obstructed segment (P5) was measured by the inner lumen of a wedged catheter. The phase angle (6) by which Ps lagged behind transpulmonary pressure was given by: e = tan-1 an Tcoll. The equation was rearranged as: Tcoll = tane/(an). The Tcoll calculated by this method was in close agreement with that measured by Hilpert (22). All these methods revealed that Tcoll in the dog lung was inversely proportional to lung volume. Also, increasing the size of the segment increased Tcoll in the dog but not in the human (87). In addition, it has recently been determined that Tcoll is dependent on lobar anatomy. Robinson and Milar (70) found in excised dogs' lungs that Tcoll was longest in the right middle and left cranial lobes and shortest in the right and left caudal lobes. Pressure-Flow Relationships Basic Concepts of Fluid Mechanics The pressure required to cause gas to flow through a tube is dependent on flow rate, geometry of the tube, density and viscosity of the gas, and flow regime (i.e., whether flow is laminar, transi- tional, or turbulent). If the tube is straight, smooth, and circular, the flow regime at low flow rates is laminar (64) having the following characteristics: (1) flow in all parts of the stream is constant and occurs in parallel laminae which slide over one another, and (2) the profile of the leading edge of flow is parabolic, the axial velocity of flow being twice as fast as the average velocity. At high flow 25 rates, the flow regime becomes turbulent (64) having the following characteristics: (1) the velocity at any point in the stream fluctuates vigorously and randomly in both magnitude and direction, and (2) the flow profile is only slightly rounded, so the axial velocity is only 1.2 times the average velocity. At moderate flow rates, flow is said to be transitional having characteristics of both laminar and turbulent flow. These P-V characteristics are illustrated in Figure 1—5 which shows the formation of a fully developed laminar flow profile in sections A-C. Gas enters the tube from the left through a smoothly rounded orifice which introduces no disturbances into the flow. Every molecule entering the tube has the same velocity, so the velocity pro- file at the entrance is virtually flat. At this point (Figure l-5A), pressure and flow are related as: p=KpW (rm where P is the driving pressure, V is flow, p is gas density, and K in equation 1-5 and in the equations that follow is a constant which includes the geometry of the system (45). It takes a finite distance and time for the gas to establish the parabolic flow profile normally associated with the laminar flow regime, and during this distance (entrance length, Figure 1-5 A-C) unequal amounts of shearing force are present between adjacent gas laminae (58). The gas in contact with the wall adheres to the wall, so that the relative velocity of the gas and solid is zero (the "no slip" condition) (64). The laminae near the 26 .Amv zepe pampangau nonopm>wc apps; can .on sop» pmcowpwmcegu .Auv zopm LeewEmp .Amv sazocm Lemar Agmucson .A mom new .Aps\sm "av xuwmcmu mom .Axv ucmpmcou a cw umcvmucou man» ms» mo anumEowm on» o» umumpmg mp Aomx so may chw Low mgzmmmga acm>wgv use .mazu Lopauewu spoosm azmwmgum m cw muwpmwgmpumgmcu zopmlmgammmga .m-_ weauwa 27 9: 3:3... 1m . me..>mkoavn~io¢x > v. n. > 311x." d x+~>x.ix+._meun_ n. no ~>m_x<.. n. m. M. . ruAMV \\“v< , \ .26.. ransom --—--q 1 59.3 3:22”.— 28 no slip layer experience a high shearing force which causes them to decelerate to decrease this shear. The laminae near the center of the tube have less shearing force which causes them to accelerate to increase their shear. This occurs until the gas has established an optimum distribution of shear between the laminae. Thus, in the entrance length, the flow profile changes from flat to parabolic. There is a thin region adjacent to the wall in which the velocity rises from zero to its value over most of the cross section (i.e., core value), V} This region is called the boundary layer (see Figure 1-5), and the central region where the flow profile is flat is called the core. The thickness of the boundary layer is defined as the distance across which the velocity rises from zero to a value indistinguishable from its core value (V). As the flow profile becomes progressively more parabolic in the entrance length, the boundary layer grows in thickness. Pressure (P) and flow (V) are related as follows during this period of boundary layer growth (45): P = K2(up)0'5 91.5 (1-6) where u is gas viscosity. When the entrance length has been established, laminar flow is fully developed. The P-V characteristics for laminar flow were first described by Poiseuille as: p “1875er (1-7) 29 where l is the tube length and r is the radius. When the dimensions of the tube are held constant, then pressure is directly proportional to flow and viscosity: p = K3uV (1-3) When the flow profile is parabolic and the flow rate is increased, the flow profile gradually departs from a parabola, first by becoming transitional where elements from adjacent laminae occasionally mix with one another (Figure l-SD). When this occurs, pressure and flow are related as follows (11): P = K pl+xu-x 92+x (1_9) 5 where x is a number between -1 and 0 (it is -1 if flow is laminar and 0 if flow is orifice flow). As flow rate continues to increase, the flow regime becomes turbulent (Figure l-5E), at which point P is proportional to gas density and flow (86): Reynolds (68) was the first to show that the transition from laminar to turbulent flow in a tube always occurs at the same value of the following dimensionless number: =22.‘ - Re W v (111) This has been confirmed with a wide range of tube diameters, flow rates, and fluid properties. Laminar flow occurs when Re is less than 2300, 30 and turbulent flow occurs when Re is greater than 3000 (61). Transitional flow occurs when Re is between 2300 and 3000. The entrance conditions of the tube play an important role in determining the critical Re (64). If there are no disturbances in the fluid before entering the tube, and if the inlet is constructed in a smooth curve, the critical Re at which turbulence develops may reach 40,000. If the incoming fluid has disturbances, then the critical Re may be reduced to 2000, but not below. Pressure-Flow Relationships in a Branched System Problems of analysis. The lung poses a unique challenge in the analysis of its P-V characteristics because it is far from being the single circular conduit of uniform dimensions described above. Rather, the lung is a complex geometrical structure consisting of irregularly shaped airways branching dichotomously and asymmetrically over twenty or more generations. The total cross sectional area of the combined airways increases from the trachea to the alveoli [it is estimated that the total cross sectional area of the bronchioles is more than one thousand times larger than that of the trachea (30)] while the radius of the individual airways from the trachea to the bronchioles decreases. This not only causes flow velocity to decrease toward the periphery [such that if the velocity in the trachea is 100 cm/sec, then the velocity in the small bronchi would be less than 0.1 cm/sec (30)], but Re decreases as well (49a). In addition, there is very little development of the flow profile until after the fifth order 31 bronchi (58). This is because eddies which form at each bifurcation are carried away from the site of their formation to disturb the flow downstream (30). The distance over which this flow disturbance occurs (the length of transition), is longer than the length of the trachea and large bronchi. Eddies thus prevent the establishment of the laminar parabolic flow profile in the trachea and large airways (10, 85). The length of transition is shorter than the length of the small bronchi, however, so that laminar flow may be established in these small airways. Thus, the driving pressure for flow through all the airways is significantly determined by the entrance effects in the . upper airways (63), and the greatest resistive pressure drop is found in these large airways due to the smaller total cross sectional area of these airways and because of the predominance of entrance length flow where the exponent for flow is greater than 1.0 as opposed to the fully developed laminar flow profile in small airways where the exponent for flow equals 1.0 (see Figure 1-5). Other factors contribute to the complexity of the P-V relationship in the lung. First, the airways are not consistently cylindrical but rather cylindrical in their proximal portion and elliptical near the branch points (24). Secondly, the dimensions of the airways change throughout the respiratory cycle, dilating during inspiration and constricting during expiration (80). Third, the flow rate in the airways varies not only in time as a result of the cyclical nature of breathing but also in space because of the increase in cross-sectional area resulting from airway branching. 32 Fourth, the airways are lined by a viscous liquid (mucus), and gas flow down the airways may cause the formation of waves on the liquid resulting in a roughened surface (9). This phenomenon of two-phase gas-liquid flow becomes significant at higher flow rates in healthy lungs or at normal flow rates in diseased lungs. Fifth, the geometry of the airway bifurcations varies throughout the lung (64). Thus, as a result of its complex geometry, the lung likely contains many Re, and turbulence may be found at a critical Re which is much lower than the value of 2300 found in a straight, smooth, cylindrical tube. The vari- ations in Re which occur during cyclic respiration support the theory that the flow regime in the lung undergoes a progressive metamorphosis as flow rate varies from low to high values (30). Rohrer's equation. Rohrer (72) was the first investigator to attempt to describe the pressure-flow relationships in the lung. He believed that the lung contained both turbulent and laminar flow which were related as follows: P = KGuV + K pv2 (1-12) 7 where K6 and K7 are constants. The term KGUV describes the laminar flow component in the lung, and the term K7pV2 describes the turbulent flow component. The constant K6 contains the geometry of that portion of the airways where flow is laminar, and K7 contains the geometry of the airways where flow is turbulent. In 1955 McIlroy et a1. (49) pointed out that Rohrer's equation provided an excellent curve fitting equation for pulmonary pressure-flow curves. However, breathing helium 33 would be expected to result in a reduction of K7 due to a decrease in gas density and a slight increase in K6 due to an increase in gas viscosity. In fact, 100 percent helium reduced K6‘ In contrast, ethane, which is only one-half as viscous as air and should have reduced K6’ produced little change in this value. McIlroy et a1. (49) also pointed out that the "constants" K6 and K7 change as the distribution of turbulence within the lung changes. As flow rate increases and turbulence increases with it, K7 should increase and K6 should diminish because more of the airways contain turbulent flow and fewer airways contain laminar flow. Similarly, because changing gas physical properties alters the distribution of Re, it also alters the distribution of flow regimes. Therefore, K6 and K7 should be different when gases of different physical properties are breathed. In contrast, K6 and K7 would remain constant only when the physical properties of the gas were changed such that the kinematic viscosity (viscosity divided by density) remained constant. Thus, even though Rohrer's equation fits experimentally determined pressure-flow curves quite well, it does not account for the site and nature of the gas flow regime in the lung. Modification of Rohrer's equations. Several investigators have attempted to modify Rohrer's equation to better account for the pressure-flow relationships in the lung and thus provide a better estimation of the flow regime. Using rigid models of human airways, Pedley, Schroter, and Sudlow (62, 63) showed that most of the resis- tance to air flow is located in the large airways. Since the flow profile through large airways is undeveloped, these investigators 34 proposed that P, V, and gas physical properties could be related as in equation 1-6 for boundary layer growth. They assume that flow regimes other than boundary layer growth contribute such a small percentage to the total resistance to flow that they may be ignored. Wood et a1. (86), however, found that this is not the case in dogs breathing air because peripheral airways with fully developed laminar flow contribute 33 percent or more of the total lower pulmonary resistance. They conclude that the equation for boundary layer growth does not describe the P-V relationships for the lower airways because other P-V regimes contribute substantially to the resistance. Thus, Wood et al. (86) propose the following equation which is designed to take into account the metamorphosis in space and time of the flow regimes: P = Kpa-l “2'3 03 (1'13) The exponent 'a' varies between 1 and 2. When flow is laminar through all the airways, a = l, and the equation simplifies to the Poiseuille equation (equation 1-8). When the flow profile is undeveloped, a = 2, and the equation simplifies to that describing orifice flow (equation 1-5). When the flow regime is different in different parts of the airway, the value of 'a' will be between 1 and 2. If there is fully developed laminar flow in those airways that contribute most to resis- tance, 'a' will be close to 1, whereas if there is orifice flow in those airways that contribute most to resistance, 'a' will be closer to 2. Thus, the flow regime in the lung may be predicted by the value of 'a'. 35 This equation proposed by Wood et a1. (86) fails in several respects to accurately describe the P-V relationships in the lung. First, since the spatial and temporal variation in Re determines the spatial and temporal variation in flow regimes which in turn determines the value of 'a', 'a' itself should be a function of Re only. However, Lisboa et a1. (41) found that the relationship between 'a' and Re is quite different during breathing 80 percent He-20 percent 02 than it is for air, suggesting that factors other than Re influence 'a'. Second, the equation predicts that when flow is laminar (i.e., a = l) and resistance is viscosity dependent, the pressure required to produce a unit of flow with He/O2 should be greater than with air because He/O2 is more viscous. In contrast, when flow is orifice flow (a = 2) and resistance is density dependent, the pressure required to produce a unit of flow should be less on He/O2 than with air because He/O2 is less dense. Lisboa et al. (41), however, found no consistent relationship between resistance with He/O2 and air and the value of 'a'. Thus, equation 1-14 does not account for the effect of gas physical properties. Third, equation 1-14 contains the same general inadequacies as Rohrer's equation (equation 1-12): that is, since the distribution of Re and thus the flow regime changes with the flow rate and gas physical properties, the geometry of the airways containing laminar and turbulent flow also changes. As a result, the constant K which contains this geometry also changes (i.e., the 'constant' is not constant). It is evident, therefore, that the pressure-flow rela- tionships in the lung are very complicated, especially considering that the flow pattern undergoes a progressive metamorphosis as flow 36 rate varies from low to high values (30). This relationship cannot be expressed by any single algebraic equation. Moody diagram. Another method used to describe the flow regime through the lung makes use of a unique relationship between the degree of turbulence as indicated by the Re and the ratio of the static pres- sure drop to dynamic pressure within the lung (i.e., the normalized pressure drop). This normalized pressure drop (Pn) is expressed as follows: AP Pn 5:5577' (1—14) where AP is the static pressure drop, and v is the velocity of gas flow. When Log Pn is plotted against Log Re (Figure 1-6), a Moody diagram (56) results, the slope of which indicates the flow regime. Three regions may be identified in the Moody diagram. At the extreme left, Re is low and flow is laminar. The slope of the curve is -1, so that pressure~ and flow are related as follows: Log 075%V7'= -Log 2&1. (1-15) At the extreme right of the Moody diagram the high Re indicates that flow is fully turbulent. The slope of the curve is 0, so pressure and flow are related as: AP _ _ Log W- C (1 17) Figure 1-6. 37 Schematic illustration of Moody diagram where the logarithm of the normalized pressure drop (Log Pn; ordinate) is plotted against the logarithm of Reynolds' number (Log Re; abscissa). The slope of the curve is determined by the gas flow regime: if the slope = -l, the flow regime is laminar (extreme left); if the slope = 0, the flow regime is turbulent (extreme right); if the slope is between -1 and 0, the flow regime is transitional (center). In the equations, AP = static pressure drop (cm H20) from the trachea to the alveoli; p = gas density (gm/ml); u = gas viscosity [gm/(sec- cm)]; v = gas velocity (cm/sec); and d = a reference cross sectional area (cmz). 38 Transitional ll 2! Fully Turbulent —q é LogRE Figure 1-6 m 39 where C is a constant. Between these two extremes the slope changes from -1 to 0. The intermediate Re's indicate that flow is transitional, and pressure and flow are related as: Log 0.5:v = x Log Qfi—- (1—18) This equation applies only to a small section of the curve where x (the slope of the curve in log-log coordinates) can be considered constant (11). The Moody diagram is a classical fluid dynamic approach for characterizing the flow regime in a single conduit (56). This graphical method assumes that when geometry is constant, the change in Pn with V is a function of Re only. A Moody diagram may be determined experimentally for a single conduit or for more complex systems such as the branching system of airways (11, 27-30, 42, 58, 60, 74). In the latter case, what is measured is the predominant flow regime through the lung as a whole rather than a single specific flow regime through any particular airway. In other words, the entire lung is considered as a single rough tube (11). This method finds that the critical Re (indicating the loss of laminar flow) in the lung differs greatly from that determined for a straight, smooth, circular tube. Slutsky et a1. (74) found that the predominant flow in a model of the human lung was laminar at Re less than 500, transitional between Re of 500 and 5000, and turbulent when Re exceeds 5000. Isaby and Chang (27), using a similar model, found laminar flow at Re less than 200, transitional flow at Re between 500 and 1500, and turbulent flow at Re greater than 4000. These critical 40 Re of 200 (27) and 500 (74) for models of the human lung would tend to be even lower in the living lung where two-phase gas-liquid flow predisposes to turbulence (9). Pressure—Flow Relationships in a Sublobar Lung Segment Collateral channel resistance (Rcoll) is measured in sublobar lung segments (22) by advancing a polyethylene catheter having a flared tip through the trachea or lobar bronchus until it becomes wedged in a small bronchus (Figure 1-4). This procedure obstructs a sublobar seg- ment which is the only region which inflates when gas is blown through the catheter. The gas which inflates the segment escapes by flowing through collateral pathways which direct the gas to airways outside the segment. When both the driving pressure for gas flow and the flow rate are known under a variety of conditions, the P-V relationships (and thus the predominant flow regime) in the segment can be estimated. Olson et a1. (60) recently pointed out that the segment flow regime may be an important determinant of Rcoll. To explain why segment inflation increases Rcoll and excised lungs (71) but may decrease Rcoll in intact lungs (32, 35, 36), Olson et a1. (60) theorize that the higher gas flow rates used to inflate segments in excised lungs may result in a turbu- lent flow regime which increases Rcoll. They used a Moody analysis to characterize the segment flow regime in anesthetized vagotomized dogs under the following conditions: transpulmonary pressure (Ptp) = 0 (functional residual capacity), 2, and 4 cm H20 while segment pressure (Ps) exceeded Ptp by 3 cm H20. They repeated their measurements using gases having different physical properties. This allowed them to vary 41 gas density, viscosity, and flow rates while maintaining P5 3 cm H20 greater than Ptp. They concluded that when Re is less than approx- imately 100, flow through the segment is laminar, and when Re is between 100 and 1000, their data indicate that flow is nonlaminar (transitional, but not fully turbulent). In addition, they estimated Re for excised and intact dogs' lungs using published data and found that Re ranged from 140 to 600 in excised lungs and from 30 to 100 in intact lungs. They suggest that the different Re in excised and intact lungs may be due to differences in the flow regime. Segment-Lobar Inhomogeneity The most important determinant of airway resistance is airway diameter which in turn is determined primarily by the compliance of the surrounding lung parenchyma which exerts radial traction on the airways (57). Since collateral channels connect an obstructed segment with the lobe, it follows that Rcoll is strongly influenced by both segment compliance (CS) and lobar compliance (CL). Although CS = CL during homogeneous lobar inflation (i.e., when lobar distention pressure is distributed uniformly throughout the segment and lobe), CS is less than CL during inhomogeneity (i.e., when the segment distention pressure is greater than lobar distention pressure) (18, 50, 51, 53, 69, 70, 75, 87). The latter occurs during measurement of Rcoll using Hilpert's technique (22) (Figure 1-4) because of interdependence in which tissue attachments common to the segment and lobe tend to prevent segment expansion, causing a smaller increase in segment volume per unit 42 increase in segment pressure. Thus, since inhomogeneity influences CS and CL’ inhomogeneity should likewise influence Rcoll. A second mechanism by which inhomogeneity may influence Rcoll is by distortion of collateral channels as they pass through the zone of parenchymal tissue distortion at the segment-lobar inter- face (Figure 1-2). Blood vessels in the interface are also likely distorted, and the resultant increased vascular resistance has been suggested as a cause of the decreased segment bloow flow measured during inhomogeneity (14, 15). The effect of inhomogeneity on collateral ventilation is uncertain. Woolcock and Macklem (87) propose that inhomogeneity enhances collateral ventilation because the reduced CS would also reduce the time constant for collateral ventilation (Tcoll = Rcoll- CS). Direct measurements, however, indicate that Tcoll changes little during inhomogeneity (71). Other studies have measured the effect of inhomogeneity on Rcoll and have yielded conflicting results. Using Hilpert's technique (Figure 1-4), Robinson and Sorenson (71) reported that inhomogeneity increased Rcoll in excised dogs' and horses' lungs while Kaplan et a1. (32) calculated a decrease in Rcoll in intact dogs' lungs. In contrast, Baker and Daly (4) measured the volume of collateral gas flow during inhomogeneity and found that the volume was directly related to the intersegmental pressure gradient, suggesting that Rcoll remained constant as inhomogeneity increased. Baarsema and Dirken (3) obtained similar findings in intact rabbits' lungs. Thus, the studies describing 43 the effect of inhomogeneity on collateral channels provide inconsistent results and demonstrate the need for further investigation. Since Rcoll is influenced by the mechanical behavior of both the segment and the lobe, it is of interest to know which region influences Rcoll more. In the only report on this issue, Kaplan et a1. (32) found that Rcoll fell more during lobar inflation than during inhomogeneous segment inflation, suggesting Rcoll was influenced more by the lobe than by the segment. This study was not well controlled, however, because comparisons were not made at equal degrees of inhomo- geneity during each mode of inflation. Also, a narrow range of lobar and segment inflation pressures were used. Other data suggest that the lobe has a stronger influence because raising lobar volume undisputedly decreases Rcoll (23, 25, 52, 71, 82, 87), whereas the evidence for the effect of segment volume on Rcoll is inconclusive. Well controlled studies are needed which make direct comparisons of segment and lobar influences on Rcoll. Purpose of the Present Studies Much of the knowledge of the characteristics and function of collateral channels has been obtained using Hilpert's technique (22) where an obstructed sublobar lung segment is inflated inhomogeneously within a lobe. The effect of inhomogeneity on Rcoll is uncertain, however. The purpose of Chapter II is to determine the effect of inhomogeneity on Rcoll under a variety of measurement conditions commonly employed by Hilpert's technique, and also to determine whether the segment of the lobe has the stronger influence on Rcoll. 44 The purpose of Chapter III is to determine the mechanism of the effect of inhomogeneity on Rcoll. Chapter IV presents findings on the anatomy of the airways in the segment-lobar interface and also describes the route for collateral gas flow. 10. 11. LIST OF REFERENCES Chapter I Adriani, quoted from Miller, W. S. The Lung. Springfield, 111.: Thomas, 1937, pp. 64-68. Alley, R. D. and G. E. Lindskog. Pharmacological factors influencing collateral respiration; possible relation to the etiology of pulmonary complications. App, Surg. 128: 497-508. Baarsema, P. R. and M. N. J. Dirken. Collateral ventilation. J, Thoracic Surg. 17: 238-251, 1948. Baker, 0. H. and W. J. Daly. Collateral ventilation demonstrated by helium transfer. J, Appl. Physiol. 26: 321-325, 1969. Boyden, E. A. Notes on the development of the lung in infancy and early childhood. Am, J, Physiol. 121: 749-762, 1967. Boyden, E. A. The structure of the pulmonary acinus in a child of six years and eight months. Am, p, Physiol. 132: 275-300, 1971. Call, E. P., Jr., G. E. Lindskog, and A. A. Liebow. Some physiologic and pharmacologic aspects of collateral ventilation. J, Thoracic Cardiovas. Surg. 49: 1015-1025, 1965. Chang, H. K. and J. P. Mortola. Fluid dynamic factors in tracheal pressure measurement. 9, Appl. Physiol.: Respirat. Env1ron. Exercise Physiol. 51: 218-225, 1981. Clark, S. W., J. G. Jones, and D. R. Oliver. Resistance to two-phase gas-liquid flow in airways. J, Appl. Physiol. 29: 464-471. 1970. Dekker, E. Transition between laminar and turbulent flow in the human trachea. 9, Appl. Physiol. 16: 1060-1064, 1961. Drazen, J. M., S. H. Loring, and R. H. Ingram, Jr. Distribution of pulmonary resistance: effects of gas density, viscosity, and flow rate. 9, Appl. Physiol. 41: 388-395, 1976. 45 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 46 Drazen, J. M., S. H. Loring, and R. H. Ingram, Jr. Localization of airway constriction using gases of varying density and viscosity. 9, Appl. Physiol. 41: 396-399, 1976. Duguid, J. B. and M. W. Lambert. The pathogenesis of coal miners pneumoconiosis. J, Path. Bact. 88: 389-403, 1964. Enjeti, S., J. T. O'Neil, P. 8. Terry, H. A. Menkes, and R. J. Traystman. Sublobar atelectasis and regional blood flow. J, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 47: 1245-1250, 1979. Enjeti, S., P. 8. Terry, H. A. Menkes, and R. J. Traystman. Mechanical factors and the regulation of perfusion through atelectatic lung. J, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 52: 647-654, 1982. Flenley, D. C., J. Picken, L. Welchel, F. Rufff, P. M. Corry, and P. T. Macklem. Blood gas transfer after small airways obstruction in the dog and minipig. Resp. Physiol. 15: 39-51. 1972. Flenley, D. C., L. Welchel, and P. T. Macklem. Factors affecting gas exchange by collateral ventilation in the dog. Re p. Physiol. 15: 52-69, 1972. Ford, G. T., C. A. Bradley, and N. R. Anthonisen. Forces involved in lobar atelectasis in intact dogs. 9, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 48: 22-33, 1980. Gamsu, G., W. M. Thurlbeck, P. T. Macklem, and R. G. Fraser. Peripheral bronchographic morphology in the normal human lung. Invest. Radiol. 6: 161-170, 1971. Gertner, A., B. Bromberger-Barnea, L. Kelly, R. Traystman, and H. A. Menkes. Local vagal reflexes in the lung periphery (abstract). Egg, Proc. 42: 1094, 1983. Henderson, R., K. Horsfield, and G. Cumming. Intersegmental collateral ventilation in the human lung. Resp. Physiol. 6: 128-134, 1968/1969. Hilpert, P. Collaterale Ventilation Habilitationsschrift. Ans der Medizinischen (Thesis). Tubingen: Universitatsklinik, 1970. Hogg. J. C., P. T. Macklem, and W. M. Thurlbeck. The resistance of collateral channels in excised human lungs. J, Clin. Invest. 48: 421-431, 1969. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 47 Horsfield, K., G. Dart, D. E. Olson, G. F. Filley, and G. Cumming. Models of the human bronchial tree. J, Appl. Physiol. 31: 207-217, 1971. Inners, C. R., P. 8. Terry, R. J. Traystman, and H. A. Menkes. Collateral ventilation and the middle lobe syndrome. Am. Rev. Resp. Dis. 118: 305-310, 1978. Inners, C. R., P. 8. Terry, R. J. Traystman, and H. A. Menkes. Lung volume and resistance to collateral ventilation in man (abstract). Egg, Proc. 37: 927, 1978. Isaby, D. and H. K. Chang. Steady and unsteady pressure-flow relationships in central airways. J, Appl. Ph siol.: Respirat. Environ. Exercise Physiol. 5 : 1338- 348, 1981. Jaffrin, M. Y. and P. Kesic. Airway resistance; a fluid mechanical approach. g, Appl. Physiol. 36: 254-361, 1974. Jaeger, M. J. and H. Matthys. The pattern of flow in the upper human airways. Resp. Physiol. 6: 113-127, 1968. Jaeger, M. J. and H. Matthys. The pressure flow characteristics of the human airways. In Airway Dynamics, Physiology and Pharmacology. edited by A. Bouhuys. Springfield, 111.: Thomas, 1970, pp. 21-32. Johnson, R. M. and G. E. Lindskog. Further studies on factors influencing collateral ventilation. J, Thoracic Surg. 62: 321-329, 1971. Kaplan, J., R. C. Koehler, P. 8. Terry, H. A. Menkes, and R. J. Traystman. Effect of lung volume on collateral ventilation in the dog. g, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 49: 9-15, 1980. Krahl, V. E. Microscopic anatomy of the lungs. Am. Rey. Resp. Dis. 80: 24-44. 1959. Lambert, M. W. Accessory bronchiole-alveolar communications. J, Path. Bact. 70: 311-314, 1955. Levy, H. L. and L. D. H. Wood. Effect of gas physical properties on resistance to collateral ventilation (abstract). Egg. Proc. 35: 231, 1976. Levy, H. L. and L. D. H. Wood. Effect of pulmonary interdependence on channels of collateral ventilation (abstract). Ihg_ Physiologist. 20: 56, 1977. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 48 Liebow. A. A. Recent advances in pulmonary anatomy. In Ciba Found. Symp. Pulmonary Structure Function, edited by A. V. S. DeReuck and M. O'Conner. London: Churchill, 1962. Lindskog, G. E. Collateral respiration in the normal and diseased lung. Yale g, 819, Med, 23: 311-316, 1950/1951. Lindskog, G. E. and H. H. Bradshaw. Collateral respiration. The chemical composition and volume of the collateral respired gases. Am, J, Physiol. 108: 581-592, 1934. Lindskog, G. E. and C. M. VanAllen. The aerodynamics of bronchial obstruction. Arch. Surg. 24: 204-230, 1932. Lisboa, C., W. R. 0. Ross, J. Jardim, and P. T. Macklem. Pulmonary pressure-flow curves measured by a data averaging circuit. J, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 47: 621-627, 1979. Lisboa, C., L. D. H. Wood, J. Jardim, and P. T. Macklem. Relation between flow, curvilinearity, and density dependence of pulmonary pressure-flow curves. J, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 48: 878-885, 1980. Loosli, C. G. Interalveolar communications in normal and pathologic mammalian lungs. Arch. Path. 24: 743-776, 1937. Macklem, P. T. Airway obstruction and collateral ventilation. Physiol. Rev. 51: 368-436, 1971. Macklem, P. T. The paradoxical nature of pulmonary pressure-flow relationships. Egg, Proc. 39: 2755-2758, 1980. Macklem, P. T., A. J. Woolcock, J. C. Hogg. J. A. Nadel, and N. J. Wilson. Partitioning of pulmonary resistance in the dog. 9, Appl. Physiol. 26: 798-805, 1969. Martin, H. B. The effect of aging on the alveolar pores of Kohn in the dog. Am, Rev. Resp. Dis. 88: 773-778, 1963. Martin, H. 8. Respiratory bronchioles as the pathway for collateral ventilation. 9, Appl. Physiol. 21: 1443-1447, 1966. McIlroy, M. B., J. Head, N. J. Selverstone, and E. P. Radford. Measurement of lung tissue viscous resistance using gases of equal kinematic viscosity. J, Appl. Physiol. 7: 485- 490. 1955. 49 49a. Mead, J. Mechanical properties of lungs. Physiol. Rev. 50. 51. 52. 53. 54. 55. 56. 57. 57a. 58. 59. 60. 41: 281-330. 1961. Mead, J., T. Takishima, and D. Leith. Stress distribution in lungs: a model of pulmonary elasticity. A, Appl. Physiol. 28: 596-608. 1970. Menkes, H., G. Gamsu, R. Schroter, and P. T. Macklem. Interdependence of lung units in isolated dog lungs. A, Appl. Ehysiol. 32: 675-680, 1974. Menkes, H., 0. Lindsay, G. Gamsu, L. Wood, A. Muir, and P. T. Macklem. Measurement of sublobar lung volume and collateral flow resistance in dogs. 9, Appl. Physiol. 35: 917-921, 1973. Menkes, H., 0. Lindsay, L. Wood, A. Muir, and P. T. Macklem. 9, Appl. Physiol. 32: 681-686, 1972. Menkes, H. A. and R. J. Traystman. State of the art. Collateral ventilation. Am, Rev. Resp. Dis. 116: 287-309, 1977. Metcalf, J. F., P. D. Wagner, and J. 8. West. Effect of local bronchial obstruction on gas exchange in the dog. Am, Agy, Resp. Dis, 117: 85-95, 1978. Moody, L. F. Friction factors for pipe flow. Trans. ASME. 66: 671-678, 1944. Murray, J. F. The Normal Lung. Philadelphia, Pa.: Saunders, 1976, p. 99. Ogawa, C. Contribution to the histology of the respiratory spaces of the vertebrate lungs. Am, g, Anat. 27: 333-393, 1920. Olson, D. E., G. A. Dart, and G. F. Filley. Pressure drop and fluid flow regime of air inspired into the human lung. g, Appl. Physiol. 28: 482-494, 1970. Olson, L. E. and N. E. Robinson. Effect of vagal stimulation on collateral flow resistance in dog lungs. 9, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 49: 289-293, 1980. Olson, L. E., J. R. Rodarte, and N. E. Robinson. Pressure-flow relationships in a collaterally ventilating dog lung segment. 9, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 54: 956-960, 1983. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 50 Patel, V. and M. R. Head. Reversion of turbulent to laminar flow. 9, Fluid Mech. 34: 371-392, 1968. Pedley, T. J., R. C. Schroter, and M. F. Sudlow. Energy losses and pressure drop in models of human airways. Resp. Physiol. 9: 371-386. 1970. Pedley, T. J., R. C. Schroter, and M. F. Sudlow. The prediction of pressure drop and variation of resistance within the human bronchial airways. Resp. Physiol. 9: 387-405, 1970. Pedley, T. J., R. C. Schroeter, and M. F. Sudlow. Gas flow and mixing in the airways. In Bioengineering Aspects of the Lung, edited by J. 8. West. New York: Marcel Dekker, Inc., 1971. Pp. 163-266. Pump, K. K. The morphology of the finer branches of the bronchial tree of the human lung. Dis. Chest. 46: 379-398, 1964. Raskin, S. P. and P. G. Herman. Interacinar pathways in the human lungs. Am, Rev. Resp. Dis. 111: 489-495, 1975. Reid, L. The secondary lobule in the adult human lung, with special reference to its appearance in bronchograms. Thorax. 13: 110-115, 1958. Reynolds, 0. An experimental investigation of the circumstances which determine whether the motion of water shall be direct or sinuous, and the law of resistance in parallel channels. Phil. Trans. A, Soc. London. 174: 935-982, 1883. Robinson, N. E. Lobar variations in the mechanics of collateral ventilation in intact dog lungs. Am, Rev. Resp. Dis. 124: 68-71. 1981. Robinson, N. E. and R. Milar. Lobar variation in collateral ventilation in excised dog lungs. Am, Rev. Resp. Dis. 121: 827-834, 1980. Robinson, N. E. and P. R. Sorenson. Collateral flow resistance and time constants in dog and horse lungs. 9, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 44: 63-68, 1978. Rohrer, F. Der Stromungswiderstand in den menschlichen Atemwegen. Pflugers Arch. Gesamte. Physiol. 162: 225-259, 1915. Sasaki, H., T. Takishima, and M. Nakamura. Collateral resistance at alveolar level in excised dog lungs. g, A 1. Physiol.: Respirat. Environ. Exercise Physiol. 48: 9 - 90, 80. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 51 Slutsky, A. S., G. G. Berdine, and J. M. Drazen. Steady flow in a model of human central airways. 9, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 49: 417-423, 1980. Sylvester, J. T., H. A. Menkes, and F. Stitik. Lung volume and interdependence in the pig. 9, Appl. Physiol. 38: 395-401, 1975. VanAllen, C. M. A dilatable bronchial cannula. ale 9, 819, App, 2: 295-296, 1930. VanAllen, C. M. Obstructive pulmonary emphysema and collateral respiration. Surg, Gynecol. Obst. 55: 303-307, 1932. VanAllen, C. M. and F. Adams. The mechanism of obstructive pulmonary atelectasis. Surg. Gynecol. Obst. 50: 385-396, 1930. VanAllen, C. M. and T. S. Jung. Postoperative atelectasis and collateral respiration. g, Thoracic Surg. 1:3-14, 1931. VanAllen, C. M. and G. E. Lindskog. Obstructive pulmonary atelectasis. Problems of pathogenesis and clinical management. Arch. Surg. 21: 1195-1213, 1930. VanAllen, C. M. and G. E. Lindskog. Collateral respiration in the lung. Role in bronchial obstruction to prevent atelectasis and to restore patency. Surg. Gynecol. Obst. 53: 16-21, 1931. VanAllen, C. M., G. E. Lindskog, and H. G. Richter. Gaseous interchange between adjacent lung lobules. Yale g, 819, Mpg, 2: 297-300, 1930. VanAllen, C. M., G. E. Lindskog, and H. G. Richter. Collateral respiration. Transfer of air collaterally between pulmonary lobules. g, Clin. Invest. 10: 559-590, 1941. VanAllen, C. M. and Y. C. 500. Collateral respiration. Spontaneous reinflation of an atelectatic pulmonary lobule by collateral respiration. 9, Clin. Invest. 12: 171-179, West, J. B. and P. Hugh-Jones. Patterns of gas flow in the upper bronchial tree. 9, Appl. Physiol. 14: 753-759, 1959. Wood, L. D. H., L. A. Engel, P. Griffin, P. Despas, and P. T. Macklem. Effect of gas physical properties and flow on lower pulmonary resistance. 9, Appl. Physiol. 41: 234-244, 1976. Woolcock, A. J. and P. T. Macklem. Mechanical factors influencing collateral ventilation in human, dog, and pig lungs. 9, Appl. Physiol. 30: 99-115, 1971. CHAPTER II THE EFFECT OF REGIONAL INHOMOGENEITY ON COLLATERAL AIRWAY RESISTANCE Introduction Regional inhomogeneity occurs when a sublobar lung segment is inflated to a different relative volume than the remainder of the lobe. The segment has a low compliance due to its interdependence with the adjacent lobar parenchyma (3, 10, 11, 13, 20), and it has been proposed that this low compliance promotes collateral ventilation by reducing the time constant for collateral air flow (20). Blood flow to a segment is not promoted by regional inhomogeneity, however, for Enjeti et a1. (2) demonstrated that regional blood flow decreases more when a sublobar than a lobar region is made atelectatic. They theorize that inter- dependence between the sublobar region and the lobe causes additional mechanical distortion of vessels which is not likely to occur in the case of lobar atelectasis (l, 2). If collateral channels behave like blood vessels, then collateral ventilation may be decreased during regional inhomogeneity rather than increased as previously suggested (20). Indeed, collateral channels passing through the parenchymal interface between the segment and the lobe are subjected to abrupt changes in distending pressure which may distort the channels, and the degree of distortion should be determined by the pressure 52 53 difference between the segment (Pct) and the lobe as measured at the airway opening (Pao). It is generally accepted that when Pct-Pao is held constant, raising transpulmonary pressure decreases collateral resistance (Rcoll) (5, 12). Although this is probably due primarily to the volume-dependency of airway resistance (9), the degree of interface distortion produced by a given Pct-Pao may vary as the lung inflates because of changes in the degree of interdependence with lung inflation (12, 19). Thus, Rcoll and interface distortion resulting from regional inhomogeneity enjoy a complex interrelationship which may depend on both Pct and transpulmonary pressure, and this interrelationship has not been adequately described in the literature. It has been reported that raising Pct-Pao from two to ten cm H20 increases Rcoll in excised dog and horse lungs (17). In contrast, raising Pct-Pao from one to three cm H20 decreases Rcoll in intact dog lungs held at functional residual capacity (8). This latter report did not explore the possibility, however, that larger changes in Pct-Pao using a wider range of transpulmonary pressures may increase Rcoll. In the present study, combinations of Pct and transpulmonary pressures were assembled into a block design so as to produce a variety of Pct and Pct-Pao at several transpulmonary pressures. Experiments were performed in isolated segments in the left cranial and right caudal lobes of excised and intact dogs' lungs to determine if differences in segment geometry or the presence of a chest wall altered the response of the collateral channels to regional inhomogeneity. Data were inter- preted on the basis of there being collateral channels and airways 54 located not only in the body of the segment but also passing through the segment-lobar parenchymal interface, so that changes in Rcoll are determined by which population of channels changes its resistance in greatest magnitude. Methods In experiments using excised lungs, mixed-breed dogs were anesthetized with sodium pentobarbital and exsanguinated. The lungs and attached trachea were removed from the chest cavity, and the pleural surfaces were kept moist with saline throughout the experiment. The lungs were suspended by the trachea which was attached to one arm of a 3-way connector. A second arm of the connector was attached to a variable speed blower which was used to inflate the lungs. The third arm of the connector was used for passage of a double lumen catheter into the tracheobronchial tree. Transpulmonary pressure was tracheal pressure (Pao), measured by a transducer (model no. PM131, Statham, Hato Rey, Puerto Rico) connected to the trachea by means of a polyethylene catheter and needle. Collateral resistance was calculated by the method of Hilpert (5) (Figure 1-4). A double-lumen catheter (o.d. 3.2 mm) with the distal end flared to an outside diameter of 5 mm was wedged in a bronchus while the lung was inflated to total lung capacity (Pao = 30 cm H20). When the lung was deflated, gentle traction on the catheter depressed the lobar pleural surface, and considerable force was required to remove the catheter, suggesting the catheter securely obstructed the airway. The inner lumen (o.d. 1.7 mm) of the catheter was connected to one side 55 of a second differential pressure transducer (model no. PM131, Statham, Hato Rey, Puerto Rico). The other side of the transducer was connected to the trachea with a catheter and needle so that the pressure differ- ence (Pct-Pao) between the tip of the wedged catheter (Pct) and the trachea could be measured. A photorecording oscilloscope (Model DR8, Electronics for Medicine, White Plains, New York) was used to record Pao and Ps-Pao. During measurement of Rcoll the lobe was inflated to the desired Pea, and air flow from a compressed air tank (Vcoll) was directed through the outer lumen of the wedged catheter and into the isolated segment; it passed through collateral channels and left the lung via the trachea. Flow was adjusted by means of a rotameter (Model 7431T, Matheson, East Rutherford, New Jersey) until the desired steady-state pressure gradient (Pct-Pao) was obtained. Collateral resistance was calculated as (Pct-Pao)/Vcoll. In one series of experiments, gas was withdrawn from the segment at a constant flow rate adjusted by the rotameter connected to a vacuum source. Collateral resistance was calculated as above. A minimum of two sequential Rcoll measurements were taken at each Pct-Pao. The lungs were inflated to total lung capacity preceding each measurement to ensure a constant volume history. In the experiments using intact lungs, Rcoll was measured as described by Inners et a1. (7) using a bronchoscope instead of a double- lumen catheter. Mixed-breed dogs were anesthetized with sodium pento- barbital (33 mg/kg), and a femoral vein was cannulated for delivery of supplemental anesthetic. A tracheostomy was performed, and the trachea 56 was connected to one arm of a 3-way connector. A second arm of the connector was attached to a Harvard pump (model 607, Harvard Apparatus, Dover, Massachusetts), frequency and tidal volume being adjusted to maintain the dogs' end-expired CO2 concentration between 4.5 and 5.0 percent. The dog was placed in sternal recumbency, and the Harvard pump was briefly turned off at end expiration, during which time a bronchoscope of 5 mm o.d. (model BF2, 82, Olympus Corp., New Hyde Park, New York) was passed through the third arm of the connector and wedged in a peripheral airway of either the left cranial or right caudal lobe. A rubber cuff placed around this third connector arm provided a seal to prevent air leaks during ventilation. An esophageal balloon, passed through the nares and into the esophagus, positioned to measure pleural pressure (Ppl), was attached to one side of a differen- tial pressure transducer. The other side of this transducer sensed tracheal pressure (Pao), providing a measurement of transpulmonary pressure (Ptp = Pap-Ppl). The esophageal balloon was positioned such that spontaneous breathing against a closed airway produced minimal fluctuations in Ptp. This resulted in a Ptp of 3 to 4 cm H20 at functional residual capacity in these dogs. A catheter was inserted into the suction/biopsy part of the bronchoscope and connected to another differential pressure transducer to measure Pct, and the other side of this transducer measured Pao. The remainder of the suction/biopsy channel was connected to a rotameter and a compressed gas source of 95 percent 02 and 5 percent C02 to provide Vcoll into the isolated segment while maintaining a physiologic alveolar CO2 tension of 5 percent. 57 I measured Rcoll at several different Ptp. The ventilator was switched off at end-expiration (functional residual capacity), and the lungs were then inflated to the desired Ptp with a variable speed blower attached to the expiratory port of the ventilator. The Vcoll was adjusted to provide a constant Pct-Pao, and Rcoll was calcu- lated as: Rcoll = (Pct-Pao)/Vcoll. The mean of a minimum of two sequential Rcoll measurements taken at each Pct-Pao in all experiments were averaged, and the lungs were inflated to total lung capacity before each measurement to ensure a constant volume history. Three series of experiments, each using a different group of dogs, were performed to determine the effect of regional inhomogeneity on Rcoll. All measurements were taken in the left cranial and right caudal lobes ventilated with room air, and the specific transpulmonary pressures, Pct, and Pct-Pao used are summarized in Table 2-1. In each series, transpulmonary pressure was held constant while Pct-Pao was raised to the values shown in the table. Specific combinations of Pct-Pao and transpulmonary pressures were chosen to allow Pct to remain constant over a range of transpulmonary pressures. This enabled me to determine not only the effect of Pct on Rcoll, but also the effect of Pct-Pao on Rcoll. In series one, the lungs of five dogs were excised. In series two, experiments were performed on the intact lungs of seven closed chest anesthetized dogs. Left cranial lobe data were taken from all seven dogs, and right caudal lobe data were taken from six dogs. In series three, lungs from six dogs were excised, and regional inhomogeneity was created by withdrawing air from the Tab1e 2-1. Protocols for series one, series two, and series three The combinations of transpulmonary pressures O) and segment-lobar pressure differences (Pct-Pao; Numbers in the table indicate segment experiments. (cm H cm H2 8 ) are shown. 58 pressure (Pct; cm HO for series one and Ewo an The Pct- Pao values are positive d negative for series three. Transpulmonary Pressure (cm H20) Pct-Pao Series 1 Series 2 Series 3 (cm H20) 2 3 4 5 5 6 7 8 4 5 6 1 6 9 3 2 6 7 9 10 2 3 3 6 7 8 9 10 11 1 2 3 4 6 7 8 9 9 10 ll 12 l 2 5 7 8 9 10 10 11 12 l 6 8 9 0 11 12 7 9 10 12 8 10 59 segment at a constant flow rate, thus deflating the segment relative to the lobe. Left cranial lobe data were taken from five dogs, and right caudal lobe data were taken from six dogs. In this series only, Rcoll was calculated as (Pao-Pct)/Vcoll. The effect of Pct-Pao on Rcoll in series one and two, Pao-Pct on Rcoll in series three, and transpulmonary pressure on Rcoll in series one-three was analyzed using a randomized complete block design analysis of variance, and significant differences between means were tested using the Student-Newman-Keuls' procedure (18). Means were considered significantly different at the p<=.05 level. Results In both excised and intact dogs' lungs, regional inhomogeneity (i.e., increasing Pct-Pao) increased Rcoll at higher Pao and had little effect on Rcoll at the lowest Pao. Figure 2-1A shows the results of series one experiments in excised lungs. In both the left cranial and right caudal lobes Rcoll is plotted against Pao at five different levels of Pct. At a constant Pct, raising Pao decreased Rcoll. In both lobes, all changes occurring at each Pct as Pao was raised were statistically significant. When Rcoll was plotted as a function of Pct-Pao at several levels of Pao (Figure 2-18), raising Pct-Pao increased Rcoll at Pao = 3, 4, and 5 cm H20 but had little effect on Rcoll at Pao = 2 cm H20. There was a significant interaction between Pct-Pao and Pao. In Figure 2-2A the results from series two experiments (closed chest dogs) are shown where Rcoll is plotted against Ptp at several Figure 2-1A. 60 Effect of transpulmonary pressure (Pao; cm H20; abscissa) on collateral resistance [Rcoll; cm H20/ (ml/sec); ordinate] in left cranial lobes (left panel) and right caudal lobes (right panel) at different levels of segment pressure (Pct; cm H20) in five excised dogs' lungs ventilated with air. cm H20 ml/sec Rcoll LEFT CRANIAL RIGHT CAUDAL LOBE LOBE Ci '3 Pci 8 IO 7 \ 9 6 8 7 6 n=5 n=5 l l g! 23452345 PaocmHZO Figure 2-1A 62 .mc cmxgme «mos» Lee unmuxm ucmgmwwwu zpucouwwwcmmm one :mpmcou m? can ems: mucwoa acmumwuo FP< .mmcap .mmou uwmwuxm m>F$ cw Ao : Eu mommv mesmmmca xgmcoepzamcmcu mo mmspw> acmemwuwu we muecvugo "Aumm\Psv\omz so “FFoumu mucmamwmwe Fogmumppou co Ammmvumnm no : Eu womauauqv oucwgmwmru wgzmmmga LmaoF1ucmEmmm Lo pummwm mcwzozm <~1N wgzmwa av me name msmm .mP-N «Lamwa 63 mP-N «cameo on... E0 com Lon. c. m“ _ mw ~. 1 [I no woo... 4 = Pct -Poo Poo Rcoll ——Vcoll = Pci- Ps Rs Vcoll - = .Ps-Poo R' Vcoll s. opsule it | capsule 4* 2 Figure 3-l 88 between the segmental alveoli and the airway opening (Ps-Pao). Gas flow (Vcoll) was adjusted with a rotameter (Model 743lT, Matheson, East Rutherford, New Jersey) until a desired stady state pressure gradient between the catheter tip and the airway opening (Pct-Pao) was achieved. The following calculations were then made: resistance through the entire segment [Rcoll = (Pct-Pao)/Vcoll], resistance between the catheter tip and the segmental subpleural alveoli [segment resistance, Rs = (Pct-Ps)/Vcoll], and resistance between the segmental subpleural alveoli and the airway opening [interface resistance, Ri = (Ps-Pao)/Vcoll]. The position of the capsule on the segment's pleural surface did not influence Ps, for when two capsules were glued to a distended segment and Vcoll was varied, the two Pct-Ps pressure gradients were the same (see Figure 3-l). Resistances were calculated during inhomogeneous segment inflation to seven different pressures (Pct-Pao = l, 2, 3, 4, 5, 6, and 7 cm H20), during which the lobe was held at each of 5 different transpulmonary pressures (Pao = 2, 3, 4, 5, and 6 cm H20). This was done in the following manner: the lobe was slowly inflated to Pao = 30 cm H20 and then deflated to the desired Pao where it remained for approximately one minute. During this time, Vcoll was adjusted to raise Pct-Pao sequentially in step increments to l, 2, 3, 4, 5, 6, and 7 cm H20, and each Vcoll was maintained for several seconds to obtain steady state conditions. The Pao and Pct-Pao pressure tracings were recorded on a photorecording oscilloscope (Model VRl2, Electronics for Medicine, White Plains, New York), and 89 the Vcoll recorded at each Pct-Pao was used to calculate Rcoll, Rs, and Ri. Measurements were made in duplicate, and the two values were averaged. During all measurements, air was used to ventilate the lobe. This entire protocol was performed three times, each time using a dif- ferent gas to inflate the segment (first air, then helium, then SF6). The physical properties of air, helium, and SF6 are shown in Table 3-l. When using either helium or SF6, the segment was initially flushed with approximately 300 ml of this gas at each Pao before making any measure- ments, ensuring that the concentration of the gas in the segment was not diluted by the room air used to inflate the lobe. Air and helium data were obtained from six dogs and SF6 data from five dogs. 'Table 3-l. Physical properties of gases [helium (He), air, and sulfurhexafluoride (SF6)] used in calculation of collateral resistance, He Air , SF6 u[gm/(sec- cm)] 1.94 x 10'4 1.90 x 10'4 1.33 x 10‘4 p(gm/ml) l.80 x 10"4 1.14 x 10'3 6.60 x 10"3 p = viscosity. p = density. 90 The effect of gas physical properties of Rcoll, Rs, and Ri was initially assessed by comparing these resistances when using the three different gases. The flow regime in the segment was further assessed using a fluid mechanical approach described by Moody (8). This consists of analyzing the pressure-flow relationships in the segment by constructing a diagram (referred to as a Moody diagram, Figure l-6) of the logarithm of the ratio of the static pressure drop to dynamic pressure (i.e., the normalized pressure drop, equation l-lS) against the logarithm of Reynolds' number [equation l-ll, Re = (pdv)/ u]. where AP = Pct-Pao (cm H20), v = Vcoll/A (cm/sec), p is gas density (gm/ml), u is gas viscosity [gm/(sec- cm)], and d and A are the diameter (cm) and cross sectional area (cm2) of the wedged airway. I assumed that the diameter of the wedged airway equalled the diameter of the flared tip catheter (0.4 cm). The Moody analysis assumes that when the geometry of the airways is constant, Pn is solely a function of Re. Varying Re should result in a single curve with a slope of -l if flow is laminar, 0 if flow is turbulent, and between -l and 0 if flow is transitional (see Figure l-5). I assumed that identical segment geometry resulted when the segment was inflated to the same pressure with each of the three gases. The effects of increasing Pct-Pao and Pao on Rcoll, Rs, and Ri, respectively, were analyzed using a randomized complete block analysis of variance (ll). Means were compared using the Student- Newman-Keul's procedure, and differences were considered statistically significant at p< .05. 91 Results Figure 3-2 shows the effect of inhomogeneous segment inflation (raising Pct-Pao at constant Pao) on Rcoll, Rs, and Ri at five different Pao. In general, raising Pct-Pao increased Rcoll, Rs, and Ri measured when air and SF6 were the inflating gas but had no effect on these resistances measured when helium was used. Figure 3-3 uses the same data as in Figure 3-2 to illustrate the effect of Pao on Rcoll, Rs, and Ri at seven different Pct-Pao. Increasing Pao decreased Rcoll, Rs, and Ri measured with each gas at all Pct-Pao. Table 3-2 shows Rs expressed as a fraction of Rcoll (Rs/Rcoll). During helium inflation, Rs/Rcoll was unaffected by increasing Pct-Pao. During air inflation, Rs/Rcoll tended to increase with increases in Pct-Pao, but these changes were statistically insignificant at the p< .05 level. During SF6 inflation, however, Rs/Rcoll increased as Pct-Pao increaSed from l-7 cm H20 at a constant Pao. Raising Pao at a constant Pct-Pao caused no significant changes in Rs/Rcoll during both air and helium inflation but decreased Rs/Rcoll during SF6 inflation. In Figure 3-4, Log Pn is plotted as a function of Log Re (Moody diagram; see Figure l-6). Data points obtained when the segment was inflated with He are shown as circles, air as triangles, and SF6 as squares. Each panel represents data obtained at a constant 2, 3, 4, 5, and 6 cm H20. All helium data lie on a line with a slope of -l indicating laminar flow, all air data lie on a line deviating slightly from a slope of -l indicating transitional flow, and all SF6 data lie on a line with a slope closer to 0 indicating transitional to Figure 3-2. 92 Effect of segment inflation by raising the segment-lobar pressure gradient (Pct-Pao; cm H20; abscissa) on total collateral resistance [Rcoll; cm H20/(ml/secE; ordinate], top panel; intrasegmental airway resistance Rs; cm H20/ (ml/sec); ordinate], middle panel, and intersegmental airway resistance [Ri; cm H20/(ml/sec), ordinate], bottom panel, at five transpulmonary pressures (Pao = 2-6 cm H20). The lobe was inflated with air while the segment was inflated with helium, air, or sulfurhexafluoride (SF6). cmflto mince Rcoll cmwo ml/Ioc Rs CMH‘O nil/soc R1 20 FELILM MR SF. 10< 3 ll 3 . os« ‘ 08‘ "? 07. / Poo °6< /\\_2 / ,0": / 5 051 / / / 5 04‘ ‘ M3 /5 02. / :22: Oh 01 123456? 1233567 1234367 1-Poo cmHgo “um 20‘ He A1! SF5 P2 l5- 2 IO F i ‘8‘ 7< 3 s, 5' 4 4. En 3< 2 2 /5 .4 3 Poo 6 4 H N? X: / OJ Ms; W iéiiSs‘r 1233587 i'zéééé‘r Pct-Poo camp hum Alr SF 204 He 6 |5< l0< I 91 a. 7. . £99 £92 6 2 2 " 6 2. 4 45:53 % cam-:5 ' .l 6 0‘ ié'aiéé’r 3553565 123356? ct-Poo amigo Figure 3-2 Figure 3-3. 94 Same data as Figure 3-2 showing the effect of transpulmonary pressure (Pao; cm H20; abscissa) on total collateral resistance [Rcoll; cm H 20/(ml/sec), ordinate], top panel, intrasegmental airway resistance [Rs; cm H 0/(ml/sec), ordinate], middle panel; and intersegmenta airway resistance [Ri; cm HZO/(ml/sec), ordinate], bottom panel; as the segment-lobar pressure gradient was held constant at seven values (Pct- Pao= l- 7 cm H 20). The lobe was inflated with air while the segment2 was inflated with helium, air, or sulfurhexafluoride (SF6 ). cmH 0 mI/uc Rcdl CMHQ MU SOC m zfi ISI l0< 201 l5< IO< Hehum w—v—v—F-F— 23456 Hdmm Hdmm l~7 q—v—v—vw— 23456 Figure 3-3 95 Au I—v—v—v—v—F—- 23456 Poo 1:an Aw A" q—vww—v— 23456 Hm map SFs 23456 $2 'TSTEF 96 Table 3-2. Intrasegmental airway resistance (Rs) expressed as a fraction of collateral resistance (Rcoll) (Rs/Rcoll, R': SEM) as the segment was inflated to a segment-lobar pressure difference (Pct-Pao) = 1-7 cm H 0 at transpulmonary pressure (Pao) = 2-6 cm H20 in the left caudal lobes of excised dogs' lungs. Pct-Pao Helium Air SF6 Pao = 2 l 18: .09 16: .05 .49: .ll 2 20: .09 16: .04 .59: .11 3 .20: .06 .20: .04 ' .64: .08 4 .18: .05 .20: .04 .64: .10 5 18+ .04 .22: .05 .64: .09 6 18+ .04 23: .05 .65: 09 7 17: .04 24: .06 .66: 09 Pao 3 1 13: 02 20: .04 45: 10 2 14: 02 21:.05 47: 09 3 .15: .02 .23: .05 .50: .09 4 .15: .02 .24: .04 .52: .09 5 .16:.03 .25:.06 .55: .09 6 .16: .03 .25: .06 .56: .09 7 .16: .03 .26: .07 .56: .09 Pao 4 l .17: .05 .19: .04 .42: .08 2 .18: .05 .21 : .04 .46: .08 3 .19: .06 .25: .06 .50: .08 4 .19: .06 .23: .06 .52: .09 5 .20: .05 .24: .06 .54: .09 6 .21:.05 .24:.07 .56: .09 7 .19:.04 .25:.07 .56:.09 Pao 5 l .l3:.03 .ll:.03 .40:.06 2 .12:.02 .l4:.03 .4l:.06 3 .12: .02 .16: .03 .43: .06 4 .l3:.02 .18:.04 .46:.06 5 .13:.02 .18:.04 .47:.06 6 .14: .02 .19: .05 .48: .06 7 .14: .02 .20: .05 .48: .07 Pao 6 1 .ll:.02 .10:.03 .24:.07 2 .10:.01 .l3:.03 .30:.08 3 .10:.01 .l7:.03 .32:.08 4 .10:.01 .18:.04 .34:.09 5 .ll:.01 .20:.04 .36:.09 6 .ll:.01 .21:.04 .39:.10 7 .12:.02 .22:.05 .39:.10 97 .zopm gmcwsmp mcwucmmmgamc .p- u maopm mm; warp cognac amp .8?“ cmumgaau amaze: mg» we omega mg: meowmcmsmv zngwm oucmcm$mg use .Ao So u." u coauuuqv mmapm> cm>mm op acmwumgm mezmmmga canop1pcwEmmm wsu mcpmwwg xn Ammgmacm "mumv muwgoap$mxng=mpzm Lo .Ammpmcmwguv Lvm .Amwpugpuv anpm; guy: umumpmcp mm; acmsmmm mg» mpwzz om: Eu m1~ um ucmumcou vag was Aomav «Lammmca xgmcospaamcmgu mpmcmn m>wm mg» mo comm :H .Am1p mszmwmv Emgmmru xuooz a mu op umgcmmog m? can Ammmwumnm "mm mogv conga: .mupoczmm $0 Egupgmmop esp mo cowuuczm a mo czogm m? Ampmcwugo "cm mogv aocu mesmmmga vaFpoEEo: on» ma Egupgmmop ugh .e1m mgsmpu 98 um 00.. // i. J“// '0“ C b toga/V u. 0 V v. // / v.8“. ,5 ¢1m «gnaw: am 93 r m w I: / C .Q 00 / vow. O a o / u. 0 1. 8.8,. .N at 3.. erl n w O / .1 / / "I l / .0m ’14 / M d. O v. ?2: /%~ um 90.. w n w w _- 'I x I I 100 / YO no: ‘0 b O . / a. O ._ 008‘ vN at 81. I. m m . o - / . / / .1 / .o M 1:11: / H 44 d / fi : O O 0 /.~ 99 turbulent flow. The Re at which the slopes deviated from -1 increased as Pao increased: at Pao = 2 cm H20 this change occurred at Re = 45 (Log Re = 1.65) whereas at Pao = 6 cm H20 the change occurred at Re = 150 (Log Re = 2.18). Data from Figure 3-5 are the same as those in Figures 3-4 with Log Pn plotted as a function of Log Re. Each panel shows data points when the segment was inflated to a constant Pct-Pao with He (circles), air (triangles), and SF6 (squares). Data obtained at each Pao (2, 3, 4, 5, and 6 cm H20) are connected by lines. Laminar and transitional flow were present when Pct-Pao = 1-3 cm H20, but transitional and turbulent flow were preSent at Pct-Pao = 4-7 cm H20. Furthermore, at constant Pct-Pao, raising Pao resulted in multiple curves and not in a single curve as occurred when Pct-Pao was raised at constant Pao (Figure 3—4). Discussion Segment volume was increased in this study by raising either Pao or Pct-Pao. To distinguish between these two methods of raising segment volume, I refer to "segment inflation" as raising Pct-Pao at constant Pao, and "lobar inflation" as raising Pao at a constant Pct- Pao. I have confirmed my previous observation that segment inflation and lobar inflation have opposing effects on Rcoll (Chapter II) (2), and I have attempted to determine the mechanism of this effect. One of my hypotheses is that Rcoll is the sum of the resistances of two populations of airways in series: intrasegmental airway resistance (Rs) contributed by airways lying in the body of the segment and 100 .zopw gnawEmp mcwacmmwgnmg .p- u mnopm was mcpp umsmmo .8?» Lmuwsumo ammumz mg» co mmogu use mcommcmewu zmzcwm mucmgmwmg one .Ammcmsam mommv muwgozpwmxmggawpzm Lo .Ammpmcmwcuv gpm .Ammpucwuv saw—m; saw: umumpwcw mm: ucwEmwm as» can: on; pcmumcou ucmmmgamg mu=_oa mg» mcwpumccou mmcwd .ow: Eu o1~ socm ummmmgucw Aomav mmgzmmoga zgmcoE—samcmgu mpmgz accumcou upon mo: Mom: Eu moma1uumv corumpwcv newsmmm mpmcma cm>mm as» 4o scam :H .Amup mg: wmv Emgmmwv xuooz a mm on umgcmcmg m? ucm Aommpumnm "mm mobv Lmnsac .mvpocxmm we Ecuwgmmop on» we :owuuczw a ma czocm my Amumc_ugo "ca mobv acct mgammmga umNVFmELo: one we Enumgmmop use .m-m mgamw: f? C? N M Q 10 £0 ’ F. >10 . . ’ fl :1 ' 7’ 8 , ’ :~& 3 J‘ / 3 111‘ , ‘ , . / a 4 ' -° lk§501 TV 0. no r r T ,/ ’ I, I /j/Ji ? »/‘/ ”TI." /-N& 8 . a 3 E / // h- // g, .fi v _53 "d 001 81 am: n v 1010 f M T g “f I ’0‘" 8 4; , f g ,4 / 3 o" ,’ <0 ‘5 5’ Na 601 101 [v E‘N ”V 1010 V "f r a '” T 1% ‘9 ,8 ” 1’ n3. ’ z ‘ W“ 2' ‘ / ‘ / / >- / / j/ “r, .' 6 __O Nd 601 AV EIN «w mo P 77' If "" r /' «1 ///{/ ’6? g /{ ,7 _~ 01’ /" / / J E z / // .- / / RI iK' -" ad 601 ft c§ru 10 v rub /F ho : / “.- / 2 // , 5 / / // / - a, .l ‘5 -J’ “001 g. 1" (LIN "1Q 1049 - r F 9 l, "a ~ x 2 // 1‘ ‘. FN é ' a I ' .6 Figure 3-5 102 intersegmental airway resistance (Ri) contributed by airways coursing through the segment-lobar interface. Accordingly, segment inflation dilates intrasegmental airways and decreases Rs while this inflation distorts intersegmental airways and increases Ri. These opposing effects would tend to minimize any change in Rcoll. In this study I used a pleural capsule to partition Rcoll into R5 and Ri, and I made the following assumptions: (1) pressure is uniformly distributed throughout the pleural surface alveoli of the segment (this is supported by the negligible pressure gradient between two capsules glued to dif- ferent points on the segmental pleural surface, Figure 3-1); (2) capsule pressure (Ps) represents the pressure in the most distal intrasegmental airway in which flow occurs; (3) Vcoll is uniformly distributed through- out the segment; and (4) the distribution of Vcoll within the segment remains constant during segment inflation or lobar inflation. Irrespective of the gas used, segment inflation caused the same directional changes in Rcoll, Rs, and Ri, not the opposing changes in Rs and Ri predicted by my hypothesis. Since lobar inflation decreases airway resistance (5, 6), I expected both segment inflation and lobar inflation would dilate intrasegmental airways and decrease Rs. Although Rs was decreased by lobar inflation (Figure 3-3), Rs surprisingly was not decreased by segment inflation (Figure 3-2) even when helium min- imized nonlaminar flow caused by the higher Rcoll. This observation suggests that lobar inflation and segment inflation may have different effects on intrasegmental airway geometry. Bronchograms of a segment in a homogeneously inflated lobe and of a segment inflated nonhomo— geneously demonstrate that intrasegmental airways dilate under both 103 conditions (7, 12). When the lobe is inflated homogeneously (i.e., when lobar distending pressure is uniform throughout the lobe, including the segment), the segment changes shape from a short fat cone to a long thin cone as intrasegmental airways lengthen and branching angles decrease (12). During segment inflation, however, the segment becomes more spherical, and branching angles appear to increase. In the present study, the increased branching angles occurring during segment inflation would tend to promote nonlaminar flow (9) and override any decrease in Rs resulting from airway dilation. In contrast, if the geometrical changes occurring during inflation in the present study are similar to those occurring during homogeneous lobar inflation (7, 12), the decreased branching angles would contribute to the measured decrease in Rs. Examination of the fraction of Rcoll comprised by Rs (Rs/Rcoll, Table 3-2) supports nonlaminar gas flow in intrasegmental airways as one cause of the increase in Rcoll during segment inflation. A com- parison of Rs/Rcoll data during air and He inflation reveals that at each Pao, Rs/Rcoll with air and Rs/Rcoll with helium are similar at low Pct-Pao and progressively diverge as Pct-Pao increases. This is due to the trend of Rs-Rcoll with air to increase during segment inflation, indicating a shift in the distribution of Rcoll from Ri to Rs, whereas Rs/Rcoll with helium remains constant. This correlates with the observation in Figure 3-2 that air and He Rcoll are similar at low Pct-Pao, but Rcoll with air exceeds Rcoll with helium at high Pct-Pao. At low Pct-Pao, Vcoll is low and, Moody plots shown in 104 Figure 3-4 indicate that flow is laminar throughout the segment with both air and He. The Pct-Pao pressure loss is thus viscous dependent, and since the viscosities of air and He are similar, the absolute magnitudes and the distribution of their respective resistances are similar. At high Pct-Pao, however, Vcoll is high, so flow with air is probably nonlaminar, and resistance is density dependent. Thus, Rcoll with air exceeds Rcoll with helium. The greater nonlaminar flow occurring with air increases Rs and therefore increases Rs/Rcoll with air. The increase in Rs as turbulence develops is evident from data when SF6 was used as the inflating gas (Table 3-4). Even at Pct-Pao = 1 cm H20, Rs/Rcoll with SF6 is much greater than Rs/Rcoll with air, and as Pct-Pao increases, Rs/Rcoll with SF6 increases even further. Whereas segment inflation with SF6 increased Rs/Rcoll, lobar inflation with SF6 shifted the distribution of Rcoll from Rs toward Ri. Although vco11 and thus Re increased during both modes of segment inflation, the possible differences in segment geometry suggested earlier (i.e., branching angles increasing during segment inflation but decreasing lobar inflation) may cause increasing nonlaminar flow in the intrasegmental airways during segment inflation, thereby increasing Rs, and decreasing nonlaminar flow during lobar inflation, thereby decreasing Rs. Analysis of data in Table 3-2 reveals that during segment inflation with SF6, Rs exceeds Ri, but during segment inflation with air and helium, Ri greatly exceeds Rs. This indicates that when flow is transitional to turbulent, Rcoll is principally determined by 105 intrasegmental airways, but when flow is laminar to transitional, Rcoll is principally determined by intersegmental airways. I used a Moody analysis (8) to more specifically characterize the flow regime in the sublobar segment. During segment inflation flow was always laminar with helium, transitional with air, and transi- tional or turbulent with SF6 (Figure 3-4). Interestingly, segment inflation resulted in all data points on one smooth curve, indicating that at any given Pao, Pn was solely a function of Re. If segment airway diameter increased during segment inflation, the curve for each consecutively higher Pct-Pao would be shifted in a parallel direction downward and to the right. Since this shift did not occur with segment inflation, several inferences may be made. First, the dimensions of the segment airways may have remained constant during segment inflation, explaining why Rcoll did not change when nonlaminar flow was minimized (see He curves in Figure 3—3). Second, Rcoll may have been determined primarily by a high resistance population of airways whose dimensions were fixed, and secondarily by one or more populations of low resistance airways whose dimensions may have varied during segment inflation. The change in Rcoll resulting from the dimensional changes in this secondary population of airways would not have been measured since their contribu- tion to total Rcoll would have been small. According to the data in Table 3-2 for air and He (where Rs/Rcoll is small), Rs might represent the low resistance airways, and Ri might represent the high resistance airways. 106 In contrast to the data shown in Figure 3-4 obtained during segment inflation (where all points at constant Pao lay on one smooth curve), Figure 3-5 shows that lobar inflation to five different Pao resulted in five separate curves, indicating that Pn was a function of Re and one or more other variables. The curves in Figure 3-5 were constructed assuming airway diameter and cross sectional area were equal to the dimensions of the wedged catheter tip. However, data in Figure 3-3 show that lobar inflation decreases Rcoll, probably by dilation of intra- and intersegmental airways since Ri and Rs also decrease. This airway dilation was not accounted for in the curves in Figure 3-5 and may be the reason why these data do not fall on a single curve. If the spread of these curves is due solely to the unaccounted for changes in segment airway diameter occurring during lobar inflation, then compensation for these diameter changes should result in approximation of these curves into a single smooth curve. If, in contrast, other variables in addition to diameter changes cause the spread of these curves, then no amount of compensation for diameter changes would lead to a single smooth curve. I tried several methods to account for segment airway diameter changes which might occur during lobar inflation. First, I assumed that segment airway diameter changed with the cube root of lobar volume (4). Values of lobar volume and compliance from excised dogs' lungs were obtained from data reported by Frank (1). The diameter of the segment airway at Pao = 30 cm H20 was assumed to be the diameter of the catheter tip (4 mm) since this is the Pao at which the catheter 107 was wedged in the airway. As the lobe was deflated from Pao = 30 cm H20, segment airway diameter was scaled to the cube of lobar volume, and the results of this recalculation are shown in Figure 3-6. Clearly, the data do not approximate a single curve, although closer approximation is evident when compared to the data shown in Figure 3-5. I next assumed that segment airway diameter is scaled to transpulmonary pressure in the manner described by Hughes et a1. (4) for excised dogs' lungs. The results of this recalculation are shown in Figure 3—7. Again, the data do not approximate a single curve, although closer approximation is evident when compared to data in Figure 3-5. Since segment airway diameter does not scale either with the cube root of lobar volume or with Pao during lobar inflation, I then determined if a single power function would relate lobar volume to segment airway diameter during lobar inflation. If this single power function exists, and if Pn is solely a function of Re, then all the data shown in each panel of Figure 3-5 should fall on a single curve. In contrast, if multiple power functions are required to unite these separate curves into one single curve, then one or more variables other than segment airway diameter would be necessary to explain the relationship between Pn and Re. Using the data in Figure 3-5, I adjusted the value of segment airway diameter in such a way as to approximate the curves as closely as possible. I then calculated the power function relating lobar volume to the adjusted diameter. 108 .3epw gecwEew mewueemegeec .p- u eeewm mw mew, eesmeo .Amegeeem "mumv eewceepwexegcewpem we .Amepmeewguv ewe .Amepeewev Eewpe; guwz eeuep :w me: peesmem esp ems: eee weeumeee «cemeseee muewee esp mcwuueeeeu mecwe .o : Eu m1~ n ee ea veneegucw we: Aeeev egemmeee xgeceepeemeeeu epwez weeumeee ewe; we: Ac : Eu weee1peev eewuepwew weesmem .mweeee ce>em may we come cm .m1m eeemwu cw eepegumeppw msegmewe aeee: an» cw euee ce esepe> Leaep we gees ease eu Levesewe zeznwe acmsmem mew—mum we uemwwu .e-m eczewe be 8: m me «no ’ >10 . , / n 74 i s 41’ .. e , .~¢ :13 ‘ _, ‘. v‘ v 0 115 601 )c Elm me 1000 5” .10 N 3 / , r~ .1_ / o" /’ // J; l 6 €- No “:1 601 .c /"° 3 ... +— / / o i 6 -I ‘1' Id 601 Log Rt 109 1v 81m no mu.) m 7' I if />n It.“ / 19 J? l U . W“ E o 5 f o 7.- No 16601 ‘ w ‘9 4 E -1..\ o. -| 4 -2 114 601 My / v , -, ,. g at / Na: 1 (/"( // 3 E / / D- ei . o - O - N Id 601 ' Pei -Poo-7 EIN nv run >' . f [At] / o/‘l / mg / / 3 / h- z - ”a Figure 3-6 l . I1||1I1II (Illa I' 110 .3epw LeewEep mcwueemeceeg .p- u eeepm mw ecwp eesmeo .Amegeeem mommy mewgeerexezgewpem Le .Amewmcewguv ewe .Amepeewev Eewpe; new: team—wew we: aceEmem men can: eee penumeee Hemmegeec muewee esp mewuueecee meewe .ou: Eu o1~ ee eemeeeeew we: eee epwzz penumeee e—e; we: wow: Eu meee1ueev :ewuepwcw aceseem .mpeeee ee>em me» we seem cm .m1m egemwu cw eeaegumeppw maegmewe aeeez ecu cw euee :e Aeeev weemmece ageeespeemcee» e» LeueEewe xezgwe aceEmem mewpeem we uuewwm .~-m eeamwe Pei ~Pm- 3 24 Rmfiwe «'1 w El mnvnw 1n _ f i /1~£ 2 / 3 / b- / / 4/ fi 1 6 .' u4h1 111 )v :1 M.,... M O E g f// //~§ E / / ’_ / / 81 5 -° HM ' W “Co: N mm m I) 1! 4 / /“3 I / / b- / / . / u a -' Iain gm nc no I / / >- / / fi 6 - .4001 Rbfind' 21 N I m C) $- 3 5'1 00- LL. 112 I found that at each Pct-Pao, a range of power functions was required to approximate the curves during lobar inflation (see Appendix B). Perhaps all segment airways do not experience the same proportional dimensional changes during lobar inflation. For example, if segment inflation decreases intersegmental airway diameter due to distortion of these airways at the segment-lobar interface (2), the resultant changes in intersegmental airway diameter would be unaccounted for in these computations. In contrast, Pn may be a function of one or more variables in addition to airway diameter and Re. In conclusion, I have characterized the gas flow regime in a nonhomogeneously inflated sublobar lung segment during a variety of measurement conditions. I have shown that Rcoll is increased by segment inflation and decreased by lobar inflation. Although non- laminar gas flow accounts for the increase in Rcoll during segment inflation, I cannot explain why segment inflation fails to decrease Rcoll in the absence of nonlaminar flow. Perhaps an increase in branching angles of intrasegmental airways or airway distortion at the segment-lobar interface during segment inflation contributes to this paradox. Several investigators have compared Rcoll and airway resistance to estimate if a segment in a healthy lung is ventilated primarily by collateral channels or airways, but one of the inherent methodological problems in these comparisons is that Rcoll is measured during segment inflation while airway resistance is measured during homogeneous lobar inflation. If segment inflation increases Rcoll, then the comparisons 113 will overestimate the difference between Rcoll and airway resistance. Collateral resistance can be estimated under homogeneous conditions (i.e., when Pct-Pao = 0 cm H20) by using data in Figure 3-2 and extrapolating the curves at each Pao to the ordinate. Assuming the segment occupies 5 percent of the lung, I estimated lobar Rcoll during homogeneous lobar inflation. Figure 3-8 compares these values of homogeneous Rcoll to total lung resistance, peripheral airway resistance, and central airway resistance in open chested vagally intact dogs' lungs (5), and to total lung resistance in vagally denervated dogs' lungs (6). Homogeneous Rcoll is substantially higher than the other resistances at all lung volumes and decreases to a greater extent as the lung is inflated, suggesting that a segment in the healthy lung is ventilated primarily by airways rather than collateral channels. Figure 3-8. 114 Comparison of the effect of raising lobar volume [plotted as percent vital capacity (VC); abscissa] on the following five resistances (cm H20/LPS, where LPS==liters per second): homogeneous collateral resistance [Rcoll (homogeneous) obtained by extrapolating the curves at each Pao in the center panel of Figure 3-2 to the ordinate]; total lung resistance (RL) and central (Rc) and peripheral (Rp) airway resistance in vagally intact dogs as reported by Macklem and Mead (5); and RL in vagotomized dogs as reported by Macklem et al. (6). _' .1 ,___—.—_‘-—-‘__——_———— Resistance (cmH20/LPS) 115 25- 20« XRcoll (homogeneous) 0 RL (vogus inioci) A Rc (vogus intact) 0 R1. (vogoiomy) '5' I Rp (vogus inioci) I 0‘ 5 . ()J IF—iI——4l——4l——|| 0 1'0 2'0 3'0 4'0 5'0 6'0 7'0 % VC Figure 3-8 10. LIST OF REFERENCES Chapter III Frank, N. R. A comparison of static volume-pressure relations of excised pulmonary lobes of dogs. J, Appl. Physiol. 18: 274-278, 1963. - Fuller, S. D. and N. E. Robinson. The effect of regional inhomogeneity on collateral airway resistance. (Accepted, J, Appl. Physiol.: Respirat. Environ. Exercise Physiol.) Hilpert, P. Collaterale ventilation Habilitations-schrift, Ans. der Medizinischen (Ph.D. Thesis). Tubingen, W. Germany: Universitatsklinik, 1970. Hughes, J. M. B., F. G. Hoppin, Jr., and J. Mead. Effect of lung inflation on bronchial length and diameter in excised lungs. J, Appl. Physiol. 32: 25-35, 1972. Macklem, P. T. and J. Mead. Resistance of central and peripheral airways measured by a retrograde catheter. J, Appl. Physiol. 22: 395-401, 1967. Macklem, P. T., A. J. Woolcock, J. C. Hogg, J. A. Nadel, and N. J. Wilson. Partitional of pulmonary resistance in the dog. J, Appl. Physiol. 26: 798-805. 1969. Menkes, H., G. Gamsu, R. Schroter, and P. T. Macklem. Inter- dependence of lung units in isolated dog lungs. J, Appl. Physiol. 32: 675-680, 1972. Moody, L. F. Friction factors for pipe flow. Trans. ASME 66: 671-678, 1944. Olson, D. E., G. A. Dart, and G. F. Filley. Pressure drop and fluid flow regime of air inspired into the human lung. J, Appl. Physiol. 28: 482-494, 1970. Sasaki, H. T. Takishima, and M. Nakamura. Collateral resistance at alveolar level in excised dog lungs. J, Appl. Ph siol.: Respirat. Environ. Exercise Physiol. 48: 982-990, 80. 116 117 11. Steel, R. G. D. and J. H. Torrie. Principles and Procedures 9f Statistics. New York: McGraw-Hill, 1960. Pp. 110-111. 12. Sylvester, J. T., H. A. Menkes, and F. Stitik. Lung volume and interdependence in the pig. J. Appl. Physiol. 38: 395-401, 1975. CHAPTER IV PATHWAYS CONNECTING OBSTRUCTED AND NONOBSTRUCTED SUBLOBAR REGIONS IN THE DOG LUNG Introduction Inhomogeneous inflation of a sublobar lung segment within a lobe likely creates an interface of tissue distortion between the segment and the lobe. Airways penetrating the interface may also be distorted and contribute to the increased collateral airway resistance (Rcoll) measured during inhomogeneity (3). Little is known about the arrangement of airways at the interface, however, and how this arrangement not only influences measurements of Rcoll but also determines the flow route followed by gas leaving an obstructed segment. A common diagram of the lung suggests that gas leaving an obstructed segment flows through collateral airways and is distributed diffusely throughout the remainder of the lobe (6). An unpublished but common observation does not support this model, however. I have observed that when gas containing no 02 (e.g., sulfurhexafluoride, SFG) flows into a catheter wedged in a subsegmental bronchus of an excised lung lobe, the obstructed segment's pleural surface, previously pink in color, turns blue while the lobar surface remains pink; substitution of 100 percent 02 for SF6 turns the segment the same homogeneous pink 118 119 color as in the lobe. Although this color change is likely do to desaturation and then oxygenation of hemoglobin in segmental capillary blood, this observation suggests that most of the gas leaving an obstructed segment does not enter the remaining lobe, requiring a quite different arrangement of collateral airways than in the model of Menkes and Traystman (6). Do collateral airways provide several different routes for collateral flow, or are there several types of airways in the interface which provide alternate routes for collateral flow? The purpose of this study is to define the arrangement of airways in the segment-lobar interface and the resulting flow route followed by gas exiting an obstructed sublobar lung segment. Methods Freshly excised left caudal lobes were obtained from mixed- breed dogs. The anatomy of the airways in the segment-lobar interface was examined in two series of experiments by making corrosion casts of individual (series one) and several adjacent (series two) sublobar segments. In a third series of studies, I measured the distribution of gas flowing out of an obstructed segment to estimate if gas follows the same flow route suggested by analysis of the resin casts. Series One Corrosion casts were made of individual subsegmental and bronchopulmonary segments. In five lobes a catheter with a flared tip (5 mm o.d.) was wedged in a subsegmental bronchus to isolate a subsegmental segment, and in five additional lobes a catheter was 120 ligated in a segmental bronchus to isolate a bronchopulmonary segment. The lobes were filled with water to reduce the occurrence of gas bubbles in the casts, following which they were laid horizontally in a water bath to minimize airway distortion. Segment casts were prepared by slowly injecting Batson's solution (Polysciences, Warrington, Pennsylvania) into the segment catheters. After the compound hardened (approximately one hour), lung tissue was corroded away by submerging the specimens in concentrated potassium hydroxide for four to five days. Series Two Corrosion casts were made of adjacent subsegmental and adjacent bronchopulmonary segments. Catheters were wedged in three to four adjacent subsegmental segments in five lobes and ligated in three to four adjacent bronchopulmonary segments in an additional five lobes. All the injections occurred as the lobe lay horizontally in a water bath. Batson's solution was slowly injected simultaneously into the catheters wedged in each segment, and each segment contained a differ- ent color. The injection was discontinued when solution of mixed colors rapidly poured out of the lobar bronchus. Vinylite solution (Wards Natural Science Establishment, Rochester, New York) was sim- ilarly injected into the catheters ligated in bronchopulmonary segments. Following injection, the resins hardened, lung tissue was corroded away, and the parenchymal interface separating adjacent segments was dissected with the aid of a binocular dissecting microscope. Casts of acini were photographed by scanning photomacrography (2). 121 In pilot experiments, I noted that the interface separating adjacent subsegmental segments not only contained parenchymal tissue from each segment, but also contained large airways coursing between the segments. In contrast, the interface separating adjacent broncho- pulmonary segments contained only parenchyma and did not contain large airways. I also found I could best demonstrate large airways using Batson's solution and parenchymal structures using Vinylite solution. Therefore, I used Batson's solution to inject adjacent subsegmental segments and Vinylite solution to inject adjacent bronchopulmonary segments. Series Three: Distribution of Gas LeaVingJan Obstructed Segment To estimate if gas follows a similar flow route suggested by the resin casts, I measured the distribution of gas exiting an obstructed segment. Excised left caudal lobes were obtained from 6 mixed breed dogs. Catheters (3.17 mm i.d.) were secured in the following segmental bronchi: superior, posterior, lateral, and the common bronchus supplying the medial and anterior segmental bronchi. These catheters completely obstructed the entire lobe, such that any gas entering or leaving the lobe did so through one or more of the four catheters. All catheters were of equal length and diameter. The lobe was suspended by the lobar bronchus in an airtight box, and the four catheters were attached to ports leading to the outside, each of which in turn were connected to port #1 of a 3-way connector (see Figure 4-1). Port #3 of the 3-way connector supplying one of the 122 .egesemeEue e» eeee we: nee m>wuwmee omen: ece egemmege xee eeeemeee eee e>wuemec ewes: eeeeemcegu eeemmece Pewueegewwwe e new: easememe we: wow: Eu neuev easemege zgeeespeemeeew .xee mg“ cw “Lee cuxwm e e» emceepue emceepe Eeeee> eewwecucee1ueumeege e ew> eeuepwcw me: weep enw .eeaeszeww ecu eugeem ewe eemmecesee e ea emceeuue me: eeuueccee seesaw ecu ece .mmee :ewueeppee new e» eecueuue ego: mxeueeceee es» we emcee .Lepeeeeeu aezum e ea eugeeape we: «gee some new .xee on» eewmuee e» meweeep mueee e» eezeeaue ecu: meeuecuee geew esp we meee Peswxege esp .A<\zv gewgeuee1peweee ece .Aev pageaep .Aev gewgeumee .Amv gewweeem "mueesmem zgeeeepeeezueege mewze—Pew ecu mcwxpeeem wgeeege FeueeEuem cw vegeumm wee: mgeuezuee ewe eeeepu .xee unawagwe :e cw Aczegm ueev meseeege Leeep an eeeeeemem weep peeeee awep eemwexe we segmewe ewaesesem .F-e eezewe 123 F-e eesewe g kgmowxm mwzgo 2:304) E4 owwmwmazoo WM mmemeemm Aha $5.233... é 124 segments (henceforth called the "inflow segment") was attached to a compressed air source via a flowmeter, and port #3 of the connectors supplying the three remaining segments (henceforth called the "outflow segments") was attached to a 3.0 liter plastic bag for collection of air. The box was also attached to a vacuum cleaner and rheostat which allowed controlled reduction in box pressure for lobar inflation. Transpulmonary pressure (Ptp) was measured with a differential pressure transducer (PM131, Statham, Hato Rey, Puerto Rico) whose negative side measured box pressure and whose positive side was open to atmosphere. The pressure tracing was displayed on a photorecording oscilloscope (Electronics for Medicine, Model DR8, White Plains, New York). Port #3 of all 3-way connectors was closed while port #2 remained open to atmosphere. The lobe was then slowly inflated to total lobe capacity (Ptp = 30 cm H20) and deflated to Ptp = 0 cm H20 twice to remove atelectasis and provide a constant volume history. The lobe was fully inflated a third time and deflated to Ptp = 0 cm H20. This pressure was held constant while timed collections of gas were obtained. Port #2 of all 3-way connectors was closed, and port #3 was opened. Compressed air was delivered at a constant rate into the inflow segment. Air flowed out of the inflow segment and into the outflow segments to enter the collection bags. Flow was stopped after one minute, and the air volume in each collection bag was measured using a wedge spirometer (Med-Science Electronics, Inc., Model 570, St. Louis, Missouri) attached to an amplifier calibrated for volume at ambient temperature. These timed collections were performed using 125 inflow rates of one and two liters per minute and at Ptp of O and 5 cm H20 to determine if inflow rate and Ptp influenced the distribution of outflow. Between each timed collection the lobe was reinflated to repeat the volume history as described above. This entire protocol was repeated four times, each time using a different segment for inflow while outflow was collected from the 3 remaining segments. The effect of inflow segment, inflow rate, and Ptp on the distribution of outflow was analyzed using a 3x2x2 factorial design analysis of variance (9). Differences between means of simple effects were tested with the Student-Newman-Keul's procedure and were considered significant at p< .05. Results Series One Figure 4-2A illustrates the cast resulting when Batson's solution was injected into an individual bronchopulmonary segment. Note that the parenchymal surface is overlayed by a large airway branching into several generations, each sending smaller branches into the segment. Casts of individual subsegmental segments were similar in appearance (Figure 4-2B). When segments were located in the center region of the lobe rather than near a pleural surface, the segments were bounded by two parenchymal surfaces, each of which was overlayed by large airways with branches going into the segment. 126 .mpeeEmem we» eucw mezeeeee eeem sewn: eee meeewgem Feexceceeee .mueesmem esp aepge>e sews: maezwwe emwew e» newee mxewg< .Aeuwzz mmN1< egeowmv “seamen Feueesmemeem use wee; m<~1¢ exemwev peesmem xgeeesweeegecege Feeew>weew we umee :ewmegeeu .~-e eeaewe 127 mN-e eeemwe mm msm: msmsw .musmEmmm mews use ums seen eusw Am:esse FweEm ms» :e umueewuswv mmsesese mmuw>ese sews: use meewsmusw mewe1ums ms» e» peppesee use swspw: mswmseem .xe:swe :muewsmuswe se Esme H sews: .xe:swe mesep e ea muswee :esse mesep msw .Amm1u msemwmv mze:swe mswstmuse ms» mmeexm e» umuemmmwu appewusee me: musmsmmm mepe use ums ms» mswpeseemm meewsmusw msw .musmEmmm ms» mswueseemm meewsmusw umswwmu1P—m: mg“ on uswee m:ess< .mmsmEmmm Pepsmsmmmeem «smeewue mmssp we umeu sewmesseu .mm1u mssuwm .Fe mews saw: maeuwmwusmusw e» meewsmusw ms» mmese A:essev mmpewsesese aseaeswemms um: Fesewuwuue e3» .msemww ms» we smusmu mse sH .wpem>_e mepe me mueswssmu use usmEmmm mepe ms» smusm e» meewsmusw ms» mmmmese A:essev mpewsesese ageueswemms e .uwmw sm:ep ms» s“ .ue1e msemws Eesw smewumem msem msu we :mw> usmsmwwwu < .wwem>we use mze:swe mepe sew: mueawmwusmusw ea msmsemm mepe mswseesowms ms» epsw mmseee sews: Am:essev mmsesese m—ewsesese aseueswemms mmuw>ese usmsmmm ums e Eesw m—ewsusege aseueswemms uses ums < .emee sewmessee smsuese Eesw meewsmusw msu we sewusee < .mpewsesese mepe ms» xe umwpeeem mmseueesum sepwswm sew: mueewmwusmusw mpewsesese asepeswemms ums msu xe umwweeem wpem>we use mxe:swe msw .mm_ewsesege xseueswemms mepe use A:essev ums eusw mmuw>wu smsusew sews: mpewsesese aseueswemms mepe e eusw mmsesese mpewsusese ageueswemmsses mepn e uwmp memsuxm ms» u< .Amu1u msemwmv mze:swe mswxpsmuse ms» Fem>ms e» umuemmmwu me: meew 1smusw mso .musmEmmm mse mswueseemm meewsmusw umswwmu1ppm: ea uswee m:ess< .musmsemm ageeespeeesesese usmeenue mmssu we umee sewmesseu .ee-e meaews .ou1u mssmwm .ee-e meaeem . 1.5.9. .x. IOO-I Figure 4-5 137 mm unwmmmm “WP? am I. mm mm w;r .e. m .. mm mm Hem .0... ... an e s e... m nm n.0,“. mm R m r... We a 1 . won ea mm wee V////////.u w dl I d d I d J S mxmmmmeewmmo m§340> 1_<._.O._. om Figure 4-5 138 Egg??? Egg??? * * 7//////////////////////////////////.///////////////% No d1fferenoe between SUPERIOR OPOSTERIOR 2? Low High PTP'O E3 PTP'5 xi 1 meme»: m_231_0> 44.—.9. .x. IOO END: 80-4 Low High High Flow Low Low High 0 0 Flow SUPERIOR POSTERIOR LATERAL MED/ANT Flow Flow Figure 4-5 139 more uniformly to the outflow segments; at the higher Ptp, less volume was collected from the outflow segment nearest the inflow segment while more volume was collected from the more distant outflow segments. There was no effect of inflow rate on the distribution of outflow. Further correlation of the similarity of the routes followed by gas and resin was obtained. A catheter was wedged in a subsegmental bronchus of several caudal lobes of excised dog lungs. The isolated segments were inflated with air, and their pleural borders were out- lined with a marking pen. Resin was then injected into the wedged catheters as the lobe floated horizontally in a water bath, and the segments distended up to, but not beyond, the pleural borders outlined when the segments were gas filled. The resulting casts contained large airways overlaying the parenchymal surfaces which I believe likely served as exit routes for both gas and resin. Discussion This study was undertaken to define the arrangement of airways in the segment-lobar interface and the route followed by gas leaving an obstructed segment in the dog lung. Corrosion casts of single and adjacent segments demonstrate three possible routes. The first route is via the bronchus which overlays an individual segment's parenchymal surface and which sends branches into the segment (Figures 4-2 and 4-3). I call this bronchus an "interface airway" because it is found in the segment-lobar parenchymal interface. I propose the model in Figure 4-6 which demonstrates a possible course of resin flow through the indi- vidual segment and interface airway. Resin from the wedged catheter Figure 4-6. 140 Model illustrating possible course of resin or gas flow through an obstructed segment and interface airway. A, Resin enters the segment through a catheter wedged in a subsegmental bronchus and initially fills all the airways and alveoli originating from the bronchus (shaded area). B, Resin continues to flow through collateral channels and into parenchymal tissue supplied by adjacent bronchi. This expands the segment's borders beyond their anatomic limits, and the expansion ceases when the nearest large airway (i.e., interface airway) is reached (shaded area). The resin enters the interface airway and flows retrograde to directly leave the lobe rather than flow across the interface through smaller, higher resistance branches to enter the lobe. 141 Figure 4-6 142 initially fills all the airways and alveoli originating from the bronchus containing the wedged catheter (shaded area, Figure 4-6A). Resin continues to flow through collateral channels and into paren- chymal tissue supplied by adjacent bronchi. This expands the segment's borders beyond their anatomic limits, and the expansion ceases when the nearest large airway (i.e., interface airway) is reached (shaded area, Figure 4-6B). The resin enters this interface airway and flows retro- grade to directly leave the lobe rather than flow across the interface airway through smaller, higher resistance branches to enter the lobe; these smaller branches form the second potential flow route. This model explains why after the segment filled, resin rapidly poured out of the lobar bronchus while the remaining lobe received no resin. In casts of individual subsegmental segments, the interface airway is either a branch of the bronchus containing the wedged catheter or a branch of a neighboring bronchus. In casts of individual broncho- pulmonary segments, the interface airway is always a branch of a neighboring segmental bronchus. The size of the segment formed with a wedged catheter is therefore not determined by the anatomic limits of the segment but by the location of the nearest interface airway. The third possible route followed by gas leaving an obstructed segment is likely between segment and lobar acini which interdigitate with one another at the interface. These interdigitating acini were clearly observed in the interface of Vinylite casts of adjacent bronchopulmonary segments (Figure 4-4). Examination of the interface reveals that at certain discrete points along the interface, respiratory 143 bronchioles from one segment penetrate the interface, enter the adjacent segment, and divide into alveoli and alveolar ducts [a pattern also described in the human lung (4, 7)]. These alveolar ducts may be connected to one another via collateral channels which I define as channels connecting acini originating from two separate airways. Potential collateral pathways in the dog lung are the inter- alveolar pore and interacinar duct (1, 4, 8). I failed to demonstrate the connections between interdigitating acini because casts of small groups of acini were too fragile for dissection. It is possible that, similar to resin flow, most of the gas flowing out of an obstructed segment follows the large interface airway to directly exit the lobe rather than flowing through collateral chan- nels to enter the lobe. This is suggested by the data in Figure 4-5 which illustrates that most of the gas exiting an obstructed broncho- pulmonary segment was collected from the adjacent segments and that little gas was collected from the more distant segments. Furthermore, when a segment is inflated, a well defined boundary between the segment and the lobe is seen on the lobar pleural surface, and examination of the interface shows that the segment's boundaries expand up to, but not beyond, the nearest interface airway. If collateral channels were the major flow route, one would not expect the segment to expand beyond its anatomic limits (which it must do to reach the interface airway), and it would be a surprising coincidence for the interface airways to always be found directly beneath the segment's pleural boundary. 144 I recently reported that inhomogeneous inflation of a subsegmental segment with air increases Rcoll (3), and I proposed that distortion of airways at the segment-lobar interface may contribute to this effect. The model in Figure 4-6 supports this concept of airway distortion. Such distortion may occur because the interface airway, being partially contained in the remaining lobe, would be held fixed as its branches move freely during segment inflation. Gas exiting the segment through the interface airway branches and interface airway would encounter a tortuous pathway with increased resistance, thus contributing to the increased Rcoll. Also, a scanning electron micrograph (Figure 1-2) show parenchymal tissue distortion, probably resulting from the inflated segment bulging into the lobe while at the same time the interface is lengthened due to segment expansion. This would cause severe distortion of collateral channels and increase resistance to flow, contributing to increased Rcoll. LIST OF REFERENCES Chapter IV Boyden, E. A. The structure of the pulmonary acinus in a child of six years and eight months. Am, J, Anat. 132: 275-300, 1971. Dale, 0. Scanning photomacrography. Functional Photo. May/June 18-21, 1982. Fuller, S. D. and N. E. Robinson. The effect of regional inhomogeneity on collateral airway resistance. (Accepted, J, App], Physiol.: Respirat. Environ. Exercise Physiol., 1984. Henderson, R., K. Horsfield, and G. Cumming. Intersegmental collateraal ventilation in the human lung. Resp. Physiol. 6: 128-134, 1968/1969. Lambert, M. Accessory bronchiole-alveolar communications. J, Path. Bact. 70: 311-314, 1955. Menkes, H. and R. J. Traystman. State of the art. Collateral Ventilation. Am, Rev. Resp. Dis. 116: 287-309, 1977. Pumo, K. K. The morphology of the finer branches of the bronchial tree of the human lung. Dis. Chest. 46: 379-398, 1964. Raskin, S. P. and P. G. Herman. Interacinar pathways in the human lungs. Am, Rev. Resp. Dis. 111: 489-495, 1975. Steel, R. G. D. and J. H. Torrie. Principles and Procedures pf Statistics. New York: McGraw-Hill, 1960, pp. 194-211. 145 CHAPTER V DISCUSSION AND SUMMARY The findings in Chapter II that segment inflation increases collateral resistance (Rcoll) while lobar inflation decreases Rcoll (Figures 2-1, 2-2, 2-3) appear conflicting. I expected Rcoll to decrease during both modes of inflation because both segment and lobar inflation increase segment volume, and because airway resistance is inversely proportional to lung volume (2). To explain the effect of segment inflation on Rcoll, I proposed two hypotheses: (1) segment inflation oppositely affects two populations of airways, such that the resistance through intrasegmental airways (Rs) decreases while the resistance through the intersegmental airways (Ri) increases, the net effect being determined by whether Rs or Ri changes more; (2) the high collateral flow rates resulting from segment inflation cause nonlaminar gas flow which increases Rcoll. These hypotheses were tested in Chapter III where Rcoll was partitioned into its Rs and Ri components. Segment inflation with air increased Rcoll, Rs, and Ri and did not cause the opposing changes in Rs and Ri as pre- dicted by the first hypothesis (Figure 3-2). In addition, the increase in Rcoll, Rs, and Ri during segment inflation with air was accentuated by segment inflation with SF6 and eliminated by segment inflation with helium, suggesting that nonlaminar flow occurred in the segment. 146 147 I concluded, for several reasons, that most of the nonlaminar flow occurred in intrasegmental airways. Firstly, Rs exceeded Ri during segment inflation with SF6 (where, according to the Moody plots in Figure 3-4, flow was transitional to turbulent), but Ri exceeded Rs during segment inflation with helium (where flow was laminar) (Figure 3-2). Secondly, the fraction of Rcoll attributable to RS (Rs/Rcoll) increased as nonlaminar flow increased (by segment inflation with SF6), but Rs/Rcoll did not change during laminar flow when the segment was inflated with helium (Table 3-2). Moody plots were constructed to more specifically characterize the flow regime in the segment. During segment inflation, the highest Reynolds' number (Re) at which laminar flow occurred ranged from 45 (at Pao = 2 cm H20) to 150 (at Pao = 6 cm H20) (Figure 3-4), suggesting that nonlaminar flow may occur in an obstructed segment during measure- ments of Rcoll at Re far below the value of 2000 required for such flow in straight smooth tubes. In addition, the Moody plots for segment inflation (Figure 3-4) differed from those for lobar inflation (Fig- ure 3-5). During segment inflation, all points in each plot lay on one smooth curve, so that Pn was solely a function of Re. Since an assumption of the Moody analysis is that airway dimensions are fixed as Pn and Re are varied, this finding indicates either that the dimen- sions of the segmental airways did not change during segment inflation. or that Rcoll may be determined primarily by one or more populations of high resistance airways whose dimensions are fixed and secondarily by one or more populations of low resistance airways whose dimensions 148 any vary during segment inflation. In contrast, lobar inflation to five different transpulmonary pressures (Pao) resulted in five separate curves, indicating that Pn was a function of Re plus one or more other variables. One of these variables is likely airway dimension since segment airways probably dilate during lobar inflation (references 5, 6, 7, and 12, Chapter 13). If this change in airway dimension could be compensated for in the Moody plots, the five separate curves in each plot should approximate a single curve. I attempted to compensate for the changes in airway diameter by scaling wedged airway diameter to the cube root for lobar volume and also to transpulmonary pressure as reported by Hughes et a1. (1). The resultant compensated curves (Figures 3-6 and 3-7), although somewhat less separated than the noncompensated curves (Figure 3-5), still were not superimposed as in Figure 3-4. I then adjusted the value of segment airway diameter as needed to force the curves to approximate as closely as possible a single curve, and I calculated a power function relating lobar volume to diameter. In each Moody plot, a range of power functions was required to approximate the curves obtained during lobar inflation (Tab1e A-l). Two conclusions may be made from the latter observation: (1) changes in wedged airway diameter may not reflect the dimensional changes in other intra- or intersegmental airways during lobar infla- tion. For example, if intersegmental airways are distorted by segment inflation, the resultant changes in intersegmental airways diameter would be unaccounted for in my computations. (2) Other variables in addition to airway diameter may contribute to the separation of the curves . 149 ‘The arrangement of the airways in the segment-lobar interface was examined in Chapter IV. I identified two types of airways in the interface: (1) large airways (termed interface airways) coursing within and parallel to the interface and providing branches to both the segment and the lobe (Figure 4-38); and (2) small airways (res- piratory bronchioles and/or alveolar ducts) from the segment and the lobe interdigitating with one another where segment and lobar paren- chyma abut (Figures 4-4B, 4-4C, and 4-40). I suggested that segment and lobar acini may be connected by collateral pathways at points of interdigitation, but I was unable to demonstrate such pathways in the resin casts. 0n the basis of these findings, there are three possible flow routes for gas exiting an obstructed segment: (1) through the interface airway to flow directly out of the lobe; (2) through the interface airway and then interface airway branches to enter the lobe; (3) from segment acini to lobar acini at points of interdigitation to enter the remaining lobe. I hypothesized that gas flowed primarily through the interface airway to directly exit the lobe rather than flowing through the higher resistance interface airway branches or respiratory bronchioles to enter the lobe. To test this hypothesis, experiments were performed in which gas flowing into one bronchopul- monary segment was collected from the three remaining bronchopulmonary segments, and the volumes were measured to determine the distribution of gas flow exiting an obstructed segment. The majority of gas flow was distributed to the adjacent segments, and only a small amount was collected from the more distant segments (Figure 4-5), suggesting that the interface airway may have served as the primary flow route. 150 Distortion of the airways at the interface may be partly responsible for the failure of segment inflation to decrease Rcoll in the absence of nonlaminar gas flow. Such distortion may occur because the interface airway, being partially contained in the remaining lobe, would be held fixed as its branches move freely during segment infla- tion. Gas exiting the segment first through the interface airway branches and then through the interface airway (Figure 4-6) would encounter a tortuous pathway which may increase resistance and promote nonlaminar flow. Furthermore, Figure 1-2 indicates there is parenchymal distortion at the interface which likely distorts the interdigitating respiratory bronchioles and collateral channels, causing an increase in their resistance. To relate this distortion to increased flow resistance, studies are required which quantify the geometrical changes undergone by the airways in the interface during segment inflation. LIST OF REFERENCES Chapter V 1. Hughes, J. M. B., F. G. Hoppin, Jr., and J. Mead. Effect of lung inflation on bronchial length and diameter in excised lungs. J, Appl. Physiol. 32: 25-35, 1972. 2. Macklem, P. T., A. J. Woolcock, J. C. H099, J. A. Nadel, and N. J. Wilson. Partitioning of pulmonary resistance in the dog. J, Appl. Physiol. 26: 798-805, 1969. 151 CHAPTER VI CONCLUSIONS Collateral resistance is lower in caudal lobes than in cranial lobes. Collateral resistance is greater in intact than in excised lungs, and the difference is greater in the left cranial lobe than in the right caudal lobe. Collateral resistance is influenced more by transpulmonary pressure than by segment pressure; ie.e., collateral resistance is influ- enced more by the lobe than by the segment. Collateral resistance is increased by segment inflation with air and decreased by lobar inflation. Collateral resistance is increased more when segment-lobar inhomogeneity is created by segment deflation than by segment inflation. The increase in collateral resistance during segment inflation with air is due to nonlaminar gas flow, but the reason why segment inflation fails to decrease collateral resistance is undetermined. The hypothesis that segment inflation affects collateral resistance via directionally opposite changes in intrasegmental and interseg- mental airway resistance is not supported. 152 10. ll. 12. 153 The major site of collateral resistance measured during segment inflation is intersegmental airways when collateral flow is laminar and intrasegmental airways when flow is turbulent. The critical Reynolds' number for nonlaminar gas flow in a sublobar segment in excised dogs' lungs during segment inflation is 45 at Pao = 2 cm H20 and 150 at Pao = 6 cm H20, far below the value of 2000 required for nonlaminar flow in straight smooth tubes. Moody plots for segment inflation form a single smooth curve at each Pao, indicating that Pn is solely a function of Re. This suggests that: a. segment airways do not change dimension during segment inflation, or b. collateral resistance may be determined primarily by one or more populations of high resistance airways whose resistance is fixed during segment inflation and secondarily by one or more populations of low resistance airways whose dimensions may vary during segment inflation. Moody plots for lobar inflation form separate curves at each Pct-Pao, indicating that Pn is a function of Re plus one or more other variables. One of these variables is probably airway dimension. During lobar inflation at a constant positive Pct-Pao in excised dogs' lungs, segment airway diameter is not scaled to the cube root of lobar volume or to transpulmonary pressure. 13. 14. 15. 16. 17. 154 Collateral resistance in excised dogs' lungs measured during homogeneous conditions (i.e., Pct-Pao = 0 cm H20) is substantially higher than total lung resistance, central airway resistance, or peripheral airway resistance in intact lungs and decreases to a greater extent than these other resistances as the lung is inflated. When a segment is inflated by flowing gas into the segment through a wedged catheter, the segment expands beyond its anatomic limits until it reaches the nearest large airway (termed an interface airway). ’ There are three possible flow routes through which gas may exit an inflated segment: a. through the interface airway to directly leave the lobe. b. through the interface airway and then through interface airway branches to enter the lobe. c. from segment acini to lobar acini through connections (likely collateral channels) at points of interdigitation. Most of the gas exiting an obstructed bronchopulmonary segment can be collected from the adjacent bronchopulmonary segment(s), but only a small portion of the gas reaches the moe distant segments. Gas exiting an obstructed sublobar lung segment is distributed more uniformly throughout the lobe at higher transpulmonary pressures. APPENDICES APPENDIX A MEAN (: SEM) VALUES OF COLLATERAL RESISTANCE FOR CHAPTER II DATA Table A-1. APPENDIX A MEAN (: SEM) VALUES OF COLLATERAL RESISTANCE FOR CHAPTER II DATA Data points used in Figures 2-1, 2-2, and 2-3 in Chapter II showing collateral resistance [Rcoll; cm H 0/(ml/sec); 2': SEM] in left cranial and right caudal dogs), and series three (excised lungs with the segment deflated relative to the lobe). obes in series one (excised lungs with the segment inflated relative to the lobe), series two (lungs of closed chest anesthetized Series One Pct-Pao (cm H20) Pao (cm H20) 1 2 3 4 5 6 7 8 Left Cranial Lobe 2 .403 .422 .423 .438 .440 1.114 1.114 1.108 :.120 :.120 3 .336 .343 .350 .372 .386 :.096 :.096 :.102 :.108 :.114 4 .268 .292 .312 .330 .342 :.078 :.084 :.090 :.096 :.102 5 .209 .235 .262 .289 .304 :.060 :.066 :.078 :.084 :.096 Right Caudal Lobe 2 .366 .378 .385 .392 .377 :.090 :.090 :.090 :.084 :.084 3 .281 .296 .309 .299 .322 :.084 :.078 :.078 :.078 :.084 4 .226 .242 .259 .263 .270 :.072 :.066 :.072 :.072 :.078 5 .175 .193 .212 .238 .247 :.066 :.066 :.072 :.072 :.072 155 156 Table A-l--Continued Series Two Pct-Pao (cm H20) Pao (cm H20) 1 2 3 4 5 6 7 Left Cranial Lobe 5 .815 .812 .848 .773 i.174 :.174 :.186 :.150 6 .581 .617 .631 .632 :.114 :.126 :.132 :.126 7 .422 .449 .468 .488 :.108 :.114 :.120 :.132 8 .349 .406 .418 .484 :.090 :.108 :.102 :.108 Right Caudal Lobe 5 .371 .364 .416 .333 :.186 :.174 :.156 :.132 6 .247 .286 .296 .296 :.072 :.084 :.084 :.084 7 .189 .209 .230 .242 :.054 :.048 :.054 :.060 8 .140 .160 .185 .211 :.036 :.036 :.042 :.060 157 Table A-1--Continued Series Three Pct-Pao (cm H20) Pao (cm H20) 1 2 3 4 5 Left Cranial Lobe .250 .435 .581 :.108 :.210 :.270 .299 .437 .507 :.120 :.186 :.222 .350 .430 .521 :.138 :.174 :.210 Right Caudal Lobe .500 .662 .845 :.300 :.354 :.426 .409 .535 .633 :.162 :.204 :.246 .372 .412 .484 :.156 :.168 :.204 Collateral experiments. Pct = segment pressure (cm H20). Pao = transpulmonary pressure (cm H (series one and three). Ptp = transpulmonary pressure (cm H anesthetized dogs (series two). Pct-Pao one and two. three. Pao-Pct resistance (R': SEM) from series one, two, and three 20) in excised dogs' lungs 20) in lungs of closed chest segment-lobar pressure difference (cm H20) in series segment-lobar pressure difference (cm H20) in series APPENDIX 8 RELATIONSHIP OF SEGMENT AIRWAY DIAMETER T0 LOBAR VOLUME DURING LOBAR INFLATION FOR CHAPTER III DATA APPENDIX 8 RELATIONSHIP OF SEGMENT AIRWAY DIAMETER TO LOBAR VOLUME DURING LOBAR INFLATION FOR CHAPTER III DATA If the separation of the curves shown in Figure 3-5 is due solely to the changes in segment airway diameter occurring during lobar inflation, then compensation for these diameter changes should lead to a singular relationship between Pn and Re. Although diameter is reported to be related to the cube root of lobar volume or to trans- pulmonary pressure during homogeneous lobar inflation (reference 4, Chapter III), data in Figures 3-6 and 3-7 indicate that such a rela- tionship may not occur during lobar inflation (raising Pao at a constant positive Pct-Pao). I determined if a single power function would relate lobar volume to segment airway diameter during lobar inflation, making Pn a single function of Re and indicating that the spread of the curves shown in Figure 3-5 was due to unaccounted for changes in segment airway diameter. In contrast, if multiple power functions are required to approximate these curves, then one or more variables other than segment airway diameter would be controlling influences on the relationship between Pn and Re. In calculations required to produce Figures B-1 and B-2, I adjusted the value of segment airway diameter in such a way as to force the curves in Figure 3-5 to approximate a single curve. In Figure B-l, in the plot for Pct-Pao = l, I adjusted the diameter to cause the Pao = 2-5 curves to empirically approximate the Pao = 6 curve. I 158 159 then used the diameters obtained to construct the Moody plots in Figure B-l. These plots indicate that the diameters used to approx- imate the curves for the Pct-Pao = l plot clearly do not apply to the Pct-Pao = 2-7 plots, as evidenced by the progressive spreading of the curves as Pct-Pao increases. In addition, Table B-1 shows that multiple power functions may be required to relate segment airway diameter to lobar volume. In Figure B-2, I adjusted the value of segment airway diameter as described above to empirically approximate the curves for Pct-Pao = 7 from Figure 3-5. These diameters were used to construct the plots in Figure B-2. The data indicate that the diameters used to approximate the curves for the Pct-Pao = 7 plot do not apply to the Pct-Pao = 1.6 plots. Also, Table B-1 shows that multiple power functions may be required to relate segment airway diameter to lobar volume. These results indicate that during lobar inflation, Pn is a function of Re, airway diameter, and probably one or more other variables. 160 .:eww seswEeP mswusmmmsems .F- n meewm mes msw— umsmee .msemww mwsp sw mpewe wwe possumsem ea umm: msm: msmumEewu mmmsw .m>s=e o n ees ms» mueewxeseee awweewsweem ea mm>s=e m-~ u eee ms» mmeee e» umumenue me: smumEewu xe:s.e usmEemm we mepe> ms» .m1m mseewm sw pews _ u eesnues ms» sH .Ammseeem mommv muwseepwexmssewpem se .Ammpmsewsuv ewe .Ammwmswev Eewwms sew: umuewwsw me: pcmsmmm ms» sms: eem uceumseu usmmmseme muswee ms» mswpemssee mmswe .ou: Ee 1N Eesw ummemsmsw Aeesv msemmmse aseseeweemsesu mwws: usepmsee uwms me: Ac 2 so meee1aeev sewuewwsw msmsmmw sms: ewsmsewuewms ms» m:esm mpmsee sm>mm ms» we seem .Aemmwumee “mm memv smeses .muwesxmm we asuwseme— ms» we sewuesew e we s:esm mw Amueswuse mss eeev eesu msemmmse umewweees msu we Espwsemew msw .m>s:u sneeze mwmswm e easw mum meemws we Fmsee seem sw mm>see ms» maeswxeseee x_ewesew ea peempu< .F-e mazes: 1’ /’ g 1' , >~E . I 5: .1 / 3 /, Id 601 ' w ammomw Ql ’ )rn z» N 6 »~& a g. 5 -! . : o - o 7 «y 1:1501 w Pct-Peed LogRl 161 Pct-Fee's // / o - 6 3 m Id 001 +0 ,8” 0 ' 1E E '“s .- .l d! // / >— // // .1: l 6 _' 9° ~0001 w Wmnvnm 0.. , /m /’ z . / «u: / // :5 /‘ 3 x/ /’ __. / // A T T o - o .7 a! Id°°1 Pei-P000 O O Q” If) a‘“ 0 Figure B-l r? Log RI 162 .:eww seswsew mswesmmmsems .w- u meewe mes msww umsmeo .msemww mse sw weeps wwe eeesemsee ee umee msm: mememEewu mmmsw .m>see m u ees mse meeewxeseee ee em>s=e m1N u eee mse mmeee ee umemenue me: smemEewu xe:swe esmEemm we mepe> mse .m1m msemws sw news A u eee1eee ms» sH .Aemseeem mommy muwseewwexmssewpee so .Aemwmsewsev swe .Ammweswev Eewpms sew: umeewwsw me: esmEmmm mse sms: eee «seemsee esmmmsems eeswee mse msweemssee mmswe .omz Ee mum Eesw ummemsesw Aeeev mseeemse xseseeweemsese mwws: eseemsee upms me: Aomz se meee1eemv seweepwsw esmEmme sms: ewsmseweewms mse m:esm mpmsee sm>mm mse we seem .Aemmwemse "mm meev smeses .muweszmm we Esewsemew ms» we seweesew e we s:esm mw Ameeswuse mse meev eesu msemmmse umewesses mse we Esewsemep msw .m>see seeese mpmswm e eesw m1m msemws we Fmsee seem sw mm>s=e ms» meeswxeseee apswesew ee eeemee< .N-e meeews 163 Pci- Poo-2 "' v 8'C‘JF‘VIDQ r P f /1-n [”10 4 // V '2 on 3 TN; 3 PNE a c a / 6 -' (yo —' - r0 )5” / ‘1' e :1 “3 i“ E 5 (D .1 $- £ 3 0'1 / .P >- , LI. / / / (5 - N0 14 0' “601 “001 I Q r‘ glmvmmw A' 111 /"’ m a / /"° / :2: °“= : e / N h / a e“ 5 : "“5 3 “‘3' E I? .. / / 11.1001 ' - T “1'0 '1 'o 164 Tab1e 8-1. Re1ationship of segment airway diameter to 1obar vo1ume when the curves in each pane1 of Figure 3-5 are forced to approximate a sing1e curve. Paoa v1." d(1)c x(1)d am" x(7)F 2 28 1.96 0.56 1.70 0.67 3 36 2.22 0.58 1.96 0.71 4 45 2.48 0.60 2.28 0.70 5 54 2.60 0.69 2.68 0.64 6 62 2.78 0.76 2.78 0.76 aPao = transpu1monary pressure (cm H20). bVL = 1obar vo1ume (percent vita1 capacity) obtained from data reported by Frank reference 1, Chapter III). cd(1) = segment airway diameter (mm) required to approximate the Pao = 2-5 curves in the Pct-Pao = 1 pane1 of Figure 3-5 to the Pao = 6 curve. dx(1) = power function re1ating VL to d1 according to the fo11owing examp1e: d(1) at Pao d(1) at Pao 6 2 = VL at Pao 6 VL at Pao 2)x(1) ed(7) = segment airway diameter (mm) required to approximate the Pao = 2-5 curves in the Pct-Pao = 7 pane1 of Figure 3-5 to the Pao = 6 curve. fx(7) = power function re1ating VL to <17 according to the fo11owing examp1e: d(7) at Pao d(7) at Pao 2 _(VL at Pao 2 x(7) 6' VL at Pao ) ° 6