+ All Categories
Home > Documents > articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary...

articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary...

Date post: 27-Jun-2020
Category:
Upload: others
View: 1 times
Download: 0 times
Share this document with a friend
10
Thorax 1987;42:1-10 Review articles Solute permeability of the alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways of a dog and discovered that the drug pro- duced muscular paralysis. He also concluded that the walls of the bronchi were far less permeable than the alveolar walls. Since this original observation a great deal of interest has been shown in the ability of soluble compounds to cross the alveolar-capillary barrier, and this interest has accelerated in recent years with the development of new techniques using radio- labelled materials. Increasingly the inspiration behind this research has been the hope of elucidating the nature of changes occurring in the alveolar-capillary membrane in the course of disease. This article out- lines some salient features of the anatomy and phys- iology of the barrier between gas and blood in the lung, summarises methods available to assess solute permeation of this barrier, and concentrates particu- larly on the technique using measurement of the rate of pulmonary clearance of technetium labelled diethylene triamine penta-acetate (99mTc-DTPA). Anatomy and physiology The histological detail of the alveolar-capillary barrier remained inaccessible to anatomists until the advent of the electron microscope. This has permitted Weibel and others2 to demonstrate most beautifully the fine structure of the lung at the alveolar level. This is illustrated schematically in figure 1. Capillaries can be seen lying in a slightly asymmetrical fashion within alveolar walls, so that the alveolar-capillary mem- brane is on one side relatively thick and on the other side thin. The latter measures less than 0 5 um in cross section, and comprises an epithelial layer of type I alveolar cells, capillary endothelium, and between these two cell types a single basement membrane, which is a continuation of the alveolar basement membrane.3 This delicacy of structure reflects the need for rapid gas transfer. On the other (thicker) side of the capillary, endothelium adheres to the capillary basement membrane, but there is a space between this and the alveolar basement membrane with its adher- ent epithelium. This connective tissue space, or inter- Address for reprint requests: Dr MP Barrowcliffe, Division of Anaesthesia, Clinical Research Centre, Northwick Park Hospital, Harrow, Middlesex HAl 3UJ. stitium, is eventually drained by lymphatics, and in early pulmonary oedema can accommodate a mod- erate increase in lung water while permitting adequate gas exchange. Under such conditions the thin side of the alveolar-capillary barrier remains dry.4'5 Fig 1 Schematic cross section of a capillary lying within the alveolar wall. To the left the alveolar-capillary barrier is relatively thin, and comprises an epithelial cell layer (Ep), a single basement membrane (BM), and the endothelial cell layer (End). To the right the barrier is much thicker, and between the two cell layers with their respective basement membranes there is an interstitial space (Int). The alveolar basement membrane (ABM) is anionic, while the capillary basement membrane (CBM) is cationic. Cap-capillary lumen; End nuc-nucleus of endothelial cell; J- interepithelial cell junction; RBC-red blood cell. (After Weibel2). on July 12, 2020 by guest. Protected by copyright. http://thorax.bmj.com/ Thorax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Downloaded from
Transcript
Page 1: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

Thorax 1987;42:1-10

Review articles

Solute permeability of the alveolar capillary barrier

In 1857 Claude Bernard' introduced curare into theairways of a dog and discovered that the drug pro-duced muscular paralysis. He also concluded that thewalls of the bronchi were far less permeable than thealveolar walls. Since this original observation a greatdeal of interest has been shown in the ability of solublecompounds to cross the alveolar-capillary barrier,and this interest has accelerated in recent years withthe development of new techniques using radio-labelled materials. Increasingly the inspiration behindthis research has been the hope of elucidating thenature of changes occurring in the alveolar-capillarymembrane in the course of disease. This article out-lines some salient features of the anatomy and phys-iology of the barrier between gas and blood in thelung, summarises methods available to assess solutepermeation of this barrier, and concentrates particu-larly on the technique using measurement of therate of pulmonary clearance of technetium labelleddiethylene triamine penta-acetate (99mTc-DTPA).

Anatomy and physiology

The histological detail of the alveolar-capillary barrierremained inaccessible to anatomists until the adventof the electron microscope. This has permitted Weibeland others2 to demonstrate most beautifully the finestructure of the lung at the alveolar level. This isillustrated schematically in figure 1. Capillaries can beseen lying in a slightly asymmetrical fashion withinalveolar walls, so that the alveolar-capillary mem-brane is on one side relatively thick and on the otherside thin. The latter measures less than 0 5 um in crosssection, and comprises an epithelial layer of type Ialveolar cells, capillary endothelium, and betweenthese two cell types a single basement membrane,which is a continuation of the alveolar basementmembrane.3 This delicacy of structure reflects theneed for rapid gas transfer. On the other (thicker) sideof the capillary, endothelium adheres to the capillarybasement membrane, but there is a space between thisand the alveolar basement membrane with its adher-ent epithelium. This connective tissue space, or inter-

Address for reprint requests: Dr MP Barrowcliffe, Division ofAnaesthesia, Clinical Research Centre, Northwick Park Hospital,Harrow, Middlesex HAl 3UJ.

stitium, is eventually drained by lymphatics, and inearly pulmonary oedema can accommodate a mod-erate increase in lung water while permitting adequategas exchange. Under such conditions the thin side ofthe alveolar-capillary barrier remains dry.4'5

Fig 1 Schematic cross section of a capillary lying withinthe alveolar wall. To the left the alveolar-capillary barrieris relatively thin, and comprises an epithelial cell layer (Ep),a single basement membrane (BM), and the endothelial celllayer (End). To the right the barrier is much thicker, andbetween the two cell layers with their respective basementmembranes there is an interstitial space (Int). The alveolarbasement membrane (ABM) is anionic, while the capillarybasement membrane (CBM) is cationic. Cap-capillarylumen; End nuc-nucleus of endothelial cell; J-interepithelial cell junction; RBC-red blood cell. (AfterWeibel2).

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 2: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

2

Recent ultrastructural studies indicate a remark-able difference in the nature and distribution of fixedelectrical charge on the various components of thealveolar-capillary barrier, and this may be importantin the transport of molecules across the barrier. Theendothelial cells have homogeneous anionic surfacesites, whereas the luminal surface of the type I alveolarcell has none.6 Type II cells show extensive anionicsites but, because they make up only 5-10% of thesurface area of pulmonary epithelium, this is for themost part a non-anionic surface. This lack of chargemay be relevant to the hypothesis, put forward byHills,7 that (cationic) surfactant binds to the alveolarsurface, providing an essentially dry layer without anaqueous subphase. This state is perhaps unlikely in thenormal lung, except in the areas occupied by type IIcells, but may be more likely during the recoveryphase of lung injury. There is then a stage when typeI cells have died and been replaced with type II cells,so that the alveolar surface is largely anionic with adry surfactant layer. This could be of considerableadvantage since, as Hills has proposed, the waterrepellant properties of surfactant generate large con-tact angles, so that the surface tension of fluid dropletswill tend to "pump out" the alveolar lumen.Knowledge of the functions of the basement mem-

branes within the lung lags behind an appreciation oftheir structure. Analogy with the kidney is interesting:glomerular capillaries restrict the passage of macro-molecules on the basis of both charge and size, so thatcirculating anions such as albumin are moreeffectively retained within the circulation than areneutral molecules of the same size.8 This is largely dueto a negatively charged glomerular basement mem-brane; loss ofmembrane charge leads to the nephroticsyndrome.9 '1 Curiously, in patients with antibodiesto glomerular basement membrane haemoptyis ismuch more likely to occur if patients smoke ciga-rettes.11 Vaccaro and Brody3 noted that the glomeru-lar and alveolar basement membranes have a similardistribution of negative charge, and that the alveolarbasement membrane has about five times more fixednegative charge than the capillary basement mem-brane. In fact, in later studies Brody suggests that thelatter is predominantly positively charged.'2The alveolar basement membrane is thus likely to

inhibit diffusion of anionic molecules into the alveolarspace, whereas the capillary basement membraneshould facilitate the diffusion of such molecules intothe interstitial space. The passage of such anions fromplasma to lung lymph in normal lung is easier than isthe passage of similarly sized neutral molecules.'3 1'In damaged lungs charge effects are also prominent12:under control conditions in experimentally perfusedlungs neither cationic nor anionic ferritin was seen toleave capillaries and enter the interstitial space. Lung

Barrowcliffe, Jones

injury, however, induced by the administration of anaphthylthiourea, caused pulmonary oedema, withrapid migration of anionic ferritin into the interstitialand alveolar spaces but with no migration of cationicferritin from the capillary.The practical implications of these findings are not

immediately apparent. Albumin is anionic with amolecular size about half that of ferritin (35 nmradius compared with 6-2 nm). In fact, most plasmaproteins are negatively charged at body pH.'" Con-ceivably, the administration of cationic macro-molecules may be of therapeutic value in increasedpulmonary capillary permeability as a means ofplasma expansion and maintenance of a favourableplasma oncotic pressure. There may, however, beextensive binding to anionic surface charges. 16 More-over, within the interstitial space the charges, exclu-sive of the capillary basement membrane, are predom-inantly negative, so that negatively charged proteinsare likely to be repelled from interstitial structureswith enhancement of their movement into lymph. 12 17

Histological appearances also suggest that thealveolar basement membrane is a more importantphysiological barrier than the capillary basementmembrane, in which case the use of cationic moleculesmight promote alveolar flooding.

In contrast to gases, which are able to diffusethroughout the whole of the alveolar-capillary surfacearea, the diffusion of hydrophilic solutes is thought tobe restricted to the much smaller surface areaoccupied by the intercellular junctions. The differentjunctions between epithelial cells and between endo-thelial cells explains the dramatic difference in thesolute permeability of these two parts of the barrier.Some studies have shown that solute permeation ofepithelial junctions is about 10 times less than that ofthe endothelium; the equivalent pore radii calculatedfrom physiological studies are around 0 6-10 nm forthe epithelium, and around 4 0-5 8 nm or more for theendothelium.'4 18-2 This low permeability of theepithelium and its ability to actively transportsodium22 23 explain the very large osmotic pressures(several atmospheres) that can be developed acrossthe epithelial part of the barrier. There is, however,evidence that the high resistance of the epitheliummay suddenly break down under the mechanical stressof distension with fluid.24 25 Therefore the rate ofdiffusion of compounds from alveoli to blood, or viceversa, is much more likely to be determined by theepithelial than by the endothelial component of thealveolar-capillary barrier. It is not known whetherendothelial permeability can increase without achange in epithelial permeability, neither is it clearwhich is more "important" in various circumstances.For example, electron microscopy shows that theinitial site of damage in oxygen exposed lungs is the

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 3: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

Solute permeability of the alveolar capillary barrier

endothelium, whereas epithelial disruption is notnotable until death is imminent.5

Clinical investigation of barrier permeability to solutes

There is no single method that fully characterisessolute permeability of the components of the barrier,although several recent approaches have proved use-ful. These techniques may be divided into those thatwill assess primarily either endothelial or epithelialpermeability (fig 2).

ENDOTHELIAL PERMEABILITYChinard26 developed a method requiring multiplearterial samples after intravenous injection of a rangeof low molecular weight tracer molecules, allowingcalculation of their transit time and distribution vol-ume during a single pass through the pulmonary vas-culature. Results are naturally dependent on the dis-tribution of pulmonary blood flow, and there are

technical difficulties in taking the necessary rapidsequential arterial samples in the clinical setting,which limits the usefulness of this approach. Thismethod, however, has been used to study conditionssuch as postoperative respiratory insufficiency27 andendothelial permeability in renal failure in man.28Brigham has more recently been the principalexponent of this technique, and preliminary findingsfrom his group indicate that the analysis of lunguptake of labelled urea can provide a useful measureof increased capillary permeability in the adult respi-ratory distress syndrome.29 30

Since many workers see the clinical problem as that

99mTcDT 1 3min Transferrin\ 99mTc RBC

Fig 2 Two minimally invasive methods for measuringindices of solute permeability of the alveolar-capillarybarrier. The clearance from lung of technetium labelleddiethylene triamine penta-acetate (DTPA) given by aerosolmeasures principally epithelial permeability. The accumu-lation within the lung of indium labelled transferrin (oralbumin), standardisedfor pulmonary blood volume,provides an index ofprotein leak of the endothelium andeventually the epithelium. In both methods either a gammacamera or a scintillation probe is used to measure changinglung radioactivity. RBC-red blood cells.

3

of diagnosing and quantifying increased protein fluxfrom blood to interstitium, there have been manyapproaches based on external detection of the move-ment of radiolabelled protein. Sugarman andcolleagues3' have injected radiolabelled albumin and,using a gamma camera, subsequently calculate theratio of activity over a region of lung to that over theheart. When they observe a rising ratio, they attributethis to lung injury and increasing protein flux into thelung. Pritchard et al32 have developed a technique forexperimental purposes that measures lung water inaddition to labelled albumin flux. More recent devel-opments have been based on the approach popu-larised by Gorin and coworkers,33 who injectedindium labelled transferrin (molecular weight 76 000),and also red blood cells labelled with technetium toallow a correction for changes in blood volume in thefield of a single scintillation detector over the lung.Basran and colleagues34 have further refined this tech-nique by introducing a second scintillation detectorover the heart. Their index of protein accumulation inthe lungs of patients with respiratory distress syn-drome was significantly different from that in normalsubjects.The most rigorous scientific evaluation of this type

of approach has been undertaken by Dauber et al35 inanimals. They have found (a) an increase in protein"leak" in thiourea induced lung injury before thedevelopment of pulmonary oedema; (b) theimportance of the blood corrections; and (c) theincreased error that results from shortening the studyperiod to less than 150 minutes. Their index of proteinleakiness was not affected by hydrostatic pulmonaryoedema.

This is potentially a very useful method that mayallow an early diagnosis of increased endothelialpermeability. We believe that this technique should bedeveloped to determine the optimum molecular sizeand charge of the probe molecule, which may vary indifferent clinical conditions.EPITHELIAL PERMEABILITYMany techniques have been developed to measureepithelial permeability in animal studies. Schankerand colleagues36 instilled drugs into the lungs of rats,and after a certain time determined drug concen-tration in excised lungs, so that they were able tocalculate the half time of drug absorption from thelung. Valimaki37 injected iodinated polyvinyl-pyrrolidone intravenously and observed that in ratsexposed to high concentrations of oxygen more tracerwas recoverable by endobronchial washing. Egandeveloped a different approach, instilling radio-labelled tracer molecules into an isolated, saline filledlung segment, and sampling this saline at intervals aswell as blood. Normal alveolar epithelium was shownto be relatively impermeable to protein during haemo-

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 4: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

4

dynamic pulmonary oedema,38 and to become perme-able after the administration of alloxan39 (a lungdamaging agent).

In man, if pulmonary oedema fluid is sufficientlyprofuse, the ratio of the concentration of endogenousprotein in this fluid to its ratio in serum allows thedistinction to be made between cardiac and non-cardiac causes.40 41 Sibbald and colleagues42 havealso achieved this by injecting radiolabelled albuminand other tracers intravenously, and then measuringtheir rate of appearance in suctioned pulmonarysecretions. Although these groups used standard suc-tion catheters, bronchoalveolar lavage via thefibreoptic bronchoscope is becoming increasinglypopular.A major concern with this type of technique is that

filling an alveolus with saline may profoundly alter theproperties of the alveolar-capillary barrier. Certainlythis is so if Hills's concept7 of the way in which surfac-tant waterproofs the alveolus is correct. Furthermore,the invasive nature of the procedures may result indamage and increased solute flux across the lung.Introduction of radiolabelled tracer into the airspacesas an aerosol, and measurement of the rate of removalby external radiation detection, renders the techniquevirtually non-invasive. This procedure was first devel-oped over 20 years ago as a diagnostic procedure forestimating regional ventilation. Various solutes havebeen used, but over the past decade DTPA hasbecome established as the most widely used com-pound. It forms stable chelates with most metals andis therefore readily labelled with gamma emittingradionuclides such as indium, chromium, and tech-netium.43

This property led to interest in its use in humanplutonium poisoning, given either by intravenousinjection or by aerosol to assist excretion ofplutoniumafter inhalation exposure.44 The plutonium-DTPAcomplex was known to cross readily into the bloodand be excreted in urine. The calcium salt of DTPA isused for this therapeutic purpose. Giving unchelatedDTPA may be damaging, since calcium is animportant determinant of barrier function, and chela-tion of extracellular calcium causes retraction of adja-cent epithelial and endothelial cells away from oneanother.45 This causes an increase in permeability anda fall in electrical resistance. Restoration of a normalcalcium concentration restores mechanical and elec-trical function.46

Taplin and Effros with their colleagues in LosAngeles used 99mTc DTPA in man initially as a meansof imaging the lungs, and also measured clearanceafter inhalation of an aerosol to detect alterations inpermeability of the alveolar-capillary barrier inpatients with interstitial lung disease.4748 Simulta-neously, Jones and his colleagues at Northwick Park

Barrowcliffe, Jones

Hospital, Harrow, were evaluating the use of a closelyrelated compound, chromium labelled EDTA, as ameans of diagnosing lung injury after acid aspi-ration,49 although in later studies 99mTc DTPA wasused as the radioaerosol.50 There is now a substantialbody of experience in the use of 99mTc DTPA to studyalveolar-capillary barrier function in man and ani-mals.The pertinent characteristics of a tracer such as

99mTc DTPA used to study epithelial permeability inthis way may be considered as follows. Firstly, it musthave an extremely low lipid solubility, so that itsdiffusion is limited to aqueous pores.5' This ensuresthat a second important property, overall molecularsize and shape, is a major determinant of trans-epithelial flux. 99mTc DTPA has a molecular weight of492 daltons, with a radius of 0-6 nm, which is similarto the calculated pore size of this cell layer. Thirdly,molecular charge is assumed to be an influence: asdiscussed above, the uniformly negative alveolar base-ment membrane is likely to impede the passage ofanionic molecules. Electrophoresis indicates that99mTc DTPA is anionic (authors' unpublished obser-vations). Fourthly, there should be no active transportsystem, and none is known to affect 99mTc DTPA.Finally, the radiolabel should remain securelyattached to an inert molecule. Although the stabilityof binding of technetium to DTPA is easily assessed invitro by thin layer chromatography, it has been sug-gested that dissociation may occur in vivo. If this doesoccur, free technetium (as the pertechnetate ion)should be detectable. Since this is concentrated in thethyroid,52 it should have been seen on a gamma cam-era image of the thorax. DTPA also appears not to bemetabolised, and if metabolism were to occur, as hap-pens with radiolabelled vasoactive intestinal pep-tide," it would result in an unexpectedly rapidclearance rate.

EXPRESSING THE RESULTS OF 99mTC DTPACLEARANCE TESTSResults from the aerosol method for measuring epi-thelial permeability have been expressed in severalways. Most commonly, a plot of radioactivity on alogarithmic scale against time on a linear scale yieldsa line whose negative slope is the clearance rate,commonly referred to as the k value. We prefer toexpress our results as a clearance halftime from lungto blood (TA/2), which is the time taken for the initialnumber of counts to fall by half. The T½/2 is a morewidely used and understood measure than the k value.The two are, however, very simply related:

T1/2 0 6932= k

The k value is often arbitrarily expressed as a per-

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 5: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

Solute permeability of the alveolar capillary barriercentage by multiplying it 100 times. A correctionshould be made for radioactive decay of the radio-nuclide being used, which may be performed on indi-vidual data points or on the resulting curve. This wasfrequently omitted in earlier studies.A bald statement of the halftime with no indication

of the variation in the data from which it was calcu-lated allows little reliance to be placed on that result.The correlation coefficient is a commonly quoted sta-tistic, but this indicates only the linearity of the plot ofradioactivity versus time. Calculation of theregression coefficient (slope) together with its stan-dard error is of greater value, since this enablesconfidence limits for a particular halftime clearancerate to be calculated.54 We have obtained satisfactoryresults after initial chest counts of 20-40 000/min. Thelonger the period of recording and the higher thecount rate, the lower is the standard error and thegreater the reliance that can be placed on the result,although this imposes a greater radioactive load onthe patient.

DETERMINANTS OF PULMONARY CLEARANCEOF 99mTC DTPASite of depositionMost studies using 99mTc DTPA have used aerosolswith a particle size in the range 0-5-2 pm aerodynamicmass median diameter (AMMD), so that deposition ismaximised in respiratory bronchioles and alveoli.This minimises the effects of ciliary clearance andabsorption through conducting airway epithelium,although some deposition on the conducting airwaysis inevitable.

In animals the permeability of the conducting air-way epithelium to hydrophilic solutes is less than thatof the alveolar epithelium because absorption ofradiolabelled DTPA deposited in the nasopharynx,trachea, or alveoli was quite different, being 16%,33%, and 100% respectively.55 Furthermore, the dogshows a much slower rate of absorption of DTPAfrom the conducting airways than from the alveoli.56Elwood et al,S7 however, do not support thesedifferences in man. After rapid inhalation of a largeparticle (6-3 pm AMMD) aerosol of 99mTc DTPA tomaximise deposition in the central airways they stud-ied clearance rate and found no significant differencebetween the T'/2 values in their subjects and our pre-vious observations of T'/2 in normal subjects withparticles of 2pmAMMD particles. A potential sourceof error was their assumption that mucociliary clear-ance would not influence the lung-blood clearance of99mTc DTPA. Elwood et al57 assumed a half liferetention of insoluble tracer in the whole lung of 23hours. The retention in the lung, however, depends onthe site of deposition: insoluble particles deposited inconducting airways have a very much shorter reten-

5tion time than those deposited in peripheral airways.Barrowcliffe et al58 used aerosolised 5 pm labelledpolystyrene microspheres to quantify mucociliaryclearance and then 5 pm particle size aerosolised99mTc DTPA to measure regional lung-blood clear-ance. They found that when mucociliary clearancewas taken into account the lung-blood clearance of99mTc DTPA from the conducting airways was verymuch longer than that from terminal airways. Part ofthis slow permeation of the conducting airways maybe due to binding of 99mTc DTPA to mucus. Marriott(personal communication) has observed that thediffusion coefficient of 99mTc DTPA through humanmucus is significantly lower than that of tritiatedwater, and binding of 99mTc DTPA to mucus occursat high affinity sites. This implies that, even overseveral hours, a significant amount of 99mTc DTPA isunlikely to cross mucus layers of physiologicallyobserved thickness.Brown and Schankers1 examined the clearance of a

large range of molecules delivered to rat lungs eitheras a small bolus or as an aerosol. They found a consid-erable difference in T½/2, with the aerosol delivery con-sistently resulting in clearance that was about twice asrapid as that after bolus delivery. This implies thattracer delivered as a bolus is more likely to be held upin central airways, and its slower clearance reflects thelower permeability in this region.

Rizk et alP9 studied the effect of bronchial andpulmonary arterial occlusion on pulmonary clearanceof aerosolised 99mTc DTPA, but were unable to mea-sure permeability of conducting airway epithelium,since bronchial artery occlusion cannot distinguishbetween lack of bronchial deposition and low perme-ability of bronchial epithelium. They did, however,demonstrate that in the absence of pulmonary bloodflow, bronchial artery blood flow was sufficient for99mTc DTPA clearance to continue at a nearly normalrate. Thus the clearance rate should not be muchinfluenced by variations in pulmonary blood flow.

Lung distensionLung distension has been shown in man and animalsto cause an acute increase in solute flux from lung toblood.59-61 The clearance rate in animals is inverselyrelated to the molecular weight of the tracer solute,and when the lung is distended the increase in clear-ance remains proportional to molecular weight.62 Thechange occurs whether lung volume is increased bypositive or by negative pressure breathing and theunderlying mechanism is not known. Possible causesinclude an increase in lung surface area, change inlung surface properties, widening of intercellularjunctions, and recruitment of "leakier" lung units.Some investigators have found a more rapid increasein clearance from the top than from the bottom of the

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 6: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

6

lung, suggesting that greater stretching of lung units atthe top may be responsible.47 63

Background radioactivityNorth American workers have tended to ignore"background" accumulation of 99mTc DTPA duringthe period of study. Clearance of tracer from lungduring the first 7-10 minutes after inhalation was cal-culated by drawing a line of best fit through the datapoints obtained with the gamma camera,60 as it issuggested that at this time there is little tracer in thebackground and thus no need to perform backgroundcorrections. When half time clearance rates are rapid,however, this may not be a valid assumption. Further-more, the pharmacokinetics of 99mTc DTPA indicatethat after intravenous bolus injection the compound israpidly cleared by a biexponential process, whichreflects distribution in the extracellular space, andrenal excretion. The relative sizes of pulmonary bloodand interstitial compartments for 99mTc DTPA can beestimated from studies in which 99mTc DTPA hasbeen used as an indicator to calculate the interstitialfluid volume of the lung.64 This gives an interstitialspace volume of 30 ml/lOOg lung-which, with anintravascular pool also around 30 ml/lOOg lung,represents a large sink within the counting field. Thus99mTc DTPA absorbed across respiratory epitheliumwill distribute to the interstitial fluid space of the lung,and in subjects with pulmonary oedema this may bea potent source of error because of the increased sizeof these spaces65 and because of a larger amount ofradioactivity leaking from alveoli into blood andinterstitium.

In 1980 Jones and colleagues50 described a tech-nique whereby a correction might be made by meansof a scintillation probe over the thigh and an intra-venous injection of 99mTc DTPA that enabled thedetermination of the relative sensitivities of the lungand the thigh probes to tracer appearing via this route.Thus a proportion of the thigh activity could be sub-tracted from the chest activity, to derive activitywithin the chest corrected for 99mTc DTPA that hadalready been absorbed across respiratory epithelium.This method assumes that 99mTc DTPA activity mea-sured by the leg detector is proportional to activitydue to distribution of 99mTc DTPA to vascular andother tissues within the counting field of the chestdetector, and that a bolus of 99mTc DTPA via aperipheral vein is distributed in the same way as is99MTc DTPA absorbed after deposition on permeablerespiratory epithelium. In this respect it is importantthat sufficient time is allowed for distribution to tis-sues after the bolus injection before extrapolation andcalculation of the requisite correction factor, since theaim of the background correction method is to correctnot simply for bloodborne, recirculating 99mTc DTPA

Barrowcliffe, Jones

but also for distribution in the parenchyma of lungand in chest wall.O'Doherty et al66 have adapted this method and,

instead of recording with a second detector over thethigh, have used a gamma camera region of interestover the shoulder, suggesting that this provides a suit-able measure of background activity, although a cali-brating intravenous injection is still necessary. Theyalso found that the correction factor increased if thelungs were leaky.

Gellert et al67 have used a region of interestbetween the kidneys as a site for measurement ofbackground activity. They thereby claim to havedeveloped a technique that requires no intravenousinjection since, when such an injection was given inprelimininary studies, the increase in activity in theinterrenal region was similar to that in various lungregions. In the absence of any means for the gener-ation of suitably sized regions of interest in the lung,however, this finding would appear to represent ahappy coincidence, and requires further explanation.

Other groups have advanced reasons why no back-ground correction may be necessary. Jefferies et al,68for example, refer to studies on two neonates with noventilation to their left lungs, in whom the chest wall,30 minutes after administration ofan aerosol of 99mTcDTPA, contributed only 6% to total thoracic counts.This low figure may result from the thinness of theneonatal chest wall and hypoplasia of underlyinglung, reducing the amount of 99mTc DTPA in thebackground. Of course, these two neonates may havehad non-leaky lungs with little 99mTc DTPA in thebackground. Extrapolation from the data ofO'Doherty et at66 shows that in adults with non-leakylungs the background at 30 minutes contributes 4-5%to whole lung counts, while the corresponding figurefor leaky lungs is about 30%.

Oberdorster et at69 examined pulmonary clearanceof inhaled 99mTc DTPA in the dog, and found that 30minutes after inhalation blood activity of 99mTcDTPA accounted for less than 2% of total thoracicactivity recorded with a gamma camera. These work-ers, however, used an aerosol with a mass medianaerodynamic diameter of 4-4 tm, which is likely tohave resulted in considerable deposition of the aerosolon conducting airways as well as alveoli. Since con-ducting airways are less permeable to 99mTc DTPAthan alveoli (and Oberdorster's own data confirm thisview), the presence in the airways of a mass of 99mTcDTPA that can be only slowly cleared will naturallydepress the contribution of circulating 99mTc DTPAto total thoracic activity. This invalidates extrapo-lation of data obtained in this way to studies directedat accurately measuring alveolar permeability.Background correction not only is useful for

increasing accuracy, but also permits a longer

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 7: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

Solute permeability of the alveolar capillary barrier

duration of measurement of lung clearance of 99mTcDTPA without including background artefact. Clear-ance curves may then be resolved into different halftime clearance rates that are likely to represent non-

homogeneous solute permeability in different regionsof the lung. Disadvantages of the correction are theincreased complexity and the increased dose of radio-activity. Moreover, there is the possibility that if thedetector over the thigh is insensitive this may increasethe scatter within the corrected chest counts.

Smoke inhalationSymptomless subjects who smoke cigarettes have a

significantly faster 99mTc DTPA clearance than non-

smokers.50 66 The clearance half time bears a hyper-bolic relation to cigarette smoke inhalation as esti-mated by the carboxyhaemoglobin level.70 This effectcan be induced in non-smoking subjects within a fewdays of their taking up smoking,71 and resolvestowards normal at a rate dependent on the degree ofinitial increase.72 Rabbits acutely exposed to cigarettesmoke show an immediate increase in clearance.73 Innon-cigarette smoking firefighters who were chroni-cally exposed to dense particulate fire smoke there wasan increase in permeability unassociated with anincrease in carboxyhaemoglobin.74 This suggests a

relationship between lung injury and the particulatephase of smoke rather than the vapour phase ofsmoke, which has been confirmed by experiments inrats exposed to unfiltered smoke.75

Chronic lung diseasesPatients with various interstitial lung diseases have an

increased clearance of 99mTc DTPA from the lung60and, in contrast to symptomless cigarette smokers,their clearance curves are multiexponential.67 76

About half of patients with pulmonary sarcoidosishave increased clearance rates.60 77An increase in 99mTc DTPA clearance has been

reported after inhalation of a histamine aerosol,78which has been attributed to stimulation of either H1or H2 receptors.78 79 The tenacious binding to mucus,however, already referred to, means that 99mTc DTPAis an unsuitable tracer to test Hogg's hypothesis80 thatbronchial hyperreactivity is associated with increasedmucosal permeability. Patients with asthma or

chronic obstructive pulmonary disease have mono-

exponential clearance curves, with no evidence ofincreased 99mTc DTPA clearance.60 81 82 Inhalationof an aerosol of water ("fog") increases the rate ofpulmonary absorption of aerosolised 99mTc DTPA,83and provokes bronchoconstriction in asthmatic sub-jects. Similarly, earlier studies in our laboratoryshowed increased absorption of solute from alveoliwhen hypotonic saline was instilled into lungs.49 Thisprobably represents not increased permeability due to

7

epithelial damage but the phenomenon of "solventdrag",84 whereby water flowing paracellularly alongan osmotic gradient carries solutes along with it.

Pulmonary infectionSince pneumonia is associated with a protein richexudate into the pleural and air spaces, it would notbe surprising if the transepithelial flux of 99mTc DTPAwere facilitated in infected areas of lung. We haveobserved that the 99mTc DTPA clearance rate is con-siderably more rapid in non-smoking patientssuffering from pulmonary tuberculosis than in normalsubjects (for example, T'/2 of six minutes comparedwith 60 minutes). There are two recent reports ofincreased clearance in immunosuppressed patientswith Pneumocystis carinii pneumonia,85 86 which sug-gest that the technique may sometimes be helpful forboth diagnosis of infection and monitoring of theresponse to treatment.

Respiratory distress syndromesThe detection of increased alveolar-capillary barrierpermeability by the use of a molecule as small as99mTc DTPA, in preference to a molecule the size ofa protein such as albumin, should allow detection ofchanges in epithelial barrier properties before a stageof injury characterised by increased protein fluxacross the epithelium. The adult respiratory distresssyndrome is associated with a considerable increase inthe clearance rate of 99mTc DTPA.23 43 87 The clear-ance curve of 99mTc DTPA in this syndrome tends tobe multi-exponential rather than monoexponential,the fast component having a T'/2 of two to threeminutes. During resolution of the disorder there is agradual reduction in size of the lung compartmentwith the very rapid half time.43 Similar changes havebeen reported in neonates with the infant respiratorydistress syndrome.68 The patchy nature of lung dam-age is the likely cause of the multiexponential patternof clearance.Marks et a161 have discussed the likely distribution

of aerosol particles on the alveolar surface. They sug-gest that an aqueous droplet that lands on surfactantis promptly internalised beneath the surface film andspreads rapidly in the aqueous subphase. If this iscorrect, the solute concentration, and thus the gra-dient for diffusion across the alveolar-capillary bar-rier, would then depend on the volume of fluid liningthe alveoli. This volume is thought to be normallyextremely small, but it is likely to be larger inoedematous lungs. The surfactant layer may, how-ever, be an important barrier to the diffusion of solutefrom alveoli to blood. Conditions such as neonatalhyaline membrane disease, in which surfactant isdeficient, are associated with greatly increased soluteclearance.6888 An accompanying inflammatory dis-

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 8: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

8order could explain the increased solute permeabilitythat was found, but Robertson et al88 describe areduction in solute permeability after surfactantreplacement. Possibly the greater mechanical forceson the alveolar walls in the surfactant depleted lungmay be responsible for increased solute permeation.

Conclusion

Some 130 years after Claude Bernard's original obser-vation, it is evident that the absorption rate of radio-labelled water soluble compounds may provide usefulinformation about physiological and inflammatoryprocesses in human lung. A new phase of research hasbegun where the effects of various mediators of lunginjury may be studied not just in terms of the cellularand biochemical response but with reference to one ofthe cardinal signs ofinflammation, the permeability ofthe alveolar-capillary barrier.

MICHAEL P BARROWCLIFFEDivision of Anaesthesia

Medical Research Council Clinical Research CentreNorthwick Park Hospital

Harrow, Middlesex HAI 3UJJ GARETH JONES

(formerly MRC Clinical Research Centre)University Department of Anaesthesia

Leeds LS2 9LN

References

1 Bernard C. Lecons sur les effets des substances toxiques etmedicamenteuses. Paris: Bailliere, 1857:286-8.

2 Weibel ER. The pathway of oxygen. Structure andfunc-tion in the mammalian respiratory system. Cambridge,Massachusetts: Harvard University Press,1984.

3 Vaccaro CA, Brody JS. Structural features of alveolarwall basement membrane in the adult rat lung. J CellBiol 1981;91:427-37.

4 Cottrell TS, Levine OR, Senior RM, Wiener J, Spiro D,Fishman AP. Electron microscopic alterations at thealveolar level in pulmonary edema. Circ Res1967;21:783-97.

5 Kistler HS, Caldwell PRB, Weibel ER. Development offine structural damage to alveolar and capillary liningcells in oxygen-poisoned rat lungs. J Cell Biol1967;32:605-28.

6 Simionescu D, Simionescu M. Differential distribution ofthe cell surface charge on the alveolar-capillary unit.Microvasc Res 1983;25:85-100.

7 Hills BA. What forces keep the air spaces of the lung dry?Thorax 1982;37:713-6.

8 Cotran RS, Rennke HG. Anionic sites and the mech-anisms of proteinuria. N Engl J Med 1983;309:1050-2.

9 Vernier RL, Klein DJ, Sisson SP, Mahan JD, OegemaTR, Brown DM. Heparan sulfate-rich anionic sites inthe human glomerular basement membrane. N Engl JMed 1983;309:l001-9.

Barrowcliffe, Jones

10 Levin M, Smith C, Walter MDS, Gascoine P, BarrattTM. Steroid-responsive nephrotic syndrome: a gener-alised disorder of membrane negative charge. Lancet1985;ii:239-42.

1 Donaghy M, Rees AJ. Cigarette smoking and lung hae-morrhage in glomerular basement membrane. Lancet1983;ii: 1390-2.

12 Brody JS, Vaccaro CA, Hill NS, Rounds S. Binding ofcharged ferritin to alveolar wall components andcharge selectivity of macromolecular transport in per-meability pulmonary edema. Circ Res 1984;55:155-67.

13 Glauser FL, Fairman RP, Millen JE. Facilitated trans-port of anionic dextrans in the pulmonary micro-vasculature of sheep. Microcirculation 1982;2:305- 13.

14 Lanken PN, Hansen-Flaschen JH, Sampson PM, PietraGG, Haselton FR, Fishman AD. Passage ofuncharged dextrans from blood to lung lymph inawake sheep. J AppI Physiol 1985;59:580-91.

15 Putnam FW. Alpha, beta, gamma, omega-the roster ofthe plasma proteins. In: Putnam FW, ed. The plasmaproteins. Structure, function and genetic control. Vol 1.New York: Academic Press, 1975:57-131.

16 Gilchrist SA, Parker JC. Exclusion of charged macro-molecules in the pulmonary interstitium. MicrovascRes 1985;30:88-98.

17 Leak LV. Distribution of cell surface charges on meso-thelium and lymphatic endothelium. Microvasc Res1986;31: 18-30.

18 Taylor AE, Gaar KA. Estimation of equivalent poreradii of pulmonary capillary and alveolar membranes.Am J Physiol 1970;218:1133-40.

19 Taylor AE, Guyton AC, Bishop VS. Permeability of thealveolar membrane to solutes. Circ Res1965;26:353-62.

20 Gorin AB, Stewart PA. Differential permeability ofendothelial and epithelial barriers to albumin flux. JAppl Physiol 1979;47:1315-24.

21 Effros RM, Mason GR, Silverman P, Reid E, HukkanenJ. Movement of ions and small solutes across endo-thelium and epithelium of perfused rabbit lungs. JApplPhysiol 1986;60:100-7.

22 Olver RE, Davis B, Marin MG, Nadel JA. Active trans-port ofNa + and Cl - across canine tracheal epitheliumin vitro. Am Rev Respir Dis 1975;112:811-6.

23 Jones JG. Mechanisms and measurement of injury to thealveolar-capillary barrier. In: Jones JG, ed. Effects ofanesthesia and surgery on pulmonary mechanisms andgas exchange. Int Anesth Clin 1984;22:131-48.

24 Parker JC, Guyton AC, Taylor AE. Pulmonary inter-stitial and capillary pressures. J Appl Physiol1978;44:267-76.

25 Gee MH, Staub NC. Role of bulk fluid flow in proteinpermeability of the dog lung alveolar membrane. JAppl Physiol 1977;42:144-9.

26 Chinard FP. The permeability characteristics of the pul-monary blood-gas barrier. In: Caro CG, ed. Advancesin respiratory physiology. London: Edward Arnold,1966:106-47.

27 Gump FE, Mashima Y, Jorgensen S, Kinney JM. Simul-taneous use of three indicators to evaluate pulmonarycapillary damage in man. Surgery 1971;70:262-9.

28 Crosbie WA, Snowden S, Parsons V. Changes in lung

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 9: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

Solute permeability of the alveolar capillary barriercapillary permeability in renal failure. Br Med J1972;iv:388-90.

29 Brigham KL, Snell JD, Harris TR, et al. Indicator dilu-tion lung water and vascular permeability in humans.Circ Res 1979;44:523-30.

30 Bernard GR, Rinaldo J, Harris T, et al. Early predictorsof ARDS reversal in patients with established ARDS[Abstract]. Am Rev Respir Dis 1985;131:A143.

31 Sugerman HJ, Hirsch JI, Tatum JL, Strash AM, SharpDE, Greenfield LG. Comparative scintigraphy in oleicacid pulmonary microvascular injury. Crit Care Med1982;10:31-3.

32 Pritchard JS, Rajagopalan B, Lee G de J. Transvascularalbumin flux and the interstitial water volume inexperimental pulmonary oedema in dogs. Clin Sci1980;59: 105-13.

33 Gorin AB, Kohler J, De Nardo G. Non-invasive mea-surement of pulmonary transvascular protein flux innormal man. J Clin Invest 1980;66:869-77.

34 Basran GS, Byrne AJ, Hardy JG. A non-invasive tech-nique for monitoring lung vascular permeability inman. Nuclear Medicine Communications 1985;6:3-10.

35 Dauber IM, Pluss WT, Van Grondelle A, Trow RS, WeilJV. Specificity and sensitivity of noninvasive mea-surement of pulmonary vascular protein leak. J ApplPhysiol 1985;59:564-74.

36 Enna SJ, Schanker LS. Absorption of drugs from therat's lung. Am J Physiol 1972;223:1227-3 1.

37 Valimaki M, Kivisaari J, Niinikoski J. Permeability ofalveolar-capillary membrane in oxygen poisoning.Aerosp Med 1974;45:370-4.

38 Egan EA, Nelson RM, Gessner IH. Solute permeabilityof the alveolar epithelium in acute hemodynamic pul-monary edema in dogs. Am J Physiol 1977;233:H80-6.

39 Nelson RM, McIntyre BR, Egan EA. Solute perme-ability of the alveolar epithelium in alloxan edema indogs. J Appl Physiol 1978;44:353-7.

40 Fein A, Grossman RF, Jones JG, et al. The value ofedema fluid protein measurement in patients with pul-monary edema. Am J Med 1979;67:32-8.

41 Sprung CL, Rackow EC, Fein A, Jacob AI, Isikoff SK.The spectrum of pulmonary edema: differentiation ofcardiogenic, intermediate and non-cardiogenic formsof pulmonary edema. Am Rev Respir Dis198 1;124:718-22.

42 Sibbald WJ, Driedger AA, Moffat JD, Myers ML, ReidBA, Holliday RL. Pulmonary microvascular clearanceof radiotracers in human cardiac and non-cardiac pul-monary edema. J Appl Physiol 198 1;50:1337-47.

43 Jones JG, Royston D, Minty BD. Changes in alveolar-capillary barrier function in animals and humans. AmRev Respir Dis 1983;127:S51-9.

44 Stather JW, Smith H, Bailey MR, Birchall A, BulmanRA, Crawley FEH. The retention of '4C-DTPA inhuman volunteers after inhalation or intravenousinjection. Health Physics 1983;44:45-52.

45 Kern DF, Malik AB. Microvascular albumin perme-ability in isolated perfused lung: effects of EDTA. JAppl Physiol 1985;58:372-5.

46 Shasby DM, Shasby SS. Effects of calcium on trans-endothelial albumin transfer and electrical resistance.J Appl Physiol 1986;60:71-9.

947 Rinderknecht J, Krauthammer M, Uszler JM, Taplin G,

Effros RM. Solute transfer across the alveolar capil-lary membrane in pulmonary fibrosis [Abstract]. AmRev Respir Dis 1977;115:156.

48 Chopra SK, Taplin GV, Tashkin DP, Elam DE. Lungclearance of soluble radio-aerosols of different molecu-lar weights in systemic sclerosis. Thorax 1979;34:63-7.

49 Jones JG, Berry M, Hulands GH, Crawley JCW. Thetime course and degree of change in alveolar-capillarymembrane permeability induced by aspiration ofhydrochloric acid and hypotonic saline. Am Rev RespirDis 1978;118:1007-13.

50 Jones JG, Lawler P, Crawley JCW, Minty BD, HulandsG, Veall N. Increased alveolar epithelial permeabilityin cigarette smokers. Lancet 1980;i:66-8.

51 Brown RA, Schanker LS. Absorption of aerosolizeddrugs from the rat lung. Drug Metabolism and Disposi-tion 1983;11:355-60.

52 Harper PV, Lathrop KA, McCordle RJ, Andros G. Theuse of technetium-99m as a clinical scanning agent forthyroid, liver and brain. In: Medical radioisotope scan-ning. Vol 2. Vienna: International Atomic EnergyAuthority, 1964:33-4.

53 Barrowcliffe MP, Morice A, Jones JG, Sever PS. Pul-monary clearance of vaso-active intestinal peptide.Thorax 1986;41:88-93.

54 Armitage P. Statistical methods in medical research.Oxford: Blackwell Scientific Publications, 1977:163-5.

55 Stather JW, Smith A, James AC, Rodwell P. The experi-mental use of aerosol and liposomal forms ofDTPA asa treatment for plutonium contamination. In: Pro-ceedings of a seminar on diagnosis and treatment ofincorporated radionuclides. Vienna: InternationalAtomic Energy Authority, 1976:387. (STI/Pub/41 1.)

56 Dudley RE, Muggenburg BA, Cuddihy RG, McClellanRO. Absorption of diethylene thiamine penta-aceticacid (DTPA) from the respiratory tract of dogs. AmInd Hyg Assoc J 1980;41:5-1 1.

57 Elwood RK, Kennedy S, Belzberg A, Hogg JC, Pare PD.Respiratory mucosal permeability in asthma. Am RevRespir Dis 1983;128:523-7.

58 Barrowcliffe MP, Jones JG, Agnew J, Francis RA,Clarke SW. The relative permeability of human con-ducting and terminal airways to 99mTc-DTPA. Eur JRespir Dis (in press).

59 Rizk NW, Luce JM, Hoeffel J, Price DC, Murray JF. Siteof deposition and factors affecting clearance of aero-solized solute from canine lungs. J AppI Physiol1984;56:723-9.

60 Rinderknecht J, Shapiro L, Krauthammer M, et al.Accelerated clearance of small solutes from the lungsin interstitial lung disease. Am Rev Respir Dis1980;121:105-17.

61 Marks JD, Luce JM, Lazar NM, Ngao-Sun Wu J, Lip-avsky A, Murray JF. Effect of increases in lung volumeon clearance of aerosolized solute from human lungs.J Appl Physiol 1985;59:1242-8.

62 Barrowcliffe MP, Zanelli GD, Jones JG. A comparisonof the effects of PEEP and severe lung injury uponclearance of hydrophilic tracer molecules [Abstract].Br J Anaesth 1986;58:122-3P.

63 Mason GR, Uszler JM, Effres RM, Reid E. Rapidly

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from

Page 10: articles Solute permeability ofthe alveolar barrierSolute permeability ofthe alveolar capillary barrier In 1857 Claude Bernard' introduced curare into the airways ofa dog and discovered

10reversible alterations of pulmonary epithelial perme-

ability induced by smoking. Chest 1983;83:6-1 1.64 Parker JC, Ryan J, Taylor AE. Plasma-lymph albumin

kinetics, total lymph flow and tissue hematocrit in nor-

mally hydrated dog lungs. Microvasc Res1984;28:254-69.

65 Parker JC, Crain M, Grimbert F, Rutili G, Taylor AE.Total lung lymph flow and fluid compartmentation inedematous dog lungs. JAppl Physiol 198 1;51:1268-77.

66 O'Doherty MJ, Page CJ, Croft DN, Bateman NT. Lung99mTc-DTPA transfer: a method for background cor-rection. Nucl Med Comm 1985;6:209-15.

67 Gellert AR, Lewis CA, Langford JA, Tolfree SEJ, RuddRM. Regional distribution of pulmonary epithelialpermeability in normal subjects and patients withasbestosis. Thorax 1985;40:734-40.

68 Jefferies AL, Coates G, O'Brodovich H. Pulmonary epi-thelial permeability in hyaline membrane disease. NEngl J Med 1985;311:1075-80.

69 Oberdorster G, Utell MJ, Weber DA, Ivanovich M,Hyde RW, Morrow PE. Lung clearance of inhaled99mTc-DTPA in the dog. J Appl Physiol1984;57:589-95.

70 Jones JG, Minty BD, Royston JP. Carboxyhaemoglobinand pulmonary epithelial permeability in man. Thorax1983;38: 129-33.

71 Minty BD, Royston D, Jones JG, Hulands GH. Theeffect of nicotine on pulmonary epithelial permeabilityin man. Chest 1984;86:72-4.

72 Minty BD, Jordan C, Jones JG. Rapid improvement inabnormal pulmonary epithelial permeability afterstopping cigarettes. Br Med J 198 1;282:83-6.

73 Witten ML, Lemen RJ, Quan SF, et al. Acute cigarettesmoke exposure increases alveolar permeability in rab-bits. Am Rev Respir Dis 1985;132:321-5.

74 Minty BD, Royston D, Jones JG, Smith DJ, SearingCSM, Beeley M. Changes in permeability of thealveolar-capillary barrier in fire-fighters. Br J Ind Med1985;42:631-4.

75 Minty BD, Royston D. Cigarette smoke induced changesin rat pulmonary clearance of 99mTcDTPA. Am RevRespir Dis 1985;132:1170-3.

76 Pantin CFA, Britten A, Lawrence R, Sweatman M,Turner-Warwick M. Lung permeability in patientswith interstitial lung disease [Abstract]. Thorax1984;39:709.

Barrowcliffe, Jones

77 Jacobs MP, Baughman RP, Hughes J, Fernandez-UlloaM. Radioaerosol lung clearance in patients with activepulmonary sarcoidosis. Am Rev Respir Dis1985;131:687-9.

78 Braude S, Coe C, Royston D, Barnes PJ. Histamineincreases lung permeability by an H2 receptor mech-anism. Lancet 1984;ii:372-4.

79 Chan TB, Shelton DM, Rees PJ, Eiser NM. Histamineinduced changes in pulmonary epithelial permeabilityvia H1 receptors [Abstract]. Thorax 1985;40:720.

80 Hogg JC. Bronchial mucosal permeability and itsrelationship to airways hyperactivity. J Allergy ClinImmunol 198 1;67:421-5.

81 O'Byrne PM, Dolovich M, Dirks R, Roberts RS, New-house MT. Lung epithelial permeability: relation tonon-specific airway responsiveness. J Appl Physiol1984;57:77-84.

82 Huchon GJ, Russell JA, Barritault LG, Udavsky A,Murray JF. Chronic air-flow limitation does notincrease respiratory epithelial permeability assessed byaerosolized solute but smoking does. Am Rev RespirDis 1984;130:457-60.

83 Borland C, Chamberlain A, Barber B, Higenbottam T.Pulmonary epithelial permeability after inhalingsaline, distilled water "fog" and cold air. Chest1985;87:373-6.

84 Whittembury G, Martinez C, Kinares H, Paz-Aliaga A.Solvent drag of large solutes indicates paracellularwater flow in leaky epithelia. Proc R Soc Lond1980;211:63-81.

85 Mason GR, Effros RM, Mena I, Duane G. Pulmonaryclearance of a small solute is increased by pneu-mocystis pneumonia in patients with acquired immunedeficiency syndrome [Abstract]. Am Rev Respir Dis1985;131:A1 16.

86 Jones DK, Higenbottam TW. Pneumocystis pneumoniaincreases the clearance rate of inhaled 99mTc-DTPAfrom lung to blood. Chest 1985;88;:631-2.

87 Mason GR, Effros RM, Uszler JM, Mena I. Small soluteclearance from the lungs of patients with cardiogenicand noncardiogenic pulmonary edema. Chest1985;88:327-34.

88 Robertson B, Berry D, Curstedt T, et al. Leakage ofprotein in the immature rabbit lung: effect of surfac-tant replacement. Respir Physiol 1985;61:265-76.

on July 12, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.42.1.1 on 1 January 1987. Dow

nloaded from


Recommended