Heart 1997;78:171-176

Comparison of bronchopulmonary collaterals andcollateral blood flow in patients with chronicthromboembolic and primary pulmonaryhypertension

Jiri Endrys, Nasser Hayat, George Cherian

AbstractObjective-To compare the visualisationof bronchopulmonary collaterals andbronchopulmonary collateral blood flowin patients with chronic thromboembolicpulmonary hypertension and primarypulmonary hypertension.Setting--Referral centre for cardiology atan academic hospital.Patients-Nine patients with chronicthromboembolic pulmonary hypertensionand 17 with primary pulmonary hyper-tension.Interventions-Bronchopulmonary col-laterals were visualised by selectivebronchial arteriography or thoracic aor-tography. Bronchopulmonary collateralblood flow was estimated by injectingindocyanine green into the ascendingaorta and sampling below the mitral valvefrom the left ventricle.Results-The degree ofpulmonary hyper-tension was comparable in the twogroups. Large bronchopulmonary collat-erals were visualised in all the patientswith thromboembolic pulmonary hyper-tension who had bronchial arteriographyor aortography or both. None of the pri-mary pulmonary hypertension groupstudied by aortography had bronchopul-monary collaterals (P << 0 001). All thepatients with chronic thromboembolicpulmonary hypertension had significantbronchopulmonary collateral blood flow,which was (mean (SD)) 29-8 (18.6)% ofthe systemic blood flow. There was norecordable collateral blood flow in 11 of 15patients with primary pulmonary hyper-tension. In the remaining four patientsthe mean value was 1-1 (1l8)% of the sys-temic blood flow (P << 0.001).Conclusions-Visualisation of broncho-pulmonary collaterals by thoracic aorto-graphy or by bronchial arteriography, orthe demonstration of an increased bron-chopulmonary collateral flow, helps todistinguish patients with chronic throm-boembolic pulmonary hypertension fromthose with primary pulmonary hyperten-sion.

(Heart 1997;78:171-176)

Keywords: pulmonary thromboembolism; primary pul-monary hypertension; bronchial arteries; heartcatheterisation

The recognition of chronic thromboembolicobstruction of the central, lobar, and segmentalpulmonary arteries resulting in pulmonaryhypertension is important since this is poten-tially curable by pulmonary thromboendar-terectomy.' On the other hand, themanagement of the majority of patients withprimary pulmonary hypertension continues tobe disappointing despite advances in pharma-cotherapy and transplantation.2-5 At present,the differentiation of chronic thromboembolicpulmonary hypertension from primary pul-monary hypertension depends mainly onthe perfusion-ventilation lung scan and pul-monary angiography. 2 6 7 The risk of pul-monary angiography in patients with severepulmonary hypertension and the precautionsduring angiography need special emphasis.1 7More recently high resolution computerisedtomography,8 magnetic resonance imaging,8pulmonary intravascular imaging,9 and trans-oesophageal echocardiography'0 have beenused for the detection of pulmonary thrombi.The separation of thromboembolic pulmonaryhypertension from primary pulmonary hyper-tension is at times difficult, and made more soby the development of in situ thrombosis ofthe pulmonary arteries in patients with pri-mary pulmonary hypertension." Manychronic and acute lung diseases are associatedwith an increase in bronchopulmonary collat-erals and bronchial arterial blood flow.'2-'4Extensive bronchopulmonary collateralsdevelop to parts of the lung affected by ligationof the pulmonary artery or its branches,'516and after experimental pulmonary embolisa-tion.17 In patients with pulmonary embolism,bronchopulmonary collaterals to the affectedsegments have been demonstrated atnecropsy'8 and during life.'9-2' This report isbased on our observations of bronchopul-monary collaterals and bronchopulmonarycollateral blood flow in patients with chronicthromboembolic pulmonary hypertension andprimary pulmonary hypertension. This studywas not designed to compare the differentmethods for detecting pulmonary thrombo-embolism.

MethodsPATIENT GROUPSWe studied nine patients with severe throm-boembolic pulmonary hypertension and 17with primary pulmonary hypertension.Haemodynamic criteria for inclusion in the

Department ofMedicine, Faculty ofMedicine, KuwaitUniversity, KuwaitJ EndrysN HayatG CherianCorrespondence to: DrGeorge Cherian, Departmentof Medicine, Faculty ofMedicine, PO Box 24923,Safat 131 10, Kuwait.

Accepted for publication6 May 1997

171 on M

arch 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.171 on 1 August 1997. D

ownloaded from

Endrys, Hayat, Cherian

thromboembolic group were pulmonary arterialsystolic pressure of > 70 mm Hg; pulmonaryarterial mean pressure of > 45 mm Hg; andpulmonary vascular resistance index of > 10units. After a full clinical assessment, thediagnosis was based on the results of perfu-sion-ventilation lung scans and pulmonaryangiography using previously described crite-ria. 1 6 22-24 None of the patients in thethromboembolic group had thromboendar-terectomy.

CARDIAC CATHETERISATION AND PULMONARYANGIOGRAPHYRight and left heart catheterisation was per-formed percutaneously from the groin usingstandard techniques. Pressures were recordedfrom the right atrium, right ventricle, pul-monary artery, pulmonary wedge, left ventri-cle, and aorta, and transseptally from the leftatrium. Cardiac output was calculated by dye

AT CATH.: 3739----10.8

AO-PA3.4 mg CG A

+

AO-MV6.8 mg CGFAMP = 2

APA AO = 294.5

FAMP-Mg CG

PBF = 5.84 I/min

AAO-MV 108

FAMP-Mg CG

V AAM

CPBF AO-MVx PBF = 2.12 I/mi

AAP_AO = 57% SBFFigure 1 Series of dye dilution curves in patient withchronic thromboembolic pulmonary hypertension (CTEPH)and high bronchopulmonary collateral bloodflow (BPCF).Sequence of curves was arrangedfor graphic reasons. CurveB measured pulmonary bloodflow (PBF) by injection ofdye into the pulmonary artery (PA) with samplingfrom theaorta (AO). Curve C quantified the left to right shunt byinjection of double the amount ofdye into the ascendingaorta with double recorder sensitivity and samplingfromthe left ventricle close to the mitral valve (MV). Ratio ofthe C curve area divided by B curve area multiplied bypulmonary bloodflow (in llmin) gives the BPCF in llmin.Curve A recorded by injection into the ascending aorta(AO) and samplingfrom main pulmonary artery (PA)excludes patent arterial duct as the cause of the left to rightshunt demonstrated by curve C. In the absence ofBPCF,the curve C would have a pattern similar to curveA withlong appearance time (AT). CG, Cardiogreen; Fm,amplification factor of the recorder; 4, actual injection ofdye; 1, injection of dye correctedfor a delay caused bysampling system; APAAO, area of the B curve; AAO-MV, areaof the C curve.

dilution techniques. Congenital heart disease,such as patent arterial duct, and otheracquired cardiac diseases were excluded.Pulmonary angiography was performed selec-tively in the right and left pulmonary artery,except in the first few patients.

BRONCHOPULMONARY COLLATERALSBronchopulmonary collaterals were assessedby two methods: (1) measurement of bron-chopulmonary collateral blood flow using thedye dilution technique; and (2) angiographi-cally by aortography or selective bronchialarteriography or both.

MEASUREMENT OF BRONCHOPULMONARYCOLLATERAL FLOWBronchopulmonary collateral flow was mea-sured by a method based on the principledeveloped by Cudkowicz,'6 and used byNakamura.'5 We modified this method byusing a cuvette densitometer to avoid errorsfrom ear densitometry. Sampling for bron-chopulmonary collateral flow was from the leftventricle below the mitral valve, to ensureproper mixing of flow from the right and leftpulmonary veins in the left atrium.Dye curves were recorded using BD5 or

BD9 recorders (Kipp and Zonen, Delft, TheNetherlands) using a D-402A densitometer(Waters Instruments, Rochester, Minnesota,USA) with the measuring cuvette directly con-nected to the catheter to reduce the samplingvolume. All dye dilution curves were recordedat least twice. A constant rate of blood with-drawal through the cuvette was achieved usinga Harvard pump at 40 ml/min speed (HarvardApparatus, Southnatick, Massachusetts,USA). Blood was infused back to the patientafter each recording. Early on, the curve areaswere measured by planimetry,26 and in laterstudies using the Hewlett-Packard catheterisa-tion computer adapted by one of us JE) for online measurement of shunts. An example ofthe dye curves recorded is shown in fig 1. Thefirst dye curve (B) measured the pulmonaryblood flow by injection of indocyanine green(usually 5 mg) into the pulmonary artery, withsampling from the aorta. The second curve(C) was recorded by injection of double orquadruple amount of dye into the ascendingaorta and sampling just below the mitral valvethrough a transseptal or retrograde catheter, todetect any aorto-pulmonary shunt. In thepresence of bronchopulmonary collateral flow,the early recirculation had an appearance timeof about three seconds, while the normal recir-culation usually started after 10 seconds. Thebronchopulmonary collateral flow was calcu-lated as the ratio of the shunt curve area (C)divided by the pulmonary blood flow area (B).This value multiplied by the pulmonary bloodflow gives the bronchopulmonary collateralflow in litres per minute. The systemic bloodflow (SBF) was derived by subtracting thebronchopulmonary collateral flow (BPCF)from the pulmonary blood flow.

Then, BF x 100 = BPCF as % of SBFSBF

172 on M

arch 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.171 on 1 August 1997. D

ownloaded from

Bronchopulmonary collaterals in chronic thromboembolic pulmonary hypertension

Patient groups and haemodynamic results. Values are mean (SD)

CTEPH PPH

Men (n) 5 7Women (n) 4 10Age (mean (SD)) years 45-6 (8 3) 31-2 (12-8) P < 0-006Aortic pressure (mm Hg)

Systolic 120-9 (14-5) 117-6 (18-1)Diastolic 79 3 (6 8) 80-8 (13-6)Mean 94-4 (8 9) 96-6 (15-3)

PA pressure (mm Hg)Systolic 97-3 (5-0) 91-5 (21-9)Diastolic 33-8 (6-4) 40-6 (11-4)Mean 57-7 (8-2) 59*4 (13-7)

LA pressure (mm Hg)Mean 8-0 (6 4) 4-1 (2 5)

RA pressure (mm Hg)Mean 7-6 (2 8) 7-8 (6 2)

Cardiac index (1/min/m2) 2-02 (0 6) 2-10 (0 7)PVRI (units/m2) 21-7 (5-2) 25-0 (11 9)BPCF (1/min) 0 93 (0 49) 0-027 (0 045) P < 0 0001BPCF (% of SBF) 29-8 (18-6) 1 1 (1-8) p < 0-0001

BPCF, bronchopulmonary collateral flow; CTEPH, chronic thromboembolic pulmonaryhypertension; LA, left atrium; PA, pulmonary artery; PPH, primary pulmonary hypertension;PVRI, pulmonary vascular resistance index; RA, right atrium; SBF, systemic blood flow.

Figure 2 Pulmonary angiogram from a patient with thromboembolic pulnhypertension. All segmental branches of the pulmonary artery in the right loupper lobe are completely or subtotally occluded. Subsegmental branches of Ilobe and left lower lobe show varying degrees of obstruction. The main and jartery are dilated.

3i

2c

0U-

M PPHM CTEPH

2

0.5

0LL

LNSJci BPCF

Figure 3 Diagram offlow data showing the cardiac index values (CI) anbronchopulmonary collateralflow (BPCF). The BPCF was minimal in thepulmonary hypertension group (PPH) and markedly increased in patients zthromboembolic pulmonary hypertension (CTEPH). *P < 0-0001.

nonarywer and leftthe right upper

STATISTICAL ANALYSISThe unpaired Student t test was used. Valuesare given as mean (SD).

ResultsThe age and sex distribution and basichaemodynamic data are shown in the table.Patients with thromboembolic pulmonaryhypertension were on the average about 15years older than the group with primary pul-monary hypertension. All patients withthromboembolic pulmonary hypertension hadsegmental or larger perfusion defects thatwere normally ventilated, while those with pri-mary pulmonary hypertension had normal orsubsegmental inhomogeneity in perfusion.Pulmonary angiography in patients withthromboembolic pulmonary hypertensionshowed varying combinations of fillingdefects, abrupt narrowing or cut off of arter-ies, webs, bands, or lumenal irregularities, asin other reports. In primary pulmonary hyper-tension there was symmetrical dilatation ofproximal and hilar pulmonary arteries withrelatively rapid taper and marked peripheralarterial pruning. The pulmonary artery sys-tolic and mean pressures and pulmonary vas-cular resistance index were raised and similarin patients with primary pulmonary hyperten-sion and thromboembolic pulmonary hyper-tension (NS). One patient in the primarypulmonary hypertension group did not havepulmonary angiography. This patient had thelowest cardiac index in this report(1 -02 1/min/m2) and the highest right atrial(mean 28 mm Hg), pulmonary artery systolic(145 mm Hg), and mean pulmonary arterypressure (95 mm Hg). The pulmonaryangiogram of a patient with thromboembolicpulmonary hypertension is shown in fig 2.

right pulmonary BRONCHOPULMONARY COLLATERAL FLOWThe bronchopulmonary collateral flow wasestimated in eight of the nine patients withthromboembolic pulmonary hypertension andin 15 of the 17 patients with primary pul-monary hypertension. In the remaining three,indocyanine green was not available at thetime of cardiac catheterisation. The bron-chopulmonary collateral flow, expressed inI/min and as a percentage of systemic bloodflow, is shown in fig 3. Bronchopulmonarycollateral flow was detected in all eight of thepatients with thromboembolic pulmonaryhypertension in whom the estimation wasmade. The mean flow was 093 (SD 049)I/min, or 29X8(18*6)% of systemic blood flow.There was no detectable bronchopulmonarycollateral flow in 11 of the 15 patients withprimary pulmonary hypertension in whom theestimation was made. In the remaining fourpatients, the mean value was 0-027 (0045)I/min, which was 1-1(1 -8)% of systemic blood

* L flow. Thus there was a highly significant dif-BPCF ference in bronchopulmonary collateral flow

between patients with chronic thromboem-primary bolic pulmonary hypertension and patientswith chronic with primary pulmonary hypertension (P <<

ooO1).

173

1

on March 30, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.2.171 on 1 A

ugust 1997. Dow

nloaded from

Endrys, Hayat, Cherian

Figure 4 Selectivebronchial arteriogram ina patient withthromboembolic pulmonaryhypertension (CTEPH).The dilated bronchialartery is clearly seen as wellas mesh ofsmal bronchialvessels connected to thepulmonary arterialbranches downstream wothe embolic occlusion. V,

bronchial artery; -.,

branches ofpulmonaryartery distal to theocclusion.

Figure 5 Aortogram inthe same patient as infig4. Intercostal arteries anda dilated bronchial arterythatfeeds mesh-likecollaterals which faindyopacify the pulmonaryartery branchesdownstream of the embolicocclusion are seen. V,bronchial artery.

Figure 6 Aortogram in apatient with primarypulmonary hypertension.In spite of well opacifiedintercostal arteries, nodilated bronchial arteriesare seen.

boembolic pulmonary hypertension and allshowed bronchopulmonary collaterals. Aorto-graphy was performed in three patients withthromboembolic pulmonary hypertension andthis also showed bronchopulmonary collater-als (fig 5). Aortography (fig 6) was done in fivepatients with primary pulmonary hypertensionand bronchopulmonary collaterals were notseen.

DiscussionThromboembolic pulmonary vascular diseaseembraces many syndromes, depending on thesegments of the pulmonary arterial tree thatare primarily affected.2728 Occlusion of thesmall muscular arteries and arterioles, whichwas in the past referred to as microthrom-boembolism, is now attributed to organisedthrombi rather than emboli.28 In chronicthromboembolic pulmonary hypertension theobstruction is in the larger proximal and inter-mediate pulmonary arteries. This type ofthromboembolic obstruction can be treatedby pulmonary thromboendarterectomy.'Generally by the time the diagnosis is madethe clots become organised into the vesselwall, and interpretation of the pulmonaryangiogram can be difficult.' 7 The medialhypertrophy of pulmonary arteries in throm-boembolic pulmonary hypertension is oftenmild or absent, suggesting that vasoconstric-tion does not play an important role.27 In prox-imal thromboembolic pulmonary hypertensionthe smaller pulmonary arteries and arteriolesare generally unaffected, allowing a pressuregradient between the potential bronchopul-monary anastomoses. There is also an initialinjury to the lung parenchyma from pul-monary infarction. On the other hand, inprimary pulmonary hypertension the predomi-nant vascular lesions are in the small musculararteries and arterioles.29 The changes are bilat-eral, widespread, uniform, and not confined toa particular segment of the lung.29 30 Atnecropsy in primary pulmonary hypertensionthere is no embolic source and in one studythromboembolic lesions were present in lessthan 3% of small pulmonary arteries.30Fibrinoid necrosis and plexiform lesions,believed to be related to intense vasoconstric-tion, are regularly observed in advancedstages.2The bronchial arteries, usually two major

arteries to each lung arise from the thoracicaorta or an intercostal artery. The rightbronchial arteries usually arise from the inter-costal arteries, generally the first intercostal.The left bronchial arteries commonly arisedirectly from the aorta. Rarely the bronchialarteries originate from the internal mammaryor coronary arteries and can therefore bemissed on aortography.'4 At the level of therespiratory bronchioles, the bronchial arteries

VISUALISATION OF BRONCHOPULMONARYCOLLATERALSSelective bronchial arteriography (fig 4) was

done in seven of the nine patients with throm-

give rise to capillaries which communicatewith pulmonary capillaries.29 The bronchialveins from the extrapulmonary airway finallydrain into the right heart. The bronchial veinsfrom the intrapulmonary airways and lungparenchyma drain through the "bronchopul-

174 on M

arch 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.171 on 1 August 1997. D

ownloaded from

Bronchopulmonary coUlaterals in chronic thromboembolic pulmonary hypertension

monary collateral" flow into the left atrium.'4Thus sampling from the left ventricle belowthe mitral valve would detect bronchopul-monary collateral flow. The bronchial arterialblood flow in the human lung is of the order of1-2% of the cardiac output.29 Many chronicand acute lung diseases are associated with anincrease in bronchial blood flow. This hasbeen shown for bronchiectasis and granulo-mas, pulmonary interstitial fibrosis, bron-chogenic carcinoma, pleural disease, asthma,bronchitis, emphysema, and acute conditionslike pneumonia and adult respiratory diseasesyndrome. 14

Ligation of a pulmonary artery in the dogresults in the development of bronchopul-monary collaterals to the affected side, and thecollateral blood flow could be as much as onethird of the output of the right ventricle.'5Malik and Tracy'7 used beads to embolise thepulmonary circulation to a degree sufficient toincrease the pulmonary artery pressure andpulmonary vascular resistance to about threetimes the normal value. They observed thatthe bronchial blood flow did not change atfive minutes, decreased at six minutes, andwas greatly increased to around 300% attwo weeks postembolisation.'7 Orell andHultgren'8 carried out postmortem bronchialarteriography in 28 cases with pulmonarythromboemboli of varying ages. In 18 of the23 cases with emboli more than two weeksold, there was filling of the pulmonary arterydistal to the occlusions through bronchopul-monary collaterals. In this study, bronchopul-monary arterial anastomoses at the level of 1-2mm thick peripheral bronchi were the mostcommon. Matsuda21 performed bronchialarteriography in 23 patients with documentedpulmonary embolism. All 17 patients in thechronic stage with no radiological infiltrateshad bronchial to pulmonary arterial collateralswith antegrade filling of the distal pulmonaryarteries. By contrast only one out of six hadsuch collateral vessels when studied at a timewhen radiological infiltrates were still present.

Injection of indocyanine green in theascending aorta and sampling below the mitralvalve will detect an aortopulmonary shunt.None of our patients had a patent arterialduct. Significant bronchopulmonary collateralflow was present in all eight of our patientswith thromboembolic pulmonary hypertensionin whom the assessment was made. On theother hand there was no detectable bron-chopulmonary collateral flow in 11 of 15patients with primary pulmonary hyperten-sion. In the remaining four patients the flowwas small. The difference in bronchopul-monary collateral flow between thromboem-bolic pulmonary hypertension and primarypulmonary hypertension patients was highlysignificant (fig 3). The cardiac and pulmonaryvascular indices were comparable in thethromboembolic and primary pulmonaryhypertension patients.Normal bronchial arteries may be visualised

on routine cineangiography when contrastmaterial is injected into the descending tho-racic aorta, but bronchopulmonary communi-

cations are not seen. In our study, selectivebronchial arteriography or aortographyshowed that all patients with thromboembolicpulmonary hypertension had dilated bronchialarteries and a mesh of bronchopulmonary col-laterals. The bronchopulmonary collateralswere seen to fill the pulmonary arterialbranches downstream of the site of embolicocclusion. Aortography did not show dilatedbronchial arteries or bronchopulmonary col-laterals in any of the patients with primary pul-monary hypertension. Selective bronchialarteriography was not performed in patientswith primary pulmonary hypertension as theyhad either undetectable or minimal bron-chopulmonary collateral flow.

Pulmonary hypertension per se is clearly notthe stimulus for the development of bron-chopulmonary collaterals in thromboembolicpulmonary hypertension since the increase inpulmonary artery pressure was equally severein primary pulmonary hypertension. Thedevelopment of bronchopulmonary collateralsin several acute and chronic lung conditionssuggests that an initial stimulus may be dam-age to the lung tissue, and such damage is not afeature of primary pulmonary hypertension.Experimental and clinical observations showthat the development of bronchopulmonarycollaterals takes about 10-14 days after theembolic episode and the collaterals increasewith time.'7 1821 The short time taken for thecollateral development suggests that any dif-ference in the length of survival of patientswith thromboembolic versus primary pul-monary hypertension is unimportant. Theobstructive pulmonary artery lesions withthromboembolic pulmonary hypertension arepatchy and largely spare the small pulmonaryarteries and arterioles.27 This would allow apressure flow gradient to develop between thesystemic bronchial circulation and the pul-monary circulation distal to the block in theproximal pulmonary arteries, thus encourag-ing the development of bronchopulmonarycollaterals.

CONCLUSIONSLarge bronchopulmonary collaterals and anincrease in collateral flow were found inpatients with thromboembolic pulmonaryhypertension and not in primary pulmonaryhypertension. These bronchopulmonary col-laterals can be demonstrated on thoracic aor-tography without resorting to selectivebronchial arteriography or an estimation ofcollateral blood flow.

This study helps in our understanding ofsome of the pathophysiological differencesbetween thromboembolic pulmonary hyper-tension and primary pulmonary hypertension.The use to which this observation is put coulddepend upon the clinical situation. In the con-text of thromboembolic and primary pul-monary hypertension, the demonstration ofbronchopulmonary collaterals establishes thepresence of proximal pulmonary thromboem-bolism. At present, pulmonary angiographymay be considered essential while planningpulmonary thromboendarterectomy. The

175 on M

arch 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.171 on 1 August 1997. D

ownloaded from

Endtys, Hayat, Cheenan

demonstration of bronchopulmonary collater-als by thoracic aortography could be one wayof selecting critically ill patients with severepulmonary hypertension and suspectedthromboembolism for pulmonary angiogra-phy. The demonstration ofbronchopulmonarycollaterals is also likely to help in the differenti-ation of patients with thromboembolic pul-monary hypertension from those with primarypulmonary hypertension and in situ thrombo-sis. Bronchopulmonary collaterals would notbe expected to develop in patients with pre-existing primary pulmonary hypertension.However, we did not have any patients in thisclinical subset.

We thank Mrs Augustilia G Pinto for her help in the preparationof this manuscript.

1 Moser KM, Auger WR, Fedullo PF, Jamieson SW.Chronic thromboembolic pulmonary hypertension: clini-cal picture and surgical management. Eur RespirJI 1992;5:334-42.

2 Rich S. Primary pulmonary hypertension. Prog CardiovascDis 1988;31:205-38.

3 D'Alonzo GE, Barst RJ, Ayres SM, Bergofsky EH,Brundage BH, Detre KM, et al. Survival in patients withprimary pulmonary hypertension: results from a NationalProspective Registry. Ann Intern Med 1991;115:343-9.

4 Uren NG, Oakley CM. The treatment of primary pul-monary hypertension [editorial]. Br Heart J 1991;66:119-21.

5 Higenbottam TW, Spiegelhalter D, Scott JP, Fuster V,Dinh-Xuan AT, Caine N, et al. Prostacyclin (Epo-prostenol) and heart lung transplantation as treatmentsfor severe pulmonary hypertension. Br Heart J 1993;70:366-71.

6 D'Alonzo GE, Bower JS, Dantzker DR. Differentiation ofpatients with primary and thromboembolic pulmonaryhypertension. Chest 1984;85:457-61.

7 Nicod P, Peterson K, Levine M, Dittrich H, BuchbinderM, Chappins F, et al. Pulmonary angiography in severechronic pulmonary hypertension. Ann Intern Med1987;107:565-8.

8 Gefter WB, Hatabu H, Holland GA, Gupta KB, HenschkeLI, Palevsky HI. Pulmonary thromboembolism: recentdevelopments in diagnosis with CT and MR imaging.Radiology 1995;197:561-74.

9 Ricou F, Nicod PH, Moser KM, Peterson KL. Catheterbased intravascular imaging of chronic thromboembolicpulmonary disease. Am J Cardiol 1991;67:749-52.

10 Pruszezyk P, Pacho R, Chlebus M, Filipecki S, Zielinski J,Torbicki A. Non invasive visualization of pulmonaryartery thrombi in patients with suspected pulmonaryembolism considered for aggressive therapy. Spiral CT

versus transoesophageal echocardiography [abstract]. EurHeartJ 1996;17(suppl):525.

11 Moser KM, Fedullo PF, Findbeiner WE, Golden J. Dopatients with primary puimonary hypertension developextensive central thrombi? Circulation 1995;91:741-5.

12 Liebow AA, Hales MR, Harrison W, Bloomer W, LindskogGE. The genesis and functional implications of collateralcirculation of the lungs. Yale J Biol Med 1950;22:637-50.

13 Endrys J, Kral B, Elias J, Hamet A. Left ventricular over-load caused by collateral pulmonary circulation inchronic pneumopathies-myth or reality? Cor Vasa1981;23:121-5.

14 Deffebach ME, Charan NB, Lakshminarayan S, Butler J.The bronchial circulation. Small, but a vital attribute ofthe lung. Am Rev Respir Dis 1987;135:463-81.

15 Liebow AA, Hales MR, Bloomer WE, Harrison W,Lindskog GE. Studies of the lung after ligation of thepulmonary artery. II. Anatomical changes. Am J7 Pathol1950;26: 177-95.

16 Cudkowicz L, Abeimann WH, Levinson GE, KatznelsonG, Jreissaty RM. Bronchial arterial blood flow. Clin Sci1960;19:1-15.

17 Malik AB, Tracy SE. Bronchovascular adjustments afterpulmonary embolism. JAppl Physiol 1980;49:476-81.

18 Orell SR, Hultgren S. Anastomoses between bronchial andpulmonary arteries in pulmonary thromboembolic dis-ease. Acta Pathol Microbiol Scand 1966;67:322-38.

19 Cabrol C, Cabrol A, Acar J, Gandjbakhch I, Guiraudon G,Laughlin L, et al. Surgical correlation of chronic postem-bolic obstructions of the pulmonary arteries. J ThoracCardiovasc Surg 1978;76:620-8.

20 Mills SR, Jackson DC, Sullivan DC, Moore AV, HeastonDK, Wolfe WG, et al. Angiographic evaluation of chronicpulmonary embolism. Radiology 1980;136:301-8.

21 Matsuda A. Bronchial arteriography in patients with pul-monary embolism. Chest 1984;85:767-73.

22 Fishmann AJ, Moser KM, Fedullo PF. Perfusion lung scanvs pulmonary angiography in evaluation of suspected pri-mary pulmonary hypertension. Chest 1983;84:679-83.

23 D'Alonzo GE, Bower JS, Dantzker DR. Differentiation ofpatients with primary and thromboembolic pulmonaryhypertension. Chest 1984;85:457-61.

24 Anderson G, Reid L, Simon G. The radiographic appear-ances in primary and thromboembolic pulmonary hyper-tension. Clin Radiol 1973;24:113-20.

25 Nakamura T, Katori R, Miyazawa K, Ishikawa K, YamakiM. Kobayashi Y. Aortopulmonary collateral flow inpatients with pulmonary disease. Am Rev Respir Dis1968;98:464-73.

26 YousofAM, Endrys J, Bicik V. Simple method of measuringdye curve area by planimetry using the half area method.JAppl Physiol 1974;36:611-3.

27 Wagenwoort CA, Denolin H. Pulmonary circulation-advances and controversies. Amsterdam: Elsevier, 1989:18-20.

28 Fishman AP. Pulmonary hypertension. In: Schlant RC,Alexander RW, eds. Hurst's The heart. New York:McGraw Hill, 1994:1862-3.

29 Fishman AP. Pulmonary diseases and disorders. New York:McGraw-Hill, 1988:995-6.

30 Anderson EG, Simon G, Reid L. Primary and thrombo-embolic pulmonary hypertension: a quantitative patho-logical study. J Pathol 1973;110:273-93.

176 on M

arch 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.171 on 1 August 1997. D

ownloaded from