A study of the pathology and pathogenesis of bronchiectasis.

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS221cylindrical bronchiectasis ofthe left lower lobe andlingula, with minimalchanges in the right middlelobe. Left lower lobectomywith removal of the lingulawas performed.Macroscopic Appearanceof Specimen.-The lingula issmall and firm, and has a Jshiny, thin pleural covering.Sections show the bronchi eto be dilated, thickened,filled with pus, and surroundedby interstitial pneumonia.The microscopicappearance is similar to thatof the posterior basal segment(q .v.). > v,t'-The lower lobe appearswell aerated, though thebasal segments contain some .jfirm areas. The pleuralmembrane is normal: thereare enlarged hilar lymph 4-glands.Sections of the lobe (Fig.17) show more extensive"disease than in Case 464.There is marked thickeningand dilatation of all branchesof the posterior basal bronchus,and extensive pneumoniain this segment.Changes in the middle and 4santerior basal segments areof lesser severity, and are FIG. 17-Moderate follicularcomparable wi.dh those seenin the posterior basal segment of the previous case.No abnormality is to be seen in the more pigmentedapical segment.Microscopic Features.-The appearance of lesionsin the middle and anterior basal segments is similar tothat of the previous case.In the posterior basal bronchus there is diffuse infiltrationof chronic inflammatory cells in the sub-epithelialtissues The ducts of mucous glands are funnelshapedinstead of being tubular, and lymph folliclessurround their mouths. Bronchial supporting structuresare normal except for loss of elastic tissue near thefollicles.The branches of the posterior basal bronchus aredilated, thickened, and filled with pus.The epithelium consists of ciliated columnar cells,with no areas of ulceration or squamous metaplasia.Inflammatory cells are profuse in the oedematous walls;lymph follicles are numerous and often confluent.There is a more complete loss of bronchial supportingtissues than in the previous case, only isolated fragmentsremaining.pbronchiectasis of left lower lobe (Case 758). Iron haematoxylin andvan Gieson.All bronchioles in the segment are diseased, most ofthem being partly occluded, while some appear almostobliterated. Epithelial changes follow those in the previousspecimen, but the inflammatory reaction is moreextensive, lymph follicles are more numerous, and thereis almost total loss of muscle and elastic tissue.In some fields the zones of interstitial pneumoniaare so wide that they form a confluent mass. As thereis little thickening of interlobular septa, the originallobular pattern is obliterated. No alveolar collapse ispresent.SEVERE FOLLICULAR BRONCHIECTASISCASE 497.-A 9-year-old girl, who had suffered froma productive cough since measles in early childhood,was admitted with an acute exacerbation of symptoms.She was dyspnoeic on slight exertion, showed fingerclubbing, and bronchograms revealed bronchiectasis ofthe left lower lobe and lingula. Left lower lobectomywith excision of the lingula was performed.Microscopic Appearance of Specimen.-The lingulais small and firm, and shows gross bronchiectasis with

222F. WHITWELLt!.^FIG. 18-Severe follicular bronchiectasis of right lower lobe 2 cm. frcHaematoxylin and eosin.interstitial pneumonia and collapse. The microscopicfeatures do not differ from those at the base of thelower lobe.The lower lobe is quite bulky and firm, and hasenlarged hilar lymph glands. The pleural membrane isthickened over the base of the lob,, but normal at theapex, which is partly aerated.On section (Figs. 18 and 19) there is mild bronchiectasiswith some pneumonia in the partially aerated apicalsegment, while the basal segments all show gross bronchiectasiswith pneumonia, collapse, and some fibrosis.It is mainly in the segmental bronchi and their first andsecond branchings that dilatation is seen: more peripheralbronchi look like narrow clefts in solid tissue.The disease is most severe in the lower parts of the basalsegments.Microscopic Features.-Though the gross appearanceis far removed from that of the previous two specimensthere are no qualitative differences in the histology.Changes in the apical segment are similar to thosedescribed in mild lesions of the previous cases. Thelesions in each basal segment areof equal severity.There is some dilatation but littlethickening of the basal segmentalbronchi, and while epithelial liningsare normal, there is diffuse inflammatoryinfiltration of sub-epithelialtissues with some lymph follicleW. formation. Elastic tissue is extensivelydestroyed, but the loss ofmuscle is only slight and cartilagesare normal. Mucous glands aresmall and scanty; they lie close tothe bronchial epithelium and havevery short ducts.The proximal branches are dilatedand distorted, but more peripheralones are flattened. Their histologyis similar to that described incorresponding bronchi of the406,01 previous case.In some areas the lesions aresimilar to those in severely affectedparts of previous specimens (Fig. 20).Sometimes, particularly near the4. i~; diaphragmatic surface of the lobe,it is impossible to define the outerwalls of air passages, or to distinguishbronchi from bronchioles.Granulation tissue in the bronchialand bronchiolar walls merges withthe confluent interstitial pneumoniaforming a granulomatous mass,which contains epithelium-linedclefts and patches of collapse.There are so many lymph folliclesthroughout this granulomatous tissueom hilum (Case 497). that the pulmonary lobules resemblelymph glands. No bronchial supportingtissues survive; thickened interlobular septa,however, help to identify the original outline of lobules.The close relationship of thickened branches of thepulmonary artery to the epithelium-lined clefts suggeststhat the latter were originally bronchi and bronchioles,and unlikely to be re-epithelialized pulmonary excavations.THE PATHOLOGY OFFOLLICULAR BRONCHIECTASISThese examples demonstrate that lesions mayrange from a relatively minor and inconspicuousfocus in one bronchopulmonary segment of anotherwise normal lobe, to formation of a more orless solid lobe which bears little resemblance tothe normal.In this investigation over a third of the specimensshowed features similar to the illustrated examples,with their typical microscopic lesions, the mostconstant and characteristic of which were the forma-

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS223tion of lymph follicles and interstitialpneumonia.LYMPH FOLLICLE FORMATIONSeverely diseased bronchi and bronchiolescontain many lymph folliclesand nodes within their thickenedwalls, but, although this lymphoidtissue is extremely prominent, itforms only part of an extensivemural inflammation. At this stage of 5*the disease there is widespread destructionof bronchial elastic tissue,muscle, cartilage, and mucous glands.From the examination of manyspecimens it appears that elastic tissueis the first of the supporting tissuesto be destroyed in bronchiectasis,muscle is the next to suffer, and cartilagesurvives the longest. Mucousglands are also destroyed, survivingfor about as long as cartilage. It isthought that once any of these FIG. 19.-Same lobe as in Fig. 18, 5specialized tissues has been destroyedit cannot regenerate.In areas of very mild bronchiectasis theonly abnormalities in the bronchial tree aresub-epithelial accumulations of lymph folliclesand nodes, and some oedema. Diffuseinflammation is not present and both muscle =and cartilage appear normal; special stains,however, show that there is destruction ofelastic tissue near the lymph follicles (Fig.16). It is considered that this loss of elastictissue is a fundamental lesion in follicularbronchiectasis, for once part of the bronchialtree has been damaged in this manner it ispermanently weakened and incapable ofovercoming secondary infections.Apart from their destructive influence onelastic tissue, the lymph follicles, merelybecause of their size, distort the affectedbronchial tree. Sub-epithelial follicles projectinto the lumina causing partial obstruction,while intramural follicles enlargebronchial walls, compressing the openings Fof peripheral branches of the bronchialtreeSimilarwhichbutenter themorediseasedsevere effectsbronchi.are seen in tthe bronchioles, where complete occlusion ; .4is often produced.Lymph follicles and nodes are present innormal lungs; Miller (1947) and Engel(1947) agree that lymphoid tissue occursat points of bifurcation of bronchi and' FIG. 20.- Follicular broncl- .- - WI .1.cm. from hilum.-hiectasis. Haematoxylin and eosin x 10.

224F. WHITWELLbronchioles, and between bronchioles and branchesof pulmonary arteries which accomp.any them. Astudy of normal lung segments has confirmed this,and has shown that this lymphoid tissue is outsidethe bronchial tree, where no alteration of elastictissue is to be seen.Abnormal lymphoid accumulations in the lungare sometimes seen in other forms of bronchiectasis,around chronic lung abscesses, and in tuberculosis.In these conditions the number of follicles is muchsmaller and they are mostly situated outside thebronchial tree, but local destructive changes dooccur near the occasional ones found in the bronchi.Many authors (Robinson, 1933; Ogilvie, 1941;Allison, Gordon, and Zinnernann, 1943) havementioned the presence of these bronchial lymphfollicles in bronchiectasis, but their significance hasnot been studied, and only Engel (1947) has notedthe associated loss of elastic tissue. He has calledthis condition " nodal bronchitis and bronchiolitis"and has described thickened collapsed bronchi andbronchioles where the walls contained lymphnodules but no diffuse cellular infiltration. Amarked loss of elastic tissue occurred near theJK3&NORMALEARLY CHANGA.9 ,LATE CHANGESLOBLFIG. 21.-Development of bronchiolar lesions and secondary lobular cfollicular bronchiectasis.rSESnodules, and this was noted too often for it to befortuitous. Engel commented that he was unableto decide whether this lesion was caused by avirus. an unusual infection, or some other influence.INTERSTITIAL PNEUMONIAInterstitial pneumonia is to be seen in otherpulmonary conditions, but it is a constant featureof follicular bronchiectasis. The pneumonia isfound in the parenchyma around affected bronchioles,and its extent is dependent upon theseverity of the bronchiectasis.In normal lung tissue the respiratory part of theorgan appears in section as a lace-work of largespaces separated from one another by thin-walledsepta (Maximow and Bloom, 1942). These fineinter-alveolar septa contain capillaries, elasticfibres, and a fine reticulum of connective tissue;and in normal parenchyma such interstitial tissuesare inconspicuous.In interstitial pneumonia, however, the peribronchiolarand inter-alveolar connective tissues areoedematous, and contain infiltrations of plasmacells, mononuclears, and lymphocytes, togetherwith occasional lymph follicles, dilatedcapillaries and lymphatics. The vastincrease in bulk of interstitial tissues isassociated with a reduction in volumeof the alveoli, and an alteration of theirlining cells.Cells lining normal alveoli are/- inconspicuous, but in interstitialpneumonia the alveoli are lined byan epithelium of small cuboidal cells.The origini of this epithelium is obscure;some think it is a down-growth ofbronchiolar epithelium, others regardDv. it as a metaplasia of cells normallyfound lining alveoli, while Watts and___2fistMcDonald (1948) think that both these',4RX, processes occur.In mild bronchiolar lesions only thosealveoli actually in contact with thebronchiolar walls appear to have undergonethis epithelialization; in fact,in many instances only parts of thesealveoli are affected. As these alteredalveoli are often only found in theperiphery of pulmonary lobules, itseems unlikely that the lining cells area bronchiolar down-growth, and moreULE probable that local metaplasia hasoccurred as a result of interstitial insollapeinflammation.

PA THOLOG Y AND PA THOGENESIS OF BRONCHIECTASIS225CLINICAL FEATURES OF IFOLLICULAR BRONCHIECTASISThe ages of all the patients were obtained fromthe register, but only two-thirds of the case-recordswere available for analysis.The age distribution of all the cases of follicularbronchiectasis is shown in Fig. 4. Other detailswere obtained from the 28 case-records, and are asfollows:Under 6 yrs.6-8 yrs. .9-1l yrs. ..12-14 yrs. .15-17 yrs. .18-20 yrs. .Over 20 yrs.IUnder 3 yrs.3-6 yrs. . .7-9 yrs. ..10-13 yrs. . .14-18 yrs. . .Over 18 yrs.AGE AT OPERATIONAGE AT ONSFT OF SYMPTOMS1 patient9 patients84 ,4 ,,2 ,0 -1 5 patients7 ,,2 ,,2 ,,2 ,,0 ,,PRE-OPERATIVE DIURATION OF SYMPTOMSUnder 1 yr. 0 patients1-3 yrs. . 7 ,,4-5 yrs. . 76-10yrs... 1011 5Iyrs. 2Over 15 yrs. 2NATURE OF ILLNESS AT ONSET OF SYMPTOMSMeasles and!or whooping-cough13 patientsPrimary bronchopneumonia ..6Insidious onset ..7Tonsillectomy ..1 patientCrush injury of chest ..I ,,PresentDoubtfulAbsentNot notedPresentDoubtfulAbsentNot notedFINGER CLUBBINGPARANASAL SINUS INFECTIONS10 patients410 420 patients43I patientThe severity and extent of the disease appearedto be unrelated to the nature or severity of theinitial illness, or to the duration of the symptoms.Patients with finger clubbing did not have themore extensive or severe pulmonary lesions; therewas also no correlation between finger clubbingand duration of symptoms.There was obvious correlation between thenature of the patients' sputa and the degree ofinflammation in the bronchi.PATHOGENESIS OF FOLLICULAR BRONCHIECTASISThe main lesions occur in smaller bronchi,bronchioles, and in alveoli, while minor secondarychanges take place in the larger bronchi. The moresevere the involvement of the lobe, the moreproximally does each of these processes extend, withobliteration of the distal bronchial tree. Secondarychanges in the larger bronchi include the destructionof supporting tissues around the mucous glandducts, which then enlarge and form part of thebronchial wall.Small bronchi and bronchioles are affectedsimilarly; only the bronchiolar changes will bedescribed as these are more closely associated withthe parenchymal disease. By studying lesions ofdifferent severity it is possible to visualize theevolution of the advanced condition (Fig. 21).The early changes consist of thickening ofbronchiolar walls due to oedema and lymph follicleformation, with destruction of elastic tissue nearthe follicles. These bronchioles are usually ratherflattened, and are surrounded by narrow rims ofinterstitial inflammation.More advanced lesions show a greater number offollicles and a diffuse mural bronchiolitis, withdestruction of all elastic tissue and most of themuscle. The bronchiolar lumina are reduced andoften appear stellate on cross-section. Interstitialinflammation involves wider areas of the parenchyma,and there is fibrosis around branches of thepulmonary arteries. Often the interlobular septaand pleural membrane are thickened. ~FAlveol,ar collapse has been seen only in specimenswhere there was thickening of interlobular septaand pleural membrane. Microscopically this collapseis of the type produced by compression, andits distribution is lobular.The cause of the alveolar collapse can be foundin the lobules (Fig. 21). Thickened interlobularsepta and pleural membrane keep the outline of alobule rigid, but as the centre contains an expandinggranulomatous mass (which has arisen from thebronchiole and peribronchiolar tissues) pressurewithin the lobule becomes raised. This tensionproduces both flattening of the weakened bronchiolarwalls, which reduces the lumen to a narrowcleft, and compression collapse of the alveolioutside the area of pneumonia.THE AETIOLOGY OF FOLLICULAR BRONCHIECTASISThe important clinical facts are that the conditionusually begins in early childhood and is asequel to whooping-cough, measles, or primarybronchopneumonia. This bronchiectasis may alsohave an insidious onset.The essential lesions are destructive muralbronchitis, bronchiolitis, and interstitial inflammation,and the pathology appears very closely relatedto the acute interstitial pneumonias which werereported by MacCallum (1940) in fatal cases ofmeasles and influenza and whooping-cough.MacCallum described the lesions found in servicepersonnel, but Engel (1947) has reported a similar

226F. WHITWELLFIG. 22.-Saccular bronchiectasis of right lower lobe (Case 395)2.5 cm. from hium. Haematoxylin and eosin >' 1.5.multifocal condition in children, under the nameof destructive mural bronchiolitis. MacCallum saysthat this condition always follows a virus infectionand haemophilic organisms can usually be grownfrom the lungs, and Engel has suggested a viralorigin for his " nodal bronchiolitis."Follicular bronchiectasis is very like rodentbronchiectasis (Cruikshank, 1948), which has beenshown by Nelson (1946) to be of viral origin.Follicular bronchiectasis is probably the sequelof acute viral infection of the lungs contracted inearly childhood. Usually this infection occurs withwhooping-cough or measles, or, when influenzal itis described as primary bronchopneumonia. Insome cases it may be relatively symptomless, likeother virus pneumonias, and pass unnoticed byparents. The original inflammation is probablymultifocal, and normally clears up without anyresidual damage, but in bronchi which are inadequatelydrained of secretions the permanent changesof follicular bronchiectasis may arise.SACCULAR BRONCHIECTASIS"Saccular " is an apt word because it fits boththe anatomical and the bronchographic appearance,though many specimens which did not show thisFIG. 23.-Saccular bronchiectasis of right lower lobe 4 cm. fromhilum (Case 395). Haematoxylin and eosin x 1.5.type of pathology could have been described assaccular from the bronchograms.Twenty-three examples of this condition occurredin the present series. Two of them will be describedin some detail, and the pathology, clinical features,and pathogenesis of saccular bronchiectasis will bedi-cussed.CASE 395.-A 33-year-old woman, previously healthyapart from bad head colds, had suffered from a troublesomecough for three years, and her symptoms hadbecome more severe recently. Her sputum was copious,purulent, and occasionally blood-stained. She hadfinger clubbing and was dyspnoeic on slight exertion.Bronchograms showed bronchiectasis limited to theright middle and lower lobes. Right lower lobectomywas performed.Macroscopic Appearance of Specimen.-The specimenis an aerated, pigmented lobe with pleural thickeningover the apical and diaphragmatic surfaces.On section (Figs. 22 and 23) the apical segment showssome collapse, emphysema, and interstitial pneumoniabut the bronchi appear normal. The base of the lobeshows gross bronchiectasis, with the main bronchi ofnormal calibre, though slightly thickened and occludedby folds of hypertrophic epithelium, while more peripheralbronchi form pus-filled cavities. These sacculesare surrounded by aerated alveoli, and there isfibrous thickening of the interlobular septa.

X * a! tvo,,,, s^

~~~~~~~~~~~~~~~~~~~~~~.228F. WHITWELLThe pre-saccular bronchishow marked polypoid changesin the epithelium. In Fig. 25polyposis of the basal segmentalbronchi can be compared withthe smooth epithelial lining ofthe normal apical bronchus.The epithelial lining of thesaccules is flatter than in theprevious specimen, and areas ofsquamous metaplasia are moreextensive. No connexion wasfound between the saccules andany distal bronchial tree-infact no intact distal bronchi arepresent, though respiratory bronchiolesoccur in the parenchyma.The base of the lobe is wellaerated except for a narrow zoneof collapse around and betweenthe saccules. In the areas of collapsethe alveoli are flattened concentricallyaround the saccules.There is no inflammatoryreaction in the parenchyma, andonly within the peribronchial,perisaccular, and interlobularsepta is an increase of fibroustissue to be found.THE PATHOLOGY OF SACCULARBRONCHIECTASISFIG. 26.-Saccular bronchiectasis 5 cm. from hilum (Case 451). Haematoxylin and eosin x 1.5. These two examples arethought to be the end-resultconspicuous by thickening of interlobular fibrous septa. of a distinctive t: ;ype of bronchiectasis. AlthoughWithin these lobules there are some respiratory bron- only 17% of s,pecimens show these lesions,chioles, but terminal bronchioles are replaced by fibrous some of the unclatssifiedlobes are probably earlierscars. stages of the ssame condition. CharacteristicCASE 451.-An 18-year-old boy was admitted to features are grossloss of bronchial structures in thehospital because of a troublesome cough and a single saccules, normaliity of the alveoli around theslight haemoptysis. Bronchitis had occurred during thesprevious four winters, but there was no history of earlier polyposis" of the pre-saccularandsaccules, rrespiratory complaint or infectious diseases. He pro- bronchi,*etain normal supporting tissues.whlch rduced 3 oz. of purulent sputum daily, had marked finger The saccules and the pre-saccular bronchi can beclubbing, and bronchograms showed bronchiectasis considered separaltely, the former being regarded aslimited to the left lower lobe. This lobe was excised. the primary abno irmality.Macroscopic Appearance of Specimen.-The lobe is SACCULES.-Saccules are essentially fibrous struc-epithelium; they have nobulky, aerated, slightly pigmented, and has a shiny tures lined by a ciuboidalpleural membrane.elastic tissue, mus,cleor cartilage in their walls, andSections of the lobe (Figs. 25 and 26) show a normal their lining memtbranes contain no normal ciliatedapical segment but severe bronchiectasis in the whole respiratory epithelium.This absence of recogniz-saccules has led to speculationof the base. The segmental basal bronchi and their main able tissues in thiebranches are not thickened or dilated, but they are about their originpartially occluded by epithelial " polyposis";and this subject will be discussedmorelater.distal bronchi are saccular and filled with pus. The la er.alveoli of the basal segments are aerated, except imme- The constant finding of areas of squamousdiately around the saccules where there is some collapse. metaplasia in the saccular epithelium is of specialMicroscopic Features.-There is obvious similarity interest as this p] ohenomenon is uncommon in any.to the previous case, and only points of difference will of the other types of bronchiectasis examined. Thebe stressed.usual distribution of this metaplasia is shown in the

_ * _ * ........... r _ r- -[two examples (Figs. 24 and 27); sometimes itis less extensive, but in a few specimens thewhole saccular lining is replaced by squamous Iepithelium extending for some distance up thepre-saccular bronchi.Squamous metaplasia is mentioned in most_Xaccounts of bronchiectasis in the literature,but its distribution in the diseased lobes has ararely been described. From published illustrationsit is clear that the term has been applied indiscriminatelyto various epithelial patterns, suchas the layers of heaped-up, flattened cells thatrestore breaches of epithelial continuity, and thetransitional epithelial appearance seen in obliquely -~ rsectioned bronchi (Engel, 1947). In the presentaccount the term is used only when prickle cellsor keratinization have been seen; the latter isextremely rare and was found only in threeFIG. 27.-Model of basal segments of Case 451 (saccularspecimens.bronchiectasis).e§;ev16Robinson (1933, 1939) desei / l{is'>

230F. WHITWELLFIG. 29.-Saccules surrounded by rims of compression collapse (CasHaematoxylin and eosin x 7.5.monary tuberculosis; Willis's statement cannottherefore be accepted.THE PRE-SACCULAR BRONCHI.-NO dilatationoccurs in the bronchi and there is little destructionof supporting tissues, despite the severe inflammatoryreaction in the bronchial walls. The mainabnormality is polyposis of the bronchial epithelium,often extensive enough to produce considerablebronchial obstruction.In the present material this change has been seenonly in saccular bronchiectasis, but similar lesionsare to be found in the bronchi draining chroniclung abscesses and secondarily infected tuberculouscavities.The bronchoscopic and histological features ofthis polypoid change have been described byJackson and Jackson (1932), who thought that itwas an inflammatory reaction produced by continualirritation from pus, an opinion supported bythe present study. Peroni (1934) considered thatthe polyps occurred first, and then producedbronchiectasis by obstruction. Samsonj (1940) described a severe case and thought2 the cause was an individual susceptibilityto mucin. Many accounts of the pathologyof bronchiectasis mention the hypertrophicbronchial epithelium, but the site andassociation with saccules have not beenstressed.COLLATERAL AIR CIRCULATION IN. DISEASED SEGMENTS.-Tfle observation thatall bronchi in the diseased segments end inblind saccules, and that these segmentsare well aerated, is considered to be ofsome importance. It provides histologicalconfirnation of the existence of a collateralair circulation in bronchiectasis, whichchallenges the validity of certain populartheories of pathogenesis.i Hf Collateral air circulation was originallya theoretical conception based upon a* belief in the existence of alveolar pores.it, Later this theory was tested and confirmedby animal experiments (Van Allen andJung, 1931), and more recently by Baarsmaand Dirken (1948) using healthy rabbits,and by Baarsma, Dirken, and Huizinga(1948) in normal human lungs. Theseworkers considered that a collateral aircirculation occurred only in the absence ofinflammatory change in the obstructedsegments.;e 451). The importance of this mechanism indiseased lungs has only recently receivedany attention. Churchill (1949), describing theradiological appearances of isolated bronchiectaticlobes in which the bronchi had been filled withlipiodol, mentioned dilated bronchi extending intoair-containing parenchyma but in no way communicatingwith it. These segments contained trappedair, as they did not deflate when the lobes wereremoved from the body and left the bronchi open.This trapped air serves the function of filling upspace but it has no oxygenating value.In the two examples of saccular bronchiectasisthe air must have reached the basal parenchymathrough the normal apical segments. This suggeststhat the intersegmental and interlobular fibroussepta are really incomplete fibrous networks whichoffer no barrier to air. Alveolar pores have beenseen in the aerated alveoli of most specimenswhenever sought in thick sections (Fig. 31).Although collateral air circulation is most easilydemonstrated in the saccular cases, it occurs inmost forms of bronchiectasis which do not involveentire lobes.

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS 231CLINICAL FEATURES OF SACCULAR BRONCHIECTASISThe age distribution of all these patients is shownin Fig. 4. Sixteen case-records were available for \\NORMAL APICALSE\.analysis, and the details are tabulated.XAGE AT OFERATIONUnder 5 yrs. 0 patients 'il15 0yrs. 5 ,\_26-30yrs.7., 21-25yrs.52| 1 9 $ 9 ''_31 35 Nrs.. . I36-40yrs.I patientOver 40 yrs.AGE AT ONSET OF SYMPTOMS ORMAL BASAL SEGACough " all life time "...patientsUnder 9 vrs. (except above) 09-12yrs .. .. I patient13-18 yrs. .. 6 patients19-25yrs...Over 25 yrs. ..I patientPRE-OPERATIVE DURATION OF SYMPTOMSUnder 2 yrs.3 patien is2-6yrs. 87l1O0 rs. 211-15yrs.Ipatient16 20yrs. IOver 20yrs.INATURE OF ILLNESS AT ONSET-OF SYMPTOMS )'Measles and, or whooping cough0 patientsBronchopneumonia .. 9Fairly recent insidious onset 4 -I'Cough all life but no severe childhoodillnesses.3FINGER CLtJBBINGPresent.10 patients FIG. 30.-Development of collateral air circulation in saccularDoubtful 2 ,. bronchiectasis.Absent. 2Not recorded.2 All patients had copious purulent sputum.PARANASAL SINUS INFECTIONS No correlation was found between the extent,Present*9 patients severity, and duration of the disease, and theDoubtful .. 2Absent .. 3 " presence of finger clubbing.Not recorded .. 2It is impossible to correlate theulj|jdegree A of polyposis and squamousmetaplasia to the duration of illness.PATHOGENESIS OF SACCULARBRONCHIECTASISAs the changes in pre-saccularbronchi are considered secondaryto the saccular lesions any discussionon pathogenesis amount,tothe question, "From what and howdo the saccules arise?""SrrE OF THE SACCULES. -Writershave variously suggested that theW iw dilatations ^ w of bronchiectasis originatefrom the bronchi (Robinson,1933; Ogilvie, 1941; Mallory, 1947),the bronchioles (Lander and Davidson,1938), and the parenchyma(Opie, 1928; Erb, 1933; McNeil,~ _ 4 R Macgregor, and Alexander, 1929;IFIG. 31.-Alveolar pores seen in thick sections. Elastic stain with pyronin x 600. Lisa and Rosenblatt, 1943).

232F. WHlTWELLNeoprene casts, reconstruction models, anddissections of the bronchial tree show that sacculesare the bulbous terminations of the first to thirdbranchings of the segmental bronchi. As adultbronchioles arise only after about 18 branchingsfrom the segmental bronchi (Broman, 1923; Reid,1950). the saccules clearly cannot be of bronchiolarorigin.Saccules are the direct continuations of largebronchi, they are encased within extension of'theperibronchial fibrous septa, and they are placedclose to and parallel with branches of pulmonaryarteries. Their situation suggests that they wereoriginally bronch:, and not bronchopulmonaryexcavations.CAUSE OF DILATATION. First, it is possible toeliminate certain of the factors that are oftensuggested as causes of dilatation. Alveolar collapseis negligible in these specimens and cannot beconsidered as a possible cause, in spite of absoluteperipheral bronchial obstruction. Fibrous tissuefound in the saccular walls occurs only as a replacementof original bronchial structures. This fibrosisis concentric with the saccules, and not in a planewhere, by contracting, it could exert any dilatingforce on the bronchi. Most writers link togetherfibrosis and contraction, especially when discussingbronchiectasis. However, the behaviour of fibroustissue is perverse, and many examnples occur of itsproperty of ungoverned stretching, e.g., aorticaneurysms and in repaired hernias. Probably thestretching of fibrous tissue in saccular walls is afactor in the production of dilatation.The clinical findings of this group of patientsshow that while in many cases the bronchiectasisdated from an attack of bronchopneumonia, innearly half the patients the disease had an insidiousonset which escaped notice.It is considered that saccular bronchiectasisbegins as a chronic mural inflammation of themedium-sized bronchi. This condition destroysbronchial wall structures and obliterates peripheralbranches (Fig. 30), but is not sufficiently acute ordiffuse to produce parenchymal scarring or tointerfere with the establishment of a collateral air.circulation, provided all segments of a lobe arenot affected.The saccular shape of the diseased brolnchi isprobably produced after the destruction of supportingtissues and occlusion of peripheral branches,and results from distension by contained pus, thepressure of which is increased by the partial occlusionof pre-saccular bronchi (see Fig. 28). Support islent to this theory by the compression-collapseseen in several specimens in perisaccular alveoli(Fig. 29) suggesting that an expansile force iscentred within the saccules.The theory that dilatation is produced throughdistension of the bronchi by pus was originally putforward by Laennec, but only a few (Riviere, 1905)have shared his opinion. As early as 1838 Williamsdiscredited Laennec's theory by arguing that mostpatients with copious sputum do not developdilatation, and many patients with bronchiectasishave little sputum. Williams's argument was falsebecause it did not take into account the conditionof the bronchial walls. Andrus (1937) discussedthis subject and said:" As a problem in physics it is sufficient to noteat this time that in order to exert a dilating force itwould be necessary that the secretion occupy the grosssection of the lumen of the bronchus; such a conditionwould, however, necessarily result in atelectasis,and its possible effects be indistinguishable from thelatter."His theoretical aerodynamics did not allow forcollateral air circulation, and the illustrated examples.do not support his statement.ATELECTATIC BRONCHIECTASISIn recent years there have been many accountsof the association of lobar and segmental collapsewith bronchiectasis, and it has become widelyaccepted that the collapse leads to the bronchiectasis.In this condition, which has been calledatelectatic bronchiectasis (although it is recognizedto be an acquired disease), the collapse is usuallythought to be caused by peripheral bronchialobstruction. Many writers have gone even furtherand stated that all bronchiectasis, even whencongenital, is caused in this way.The specimens which have here been classifiedas follicular and saccular rarely showed collapse,even on microscopic exa-mination; where collapsewas found it appeared to be consequent to thebronchiectasis. However, some others of thespecimens did show severe alveolar collapse, sathey have been examined as a group in an attemptto find out whether the collapse or the bronchiectasiscomes first, and how this condition arises.The specimens lack homogeneity; in some thebronchi are collapsed, in others there is grossdilatation, and in most of them the bronchi arethickened but only moderately dilated (Fig. 32).However, the grouping has been of some valuebecause it has shown that " atelectatic bronchiectasis"differs from other forms of the disease inits lobar and segmental distribution, and in the

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS 233r-2L ~ ~~~~~ '.1 ~ ~~~~ S+. 1*2-.X~~~~~~\&"AI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ J%--4s~'''| ~ 1.4'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'- '>.P'w^* wq. -1 f\0ziSl v*) FE p 1-4-0-1M#vi *ffhQ\Bs -#--pqee"4 "'" ~*specimens form only a quarter of the whole series,'they represent three-quarters of the collapsed tspecimens.*Specimens consisting of middle with lower lobes 5 .'-^ . ;>r *-3 ido not always show atelectatic bronchiectasis in FIG. 34.-Atelectatic bronchiectasis with intra-bronchiolarboth lobes. The findings are as follows. and intra-alveolar haemorrhage with patent bronchioles..... -..- I I --.

234Middle Lobe Lower Lobe Specimens No. ofCollapse only Atelectatic bronchiectasis 2Atelectatic bronchiectasis Collapse only IAtelectatic bronchiectasis Atelectatic bronchiectasis 3Atelectatic bronchiectasis Saccular bronchiectasis ICollapse only Follicular bronchiectasis 1Atelectatic bronchiectasis Follicular bronchiectasis 1Collapse only Unclassified bronchiectasis IF. WHITWELLSEGMENTAL DISTRIBUTION.-In all segments of anaffected lobe the bronchi show similar histologicalchanges. It is very difficult to assess calibre differencesin collapsed specimens, but usually thedegree of dilatation appeared about equal in eachsegment.BRONCHIAL OBSTRUCTION.-The lobar, bronchihave been examined for stenoses and other lesionsthat might cause bronchial obstruction. None ispresent in lobes showing bronchiectasis, but inspecimens of simple collapse the bronchi are alsocollapsed and therefore obstructed.Sections have been examined microscopically forobstructing lesions of the peripheral bronchial tree.In contrast to follicular and saccular bronchiectasis,in which peripheral obstruction is invariable, thebronchioles are patent and can be followed into thepulmonary lobules, where the air cells are partlycollapsed (Fig. 34).HILAR LYMPH-GLAND CHANGES.- In many specimensthe lymph-glands around the hilar bronchiare greatly enlarged, and on section showed nonspecificchronic lymphadenitis, often with fibroticchanges. These glands are indistinguishable fromthe hilar glands in follicular bronchiectasis.In three of the middle with lower lobe specimensthere are small partly-calcified caseous foci in theperiphery of one lobe (two in middle lobes and onein a lower lobe), and a similar focus is present inthe periphery of a left lower lobe. In two ofthese specimens the hilar lymph-glands containcalcified caseous foci. Except for two specimens oftuberculous bronchiectasis and a saccular bronchiectasiswith tuberculosis in the non-bronchiectaticpart of the lobe, these are the only tuberculouslesions seen in the whole series.BRONCHIAL TREE.-The bronchi in collapsedspecimens show similar histology throughout alobe. There may be (1) moderate inflammationwithout any destruction of supporting tissues.This was also always seen in specimens of simplecollapse. (2) Slight superficial inflammation, butbronchial walls show dense fibrous thickening, andcontain no elastic tissue, muscle, or cartilage(Fig. 33). (3) Severe inflammatory changes withepithelial ulcerations and destruction of supportingtissues.The bronchioles show similar lesions, but are lessseverely affected. In some specimens inflamedbronchial walls contain numerous lymph follicles,but interstitial pneumonia is absent.The type of histological change in the bronchiappears to be unrelated to the degree of dilatation.ALVEOLI.-In spite of the gross reduction in sizeof the lobes, and the apparent total alveolar collapse,the microscopic appearances are usuallythose of incomplete absorption collapse.Many lobes seem to be infarcted; in these thereis extravasation of red cells into interstitial tissues,and the partly collapsed alveoli, and some bronchioles,are distended with blood (Fig. 34). MostFIG. 35.-Development of atelectaticbronchiectasis (a) smallfocus in periphery of middleand lower lobe; (b) enlargedhilar glands and lobar col-(t }~ 1 tlapse; (c) atelectatic bronchiectasiswith calcified foci inhilar glands.abc

PATHOLOGY AND PA THOGENESIS OF BRONCHIECTASISof this haemorrhage consists of fresh red cells, but(leposits of haemosiderin in cells of the alveolarwalls and in intra-alveolar heart-failure cells showsthat some bleeding must have occurred beforeoperation. This intra-alveolar haemorrhage wasrarely seen in other forms of bronchiectasis. Theoperative technique was similar in all cases.In the lobes showing simple collapse elastictissue in alveolar walls appears to be thickened,probably because of its relaxed state. In atelectaticbronchiectasis the alveolar wall elastic tissue iseither represented by a few short thick curls, or itis completely absent. A variable amount ofinter-alveolar fibrosis is seen but there are noother signs of inflammation.CLINICAL FEATURES OF ATELECTATICBRONCHIECTASISThe age distribution of all cases is shown inFig. 4. The case-records of 15 patients wereavailable for analysis, and provided the followingdetails.AGE AT OPERATIONUnder 5 yrs.5-10 yrs.11-1 5 yrs.16-20yrs.21-25yrs.26-30yrs.Over 30 yrs.AGE AT ONSET OF SYMPTOMSUnder 3 yrs.3-6yrs.15yrs.21-23yrs.Not stated0 patients6 332 -I patient0 ,3 patients7I patientpatients2PRE-OPERATIVE DURATION OF SYMPTOMSUnder I yr. 0 patients1-3 yrs. 54-5 yrs. 36-10 yrs. 3II-I 5 yrs. 2Not stated 2NATURE OF ILLNESS AT ONSET OF SYMPTOMSMeasles and/or whooping cough7 patientsPrimary bronchopneumonia 3Insidious onset .. 2Pleurisy .. I patientNotstated.2 patientsFINGER CLUBBINGPresent (recorded as mild)Absent . .Not statedPresentDoubtfulAbsentPARANASAL SINUS INFECTIONS5 patients9 .,I patient3 patients48Three patients who had tuberculous foci in thelobes and hilar lymph-glands dated their symptomsfrom whooping-cough or measles in early childhood.There was a close correlation between the amountof inflammatory change seen in the bronchi and235the quantity and nature of the sputa; in manycases there was a complete absence of sputum.PATHOGENESIS OF ATELECTATIC BRONCHIECTASISThe following features, which are characteristic ofatelectatic bronchiectasis and not of other forms ofthe disease, suggest how atelectatic bronchiectasismay arise: (1) the generalized distribution ofcollapse and bronchiectasis in the lobes; (2) theabsence of central or peripheral bronchial obstructionin the lobes at the time of operation; (3) thepresence of tuberculous foci in some lobes andhilar lymph-glands; (4) the frequency of involvementof the right middle lobe, either alone or withthe lower lobe; (5) the variability of bronchialchanges, and the absence of inflammation in thecollapsed parenchyma; (6) the absence of fingerclubbing, sinus infections, and foul sputum inmany patients.The bronchiectasis must either be the cause, orthe result, of collapse. If collapse were secondaryto bronchiectasis one would expect the segmentaldistribution of the lesions, and their histology, tobe the same as in aerated forms of the disease.This was not so, so the collapse is probablythe primary condition. There were a few exceptionsto this generalization; they may have beensecondarily-collapsed follicular bronchiectases.If the theory that collapse is caused by peripheralbronchial obstruction is accepted, in these casesone is forced to conclude that simultaneous obstructionsoccurred in all peripheral bronchi; otherwisethe bronchiectasis could not affect all branches,and collateral air circulation would prevent alveolarcollapse. The improbability of such an occurrence,and the complete absence of peripheral obstructioni most of the specimens, suggests that there is nosound basis for this theory.There is far more evidence supporting the viewthat collapse is caused by obstruction of lobarbronchi. Although no obstructive lesions of theproximal bronchi were found in these specimens, orseen in pre-operative bronchograms, the frequentfinding of enlarged, fibrotic, or caseous hilar lymphglandssuggests that collapse arose from occlusionby these glands of the lobar bronchi. This theoryreceives further support from the high incidence ofright middle lobe specimens in atelectatic bronchiectasis,for Brock (1946) has demonstrated thepeculiar vulnerability of the middle lobe bronchusto compression from enlargement of the surroundinglymph glands.Recently Brock (1950) has described suppurativechanges, bronchiectasis, and collapse in middlelobes as a sequel to tuberculous hilar adenitis. He

236F. WHITWELLstates that in some cases this glandular enlargementmay be non-tuberculous, and that similar changesmay occur in other lobes. Specimens in this groupwere not examined for calcified foci radiologically,but they were dissected carefully and all hilarlymph glands were examined.Some specimens in this group showed absorptioncollapse without bronchiectasis. In these lobes thebronchi were collapsed with their walls in apposition,and no irreversible histological changes werepresent, either in the bronchi or alveoli.It is thought that this is similar to the state of alobe when its hilar bronchus had been occludedby lymph gland enlargement. When these glandsshrink the lobar bronchus becomes patent and aircan again enter the bronchial tree. In many casesthis is followed by aeration of the collapsed alveoliand the restoration of the lobe to normal, but ifsevere inflammatory changes have occurred in thebronchi, or if fibrosis has taken place in the parenchyma,the lobe is unable to expand again. Insuch a case (Fig. 35), permanently collapsed alveolisurround aerated bronchi, which dilate because of*destruction of their supporting tissues and accumulationof bronchial secretions. The slight reductionin intra-pleural pressure, which might occur aftercollapse of a middle lobe, could exert little dilatingforce on its bronchi.Bronchial dilatation in these cases may beexplained on mechanical grounds without thenecessity of there being severe inflammatorychanges, though the initial cause of the hilarlymphadenopathy is usually inflammatory. Theabsence of finger clubbing, sinus infections, andpurulent sputum in many of this group of patientssupports this theory of a mechanical disorder.DISCUSSIONAlthough in this investigation a great variety ofbronchiectatic lesions has been found, the accounthas been limited to a. description of three commonforms which were seen in operation specimens. Inno instance could the bronchial dilatation havebeen considered to be reversible; it was always theresult of destructive inflammatory processes, permanentalveolar collapse, or abnormal developmentresulting in bronchi of wide calibre.Follicular and saccular bronchiectasis havedistinctive lesions and natural histories. Atelectaticbronchiectasis is a less specific and largely amechanical disorder, which usually follows occlusionby hilar adenitis of a lobar brorchus. The causesof lymphadenopathy are not restricted to anyperiod of life, neither are the specimens frompatients of any particular age, though they aremostly from young children. It is natural that thehilar adenitis seen in follicular bronchiectasisshould, in some instances, produce total obstructionof lobar bronchi, and result in specimens whichshow a mixture of follicular and atelectatic lesions.All these lesions have been described previouslyon many occasions, but in the literature they haveusually been regarded as variations in a singledisease entity. it is not intended to review thisliterature; excellent summaries can be found inthe works of Ewart (1900), Ballon, Singer, andGraham (1931), Ogilvie (1941), and Lisa andRosenblatt (1943). It is sufficient to remark thatonly three accounts appear to be inconsistent withthe present study. Lander and Davidson (1938),reporting on 140 operation specimens, found littledestruction of bronchial wall structures, and inmany cases only slight indications of inflammation.Specimens such as these were extremely rare in thepresent series; they were not considered to be truebronchiectasis, and their removal had been dueto misinterpretation of radiological and clinicalfeatures. Lisa and Rosenblatt (1943), reporting on110 post-mortem specimens, remarked that thebronchial elastic tissues were relatively unimpaired.In the present study one of the earliest and mostconstant features of affected bronchi and bronchioleshas been the destruction of elastic tissue. Moore,Kobernick, and Wiglesworth (1949) have statedthat there is little correlation between the bronchographicappearances of a lobe and the condition ofthe bronchi; in fact, their work suggests thatsegmental resections can be of little value. Theopposite conclusion has been reached from thepresent study.Literature on the pathogenesis of bronchiectasisis even more voluminous than that on its pathology;and it is much more confusing. One reason is,perhaps, that many theories are based on pathologywhich is inferred rather than observed. For example,similar radiological appearances are described ascollapse, fibrosis, and pneumonia, by differentwriters who deduce that these various conditionsare the cause of bronchiectasis. Another sourceof confusion is the vague use of the word " bronchiectasis."For example, it is used in describingsome histological findings in fatal cases of pneumonia(Opie, 1928; McNeil et al., 1929), the gross lesionsproduced in animals by experimental methods(Weinberg, 1937; Tannenberg and Pinner, 1942),and the temporary bronchographic changes foundin 90% of university students who suffer fromrecurrent bronchitis (Ochsner, 1930). From studiesin similar rather restricted fields many writers have

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS237generalized about the pathogenesis of clinicalbronchiectasis.Fortunately, the number of theories of pathogenesisis limited; in fact, apart from the possibleinfluence of allergy and sinus infections, no newtheory has been propounded in the last hundredyears. Nearly all writers are agreed that normalbronchi do not dilate when subjected to variousmechanical stresses, and most theories accept aprimary basis of bronchial weakness, due either toinflammation or developmental defects.Developmental abnormalities and the mostcommonly suggested secondary causes of dilatationare discussed below.DEVELOPMENTAL ABNORMALITIESAbout 10% of the specimens were examples ofcongenital cystic lung, which is a developmentalabnormality. In five other specimens the bronchiwere thin-walled, and the calibre of the smallbranches was unduly wide, often, indeed, muchwider than the proximal bronchi. No inflammatory,fibrotic, or destructive lesions were present, and theparenchyma was entirely normal. In three of thesespecimens, however, the bronchial supportingtissues seemed to be underdeveloped. The widecalibre appeared to be their natural condition andnot due to dilatation.Apart from these few specimens no evidence wasfound of any developmental abnormalities; norwas it necessary to postulate any such defects inorder to explain the bronchiectasis.PRESSURE OF SECRETION IN BRONCHIThis theory, and the objections to it, have beendiscussed on page 232. The terminal dilatationsin saccular bronchiectasis probably result fromdistension of weakened bronchi by their containedsecretions. The same mechanism may account formany of the dilatations of follicular and atelectaticbronchiectasis, though this is more difficult tosubstantiate.FIBROSISThe usual theory supposes that bronchi becomedilated by the contraction of bands of fibroustissue running between them, but no such arrangementof fibrous tissue was seen in the specimens.Fibrosis occurred as a thickening of the normalfibrous structures in the peribronchial, perivascular,interlobular, and pleural connective tissues. Occasionallyconcentric fibrosis was seen in the walls ofdiseased bronchi, and it was always present in thewalls of saccules. Inter-alveolar fibrosis wasQfound in some areas of chronic interstitial pneumonia,and in absorption collapse.In most forms of bronchiectasis fibrosis is a latedevelopment. It is the stretching of fibrous bronchialwalls in saccular bronchiectasis that leads todilatation; it is the parenchymal fibrosis ofatelectatic bronchiectasis which prevents re-aeration.BRONCHO-PULMONARY EXCAVATIONIn a few lobes the dilatations appeared to bere-epithelialized abscess cavities, which had originatedin the parenchyma. However, these lesionswere uncommon and not seen in the three typesof bronchiectasis described.PERIPHERAL BRONCHIAL OBSTRUCTION ANDALVEOLAR ABSORPTION COLLAPSEThese two lesions are linked in most recentdiscussion on pathogenesis (Warner and Graham,1933; Boyd, 1935; Andrus, 1937, 1940; Landerand Davidson, 1938; Fleischner, 1940; Lander,1946; Coope, 1948). It is argued that collapsefollows obstruction, and dilatation occurs inbronchi proximal to the obstruction. These bronchiare subjected to a great dilating stress, which iscaused by the difference in pressure between thepleural space and the atmosphere in the bronchi.From the present study there appears to be nosound basis for this theory. Obstruction of peripheralbronchi was always present in follicular andsaccular bronchiectasis, but absorption collapsewas absent, probably because of adequate collateralair circulation. Absorption collapse was found in10% of the specimens but it affected whole lobes,not lobules or segments. No peripheral bronchialor bronchiolar obstruction was present in theselobes, and the collapse had probably been causedby proximal bronchial obstruction.Many other arguments can be raised against thistheory. In cases where only a few small bronchiin one segment are affected there could be nosignificant alteration of pleural pressure to create adilating force, even if collapse occurred distal tothe bronchiectasis. The absence of radiologicalcollapse in many cases has been explained in manyunconvincing ways. Andrus (1940) states thatpatchy collapse is diagnosed as pneumonia, thatcollapse is concealed by the heart shadow, and thatcollapse may cause lung injury " and then bedissipated." Perry and King (1940) think thatcollapse can "lie dormant" for many years untilpurulent bronchitis supervenes. These reasons forthe absence of radiological collapse, even if accepted,do not explain the normal aeration of so many

2)38F. WHITWELLTABLE IIIPATHOLOGY OF BRONCHIECTASISFollicular I Saccular Atelectatic84% left-sided 72% left-sided 36% left-sidedIso'ated bronchi affected Isolated bronchi affected All bronchi involvedChronic inflammatory changes in bronchi Chronic inflammatory changes in bronchi Variable degree of inflammatory change,and bronchioles, with formation of lymph and saccules, no lymph follicle formation lymph follicles sometimes seenfolliclesDestruction of elastic tissue and muscle, some Little destruction of supporting tissues except Variable amount of destruction, sometimesdestruction of cartilage and mucous glands within saccules, where none remain very littleBronchial epithelium intact and smooth, Bronchial epithelium intact but polypoid. Bronchial epithelium usually intact andsmall ulcerations in bronchioles. No Some ulcerations in saccules and promi- smooth. Few ulcerations, no squamoussquamous metaplasia nent squamous metaplasia metaplasiaBronchio'es narrowed or obliterated Bronchioles obliterated Bronchioles patentAlveoli mostly well aerated, but some inter- Alveoli well aerated, no chronic interstitial Marked alveolar collapse, uniform throughstitialpneumonia around affected bronchi pneumonia. Some compression collapse out lobe. Little inflammatory change inand bronchioles. Some compression around saccules collapsed alveoli, but some fibrosiscollapse around areas of pneumoniaClinical FindingsAge at onset of symptomsDuration of symptomsAge at time of operationIllness at onset of symptomsBronchogramsNasal sinus involvementClubbing__TABLE IVCLINICAL ASPECTS OF DIFFERENT TYPES OF BRONCHIECTASISFollicularSaccular53% under 3 yrs. 78% under 7 25% complained of symptoms allyrs. their lives. 70% began symptomsbetween 13 and 25 yrs.1-15 yrs. 50% for 1-5 yrs. 2-20yrs. 50%for2-6yrs.Mainly 6-11 yrs. old, with gradual None under 15 yrs., mostly bedecreaseup to 20 yrs. None tween 15 and 25 yrs., with a fewover 20 yrs.cases in older groups~~~~~IVague onset in 25%. Measles andy Vague onset in 45% (includingor whooping-cough in 45%. 25% who had symptoms allPrimary bronchopneumonia in their lives). Unspecified pneumoniain 56%. No association20%with measles or whooping-coughI _ .1Usually cylindrical, sometimes SaccularatelectaticOver 70%Under 50%Over 70()75%1-Atelectatic20% under 3 yrs. 70, under 7yrs.1-15 yrs. 33% for 1-3 yrs.Mostly between 5 and 10 yrs.,but no sudden fall-off of olderpatientsVague onset rare. Measles orwhooping-cough in 50%/; primarybronchopneumonia in 25%AtelectaticInfrequentInfrequentoperation specimens, such as has been seen in thepresent series, and which has frequently beendescribed in the literature.SUMMARYThe investigation has consisted of the examinationof 200 consecutively removed bronchiectatic lungsand lobes. Methods used consisted of neopreneinjection casts of bronchial trees, large histologicalsections of entire lobes, and detailed dissections ofbronchial trees with histological examinations ofselected areas.The lobar, segmental, and intra-segmental distributionof the lesions have been described (seeTables I and 11, and Fig. 10).More than half the specimens belonged to oneor other of three rather distinctive types of bronchiectasis,which have been called follicular,saccular, and atelectatic bronchiectasis, and theirpathology, clinical features, and pathogenesis hasbeen discussed (see Figs. 21, 30, and 35, andTables I II and IV).Finally, current theories of pathogenesis havebeen criticized in the light of the present enquiry.This work has been only made possible through theco-operation of the staff of the Surgical Chest Centre,Broadgreen Hospital. In addition I would like to thankMr. H. Morriston Davies, Mr. F. R. Edwards, and Dr.Robert Coope for many valuable discussions on medicaland surgical aspects of bronchiectasis. I wish also toacknowledge my indebtedness to Dr. Rachel M. Rawcliffefor her assistance in the preparation of neoprenecasts and in the tabulation of case-records, and toDr. P. J. Taylor for correction of the MS.

PATHOLOGY AND PATHOGENESIS OF BRONCHIECTASIS239REFERENCESAllison, P. R., Gordon, J., and Zinnemann, K. (1943). J. Path. Ba t.,55, 465.Andrus, P. M. (1937). Amer. Rev. Tuberc., 36, 46.(1940). Ibid., 41, 87.Baarsma, P. R., and Dirken, M. N. J. (1948). J. thorac. Surg., 17,238.- and Huizinga, E. (1948). Ibid., 17, 252.Ballon; H., Singer, J. J., and Graham, E. A. (1931). Ibid., 1, 154.Boyd, G. L. (1935). J. Amer. med Ass., 105, 1832.Brock, R. C. (1946). The Anatomy of the Bronchial Tree. London.-(1950). Thorax, 5, 5.Broman, 1. (1923). Anat. Anz., 57, Suppl. (Verh. anat. Ges.), p. 83.Quoted by Engel.Churchill, E. D. (1949). J. thorac. Surg., 18, 279.-and Belsey, R. (1939). Ann. Surg., 109, 481.Coope, R. (1948). Diseases of the Chest, 2nd ed. Edinburgh.Cruikshank, A. H. (1948). J. Path. Bact., 60, 520.Engel, S. (1947). The Child's Lung. London.Erb, I. H. (1933). Arch. Path., 15, 357.Ewart. W. (1900), Allbutt and Rolleston's System of Medicine, 2nded., vol. 5, p. 127 seq. London.Fleischner, F. (1940). Amer. Rev. Tuberc., 42, 297.Jackson, C., and Jackson, C. L. (1932). J. Amer med. Ass., 99, 1747.Laennec, R. T. H. (1819). Dc L'Auscultation Mediate. Paris.Lander, F. P. L. (1946). Thorax, 1, 198.and Davidson, M. (1938). Brit. J. Radiol., 11, 65.Lisa, J. R., and Rosenblatt, M. B. (1943). Bronchiectasis. London.MacCallum, W. G. (1940). Textbook of Pathology, 7th ed. Philadelphia.McNeil, C., Macgregor, A. R., and Alexander, W. A. (1929). Arch.Dis. Childh., 4, 170.Mallory, T. B. (1947). New Engi. J. Med., 237, 795.Maximow, A. A., and Bloom, W. (1942). Textbook of Histology,4th ed. Philadelphia.Miller, W. S. (1947). The Lung, 2nd ed. Springfield, Illinois.Moore, J. R., Kobernick, S. D., and Wiglesworth, F. W. (1949).Suirg. Gynec. Obstet., 89, 145.Nelson, J. B. (1946). J. exp. Med., 84, 7 and 15.Ochsner, A. (1930). Amer. J. med. Sci., 179, 388.Ogilvie, A. G. (1941). Arch. intern. Med., 68, 395.Opie, E. L. (1928). Arch. Path., Chwago, 5, 285.Overholt, R. H., Betts, R. H., and tods F. M. (1947). Dir. Chest,13, 583.Peroni A. (1934). Arch. Otolaryng., 19, 1. Quoted by Samson.Perry, K. M. A., and King, D. S. (1940). Amer. Rev. Tuberc., 41,531.Reid, L. McA. (1950). Thorax, 5, 233.Riviere, C. (1905). St. Bart's Hosp. Rep., 41, 123.Robinson, W. L. (1933). Brit. J. Surg., 21, 302.(1939). Amer. J. Path., 15, 638.Samson, P. C. (1940). J. thorac. Surg., 9, 679.Schmidt, H. W. (1947). Ann. Otol., St. Louis, 56, 793.Tannenberg, J., and Pinner, M. (1942). J. thorac. Surg., 11, 571.Van Allen, C. M., and Jung, T. S. (1931). Ibid., 1, 3.Warner, W. P., and Graham, D. (1933). Arch. intern. Med., 52, 888.Watts, C. F., and McDonald, J. R. (1948). Arch. Path., Chicago,45, 742.Weinberg, J. (1937). J. thorac. Surg., 6, 402.Williams, C. J. B. (1838). Lond. med. Gaz., n.s. 1, 913.Willis, R. A. (1 948). Pathology of Tumours. London.