Thorax 1984;39:185-191

Cutaneous vasculitis and immune complexes in severe

bronchiectasisANGELA M HILTON, PS HASLETON, A BRADLOW, BC LEAHY, KM COOPER, MMOORE

From the Departments of Chest Medicine and Pathology, Wythenshawe Hospital; and the Department ofImmunology, Paterson Laboratories, Christie Hospital and Holt Radium Institute, Manchester

ABSTRACT Four patients with severe bronchiectasis (chronic bronchial suppuration) are describedwho developed cutaneous lesions associated with exacerbations of their respiratory disease. Theskin abnormalities consisted of purpuric lesions in three patients and an erythematous vasculitis inone. Circulating immune complexes were present in all patients and in three skin biopsy speci-mens showed deposition of C3, IgG, and IgA in dermal blood vessels. Haemophilus influenzaehad been isolated from the sputum of all four patients and in two patients was present at the timethe cutaneous lesions appeared. It is suggested that local immune complex deposition wasresponsible for the skin lesions which occurred during acute exacerbations of bronchiectasis.

In recent years various immunological abnormalitieshave been reported in patients with chronic bron-chial suppuration (severe bronchiectasis). Theseabnormalities have included an increased incidenceof autoantibodies, raised immunoglobulin concent-rations, and circulating immune complexes.'-3 Inaddition, many immunological diseases have beenshown to be associated with severe bronchiectasis-for example, autoimmune thyroiditis, perniciousanaemia, rheumatoid arthritis, systemic lupuserythematosus, and ulcerative colitis.'4 6 In com-mon with patients with other chronic pulmonaryconditions, those with bronchiectasis have a highincidence of circulating immune complexes;' butthere have been few reports of cutaneous lesionsassociated with this disease and in a review of 35patients with cutaneous vasculitis8 only one hadbronchiectasis.We report four patients with chronic bronchial

suppuration who developed skin lesions of varyingtype and distribution in association with exacerba-tions of their respiratory symptoms. All had circulat-ing immune complexes and in three patients therewas histological evidence to suggest that localimmune complex deposition may have mediated thecutaneous lesions.

Address for reprints: Dr Angela M Hilton, Chest Clinic, Wythen-shawe Hospital, Manchester M23 9LT.

Accepted 14 October 1983

Case histories

Clinical details of the four patients are summarisedin table 1.

PATIENT 1A 68 year old man had an 18 year history of coughwith the production of large amounts of purulentsputum. The onset of symptoms was not related to aspecific illness and his symptoms had become pro-gressively worse over the last few years, withincreasing breathlessness on minimal exertion andthe production of 100-200 ml of purulent sputumdaily. He had been a lifelong non-smoker and therewas no family history of pulmonary disease or auto-immune disorders. Serial chest radiographs over a10 year period showed progressive changes consis-tent with bronchiectasis. On examination there wasno finger clubbing but he had coarse crackles overboth lower lobes, worse on the left side. In June1980 he noticed for the first time a raised, red, itchyrash affecting the index fingers and thumbs of bothhands symmetrically (fig 1). The skin lesions wouldoccur immediately before a severe exacerbation ofhis respiratory symptoms and lasted four to sevendays, gradually fading as his condition improvedwith antibiotic treatment and physiotherapy.

During the 18 months from June 1980 toDecember 1981 he had at least eight similarepisodes in which skin lesions recurred, affecting theindex fingers and thumbs of both hands symmetri-

185

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Hilton, Hasleton, Bradlow, Leahy, Cooper, Moore

Table 1 Clinical details ofthe patients studied

Patient No Age (y) Sex Duration (y) ofrespiratory Dermatological diagnosis Time (y) since onset ofskinsymptoms lesions

1 68 M 18 Vasculitis 21/22 57 M 26 Purpura 153 59 M 38 Henoch-Schonlein purpura 84 67 F 12 Vasculitic purpura 2

Table 2 Immunoglobulins and autoantibodies in the four patients (normal values shown in parentheses)

Patient No IgA IgG IgM Rhematoid Antinuclear Smooth muscle(0 9-3*3 gll)* (7-15 gll)* (0-55-2-8 gll)* factor/SCAT factor antibody

1 12-5 11-0 025 - -2 6-7 13-4 0.53 19 27-7 0-5 - +

(1/80)4 0-84 19-4 0-7 + - +

( 1/320)SCAT 1/1024

* Values in italic in the columns are above the normal range.+ indicates a positive and - a negative result.SCAT-sheep cell agglutination test.

cally. On each occasion they were associated with orimmediately preceded an acute exacerbation. Withcontrol of the respiratory symptoms the skin lesionsfaded and healed by exfoliation over a period of upto a week. In December 1981, because of repeatedsevere exacerbations and frequent hospital admis-sions during the preceding 18 months, he wasstarted on long term antibiotic treatment with co-trimoxazole 480 mg twice daily. Since that time hiscondition has been stable, with no severe acuteexacerbations, hospital admissions, or recurrence ofskin lesions.Repeated sputum cultures over several years have

frequently been sterile; but on two earlier occasionsPseudomonas aeruginosa organisms were isolatedfrom his sputum, and during the period when he was

rig I Lejt nana oypatient I snowing erytnematous rasn onthe thumb and index finger.

having recurrent episodes of vasculitis in July 1980and September 1981 Haemophilus influenzae wasgrown.Routine haematological and biochemical inves-

tigations and urine examination gave normal results.The serum immunoglobulin concentrations arelisted in table 2. These were normal apart from araised IgA concentration. The results of tests forautoantibodies were negative, as was the response tothe aspergillus precipitin test. The results of com-plement studies included: CH,. 128% (normal), C32-22 g/l (normal 0-8-2-8 g/l), and C4 normal.

PATIENT 2A 57 year old man had a 25 year history of persis-tent cough productive of purulent sputum. Therewas no obvious precipitating illness and progressivechanges of bronchiectasis had developed during 25years. He was a non-smoker and there was noassociated autoimmune disease or relevant familyhistory. He produced about 100 ml of purulentsputum daily. Fifteen years earlier he had developedpurpura affecting both legs below the knees andboth forearms. The purpuric lesions were mostnoticeable on his legs and were associated with painin ankles and knees. At that time the results ofcoagulation studies were normal. Eight years ago afurther episode of purpura occurred, associated thistime with increasing dyspnoea, cough, and increasedsputum production; for the first time he was noted tohave cor pulmonale and hypercapnia. Furtherexacerbations of his chest condition requiring hospi-tal admission occurred seven, six, and two years ago,but on these occasions no rash was noted. On hismost recent admission in December 1982 for a

186

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Cutaneous vasculitis and immune complexes in severe bronchiectasis

severe exacerbation of bronchiectasis and cor pul-monale he had central cyanosis, finger clubbing andcoarse crackles over both lower lobes; and small(<0.5 cm diameter) round purpuric lesions werenoted on both shins and forearms. The rash fadedbut did not entirely disappear with improvement inhis symptoms after physiotherapy and antibiotictreatment. His chest radiograph showed extensivebilateral cystic bronchiectasis. The serum globulinconcentration was raised at 41 g/l, with a raised IgAlevel (table 2). Platelets, the results of coagulationstudies and the Hess test, and bleeding and clottingtimes were normal. The haemoglobulin concentra-tion was 18-4 g/dl with a packed cell volume of 061.Sputum culture in 1968 had shown Haemophilusinfluenzae but all subsequent sputum specimenswere bacteriologically sterile. Complement studiesgave normal results.

PATIENT 3A 59 year old man presented in 1975 with a historyof two episodes of pneumonia in his late teens. Sincethen he had regularly produced purulent sputumand at the age of 21 bronchiectasis had been diag-nosed after a bronchogram. His condition remainedstable, with regular sputum production and symp-toms of mild airflow obstruction, until 1974. Overthe subsequent nine years he had graduallydeteriorated, with an increased volume and purul-ence of sputum and increasing dyspnoea.On examination he had finger clubbing, central

cyanosis, and coarse crackles at both lung bases. Hischest radiograph showed evidence of cystic bron-chiectasis predominantly affecting the right middleand right lower lobes. He had been a cigarettesmoker until 1975.

In July 1976 he had had an episode of severecentral abdominal pain and a purpuric rash over thebuttocks, lower abdomen, and lower limbs. This wasassociated with an exacerbation of bronchiectasisand H infiuenzae had been cultured in the sputum.His haemoglobin concentration was 18 g/dl. Theresults of coagulation studies, bleeding time,platelets, and the response to the Hess test werenormal. Gastroscopy showed purpuric lesions in thepyloric canal and duodenum, findings suggestive ofHenoch-Schonlein purpura. Bouts of colicky painpersisted for several weeks and urine examinationshowed mild proteinuria but no evidence of redcells. From 1976 to 1979 his regular sputum produc-tion continued but there was no recurrence of therash. In July 1979 further purpuric lesions werenoted over his lower limbs, associated with anexacerbation of bronchiectasis. As the pulmonarysymptoms improved with antibiotic treatment the

rash disappeared. In January 1980 he was admittedwith cor pulmonale associated with a purpuric rashover both lower limbs. When he had been treatedwith diuretics and oxygen and improved the rashfaded. H influenzae was again isolated from hissputum. Immunological investigations at this timeshowed a polyclonal increase in immunoglobulins;the response to the antinuclear factor test wasweakly positive (table 2). The results of complementstudies were within normal limits but he had ele-vated Clq binding levels, suggesting the presence ofcirculating immune complexes (table 3).

PATIENT 4This 65 year old woman had been attending the out-patient department for 10 years with graduallyincreasing dyspnoea on exertion and a cough pro-ductive of increasing amounts of purulent sputum.She had a history of cough and sputum productiondating from a childhood chest infection. She was a

lifelong non-smoker. Her chest radiograph showedring shadows with areas of collapse and fibrosisthroughout both lung fields, particularly in the upperzones, and the appearances were consistent withbronchiectasis. Six months before hospital admis-sion in 1978 she had had two episodes of transientsymmetrical polyarthralgia affecting the ankles,knees, hands, and wrists. No inflammatory synovitiswas present but the result of the rheumatoid factor(latex) test was positive at 1/320. The antinuclearantibody test gave a negative result. A few weekslater she developed a painless rash on both legs,which lasted several weeks and was associated withincreased dyspnoea and increased sputum produc-tion. There had been no new antibiotic or other drugtreatment immediately before the development ofthe rash.On examination she had finger clubbing, splinter

haemorrhages, and nail fold infarcts. There was anextensive symmetrical vasculitic rash consisting ofraised purpura and ulcers covering both legs andfeet. Her chest was hyperinflated with coarse crack-les at both lung bases. Investigation showed anerythrocyte sedimentation rate of 82 mm in onehour; urine sterile on culture; and renal function,including blood urea, serum creatinine, andcreatinine clearance, normal. Immunological inves-tigations showed a raised IgG concentration, posi-tive responses to the sheep cell agglutination test(SCAT) and latex test (table 2), a raised plasma C4level, and considerably increased plasma Clq bind-ing activity (75%) (table 3). Numerous blood cul-tures were sterile, and sputum culture isolatedHaemophilus infiuenzae. The patient was treatedwith high doses of prednisolone (80 mg daily) andamoxycillin (500 mg four times a day), with gradual

187

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Hilton, Hasleton, Bradlow, Leahy, Cooper, Moore

Table 3 Circulating immune complex assays in the four patients (normal ranges shown in parentheses)

Patient No Clq binding activity PEGISRID assayt

1 24-2% (<4.5%) * IgG precipitate 142 mg/l (<90) +IgA precipitate 49 mg/I (<20) +IgM precipitate 19 mg/l (<40)Clq precipitate 116 mg/I (<60) +

2 28 units (<20) t IgG precipitate 52 mg/iIgA precipitate 14 mg/lIgM precipitate < 10 mg/l

3 57 units (<20) t IgG precipitate 92 mg/i +IgA precipitate 24 mg/l +IgM precipitate 28 mg/lClq precipitate 98 mg/i +

4 75% (<4.5%) * ND* Soluble phase Clq binding assay.t Solid phase Clq binding assay.t Values in italic are above the normal range.PEG-polyethylene glycol; SRID-single radial immunodiffusion; ND-not done.

improvement in her chest condition and progressiveresolution of the cutaneous vasculitis over six weeks.

MethodsMEASUREMENT OF CIRCULATING IMMUNECOMPLEXESClq binding assay (soluble phase): patients 1 and 4Clq was prepared according to the method of Vol-anakis and Stroud9 and the Clq binding assay wascarried out as described by Zubler et al.'° The resultswere expressed as the percentage of precipitatedradioactivity, the upper limit of normal being 4-5%.

Clq binding assay (solid phase): patients 2 and 3The Clq binding assay was performed as describedby Hay et al," and the results were expressed on ascale with a maximum of 100 units, the upper limitof normal being 20 units.

Polyethylene glycol (PEG) precipitation and singleradial immunodiffusion (SRID): patients 1, 2 and 3The Mercia-Brocades CIC Kit (M 202) was used.

HISTOPATHOLOGYSkin biopsy specimens were obtained from patients1, 2, and 4. 5,um sections were stained withhaematoxylin and eosin, elastic van Gieson, andMartins scarlet blue (MSB) for fibrin. The sectionsfor immunoglobulin and C3 determination were tre-ated as follows. 5,um sections were deparaffinised inxylol and taken to alcohol; endogenous peroxidasewas inhibited by treatment with 0-5% H202 inmethanol for 10 minutes. The slides were washedwell in tap water and then placed in distilled water at37°C for 10 minutes. The sections were then treatedwith 0-1% trypsin and 0- 1% CaC12 at pH 7.8 for 15minutes and washed with tris buffered saline (TBS)for 10 minutes on two occasions. The slides werethen exposed to rabbit antihuman IgA, IgG, IgM,

fibrinogen, and C3 at dilutions of 1/800 (except forfibrinogen, which was at a dilution of 1/400). Theslides were washed in TBS on three occasions for 10minutes each. Swine antirabbit IgG at a dilution of1/400 was added to the appropriate slides. The sec-tions were washed again with TBS for three periodsof 10 minutes. PAP (peroxidase-rabbit antiperoxid-ase) in a dilution of 1/500 was added for 30 minutes.The slides were washed again in TBS and the horse-radish peroxidase was visualised by using diaminobenzidine and H202 for 10 minutes. The sectionswere then washed in TBS and in running tap waterfor five minutes. The sections were counterstainedwith haematoxylin, washed for two to three minutes,differentiated in 1% acid alcohol, "blued" by wash-ing in Scott's tap water substitute for at least 10minutes, dehydrated, cleared, and mounted. Nega-tive controls were obtained by substituting TBS forthe primary positive antiserum and positive controlswere human tonsillar tissue; for C3 immunofluores-cence material from a case of membranoprolifera-tive glomerulonephritis was used.

Results

CIRCULATING IMMUNE COMPLEX ASSAYSTable 3 summarises the results of assays of circulat-ing immune complexes in the four patients. Allpatients were found to have increased Clq bindingactivity consistent with the presence of complementfixing circulating immune complexes measured byeither the soluble or the solid phase assay. In addi-tion, two patients (1 and 3) were found to have IgG,IgA, and Clq containing immune complexes by thePEG-SRID method.

Figure 2 shows sequential measurements of Clqbinding activity (soluble phase method) in patient 1over a period of two and a half years. Moderatelyhigh levels of circulating immune complexes werepresent during this time when intermittent episodes

188

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Cutaneous vasculitis and immune complexes in severe bronchiectasis

8a

6'-4

2

0

v v vv v v v v

4 4 44 4 4 4 4A

/\!-0/7.-

6 12 18 24 30Time in Months

Fig 2 Sequential Clq binding activity in patient Imeasured for two and a halfyears and its relationship to thecutaneous lesions. V indicates episodes of vasculitis and Athe start oflong term antibiotic treatment.

of "vasculitis" occurred. Since he started long termantibiotic treatment there has been a fall in theimmune complex levels and no further vasculiticepisodes have developed.

PATHOLOGYPatient I The epidermis was intact with no evi-dence of bullae or ulceration. In the dermis therewere dilated blood vessels, two of which were sur-rounded by neutrophil polymorphs. The other ves-sels were cuffed by lymphocytes and histiocytes. Nofibrinoid necrosis was present. There was dermaloedema and the endothelial cells were swollen.Immunoperoxidase staining showed IgA, IgG, IgM,C3, and fibrinogen in the vessel wall as well as in thedermis. The changes were strongly suggestive ofanaphylactoid purpura.Patient 2 The epidermis was intact and showed noulceration or dysplasia. In the dermis there wasoedema and around arterioles and capillaries therewas an acute on chronic inflammatory reaction withlymphocytes, neutrophil polymorphs, and someeosinophils (fig 3a). No intravascular fibrin orfibrinoid necrosis was seen. Immunoperoxidasestaining showed C3 on the luminal aspect of smallvessels (fig 3b), IgG deposition occurring at thesame site as well as in the vessel walls. IgG stainingalso occurred in the dermis and between epithelialcells, and IgA had the same distribution as IgG. IgMwas seen in one blood vessel wall only. Fibrinogenwas present in the dermis and in the lumen of someblood vessels. The histological features were thoseof anaphylactoid purpura.

Patient 4 The epidermis was intact and orderlywith no evidence of bullae or ulceration. Thesuperficial dermis was normal. In the deeper dermisthe arterioles and capillaries were surrounded byacute inflammatory cells spreading out into the der-mis with abscess formation in some areas. There wasevidence of fibrinoid necrosis of the dermal vessels.Oedema and haemorrhage were also present in thedermis. Medium sized muscular arteries showed noevidence of fibrinoid necrosis or arteritis. Onimmunoperoxidase staining fibrinogen wasidentified in the affected blood vessel walls and inthe dermis. IgA and IgM were seen in the lumen ofaffected and normal vessels and IgG was present inthe walls of affected vessels. C3 was seen in somevessel walls but appeared mostly in those withoutfibrinoid necrosis and where the inflammation wasleast noticeabie.

Discussion

Recent studies using three different assays haveshown a very high incidence of circulating immunecomplexes in bronchiectasis. They were detected inabout 80% of patients with bronchiectasis when allthree assays were applied to each serum sample.Since the different methods detected complexeswith differing properties in relation to their size,antibody content, ratio of antigen to antibody, andability to fix complement, it was provisionally con-cluded that complexes in patients with bronchiec-tasis were heterogeneous in nature and that differ-ent types of complex could be present at any giventime in individual patients.7The four patients we have described with cutane-

ous lesions and bronchiectasis all had circulatingimmune complexes demonstrated by Clq bindingand therefore complement fixing. Patients 1 and 3also had immune complexes detected by PEG pre-cipitation. It is possible that such complement fixingcomplexes have a greater potential for deposition intissues such as blood vessels, with resultant tissuedamage, than complexes which are independent ofcomplement. The fact that all patients had normal orraised complement levels would not exclude com-plement consumption by immune complexes. Anincrease in acute phase proteins during acuteexacerbations could mask any depletion of C3 or C4or both. The presence of C3 breakdown productswould have provided suggestive evidence of com-plement consumption but the appropriate methodswere not available when the patients were investi-gated.The high incidence of circulating immune com-

plexes in patients with chronic bronchial suppura-tion has previously led us to conclude that they

32

30

28

2624

22

S 20U' 18fC is

_ 14

V12'X 10be

.**Xsvvvoso.......................vov9@................. ..........................

189

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Hilton, Hasleton, Bradlow, Leahy, Cooper, Moore

*S.. S. 5s4.?.9

190

': 4' 4 %;J "'; ,:.d

vh4X

'8 fS*''e¢4;,* *~~~

Fig 3 (a) Small blood vessels surrounded by acute andchronic inflammatory cells in the dermis ofpatient 2.(Haematoxylin and eosin, x 190).

probably represent large amounts of antigen com-bined with antibody in the circulation resulting frompersistent chronic antigenic stimulation. The antigencould be derived from an exogenous source such asbacterial products or from products of tissue des-truction. H influenzae had been isolated from thesputum of all four patients on at least one occasion,and in cases 1 and 3 was present in the sputum dur-ing exacerbations of respiratory symptoms associ-ated with the cutaneous lesions. Possibly H influen-zae or its products could form the antigenic compo-nent of the immune complexes in these patients,their persistence in the circulation resulting from afailure of normal clearance mechanisms.

There has been little previous evidence that suchimmune complexes are of pathological significance.In the four cases described here, however, we sug-gest that local immune complex deposition aroundblood vessels in the skin was responsible for the var-ious cutaneous abnormalities which occurred inassociation with exacerbations of respiratory symp-toms. The histological evidence of immunoglobulinand complement deposition in blood vessel walls,

(b) Vessel in the dermis ofpatient 2 showing complement(C3) deposition on its luminal aspect. (Immunoperoxidase,x 190).

demonstrated in three patients by immunoper-oxidase staining, would support this.The histological changes were diagnostic of

anaphylactoid purpura or allergic vasculitis in cases2 and 4 and highly suggestive in case 1. There wasfibrinoid necrosis in case 4 and a well markedneutrophil polymorph reaction around both capil-laries and arterioles in cases 1 and 4. The sparing ofmuscular arteries excludes a diagnosis of polyar-teritis nodosa. The pattern of immunoglobulin andC3 deposition has been noted previously by Schroe-ter et al,'2 who showed deposits of IgG, C3, and insome cases IgM by direct immunofluoresence.Interestingly, in case 4 the complement was onlyseen in the vessel walls in the early stage ofinflammation. Possibly once the complementdependent leucocytic infiltration had destroyed thevessel wall complement was lost. Occasionally thishistological picture may be seen in systemic lupuserythematosus-with which bronchiectasis may beassociated. '

Renal deposition of immune complexes, similar tothose in the skin, may have mediated the Henoch-

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from

Cutaneous vasculitis and immune complexes in severe bronchiectasis

Schonlein purpura described in case 3. This mechan-ism has been suggested as a cause of Henoch-Schonlein purpura in two patients with squamous

cell carcinoma of the bronchus, where tumourantigen-IgA immune complexes were detected inrenal tissue.'3Immune complexes could be important in other

ways in patients with severe bronchiectasis. Localdeposition of immune complexes in pulmonary tis-sue could perpetuate and increase tissue damage inthe lungs. In cryptogenic fibrosing alveolitisimmunological abnormalities similar to thosedescribed in bronchiectasis have been reported-that is, raised immunoglobulin concentrations,'4increased autoantibodies,'5 an increased incidenceof autoimmune disease,'6 and circulating immunecomplexes.'" 18 In addition, in fibrosing alveolitisimmune complex deposition has been shown inalveolar walls and capillaries;" '9 and there isexperimental evidence that both the acute and thechronic changes of fibrosing alveolitis can be pro-duced by immune complex deposition in lung tis-sue.20 21

In malignant disease, where circulating immunecomplexes are frequently present,22 it has been sug-

gested that such complexes may impair cell medi-ated immunity, thus favouring the spread of malig-nant disease.23 Similar impairment of normal hostdefences could be mediated by immune complexesin patients with severe bronchiectasis, contributingto their well established susceptibility to andincreased frequency of recurrent infections, whichmust contribute to the further progression of thedisease.

We thank Drs P Barnes, L Doyle, and TB Strettonfor permission to report details of patients 2, 3, and4 and also Professor J Ball for permission to studythe sections from patient 4 and Dr RC Jennings forpermission to study the sections from patient 1. Thesolid phase Clq and PEG assays were performed bythe regional immunology laboratory, St Mary'sHospital, Manchester.

References

Hilton AM, Doyle L. Immunological abnormalities inbronchiectasis with chronic bronchial suppuration. BrJ Dis Chest 1978;72:207-16.

2 Hilton AM, Moore M, Howat JMT. Detection of cir-culating anticomplementary factors in chronic lungdiseases. Clin Exp Immunol 1978;31:237-43.

3Cooper KM, Moore M, Hilton AM. Clq binding activityin the sera of patients with chronic lung diseases. Clin

Exp Immunol 1981;45:18-28.4 Walker WC. Pulmonary infections and rheumatoid arth-

ritis. Q J Med 1967;36:239-51.Mathieu JP, Stack BHR, Dick WC, Buchanan WW.Pulmonary infection and rheumatoid arthritis. Br JDis Chest 1978;72:57-61.

6Butland RJA, Cole P, Citron KM, Turner-Warwick M.Chronic bronchial suppuration and inflammatorybowel disease. Q J Med 1981;50:63-75.

7Cooper KM, Moore M. Heterogeneity of circulatingimmune complexes in chronic lung diseases. J ClinLab Immunol 1982;7:155-61.

8 Cream JJ. Clinical and immunological aspects of cutane-ous vasculitis. Q J Med 1976;45:255-76.

9Volanakis JE, Stroud RM. Rabbit Clq: purification,functional and structional studies. J Immunol Meth1972;2:25-34.

'0 Zubler RH, Lange G, Lambert PH, Miescher PA.Detection of immune complexes in unheated sera by amodified 125 I-Clq binding test. J Immunol1976;116:232-5.

" Hay FC, Nineham LJ, Roitt IM. Routine assay for thedetection of immune complexes of known immuno-globulin class using solid phase Clq. Clin ExpImmunol 1976;24:396-400.

12 Schroeter AL, Copeman PWM, Jordan RE, Sams WM,Winkelmann RK. Immunofluorescence of cutaneousvasculitis associated with systemic disease. Arch Der-matol 1971;104:254-9.

3 Cairns SA, Mallick NP, Lawler W, Williams G. Squam-ous cell carcinoma of bronchus presenting withHenoch-Schonlein purpura. Br Med J 1978;ii:474-5.

41 Hobbs JR, Turner-Warwick M. Assay of circulatingimmuno-globulins in patients with fibrosing alveolitis.Clin Exp Immunol 1967;2:645-52.

Is Turner-Warwick M, Haslam P. Antibodies in somechronic fibrosing lung diseases. I Non-organ specificautoantibodies. Clin Allergy 1971;1:83-95.

16Turner-Warwick M. Immunology of the respiratorytract. BrJ Hosp Med 1973;9:19-26.

'7Haslam PL, Thompson B, Mohammed I. et al. Circulat-ing immune complexes in patients with cryptogenicfibrosing alveolitis. Clin Exp Immunol 1979;37:381-90.

8 Dreisin RB, Scharz MI, Theofilopoulos AN, StanfordRE. Circulating immune complexes in the idiopathicinterstitial pneumonias. N Engl J Med1978;298:353-7.

'9 Turner-Warwick M, Haslam P, Weeks J. Antibodies insome chronic fibrosing lung diseases. II Immuno-fluorescent studies. Clin Allergy 1971;1:209-19.

20 Scherzer H, Ward PA. Lung and dermal vascular injuryproduced by preformed immune complexes. Am RevRespir Dis 1978;117:551-7.

21 Brentjens JR, O'Connell DW, Pawlowski IB, Hsu KC,Andres GA. Experimental immune complex diseaseof the lung. J Exp Med 1974;140:105-25.

22 Samayoa EA, McDuffie FC, Nelson AM, Go VLW,Luthra HS, Brumfield HW. Immunoglobulin com-plexes in sera of patients with malignancy. Int J Cancer1977; 19: 12-7.

23 Theofilopoulos AN. Immune complexes in cancer. NEngl J Med 1982;307:1208-9.

191

on July 24, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.39.3.185 on 1 March 1984. D

ownloaded from