Boala Parenchimatoasa Pulmonara Difuza

8/9/2019 Boala Parenchimatoasa Pulmonara Difuza

1/30

Introduction

NomenclatureThe parenchyma of the lung includes the pul-monary alveolar epithelium and capillaryendothelium and the spaces between these

structures, together with the tissues within thesepta including the perivascular and peri-lymphatic tissues. More centrally it includesthe peribronchiolar and peribronchial tissues.Many terms have been used to encompass thelarge group of disorders that primarily aVectthe lung parenchyma in a diVuse manner. Dif-fuse parenchymal lung disease (DPLD) isincreasingly in favour worldwide as a genericterm for these disorders, in preference to termssuch as interstitial lung disease or diVuse lungdisease. Alveolar filling diseases are alsoincluded under DPLD because of the similar-ity of presentation.

Need for recommendations in themanagement of DPLDThe DPLDs are important, accounting forabout 15% of respiratory practice.1 They com-prise a very wide spectrum of pathologies,presentations, and outcomes. Despite theiracknowledged complexity, no consensus on thegeneral management of these disorders is avail-able. There is a serious paucity of evidenceabout the management of DPLDs, but themorbidity of the DPLDs themselves and thetreatments available may be high, with poten-tially serious consequences therefore for mis-management. New techniques have become

available which may have a role in the manage-ment of DLPD, but the place of thesetechniques has not hitherto been adequatelyaddressed. There is concern that DPLDs maybe poorly recognised and managed by non-specialists. Respiratory specialists are the onlygroup with appropriate training and skills todeal with the complexity of the diagnosis andmanagement of these conditions. There is,however, evidence of wide variation in themanagement of DPLDs amongst respiratoryphysicians.2 3

Against this background the British Tho-racic Society (BTS) Standards of Care Com-

mittee set up a sub-committee to formulaterecommendations on DPLD management.

The targetPart 1 of these recommendations covers thegeneral recognition, investigation, and assess-ment of a particular group of patients withDPLDnamely, those who usually presentwith diVuse shadowing on chest radiographyand with gradual or subacute onset of symp-toms. It is recognised that some patients willhave shadowing on the chest radiograph withno or few symptoms, while others will have

symptoms due to DPLD but a normal chestradiograph. Such patients are also included inthe remit of this document.

In Part 2 the treatment of a range of DPLDs

is reviewed, with particular emphasis oncryptogenic fibrosing alveolitis (CFA), sar-coidosis, and referral for transplantation.

Purpose of the recommendations+ To support improved recognition, diagnosis,

assessment, and treatment of patients withDPLD.

+ To raise awareness of the importance ofDPLDs.

+ To provide an authoritative current litera-ture review of DPLD management.

+ To provide practical, evidence and consen-sus based recommendations that will aid the

development of clinical services for patientswith DPLD.

Methods of guideline productionThe core group initially defined the target andthe purpose of the recommendations. Independ-ent Medline literature searches from 1985onwards were carried out by the Library of theRoyal Society of Medicine (on diagnosis andassessment of DPLD) and by the NHS LibraryDepartment of University Hospital Nottingham(on treatment of DPLDs). After additional handsearching of the literature, a draft document wasproduced containing recommendations basedon position papers on (1) clinical assessment,

(2) imaging, (3) lung function and exercise test-ing, (4) the role of bronchoalveolar lavage andDTPA scanning, (5) biopsy and pathology, and(6) DPLD in immunocompromised patients.The entire document was reviewed by a widergroup of respiratory physicians, thoracic sur-geons, radiologists, pathologists, and generalpractitioners (listed in Appendix 1). The recom-mendations were also reviewed by BTS mem-bers at a symposium at the BTS Meeting inDecember 1997. Modifications were then madeand an algorithm for the management of DPLDwas added.

Following agreement that the final document

should also contain recommendations on treat-ment, the group produced literature reviews ontreatment of (1) CFA, (2) sarcoidosis, (3) otherDPLDs, and (4) transplantation. The treatmentrecommendations were then agreed by consen-sus between the 11 members of the group.

The summary recommendations in Part 1and the treatment recommendations for CFA,sarcoidosis, other DPLDs, and referral fortransplantation in Part 2 were graded accord-ing to the criteria of the Scottish IntercollegiateGuidelines Network (grades A to C) which isgiven in Appendix 2.

Membership of the coregroupThe core group responsiblefor considering whetherrecommendations in DPLDare appropriate and forwriting a draft documenton diagnosis and assessmentconsisted of: Dr M Cushley,Dr A Davison, Dr R duBois, Dr C Flower, Dr AGreening, Dr N Ibrahim,Dr I Johnston (Chairman),Dr D Mitchell and Profes-sor A Pickering. Professor JGibson and Dr J Egan wereco-opted onto the groupin considering treatmentrecommendations.

Thorax1999;54(Suppl 1):S1 S1


2/30

Part 1: Diagnosis and assessment of diVuse parenchymallung disease

Summary of recommendations

Referral

+ Patients with DPLD or suspected DPLD should be under the direct or joint care of a respiratory physician. (C)

Clinical+ An accurate detailed history is vital. There should be particular emphasis on coexisting or past systemic disease,

medication, occupational and environmental exposures, travel, family, and smoking history, and the possibility ofunderlying immunosuppression, including HIV. (C)

+ In addition to careful respiratory examination, there should be evaluation of systemic disease including fever,cardiac status, rashes, eye signs, hepatosplenomegaly, arthritis and urine dipstick. (C)

+ The initial laboratory investigations should include full blood and eosinophil counts, viscosity or erythrocytesedimentation rate (ESR), serum urea and electrolytes, creatinine, calcium, liver function tests, and autoantibod-ies (antinuclear and rheumatoid factors). Depending on the clinical setting, further investigation may need toinclude additional autoantibodies (antineutrophil cytoplasmic antibody (ANCA), glomerular basementmembrane antibody (anti-GBM)), serum precipitins, electrocardiography and echocardiography. (C)

+ Previous radiographs or reports should be sought. (C)

Imaging+ Radiologists and respiratory physicians should meet regularly to jointly evaluate imaging in patients with DPLD.

(C)+ Where a radiologists report suggests DPLD on a chest radiograph, it should also raise the question of respira-

tory physician involvement. A statement such as suggest referral to respiratory physician is appropriate.(C)

+ In patients for whom a diagnosis is reached with a high degree of certainty from clinical assessment and chestradiography, high resolution computed tomographic (HRCT) scanning is not required, except in cryptogenicfibrosing alveolitis (CFA). (C)

+ In patients for whom the diagnosis is uncertain after chest radiography and clinical assessment, HRCT scanningis the next investigation of choice and should precede biopsy. (C)

+ HRCT scanning is valuable in detecting DPLD in patients with a normal chest radiograph. (B)+ In the appropriate clinical setting, appearances on the HRCT scan may be suYciently characteristic to preclude

the need for biopsy samples to be taken. (B)+ HRCT scanning also provides prognostic information in CFA and is recommended in this disease irrespective of

its diagnostic role. (B)+ HRCT scanning should be performed on appropriate equipment. At least once a week checks should be made

with a phantom to ensure the equipment gives high quality images. (C)+ Wherever possible an extended interscan interval (20 mm) and a low dose technique should be used. (C)+ The radiologist performing the HRCT scans should have expertise in the technique and recognise the strengths

and limitations of the procedure. At least one radiologist in any department should have a declared interest andbe trained in chest radiology and HRCT scanning. (C)

+ Consideration should be given to establishing a reference panel of radiologists with particular expertise in HRCTscanning. (C)

+ DTPA scanning is being evaluated in the diagnosis and assessment of DPLD but is not currently recommendedas a routine test. (C)

+ Gallium scanning is not recommended in DPLD but is of occasional value in extrapulmonary sarcoidosis. (C)

Lung function+ While most patients will have restrictive lung function, some DPLDs are associated with airflow obstruction or

occur in patients with pre-existing airflow obstruction. A finding of airflow obstruction should not by itself leadto exclusion of a diagnosis of DPLD. (B)

+ The minimum lung function assessment should be spirometric values and gas transfer factor, which together givea reasonable measure of the extent of the disease. Vital capacity and gas transfer factor are the most appropriatelung function tools for disease monitoring. (B)

+ Respiratory exercise testing should be considered in patients in whom DPLD is still suspected after a nor-mal HRCT scan and static lung function. It may also help to suggest or exclude the possibility of DPLD inbreathless patients in whom there are no clear pointers to a diagnosis of DPLD. Facilities for exercise testingshould therefore be within easy access of all respiratory physicians. (C)

+ When a diagnosis of DPLD has already been made from other assessments, there is inadequate evidence as towhether exercise testing is of further value in management. For the present, exercise testing is not recommendedroutinely. (C)

+ The BTS should establish the most appropriate methodology for exercise testing and the most clinically impor-tant parameters derived from such testing. (C)

Thorax1999;54(Suppl 1):S2S14S2


3/30

Bronchoalveolar lavage (BAL)+ BAL should be performed routinely in patients in whom malignancy or opportunist infection is being consid-

ered. BAL should be considered in other conditions such as suspected occupational lung disease or alveolarproteinosis. (B)

+ BAL should be readily available to all respiratory physicians assessing a patient with suspected DPLD. (B)+ BAL should be performed according to published guidelines. (C)+ The added value of diVerential cell counts in BAL fluid is uncertain. A network of centres with expertise in the

perfomance and analysis of BAL fluid samples should be established to answer important questions about theclinical value of this technique. (C)

Biopsy techniques and pathology+ Pathologists and respiratory physicians should meet regularly to review biopsy samples from patients with

DPLD. (C)+ It should be standard practice to take lung biopsy samples in DPLD when the diagnosis remains uncertain after

clinical and radiological assessment, unless there are patient contraindications or when the samples are veryunlikely to contribute to management. (B)

+ When small tissue samples are likely to provide a diagnosis (as in sarcoidosis and malignancy), transbronchiallung biopsy (TBLB) samples should be taken. (B)

+

In patients with suspected sarcoidosis endobronchial biopsy samples should also be taken, particularly if difficultyis found in obtaining adequate TBLB samples. (C)+ TBLB samples should only be taken by a bronchoscopist experienced in the technique, or by a trainee under the

direct supervision of such a bronchoscopist. (C)+ TBLB samples must be taken in line with previous BTS recommendations on bronchoscopy. (C)+ TBLB samples must be taken only with a normal coagulation screen and platelet count. It is a day case proce-

dure. Fluoroscopy is not required. Four to six adequate biopsy samples should be taken from one lung. (B)+ The pathological report must be considered in relation to the clinical setting. A normal TBLB sample, or one

showing non-specific inflammation or fibrosis, should be considered non-diagnostic. (B)+ In the diagnosis of conditions such as CFA, vasculitis, lymphoma, lymphangioleiomyomatosis (LAM), or

Langerhans cell histiocytosis (LCH) open lung biopsy (OLB) samples or video-assisted thoracoscopic (VATS)biopsy samples should be taken if a histological diagnosis is required. (B)

+ As with TBLB samples, open lung or VATS biopsy samples showing non-specific inflammation or fibrosis withno characteristic features of CFA should be reported as non-diagnostic. (B)

+ OLB and VATS biopsy samples are equally eVective in providing tissue for histological examination. VATS ispreferred in stable patients. OLB is the procedure of choice in ventilated patients. (B)

+ HRCT scanning should be used to guide the site of OLB or VATS biopsies. (B)+ Percutaneous biopsies are not recommended for DPLD in general, but may be appropriate when there are focal

peripheral infiltrates. (B)+ Wherever possible, biopsy samples from patients with DPLD should be examined by a pathologist experienced

in respiratory pathology. (C)+ A reference panel of pathologists with experience in the diagnosis of DPLD should be established. (C)

Training/education+ Where two or more respiratory physicians work in a team, one physician should have a declared interest in DPLD

and take the lead in developing services for such patients. (C)+ Training in the management of DPLDs should be a required part of specialist registrar training. (C)+ Consideration should be given to establishing a national network of centres with an interest and expertise in

DPLD to promote best practices and research in the field. (C)+ TBLB and BAL are specialised procedures. The BTS should give consideration to the minimum number of pro-

cedures to be performed by trainees before they are considered accredited in such techniques. (C)+ Laboratory pulmonary function testing should be performed by trained lung function technicians. The BTS

should establish minimum standards for the training of such technicians. (C)+ The BTS and ARTP have developed standards for the performance of lung function tests, and are currently

working on standards in exercise testing. All lung function laboratories should adhere to such standards and besubject to national quality control. (C)

+ Ideally respiratory physicians, radiologists, pathologists, and thoracic surgeons should meet to review the clinicalcourse, imaging, and biopsies of patients with DPLD. As a minimum there should be meetings between respira-tory physicians and radiologists, and between respiratory physicians and pathologists, as indicated above. (C)

+ VATS and OLB are procedures with considerable potential for morbidity and mortality and should be performedonly by thoracic surgeons with expertise in the field. (C)

Diagnosis and assessment of diVuse parenchymal lung disease S3


4/30

Background literature review

Clinical evaluationCLASSIFICATION

The DPLDs comprise over 200 entities4 andinclude a wide spectrum of diseases, many

uncommon and many of unknown aetiology.There is no universally agreed classification ofDPLDs. Clinically, however, the diseases havesimilar presentations with increasing shortnessof breath and widespread shadowing on thechest radiograph. While the focus of this reviewis on DPLDs that present with a subacute orchronic onset, there is overlap with thosepresenting acutely. As clinical assessment is thestarting point for the management of these dis-orders, the whole spectrum of DPLDs ispresented here.

They can be classified into:1. Acute DPLD (table 1).

2. Episodic DPLD, all of which may presentacutely (table 2).3. Chronic DPLD due to occupational or

environmental agents (table 3) or drugs(table 4).

4. Chronic DPLD with evidence of systemicdisease (table 5).

5. Chronic DPLD with no evidence of sys-temic disease (table 6).

EpidemiologyEpidemiological data on DPLD are sparse.The annual incidence of chronic DPLD inNew Mexico was found to be 31.5 per 100 000

in men and 26.1 per 100 000 in women. 5 Theprevalence of cryptogenic fibrosing alveolitis(CFA) has been reported to range from 6 per100 000 to 14.6 per 100 000.5 6 The incidenceof CFA increases markedly with age, being 160per 100 000 in those of 75 years of age andolder.5 Deaths in England and Wales attributedto CFA rose from 336 in 1979 to 1035 in1996,7 with mortality rates increasing sharply

with age.8 CFA and sarcoidosis are probablythe most common chronic DPLDs.5

Clinical assessmentThis involves a combination of history andexamination, laboratory investigations, pulmo-nary physiology, imaging, bronchoalveolar lav-age, and histological examination. A recom-mended algorithm for the diagnosis of DPLDis shown in fig 1.

History and examinationLENGTH OF HISTORY

It is first necessary to establish whether the dis-ease is acute (table 1), episodic (table 2), orchronic (tables 36). The acute diseases gener-ally have a history of less than three weeks. Thelength of time symptoms have been present istherefore very important. Every eVort must bemade to review all previous chest radiographsor, failing this, reports. Radiographic abnor-malities may have been present for muchlonger than the symptoms, thereby identifyingthe condition as chronic. Episodic shadowingon the chest radiograph narrows the diagnostic

field (table 2). However, there is considerablevariation in the time course of presentation ofmany DPLDs. For example, eosinophilicpneumonia,9 cryptogenic organising pneumo-nia (COP),10 and Wegeners granulomatosis11

may present acutely, episodically or chroni-cally. Others such as drug-induced DPLDsmay be acute or chronic.12

OCCUPATIONAL,ENVIRONMENTAL AND TRAVELHISTORY

The range of occupational exposures associ-ated with the development of DPLD is verybroad and includes avian, animal and fish pro-

teins, fungal spores, asbestos, silica, cobalt,beryllium, aluminium, isocyanates, copper sul-phate, and sodium diazobenzene sulphate.Most agents relate to specific occupations butsome, such as asbestos and silica, may beencountered in diVerent occupational settings.It is not suYcient to record an occupationalhistory as a job title. A detailed history of theoccupational process, the exposure levels thatare likely to have been experienced, and theprovision and type of respiratory protectionprovided should be taken in both working andretired patients. Actual exposure measure-ments may be available from the industry con-

cerned. Similarly, attempts should be made toobtain previous measurements of lung functionand chest radiographs from occupationalsettings or previous hospital investigations. Incases in which it is uncertain whether occupa-

Table 1 Acute DPLD

Cause Example

I nf ect io n Bact er ial (including t ub er culo sis)

Viral (e.g. chicken pox, measles)Fungal (e.g. invasive aspergillosis,

histoplasmosis. associated withimmunodeficiency, e.g.Pneumocystis and cytomegalovirus)

Allergy Drugs* (e.g. penicillin)Fungi (e.g. aspergillosis)Helminths (e.g.Toxocara)

Toxin s Dr ugs* (e.g. c ytotoxi cs, a miodarone)Toxic gases, fumes (e.g. chlorine)

Haemodynamic Left venticular failure*, fluidoverload, renal failure

Vasculitis/haemorrhage* Goodpastures syndrome, idiopathichaemosiderosis, Behcetssyndrome, systemic lupuserythematosus , Wegenersgranulomatosis, Churg-Strausssyndrome

ARDS Trauma, septicaemia

Unknown Cr yptogenic organisingpneumonia*, cryptogenicpulmonary eosinophilia*

After reference 39.*May also present with chronic disease.

Table 2 Episodic DLPD

Eosinophilic pneumoniaVasculitides/pulmonary haemorrhageChurg-Strauss syndrome

Extrinsic allergic alveolitisCryptogenic organising pneumonia

S4 Diagnosis and assessment of diVuse parenchymal lung disease


5/30

tional exposures are relevant, advice from arespiratory physician with expertise in occupa-tional disease should be considered. Hobbiesand pastimes should be reviewed, with specificenquiry as to exposure to birds. Keeping budg-erigars or pigeons is the commonest form ofexposure. Parasites may cause pulmonary eosi-nophilia and therefore a travel history shouldalso be taken.

OTHER IMPORTANT PAST HISTORY

A past history of cancer and radiotherapyshould be sought. Patients with basal cracklesdue to chronic DPLD will frequently have been

prescribed diuretics for an erroneous diagnosisof heart failure. A lack of response to diureticsis usual in chronic DPLD but does not excludecardiac failure. A cardiac history and evalua-tion is therefore crucial. A past and currenthistory of asthma and rhinitis is invariable inChurg-Strauss syndome.

Individuals with opportunist infections, neo-plasms, or other conditions relating to immu-nodeficiency (HIV positive or otherwise) canpresent subacutely with diVuse lung shadowingon the chest radiograph. Risk factors for HIVshould be sought (homo/bisexual, intravenousdrug administration, blood transfusion, bloodproducts, HIV positive mother, unprotectedsex). Patients with other forms of immunodefi-ciency (non-HIV) may have obvious predispos-ing features such as cytotoxic or immunosup-pressant therapy for transplantation, neoplasm,or collagen vascular disease. However, theremay be no obvious risk factors for HIVinfection and the less common immunodefi-ciency syndromes may not have clear clinical

pointers. A high index of clinical suspicion istherefore necessary.

FAMILY, SMOKING AND DRUG HISTORYCFA and sarcoidosis may rarely be familial.13 14

Patients with Langerhans cell histiocytosis(LCH) and Goodpastures syndrome are almostalways smokers.15 16 By contrast, patients withextrinsic allergic alveolitis and sarcoidosis areless likely to be smokers than the generalpopulation,17 18 but the diVerences are not largeenough to be helpful diagnostically.

A careful drug history is vital as many classesof drugs cause DPLD (table 4).19 20 The

variable timing of onset of symptoms fromdrug related DPLD has already been noted.Some drugs such as cyclophosphamide andamiodarone may have been taken for up to sev-eral years before such reactions develop.19 20

RESPIRATORY SYMPTOMS AND SIGNS

Most DPLDs present with shortness of breathbut cough may be prominent, particularly inlymphangitic carcinoma, sarcoidosis, COP,CFA, extrinsic allergic alveolitis, and eosi-nophilic pneumonia. Other chest symptomsare uncommon but, if present, are importantand may strongly suggest various diagnoses.

Pleurisy may occur in the course of systemiclupus erythematosus (SLE) in up to 50% ofcases and in 25% of cases with rheumatoidarthritis, but it is rare in CFA. Acute chest painsecondary to a pneumothorax occurs at somepoint in up to 40% patients with LCH,tuberosesclerosis,neurofibromatosis,or lymph-angioleiomyomatosis (LAM). Haemoptysis isoften, but not always, present in patients withpulmonary haemorrhage and vasculitis. It isunusual in cases of lymphangitic carcinomaunless there is primary lung cancer. Wheezingmay occur in eosinophilic pneumonia becausemany patients also have asthma.

Bilateral fine end inspiratory crackles arecommon in CFA, fibrosing alveolitis associatedwith connective tissue disease, and asbestosis,occurring in 60% in two series21 22 but generallyreported in at least 90% of patients with

Table 3 Chronic DPLD secondary to occupational or environmental agents

Agent inhaled Example

Inorganic dustsFibrogenic Asbestosis

SilicosisCoal workers pneumoconiosisHard metal (cobalt)Aluminium lung

Non-fibrogenic Siderosis (iron)

Stannosis (tin)Baritosis (barium)Antimony

Granulomatous/fibrogenic BerylliosisOrganic dusts (extrinsic allergic alveolitis)

Bacteria Farmers l ung (Thermoactinomycetes in mouldy hay)Bagassosis (Thermoactinomycetes in mouldy sugar cane)

Fungi Suberosis (in cork workers)Cheese workers lung (mouldy cheese)

Animal protein Bir d f ancier s lung (avian pro tein on feathe rs)Chemicals Pyrethrum extrinsic allergic alveolitis

Isocyanates

Table 4 Drug and toxin induced DPLD: classification with examples

Antibiotics Nitrofurantoin, sulphasalazineAnti-inflammatory agents Gold, penicillamine, aspirinCardiovascular agents Amiodarone

Chemotherapeutic agents Bleomycin, methotrexateDrug induced SLE HydrallazineIllicit drugs Heroin, methadone, talcMiscellaneous Oxygen, radiation, lipoid pneumonia

After reference 40.

Table 5 Examples of chronic DPLD with evidence of systemic disease

(A) Connective tissue disordersSystemic sclerosis Rheumatoid arthritisS ys te mic lup us e ryth em ato sus Polymyo sitisSjogrens syndrome Mixed connective tissue disordersAnkylosing spondylitis Behcets disease

(B) Neoplastic*Lymphoma *Lymphangitic carcinomaMicrometastates

(C) *VasculitisWeg ener s g ranulo mato sis M ic ros co pic p olyang iitis

Goodpastures syndrome(D) *Sarcoidosis(E) Inherited disorders

Tuberose sclerosis NeurofibromatosisLipid storage disease Her mansky-Pudlak syndrome

(F) Other miscellaneousIn flammator y bowel di sease HIV a ssoci atedPost bone marrow transplantation *Amyloidosis*Cryptogenic organising pneumonia Miliary tuberculosis*Langerhans cell histiocytosis *Pulmonary eosinophilia

*May be confined to the lung.

Table 6 Examples of chronic DPLD with no evidence of systemic disease or external agentexposure

Cr yptogeni c fi brosing alveol itis *Sarcoidosis*Cryptogenic organising pneumonia *Langerhans cell histiocytosis*Alveolar proteinosis Bronchocentric granulomatosis

Chronic aspiration Pulmonary veno-occlusive diseaseAlveol ar mi croli thia sis Idi opathi c pulmona ry ha emosid erosisLympha ng ioleiomyomatosis Bronchoalveola r ca rci noma*Pulmonary eosinophilia

*May be associated with systemic disease or external agents.



6/30

Full blood count (eosinophils) and ESR

Urea, creatinine and electrolytes

Liver function

ANA and rheumatoid factor

Lung function tests (VC/TLCO)

Asindicated

Serum precipitins

Autoantibodies

ECG/echocardiography

ACE

ANCA/anti-GBM

History

Past history

Lifetime occupations

Drugs

Hobbies/pastimes

Pets

Travel

Immunosuppression

Systemic diseases

Smoking

Secure clinical

diagnosis of high

probability*

Diagnosis

likely by

bronchoscopy**

Clinical

diagnosis

of CFA

No further

investigation

No

diagnosis

Transbronchial

biopsy

bronchoalveolar

lavage

Diagnosis

of high

probability****

Video-assisted

thoracoscopy/

open lung biopsy2

Age

Performance

status

Clinical

diagnosis

Secure diagnosis

* Bilateral hilar lymphadenopathy

erythema nodosum

Relevant occupational exposure

and chest radiography

Pulmonary eosinophil ia

Some drugs

Secondary cancer

**** HRCT scanning is likely to yield

a diagnosis of high probability for:

Fibrosing alveolitis/asbestosis

Lymphangitic carcinoma

Sarcoidosis

Subacute extrinsic allergic alveolitis

Silicosis

Lymphangioleiomyomatosis

Alveolar proteinosis

Langerhan's cell histiocytosis

** Sarcoidosis

Infection

Malignancy

Organising pneumonia

1Some specialist centres also use bronchoalveolar lavage at this point as an adjunct

to diagnosis and assessment of DPLD.

2Transbronchial lung biopsy may still be appropriate if HRCT scan suggests diagnosis likely

to be made by bronchoscopy (see text).

*** Fibrosing alveolitis

Langerhans' cell histiocytosis

Lymphangioleiomyomatosis

Diagnosis

uncertain

Uncertain

diagnosis or

unlikely to be

diagnosed at

bronchoscopy***

HRCT1

InvestigationsExamination and

chest radiography

Current radiograph

Previous radiograph

Urine dipstick

Figure 1 Recommended algorithm for the diagnosis of DPLD.



7/30

CFA.23 24 Crackles are much less common insarcoidosis and extrinsic allergic alveolitis(25% or less)22 25 and LCH (10%).22 Wheezesor squeaks may be heard if bronchiolitis ispresent. Finger clubbing is seen in 4966% ofpatients with CFA,2 23 in up to 75% of patientswith DPLD due to rheumatoid arthritis,26 butless often in asbestosis (43% of cases in oneseries27) and in other DPLDs. Signs of pulmo-nary hypertension should be sought, particu-larly in systemic sclerosis. In patients withsevere disease tachypnoea at rest, cyanosis, andsigns of cor pulmonale may be present.

SYSTEMIC SYMPTOMS AND SIGNS

Systemic symptoms and signs should be sought(table 7). Fatigue and some weight loss arecommon but non-specific. Significant weightloss suggests malignancy but may occur withany severe DPLD. Systemic features thatsuggest the possibility of HIV disease includedramatic weight loss, a background of general-ised ill health, unexplained diarrhoea, general-ised lymphadenopathy, oral candidiasis, andcutaneous Kaposis sarcoma.28 29

Initial blood and other testsThe minimum initial investigations are urinedipstick, full blood count and eosinophil count,urea, electrolytes and creatinine, liver functiontests, and auto-antibodies (ANF and RF). Ifvasculitis is suspected, the titres of anti-neutrophil cytoplasmic and anti-basementmembane antibodies should also be measured.If the diVerential diagnosis includes sarcoidosis,serum calcium levels should be measured as

they were found to be raised in 1118% ofpatients in large European studies.30

Serum angiotensin converting enzyme(ACE) has a diagnostic sensitivity of only 60%and poor specificity.30 31 While it has been con-sidered useful in long term management,30 itdoes not correlate with radiographic stage andhence prognosis,32 and does not add to the pre-dictive value of serial pulmonary function test-ing and chest radiography in the managementof the disease.33 The routine measurement ofserum ACE levels is not recommended.

If extrinsic allergic alveolitis (EAA) issuspected, serum should be tested for precip-

itins to environmental allergensfor example,

avianproteins,Micropolysporafaeni,and Thermo-actinomyces vulgaris. Precipitins are markers ofantigen exposure and do not seem to beinvolved in the pathogenesis of the disease.While most pigeon breeders and farmers withEAA have serum precipitins,34 the selectivity isvery poor with precipitins being found in up to50% of asymptomatic pigeon breeders35 36 and810% of asymptomatic farmers.35 The avail-able evidence does not support routine testingfor precipitins in patients with DPLD.

In the United Kingdom there is a nationalquality control system for all these investiga-tions except anti-basement membrane anti-body.

Echocardiography should be performed inpatients with systemic sclerosis to assesspulmonary hypertension and in cases wherethere is doubt as to whether heart failure maybe the cause of diVuse shadowing on the chestradiograph. Electrocardiography should beperformed in patients with sarcoidosis.

The technique of induced sputum can diag-

nose Pneumocystis cariniipneumonia (PCP) inup to 50% of HIV positive patients.37 38 Thisapproach requires meticulous technique, istime consuming, and requires special labora-tory skills. It is therefore suitable only for cen-tres handling large numbers of such patients.

Imaging in DPLDHOW USEFUL IS THE CHEST RADIOGRAPH?

Patients with suspected DPLD will have achest radiograph as the initial imaging investi-gation. In most cases this is abnormal andoccasionally the radiographic appearances are

suYciently characteristic to enable a specificdiagnosis to be made when taken in conjunc-tion with the clinical and laboratory findingsfor example, sarcoidosis, pulmonary eosi-nophilia, and some occupational lung diseases.In most patients, however, the chest radio-graphic pattern is not specific and the correctdiagnosis will be made from the first twochoices in only 50% of cases.41

In a small proportion of patients with openlung biopsy confirmation of DPLD the chestradiograph may be normal.42 Conversely, thechest radiograph may be over read as abnormalwhen there is no histological evidence of

DPLD.

43

The best kilovoltage peak (kVp) forthe assessment of DPLD on the chest radio-graph is 125.

WHAT IS THE ADDED VALUE OF HIGH RESOLUTION

COMPUTED TOMOGRAPHIC (HRCT) SCANNING?

HRCT scanning is capable of imaging the lungwith excellent spatial resolution and providinganatomical detail similar to that seen by grosspathological examination. The modificationsof the CT technique which make it one ofhigh resolution are the use of thin sections(collimation) and image reconstruction with ahigh spatial frequency algorithm. In addition,

both kilovoltage peak and milliamperes arenormally increased.The added value of HRCT scanning in

DPLD depends upon its ability to increaseconfidence of a specific diagnosis, to alter

Table 7 Systemic signs and DPLD

Fe ver Infections (including im munodeficiency), e os inophilicpneumonia, drug reactions, vasculitis, connective tissue disorders,cryptogenic organising pneumonia, extrinsic alveolitis, sarcoidosis,AIDS, lymphoma and lymphangitic carcinoma

Rash Sarcoidosis, c onnective tissue disorders, v asculitisScler itis Sarc oidos is , c onnective tiss ue disor de rs, vasculitisKeratoconjunctivitis sicca Sjogrens syndromeUve itis Sarc oidos is , Behc ets syndrome , ankylosing s pondylitisRaynauds phenomenon Systemic sclerosis,CFASystemic hypertension Connective tissue disorders, Goodpastures syndrome, vasculitisLachrymal and salivary

gland enlargement Sarcoidosis,Sjogrens syndromeLymphadenopathy Sarcoidosis,lymphoma,lymphangitic carcinoma,HIVPericarditis Connective tissue disorders,lymphangitis,lymphoma,vasculitisHepatosplenomegaly Sarcoidosis, LCH, connective tissue disorders, amyloidosis

Arthritis Connective tissue disorders,vasculitis, sarcoidosis,Goodpasturessyndrome

Haematuria VasculitisOral candidiasis Immunodeficiency (especially HIV)

CFA = cryptogenic fibrosing alveolitis; LCH = Langerhans cell histiocytosis.



8/30

patient management and, if possible, to influ-ence outcome.

Detection of DPLDHRCT scanning is able to detect DPLD not

visible on the chest radiograph. The relativesensitivities of the two techniques for thedetection of DPLD are 94% and 80%,respectively.44 HRCT scanning is also betterable than chest radiography to determinewhether the lungs are normal.45 Thus, in-creased sensitivity is matched by increasedspecificity.

Examples of the clinical usefulness of HRCTscanning in patients with symptoms of DPLDand a normal chest radiograph include thedetection of occupational lung disease (such asasbestosis and EAA), early fibrosing alveolitis,and early lymphangitic carcinoma.46 47

Characterisation of disease and extent of diseaseSeveral studies have compared the diagnosticvalue of chest radiography and HRCT scan-ning. They indicate that observer variability islower with HRCT scanning than with chestradiography and that a confident diagnosis ismore likely to be made and more likely to becorrect with HRCT scanning.45 4851 CTinterpretation is particularly accurate in fibros-ing alveolitis, lymphangitic carcinoma, sar-coidosis, silicosis, subacute EAA, and alveolarproteinosis.45 48 50 The diagnostic accuracy ofHRCT scanning is further increased by

concurrent clinical evaluation.52

Unlike chest radiography, HRCT scans pro-vide cross sectional images and the extent ofdisease is therefore much more readily appreci-ated than on the chest radiograph. HRCTscanning may also elucidate patients withinexplicable or complex lung functionabnormalitiesfor example, co-existing fibros-ing alveolitis and emphysema.53

Impact on lung biopsy samplesHRCT scanning has a high degree of accuracyin many forms of DPLD. The percentage offirst choice diagnoses made with a high level of

confidence in two studies was 82% and93%.46 48 Fibrosing alveolitis may be confi-dently distinguished from other forms ofDPLD with an accuracy of 88%.54 Using Baye-sian analysis, Grenier et al concluded that the

combination of clinical, radiographic, andHRCT findings enabled a correct diagnosiswith a high level of confidence in 6180% ofpatients with DPLD.52 On the basis of thesefindings it is evident that HRCT scanning can

prevent the need for a histological diagnosis,particularly when the HRCT appearances arecharacteristic and taken in conjunction withthe clinical findingsfor example, fibrosingalveolitis, sarcoidosis, EAA, asbestosis, lymph-angitic carcinoma, LAM, LCH (table 8).55

Some caution is required, however, as prospec-tive data on the diagnostic accuracy of HRCTscans in unselected populations are stilllacking.

In considering the need for histological con-firmation of DPLD, it is evident that the accu-racy of the CT diagnosis is highly dependenton the experience and interest of the reporting

radiologist. The extent to which HRCTscanning precludes the need for biopsy samplesto be taken in routine clinical practice is notknown.

For patients in whom lung biopsy samplesare required, HRCT scanning is better able todiVerentiate between the need for transbron-chial biopsy or open lung biopsy samples,48 andis also able to determine the most appropriateareas from which the biopsy samples should betaken.56

Assessment of disease activityResearch on the ability of HRCT scans to dif-

ferentiate between active and inactive diseasehas been mainly confined to CFA and fibrosingalveolitis associated with systemic sclerosis.There is evidence that a predominant groundglass pattern is more likely to represent activeinflammatory disease and to respond to appro-priate therapy, particularly in fibrosing alveoli-tis, EAA, and desquamative interstitialpneumonia.5761 It is still unproven that aground glass pattern precedes a reticular orhoneycomb pattern, although this seems likely.Not all ground glass change indicates cellularinflammation, however, as fine intralobularfibrosis may be indistinguishable from a cellu-

lar infiltrate on HRCT scans.60 62

The associ-ation of a ground glass pattern with tractionbronchiectasis or bronchiolectasis is likely toindicate some associated fibrosis, whereasground glass change without traction bron-

Table 8 Typical HRCT appearances of some DPLDs

+ Cryptogenic fibrosing alveolitis (CFA)Patchy abnormalities which predominate in the periphery of the lung and in the lower lobesReticular and honeycomb changes often associated with ground glass opacification and traction bronchiectasis

+ AsbestosisVery similar to CFA with abnormalities mainly in the subpleural regionReticulonodular opacities, thickened interlobular septa and honeycomb changes. Often associated pleural plaques

+ SarcoidosisAbnormalities tend to predominate in the mid/upper zonesMicronodules with a bronchovascular and subpleural distribution

Later there are conglomerate masses with lung distortionCommonly lymph node enlargement+ Lymphangitic carcinoma

Irregular thickening of the interlobular septa. Peribronchial cuYng. Thickening of fissures. No architectural distortion+ Extrinsic allergic alveolitis

Ground glass opacification and poorly defined centrilobular micronodules. Air trapping on expiratory scans. In late stagedisease there are irregular linear and reticular opacities due to fibrosis

+ Langerhans cell histiocytosis (LCH)Cysts often of bizarre shape, associated with nodules. The lung bases are usually spared

+ LymphangioleiomyomatosisThin walled cysts surrounded by normal lung. No zonal predominance.



9/30

chiectasis usually indicates active inflamma-tion.62 Reticular and honeycomb patterns onHRCT scans correlate well with histologicalevidence of fibrosis.58 63

Prediction of response to treatmentBecause of its ability to diVerentiate betweencellular and fibrotic disease with reasonableaccuracy, HRCT scanning can be used to pre-dict response to treatment and is significantlymore accurate than chest radiography in thisrespect.57 61 In treated patients with CFAimprovement in lung function in those with apredominantly ground glass pattern is signifi-cantly better than in those with a reticular pat-tern or a mixture of ground glass and reticularpatterns. This improved response rate ismatched by improved survival, the predictivevalue of the HRCT scan being independent oflung function or duration of breathlessness.57

The extent of fibrosis on HRCT scanningshowed 80% sensitivity and 85% selectivity inpredicting survival.64 In patients with rapidly

deteriorating CFA early evidence suggests thatthe appearance of peripheral as opposed tomultifocal or diVuse parenchymal opacificationon the HRCT scan predicts a better responseto steroid treatment and survival.65

Radiation doseThe skin radiation dose from HRCT scanningusing thin sections (12 mm) at 10 mmintervals is approximately 0.7 mSv. This isapproximately 10% of the dose of conventionalCT scanning using 10 mm slices at 10 mmintervals.66 If HRCT scanning is performed at20 mm intervals the eVective dose is 0.35 mSv.

In radiation terms this is the equivalent ofapproximately seven postero-anterior chestradiographs (for which the eVective dose is0.05 mSv). A significant reduction in radiationdose is also possible without substantial impacton image resolution by the use of lowermilliamperes.67 Combining a low dose tech-nique with a limited number of slices providesgreater accuracy than chest radiography withno significant increase in eVective radiationdose.68 As there is the potential for a high doseof irradiation from CT scanning, all unitsshould have regular quality assurance checksand doses should be monitored periodically

and compared with national standards.

Imaging: conclusions+ Using HRCT scanning rather than radio-

graphy, clinicians are significantly morelikely to determine the correct diagnosis forDPLD, the extent of the disease, and theoptimal site from which biopsy samplesshould be taken.

+ A combination of clinical and HRCT infor-mation enables a correct diagnosis to bemade in up to 80% of patients with DPLD.

+ In the appropriate clinical setting, lungbiopsy samples may not be required when

the appearance of the HRCT scan ischaracteristic.+ HRCT appearances are also of value in

determining disease activity and, in CFA,patients prognosis.

+ HRCT is a specialised technique. Interpret-ation of HRCT images requires particularskills on the part of the radiologist togetherwith a proper understanding of DPLD.

+ Radiologists who perform and reportHRCT examinations should ensure thatthey have the necessary expertise and meetregularly with respiratory physician col-leagues to review imaging from patients withDPLD.

GALLIUM SCANNING AND OTHER IMAGING

TECHNIQUES

Gallium scanning was used in the 1980s as aninvestigation that was thought to be sensitivefor the identification of DPLDs such assarcoidosis, and was believed to be of prognos-tic value.6971 Gallium scans can be diagnosti-cally helpful where other investigations are notdiagnostic in sarcoidosis, but its value in CFAhas never been proved. Furthermore, thepulmonary parenchyma may be positive in awide range of DPLDs and, in this context, gal-

lium scanning is not helpful. Gallium scanshave also been evaluated in pulmonary HIVdisease and add nothing to other diagnostictechniques.72 Serial gallium scans are not usefulas monitors of disease in CFA.

Gallium scanning is expensive and involvesthe patient making two hospital visits (one forinjection and one for scanning). Quantificationis often imprecise and the technique isassociated with radiation exposure. The proce-dure should be restricted to situations in whichthe diagnosis of DPLD (unspecified) is sus-pected and for which further evidence isrequired.

Imaging with radiolabelled indium-111 neu-trophils is being used in research studies todemonstrate neutrophil traYc to the lungs.Radiolabelled indium-111 transferrin is alsobeing explored as a measure of pulmonary vas-cular leak. Positron emission tomography(PET) has been used to identify increasedmetabolic activity within the lung in fibrosingalveolitis. The role of magnetic resonanceimaging (MRI) in identifying early inflamma-tion in the lung is also being investigated. All ofthese techniques may establish themselves inthe future as investigations in DPLD. Atpresent, however, they should be reserved for

research studies.

Gallium scanning: conclusions+ Gallium scanning and other imaging mo-

dalities such as PET and MRI have littleplace in the diagnosis or management ofDPLD.

+ Gallium scanning may have a role insuspected extrathoracic sarcoidosis which isnot accessible to biopsy sampling.

Lung function and exercise testing inDPLDLUNG FUNCTION TESTING AND DIAGNOSIS OF

DPLDDPLDs are usually thought to be characterisedby restrictive lung function, by which is meanta reduction in lung volumes with preservedratio of forced expiratory volume in one second



10/30

(FEV1) to forced vital capacity (FVC) togetherwith a reduction in carbon monoxide transferfactor (TLCO). However, in early disease lungvolumes and transfer factor may be within thenormal range.73 Furthermore, in sarcoidosisevidence of airflow obstruction has been foundin more than half the patients in somestudies,74 75 the prevalence increasing withradiographic stage, but in only 11% ofnon-smokers at presentation in another study.76

Airflow obstruction is found at presentation in433% of patients with LCH77 78 and in6578% of patients with LAM.79 80 Lungvolumes may be relatively preserved in smokerswith CFA,53 81 82 possibly due to coexistingemphysema,53 81 although the FEV1/FVC ratioremains normal.53 8183 These findings suggestthat it is inappropriate to use restrictive lungfunction as an exclusive diagnostic criterion.The sensitivity and specificity of routine lungfunction tests in populations of patients with adiagnosis of DPLD established by other meansare not known.

In systemic sclerosis an isolated reduction inTLCO may point to pulmonary vascular disease.In one study 19% of patients had such lungfunction, 11% of whom developed isolatedpulmonary hypertension, the risk increasingmarkedly with lower TLCO.84

In DPLD in immunodeficiency, lung func-tion tests do not generally have a diagnosticrole.85 Patients may often be too ill to performthe tests and the investigations lack the appro-priate sensitivity and specificity to be ofdiagnostic value. However, the negative predic-tive value of a normal TLCOfor the presence ofPCP in HIV infection is extremely highthat

is, if the TLCOis normal, the presence of PCPis virtually excluded.86

Of the various indices derived from exercisetesting, an increased alveolar-arterial (Aa)oxygen gradient is a more sensitive indicator ofthe presence of various DPLDs87 andsarcoidosis88 89 than conventional lung functiontests. Abnormalities of gas exchange duringexercise may be found in subjects with DPLDand normal resting lung function, and suchabnormalities are not predictable by measure-ments taken at rest.87 In HIV positive patientswith symptomatic DPLD but normal restingarterial oxygen saturation, desaturation on

simple exercise testing has good specificity forPCP.90 91 In general, however, the role ofexercise testing in the diagnosis of DPLD isuncertain, but it has been useful in thedetection or exclusion of DPLD in sympto-matic patients with a normal chest radiographand resting lung function.92 However, thesensitivity and specificity of exercise testing forthe diagnosis of DPLD in this respect areunknown, and if occult DPLD is suspected asthe cause of breathlessness in such patientsmost will currently go on to HRCT scanning.No studies have assessed the diagnostic valueof exercise testing in comparison with HRCT

scanning.

Lung function testing and diagnosis: conclusions+ The usefulness of lung function testing in

the diagnosis of DPLD is limited.

+ A restrictive pattern of lung function isprobably the commonest pattern, but a pro-portion of patients have preserved lung vol-umes or airflow obstruction. Such abnor-malities should not lead to the exclusion of adiagnosis of DPLD.

+ The literature suggests that the role of exer-cise testing in the diagnosis of DPLD is

probably limited to the detection or exclu-sion of occult DPLD as the cause of breath-lessness in some patients with a normalchest radiograph and lung function. There isno evidence as to the relative value ofexercise testing and HRCT scanning in thissituation.

SEVERITY AND PATTERN OF DISEASE

Lung function tests are conventionally used togive a global index of functional impairment.Vital capacity (VC), total lung capacity (TLC),and TLCO are most commonly used while, inthe UK, exercise testing is used relatively little.

Several studies have correlated lung functionand exercise test parameters with the degree ofpathological abnormality on lung biopsy sam-ples using the latter as the gold standard.

Fulmer et al93 found that TLC, TLCO, andoxygen tension (PaO2) at rest did not correlatewith either the degree of fibrosis or cellularityin patients with CFA, while both exerciseinduced changes in PaO2 and the Aa oxygengradient were highly correlated with fibrosisand reasonably so with cellularity. VC and pul-monary compliance correlated with fibrosis,though not with cellularity.93 These and otherdata led to the view that gas exchange on exer-

cise gave a better indication of overall diseaseextent than lung volumes or TLCO.88 However,other studies have reported diVerent findings.In untreated patients with CFA both gas trans-fer (TLCO, not KCO) and lung volumes stronglycorrelated with the extent of fibrosis and cellu-lar infiltration, and more strongly so than gasexchange on exercise.94 Other studies suggestthat TLCO is a predictor of exercise testabnormalities.95 96

In sarcoidosis TLCO (particularly) and VCcorrelate with the overall severity of pathologi-cal change, but only TLCOand PaO2on exerciseparallel both the degree of fibrosis and

granulomatous involvement.97

Neither lung function nor exercise testparameters discriminate between fibrosis andinflammation in CFA.88 93 94

Watterset al98 developed a composite clinicalradiographic physiological (CRP) score usingseven variables: dyspnoea, chest radiograph,FEV1 and FVC, TLC, TLCO/VA, resting Aagradient, and exercise oxygen saturation. TheCRP score correlated well with an overallpathology score and better than any of its com-ponent parts. The CRP score is furtherdiscussed below.

Severity of disease: conclusions+ Simple lung function testing using lung vol-

umes and gas transfer factor gives a reason-able measure of the extent of disease.



11/30

+ The evidence that exercise testing providesadditional useful information as to overallseverity is conflicting.

+ Lung volumes, gas transfer factor, and exer-cise parameters, considered either separatelyor together, do not discriminate betweeninflammation and fibrosis.

MONITORING THE COURSE OF DISEASE

There is little information on the use of seriallung function tests in managing disease or onthe value of one oV tests in predicting treat-ment response or prognosis.

In CFA Agustiet al96 found that both the car-bon monoxide transfer coeYcient (KCO) andthe increase in Aa gradient on exercise at initialassessment correlated reasonably with theincrease over three years in Aa gradient at rest.However, this study had only one outcomemeasurethat is, the ability to be able topredict the change in Aa gradient at rest (not

on exercise) over three years. All patients in thestudy had received steroid treatment, thereforeallowing no analysis of treatment eVects. Agustiet al suggested that exercise testing addedfurther information but no convincing evidenceis provided for this assertion.96 Initial VC hasbeen a predictor of survival in some studies 99101

but not in others.23 24 64 102 Likewise, TLCO haspredicted survival in some studies24 99 101 but notin the most recent study.64 KCOor exercise test-ing do not predict survival.102 In multivariateanalysis an increased FEV1/FVC ratio was themost important initial lung function predictorof survival.99 Over one year of treatment,

changes in FVC and TLCO are also stronglypredictive of survival in CFA.103

In sarcoidosis changes in VC correlate betterwith radiographic improvement than TLCO104

and the reverse is the case with radiographicworsening, with no additional informationfrom exercise testing. Lawrence et al105 studied12 patients with sarcoidosis and found that theimprovement in VC over six weeks paralleledclinical improvement better than TLC andTLCO, but this small study was mainlyconcerned with other markers of activity. VC isprobably a more sensitive index of response tosteroids overall, and there is no evidence that

blood gas analysis/compliance testing addsuseful information (see reference 104 forreview). A small study suggested that monitor-ing VC over a three week trial of steroids islikely to show which patients are responsive. 106

In the original evaluation of the CRP score,the correlation of a change in CRP score oversix months of treatment with a pathologicalscore at presentation reflecting cellularity wasbetter (though not significantly so) than any ofthe component parts of the score includingFVC and KCO.98 Recently, the CRP score hasbeen shown not to improve the ability of the

HRCT appearances to predict mortality inCFA.64 There has been relatively little pub-lished assessment of the value of the CRP scorein monitoring disease in diVerent populationsand its place remains unclear.

Lung function in monitoring DPLD: conclusions+ VC and TLCOare the most appropriate and

simplest indicators of change in DPLD.+ The evidence on whether these measure-

ments can predict survival is contradictory.+ Inadequate data are available on the ques-

tion of whether exercise testing oVersadditional information to VC and TLCO.

+ Further prospective studies are urgentlyrequired as to the most appropriate meas-ures for monitoring disease activity inDPLD.

Role of bronchoalveolar lavageBronchoalveolar lavage (BAL) to sample cellsand non-cellular material from the lower respi-ratory tract has now been used for almost 20years in the evaluation of DPLD.107109 Guide-lines as to the performance of BAL areavailable.110 111 The central current issues arewhether BAL contributes to making a diag-nosis in patients with DPLD and to the

assessment of prognosis.

BAL AND DIAGNOSIS OF DPLD

In diseases such as those due to occupationalexposures to inorganic dusts,112115 opportunistinfection,116 117 suspected malignancy, haema-tological disease and the sequelae of trans-plantation,118120 drug induced disease,121123 andsome rare DPLDs such as alveolarproteinosis124127 and LCH128130 the BAL find-ings can be diagnostic.

In HIV disease BAL oVers the mostimmediate way of establishing a diagnosis ofDPLD and 7090% diagnostic yields are

reported, especially in PCP.28 131 BAL in thissituation is performed by wedging the bron-choscope in an appropriate segmental bron-chus, usually the middle lobe, and instilling andaspirating between two and four 2060 mlaliquots of warmed sterile normal saline.

In other DPLDs characterised by theaccumulation of abnormal numbers of inflam-matory cells in the lung the diVerential cellcounts can detect subclinical disease, especiallyin the context of rheumatological disease.132137

Diagnostically, the predominant inflamma-tory profile provides an indication of the natureof the underlying disease process. An increase

in granulocytes, particularly neutrophils andeosinophils, in the BAL fluid is typical incases of fibrosing alveolitis occurring aloneor with rheumatological disease, asbestosis,and the adult respiratory distress syn-drome.107 108 111 133 138141 Increased numbers oflymphocytes are associated with granuloma-tous and drug-induced DPLD.111 119 142 Ingranulomatous disease diagnostic specificityincreases with the use of CD4:CD8 lym-phocyte ratios at the expense of sensitivity.143 144

In a recent study a discriminant diagnosticmodel was generated from BAL cell counts in apopulation of patients with sarcoidosis, EAA,

and CFA. The model was then tested on a sec-ond population and found to predict thecorrect diagnosis in 95% of cases.145 However,no studies have been performed that assesswhether BAL cell counts provide any useful



12/30

information over and above that from HRCTscanning.

BAL AND PROGNOSIS OF DPLD

Patients with CFA with increased numbers oflymphocytes in the BAL fluid, with or withoutgranulocytes, are more likely to respond tosteroids.107 146 147 In one study increased num-

bers of granulocytes without lymphocytes sug-gested a better responsiveness to cyclophos-phamide than to prednisolone,148 but serialBAL cell counts as a monitor of disease areclinically unhelpful.148

The predominant inflammatory cell is alsoan indicator of prognosis. More extensivedisease in fibrosing alveolitis is associated witha poor prognosis23 24 and BAL neutrophiliaand/or eosinophilia is associated with moreextensive disease and a poorprognosis.107 139 146153 However, in a multivariateanalysis (which did not include HRCT infor-mation as a variable), BAL cell counts did notindependently predict prognosis.151 154

BAL in diagnosis and prognosis: conclusions+ BAL is a useful diagnostic tool, particularly

in suspected infection, malignancy, andsome rare DPLDs.

+ There are inadequate data as to the addedvalue of BAL cell counts in diagnosis andprognosis in other DPLDs. Whether BALcell counts should be performed routinely istherefore unclear. Further research on BALcounts should address this question.

99mTc-diethylenetriamine penta-acetate

(DTPA) scanning99mTc-DTPA aerosol clearance is an index oflung epithelial permeability. Increased perme-ability results from inflammation in the lungparenchyma in a variety of DPLDs such asfibrosing alveolitis, dermatomyositis, EAA,sarcoidosis, radiation pneumonitis, andpneumonconiosis, 155170 often when there islittle other evidence of DPLD. An increasedDTPA clearance is a sensitive marker ofinflammation and a normal clearance certifiesabsence of inflammation.171

In addition to its potential role in the detec-tion of DPLD, DTPA scanning has been used

as a measure of prognosis in sarcoidosis, fibro-sing alveolitis, and asbestosis.158 163 164 Persist-ently rapid clearance is associated with a higherrisk of subsequent deterioration in lungfunction, but persistently normal clearancepredicts disease stability. In HIV disease abiphasic pattern of clearance is seen in PCPwith early rapid clearance of DTPA from thelung,172 but the technique lacks specificity andadds little to other diagnostic techniques. Adrawback of DTPA is that it cannot beemployed in cigarette smokers as airwayinflammation aVects the clearance of theisotope.173176

Serial studies of DTPA in fibrosing alveolitisand radiation induced fibrosis have concludedthat DTPA measurements may predict trendsthat justify treatment of patients before symp-toms develop.164 177

DTPA scanning: conclusions+ DTPA scanning may become important in

the repertoire of investigations in certainDPLDs.

+ Further studies are needed to establish itsvalue in conjunction with other investiga-tions such as HRCT scanning, particularlyin rheumatological disease as an index ofearly lung involvement and in CFA as ameasure of prognosis.

+ InsuYcient data are available to support theroutine use of DTPA scanning in the assess-ment of DPLD.

Role of lung biopsy and pathologyHistological examination is usually performedin DPLD to assist diagnosis, although in CFAit has also been recommended by some authorsto assess disease activity and the likelihood oftreatment response.21 178 179 Despite previousrecommendations that histological examina-tion should be routinely performed inDPLD,179 180 transbronchial biopsy (TBLB) or

open lung biopsy (OLB) samples were onlyobtained in 2833% and 812% of patients,respectively, in two UK studies of CFA.2 181 Inthe USA a questionnaire survey with a lowresponse rate suggested that most physicianstry to obtain a tissue diagnosis in DPLD,3 butan epidemiological survey found that, as in theUK, in only a small proportion (11%) ofpatients with CFA had an OLB sample beentaken.5

The usefulness of HRCT scanning in bothdiagnosis and assessment reduces the need forlung biopsy samples (see page S8) andinfluences the choice of biopsy technique. The

advent of this technique in recent years nownecessitates a reappraisal of the role of lungbiopsy samples.

REVIEW OF BIOPSY PROCEDURES

The decision as to the type, size, and site of thebiopsy sample is determined by the level ofprebiopsy diagnostic certainty, the suspectednature, distribution and extent of the DPLD,and the patients perfomance status. SomeDPLDs show overlapping histological featuresand, for optimal diagnostic results, close liaisonbetween the physician, thoracic surgeon, radi-ologist, and pathologist is important.

Transbronchial lung biopsy (TBLB)Although TBLB samples were first obtainedvia a rigid bronchoscope,197 198 they are nowalmost exclusively obtained via the fibreopticbronchoscope. Sampling error is a problemand the specimens are often too small to enablea definitive diagnosis.199 200 Crushing of thespecimen and failure to penetrate beyond theperibronchial sheath may also preclude histo-logical assessment. Alligator forceps producebigger specimens than large and standard cupforceps but may be more likely to obtain bron-chial wall samples and make no significant dif-

ference to the diagnostic yield.201203

Despite these problems, TBLB achieves ahigh diagnostic yield in DPLDs with centri-lobular accentuation such as granulomatousand metastatic diseases.204206 A high diagnostic



13/30

yield can also be achieved in infection, alveolarproteinosis, and eosinophilic pneumonia. Onereport suggests that TBLB can also providediagnostic material in 64% of patients withCOP.207

In patients with sarcoidosis with diVuseinfiltrates a diagnostic yield of 7589% can beexpected, and in carcinoma the yield is6468%.204 206 208 209 A diagnosis is less likely insarcoidosis (4466%) when parenchymal dis-ease is not visible on the chestradiograph.206 208 210 Endobronchial biopsy sam-ples in sarcoidosis are diagnostic in 4577% ofcases,208 209 211 but whether such biopsy samplesprovide additional diagnostic information tothat obtained from TBLB samples is not clear.The diagnostic rate is less good over the broadspectrum of DPLD with diagnostic infor-mation in 3879% of cases199 200 204 205 212 213 and,most recently, 50% in a large study.206 Inpatients with CFA the TBLB specimens aretoo small and non-representative to alloweither a reliable diagnosis or determination of

the relative degree of cellularity andfibrosis.199 200

Multiple biopsy samples are required withthe diagnostic yield rising from about 60% to90% in radiological stage 2 sarcoidosis whenthe number of samples is increased from two tofour.214 In general, 46 biopsy samples achievean overall good yield in DPLD.206 215

TBLB is a safe outpatient procedure.216

Pneumothorax has been reported in up to 10%of cases but, more generally, the figure is0.72% with about half requiring tubedrainage.198 204 206 217219 The rate of pneumotho-rax was reduced by avoiding the right middle

lobe and lingula and with the use of fluoroscopy.198 219 However, other studies haveshown that TBLB without fluoroscopyachieves similar results with regard to positivehistological findings and the incidence ofpneumothorax.201 205 Bleeding occurs in 9% ofcases216 but this only exceeds 50 ml in about1%.198 204 218 Mortality is approximately 0.1%with haemorrhage the main cause.219 Coagula-tion screening prior to TBLB is necessarypractice.217 220

Open lung biopsy (OLB)OLB by limited thoracotomy provides larger

specimens than TBLB. In a large number ofseries the diagnostic yield was 94% comparedwith 72% with TBLB, 72% with drill, and 63%with needle biopsy.200 OLB provided a diagno-sis in up to 92% of patients who did not have adiagnosis achieved by other biopsytechniques.200 221 In a prospective study of 20consecutive patients with DPLD undiagnosedby non-invasive methods, all underwent aspira-tion and cutting needle biopsy, TBLB, andOLB. There was a 94% success rate with OLBcompared with 59% with TBLB and 2953%with needle biopsy.222

The complication rate is up to 7%, with a

mortality due to the procedure of probably lessthan 1%.221 223 224 The procedure takes about 30minutes and the intercostal drain is in situ forapproximately one day,225 but with considerableindividual variation. Conventionally, more than

one specimen has been taken to includemacroscopically normal lung adjacent to andremote from obviously abnormal sites, avoid-ing areas with the greatest involvement radio-logically which more often show end stagefibrosis of no diagnostic value.226 However, arecent study concluded that a single biopsysample from an inflamed and least fibrotic areaof the radiographically most involved lobeachieves the highest diagnostic yield.227 HRCTscanning is useful to guide the selection of thebiopsy site and is more accurate than chestradiography in predicting whether TBLB orOLB is the more appropriate procedure.48 56

Video assisted thoracoscopic (VATS) lung biopsyVATS is carried out under general anaesthesiawith the patient ventilated through a doublelumen tube or with high frequency jet ventila-tion. A fibreoptic bronchoscope is passedthrough a 2 cm stab incision with two ancillaryoperating ports. Biopsy specimens are the samesize as those from OLB. The diagnostic

accuracy is comparable (8695% with VATScompared with 93100% with OLB).228 229 Theprocedure takes about 40 minutes, similar toOLB,228 and in uncomplicated patients peri-operative morbidity, postoperative pleuraldrainage, and length of hospital stay appear tobe less with VATS, though comparisons havenot been randomised.228234 In ventilator-dependent patients, however, OLB is theappropriate procedure.235

There is controversy over whether the lingulaor middle lobe (surgically easily accessible) aresuitable sites for OLB or VATS biopsy proce-dures in patients with DPLD. It has been

suggested that these sites should be avoidedbecause biopsy samples often show non-specificfibrosis and vascular changes not seen elsewherein the lungs.226 236239 However, in more recentstudies of DPLD, both in immunocompromisedand non-immunocompromised patients, thelingula and right middle lobe have given similarhistological results to other sites.227 240 241 Which-ever site is selected, it is important to obtain anadequate size of specimen as the disease processmay be focal and the diagnosis missed becauseof sampling limitation.226 241 Biopsy samplesshowing end stage lung or minimal disease giveno useful information as to the underlying diseae

aetiology.

Percutaneous biopsyA number of percutaneous techniques havebeen used to obtain lung biopsy samples.Initially these were for peripheral tumours182 183

with subsequent modification for DPLD.184

Diagnostic material in DPLD has been obtainedin up to 78% of cases.185 Morbidity was highwith pneumothorax in up to 50%.185190 Airembolism and haemorrhage were less commonsequelae but were more likely to be fatal with anoverall mortality of 0.13.1%.188 191194 Smallrecent studies have given similar diagnostic rates

but with lower morbidity and mortality.195 196

HANDLING OF LUNG BIOPSY SAMPLES

It is recommended that 46 transbronchialbiopsy samples should be obtained and placed



14/30

with minimal handling and without delay intobuVered formalin.206 242 Open biopsy samplesand those obtained by VATS should be at least4 cm in maximum diameter when inflated andinclude a depth of at least 11.5 cm. The biopsysample should be handled with minimal traumaand sent fresh to the laboratory without delay sothat it can be gently inflated with formalin usinga syringe and a small gauge needle insertedthrough the pleura.242 The technique of inflatingopen lung biopsy samples could also be carriedout in theatre by trained personnel. In DPLD,frozen sections are only of value in suspectedmalignancy. It may be appropriate to freeze asmall portion of the unfixed specimen forimmunofluorescence techniques, but not at theexpense of tissue needed for histological exam-ination. Small pieces of fresh tissue can be takenfor microbiology/virology as required. Immuno-histochemical techniques, in situ hybridisation,and polymerase chain reaction (PCR) can all beapplied to maximise the diagnostic yields.243245

ASSESSMENT OF LUNG PATHOLOGY IN CFA

The diagnosis of CFA should be made only ifcharacteristic histological features arepresent.246 To overcome interobserver variationa panel of pathologists used a semiquantitativescoring system for assessment of inflammatoryand exudative changes, fibrotic and reparativechanges, and airway alteration in CFA.247

There was agreement between pathologists inup to 64% of assessments and the results com-pared favourably with those obtained from amorphometric analysis of specific histologicalfeatures.247 248

The value of such intensive analysis is not

known. In CFA the added value of a biopsysample as an independent predictor of progno-sis has not been established. Single biopsysamples are unlikely to be representative of thedisease process because of heterogeneity of thedisease in diVerent sites.246 The relationshipbetween biopsy results and prognosis was lostafter controlling for age and sex in one study 23

and, although a pathological fibrosis score pre-dicted mortality, an HRCT fibrosis score hadgreater sensitivity and selectivity in thisrespect.64

Lung biopsy and pathology in DPLD: conclusions

+ Biopsy samples, when required, should betaken before the initiation of treatment.

+ TBLB is the initial procedure of choice inthose patients likely to have DPLDs inwhich small samples may be diagnosticforexample, conditions such as sarcoidosis andmalignancy.

+ Four to six TBLB specimens should betaken.

+ Only specimens that contain abnormal

alveolar tissue are likely to contribute todiagnosis, though such specimens may stillbe non-specific. TBLB specimens consistingof normal lung or bronchial tissue, or offibrous tissue, should be regarded as dia-gnostically unhelpful.

+ Endobronchial biopsy samples are fre-quently positive in suspected sarcoidosis,but it is not clear whether they add furtherinformation to TBLB. However, the mor-bidity of taking endobronchial biopsy sam-ples is very low.

+ TBLB showing non-specific pneumonitis,with or without fibrosis, may help in

supporting a clinical impression of DPLDbut should not be regarded as diagnostic.+ TBLB is not useful for the diagnosis or stag-

ing of CFA.+ Percutaneous needle biopsy samples should

be reserved for focal lesions.+ Certain DPLDs such as CFA, LCH, LAM,

vasculitis, and lymphoma cannot be reliablydiagnosed with TBLB. OLB or VATSbiopsy samples are necessary if histologicalexamination is required.

+ The diagnosis of CFA should be made onlyif characteristic histological features arepresent. In the absence of diagnostic fea-

tures a biopsy sample should be reported asnon-specific.

+ In patients with suspected COP, TBLBsamples should be taken initially andOLB/VATS biopsy samples considered if theTBLB sample is non-specific.

+ OLB and VATS biopsy techniques areequally eVective in providing tissue forhistological diagnosis. VATS is preferred instable patients. OLB is required if thepatient is on a ventilator.

+ Biopsy samples from the middle lobe or lin-gula may be taken provided they are of gen-erous size, contain deep alveolar tissue, and

are from a site involved by active disease asjudged by HRCT scanning.



15/30

Part 2: Treatment of diVuse parenchymallung disease

Part 2 reviews and provides recommendationson the treatment of two major DPLDsCFA

and sarcoidosis. A further section brieflyreviews several other DPLDs, some rare, forwhich only anecdotal evidence is generallyavailable on treatment. Finally, the indicationsfor transplantation are presented.

Cryptogenic fibrosing alveolitisThere is a paucity of controlled trials of drugtreatment in this disorder and none have beenplacebo controlled. Corticosteroids have beenfrequently tried and other immunosuppres-sants less so.

CORTICOSTEROIDS

Although corticosteroids have been used sincethe 1950s, there are no controlled trialscomparing them with placebo, and no data asto whether survival or quality of life isimproved.249 Since 1970 there have beenseveral retrospective reviews of treatment withsteroids in which symptomatic improvementwas reported in 4157% of patients.100 250

Objective improvement, defined as an increasein FVC of 10%100 147 250 or 15%,24 was reportedin 1637% of cases.24 100 147 250 Two more recentstudies compared prednisolone alone withprednisolone and either azathioprine or cyclo-phosphamide. In the prednisolone arm objec-

tive improvement in FVC or TLCOwas foundin 2332% of patients.251 252

There is diYculty in interpreting such stud-ies because (1) some studies included patientswith connective tissue diseases, (2) the dose ofprednisolone used varied even within the samestudy, (3) the proportion of patients followedup has been variable, and (4) there arediVerences as to how long the improvement hasbeen maintained (some studies assessed initialimprovement, others looked for improvementover one year).

Nevertheless, common features that suchstudies identified as being associated with a

steroid response are female sex100 147

and anincreased cellularity of the biopsy sample.21 24 100

A response to steroids has also been associatedwith improved survival.24 100 147 250 However, itmay be that such a response merely identifies agroup with a better prognosis irrespective of

steroids. Turner-Warwick et al also studied asubstantial number of untreated patients and

compared their survival with a treated group.The survival of the untreated group lay between

that of a group of patients responding to steroidsand a group not responding to steroids. Indeed,

in their study there was no survival diVerence

between the steroid treated group and the

untreated group after controlling for age and

sex.250 Young age,23 24 less advanced

disease,23 24 99 100 147 female sex23 99 and a cellular

biopsy sample21 23 have also been associated with

improved survival. However, in one study a cel-

lular biopsy sample was no longer related to sur-

vival after controlling for age and sex.23

Two small randomised studies reported on

the use of high dose methylprednisolone,neither finding clinical benefit at up to sixmonths compared with conventional doses of

oral prednisolone.253 254 Methylprednisolone

has recently been used in acute exacerbations

of CFA with variable success according to the

pattern of parenchymal opacification on the

HRCT scan, but without comparison against

oral prednisolone.65

Several further issues complicate the assess-

ment of the place of steroids in the treatment of

CFA.

(1) Does stability on treatment represent suc-

cess? Few studies have addressed this

issue, most defining steroid response as

improvement in symptoms and lung func-

tion. In two studies the proportion of

patients with stable FVC on prednisolone

was 23%251 252 and, in an earlier study,

1219% were stable on treatment com-

pared with 15% of an untreated group.21

(2) Does spontaneous improvement occur?

DiVerent histological subsets may have

diVerent prognoses.255 Desquamative inter-

stitial pneumonia (DIP) which is charac-

terised by uniform appearances of macro-

phages in alveolar spaces and minimalinflammation with patchy ground glass

shadowing on the HRCT scan may not be

progressive,256 and 22% were reported to

improve without treatment.21 By contrast,

no reports indicate spontaneous improve-

ment in the common type of CFA

namely, usual interstitial pneumonia(UIP)which is characterised by patchy

inflammation, fibrosis, and honeycombing.

(3) Side eVects of steroids. In four studies

serious side eVects were reportedly due to

steroids. The proportion of patients af-

fected ranged from 1926%250 252 to two

thirds21 and a large number.251 Such dif-

ferences may relate to the generally higherdoses of steroids used in the USA21 251 thanin the UK (table 9).250 252

Table 9 Regimens of prednisolone used in CFA

Author Regimen used

Stack100 2040 mg/day for 28 weeks. Then stop or reduce to 510 mgCarrington21 3060 mg daily tapering to 20 mg maintained for 1 yearTurner-Warwick250 1060 mg/dayRudd147 4060 mg for 46 weeks reducing to 1530 mg on alternate daysCrystal257 1 mg/kg/day for 6 weeks tapering to 0.25 mg/kg/dayTukiainen24 40 mg tapering to 10 mg/day over 3 months. At 6 months stop or continue

510 mg/dayJohnson252 >60 kg weight: 60 mg for 4 weeks reducing by 5 mg each week to 20 mg on

alternate days.


16/30

Corticosteroids and CFA: conclusions+ There has been no placebo controlled trial

of steroids in CFA, so there is no direct evi-dence that steroids improve survival.

+ A proportion of patients do respond interms of symptoms (about 50%) and lungfunction (about 25%).

+ Although a steroid response is associatedwith better survival, it is not clear if this isdue to the treatment itself.

+ There are no comparative data on dose regi-mens nor on the length of time for whichsteroids or other treatment should be used.

IMMUNOSUPPRESSANT THERAPY

Cyclophosphamide and azathioprine are thetwo most studied drugs. There have been nocontrolled trials of either drug against placebo.Furthermore, it is not always clear from studiesthat have used these drugs whether they havebeen used concurrently with prednisolone.

Cyclophosphamide

The only randomised controlled trial usingcyclophosphamide compared low dose pred-nisolone (20 mg on alternate days) plus oralcyclophosphamide with high dose pred-nisolone.252 The initial response rates were 24%and 32% in the two groups; after three yearsthree of 21 patients in the group receivingcyclophosphamide/prednisolone had diedcompared with 10 of 22 receiving prednisolonealone. However, at 59 years there were 15deaths in each group. The better survival atthree years was largely attributed to higher lungvolumes at the start of treatment.

In uncontrolled studies a response (>10%

increase in VC over a defined period) has beenreported in about half of patients (combiningstudies, 13 of 23),147 258 259 although there wasno response in any of seven patients in onestudy.260 A probable cure with intravenouscyclophosphamide was reported in a case offulminant CFA261 and a further case reportrecords DIP apparently responsive only tocyclophosphamide.262 Pulse intravenous cyclo-phosphamide in 33 patients was reported toimprove mean lung function in those survivingan initial six months.263 Such data are diYcultto interpret without controls. In a further studypulse cyclophosphamide was associated with

stable lung function both in survivors andthose who died, suggesting that lung functionmay be an unreliable predictor of mortality inend stage CFA.264

In another study responders to cyclophos-phamide had previously responded toprednisolone265 and, anecdotally, Lynch andMcClune266 reported that 70 kg, 120 mg/day; 6069 kg, 110 mg/day; 3000, WBCmonitored every 2 weeks for 46 weeks in first instance

S16 Treatment of diVuse parenchymal lung disease


17/30

intestinal eVects.271

In a further study 22patients receiving colchicine alone were com-pared with 22 historical controls with UIPgiven prednisolone. There was no significantdiVerence in the clinical course of these twogroups using colchicine 0.6 mg once or twicedaily.272 A recently reported randomised pro-spective trial of colchicine versus high doseprednisolone found that few patients appearedto respond to either drug, and there were nodiVerences in survival between the twogroups.273

Penicillamine

Few data are available. In one study three of 10patients improved at two years on prednisoloneand penicillamine269 and in another one of threepatients improved.147

CyclosporinIn seven patients who failed to improve onprednisolone an initial response to cyclosporinwas reported but was not maintained; all sevenpatients died within nine months.274 In afurther study five patients improved and onestabilised on cyclosporin and prednisoloneafter previous deterioration on prednisolone,but only one improved in both symptoms and

lung function.275 A substantial steroid sparingeVect was reported in 10 patients awaiting alung transplant in a further uncontrolledstudy.276

MethotrexateThere are no reports of the use of this drug inCFA.

Other treatments and CFA: conclusions+ Various treatments have been used includ-

ing penicillamine, cyclosporin A, and colchi-cine. Of these colchicine has been best stud-

ied, and appears to have a similar eVect toprednisolone on the course of CFA, but withfewer side eVects. However, as noted previ-ously, it is not known whether prednisoloneitself alters the course of CFA.

BTS RECOMMENDATIONS

When to start treatment?With the uncertain benefits of treatment inCFA, it is essential that treatment is tailored toindividual patients. Discussion with patients ofthe possible outcomes of treatment and sideeVects is strongly advised. Due account shouldbe taken of comorbidities. In certain patients itwill be appropriate not to treat with steroids/immunosuppressants in the light of, for exam-ple, the patients wishes, severe comorbidities,or knowledge of long term disease stability. It isrecommended that clinicians otherwise deter-mine when to start treatment using thealgorithm shown in fig 2. The recommenda-tions are graded AC as defined in Appendix 2.

Further points+ Instead of review at three months (fig 2,

pointd), treatment may be indicated whenthere is a marked reduction in lung functionand patient preference is for treatment. (C)

+ After one year (fig 2, point e) indefinite fol-

low up is strongly recommended.The followup interval can be extended. Open appoint-ments should be available for patients whodeteriorate symptomatically between rou-tine assessments. (C)

+ A ground glass predominant pattern indi-cates underlying pathology that has a poten-tial for spontaneous improvement. Thisoccurs in a minority of cases, however, andtreatment is therefore recommended. (B)

+ VC/TLCOshould be performed where possi-ble at each visit. (B)

+ Repeat chest radiography should be per-formed with unexpected worsening of

symptoms or new symptoms such as haemo-ptysis. (C)

Drug treatment of CFA+ Recommended initial treatment is com-

bined therapy with:oral prednisolone 0.5 mg/kg and azathio-prine 23 mg/kg. (B)

+ Patients should be re-assessed at one month.A response or stability should be followed bya slow tail of prednisolone. A decline shouldbe followed by a fast tail (where response ordecline = increase or fall in VC/TLCO of1015% or more, respectively).

+ A response to prednisolone may be seen atone month, but a longer period (up to threemonths) may be required to assess whetherprednisolone has led to stability. (C)

+ If there is a response to prednisolone buttroublesome side eVects, or there are diYcul-ties using steroids at the outset, azathioprinealone should be considered. (C)

+ If azathioprine is not tolerated cyclophos-phamide should be considered in a dose of12 mg /kg. (C)

+ Slow tail of prednisolone: a suggestedregimen to tail from 40 mg is a reduction inthe daily dose by 5 mg each week to 20 mg.

(C) If at review a response or stability ismaintained, prednisolone should be furthertailed by 5 mg each month to a maintenancedose of 10 mg daily for one year (C) withreview at least three monthly with VC/TLCO.

Figure 2 Algorithm of when to treat CFA.a Ground glass pattern should bepredominant, or g round glass on ly.b A mixed patter n of ground glass and reticularshadowing on HRCT scanning is referred to as mixed.c Decline in lung function = asustained fall in VC/TLCOof 1015% or more.d Treatment may be indicated when there ismarked impairment in lung function and patient preference is for treatment.(C) e After 1year ind efinite follow up is strongly reco mmended.(C)

Presentation

Highly symptomatic

or rapid progression

HRCT

Treat Repeat VC/TLCOno later than3 months

d Stable

3 monthly

review for

at least 1

yeare

Decline in lung functionc

Obvious increased symptoms

Ground glass

a

pattern

Mild/moderate symptoms

or long history

HRCT Reticular/mixedb

pattern

Treatment of diVuse parenchymal lung disease S17


18/30

(B) A further slow reduction may be consid-ered subsequently. (C)

+ Fast tail of prednisolone: a suggested regi-men is a reduction in the daily dose by 10 mgeach week to 10 mg daily. (C) Furtherreduction must be on an individual basis. (C)

+ Other treatments should be considered ifsteroids and immunosuppressants are nottolerated. (C) If these treatments have beentolerated but have failed, the likelihood of abeneficial response to other treatments isvery low. Colchicine is an alternative toprednisolone alone (A) but has not beencompared with steroid/immunosuppressantcombinations.

Documents

Boala Parenchimatoasa Pulmonara Difuza