diagnostic the infiltrates - bmj

Thorax, 1977, 32, 539-545

Diagnostic fibreoptic bronchoscopy in theimmunocompromised host with pulmonary infiltratesRICHARD A. MATTHAY, WAYNE C. FARMER, AND DIXON ODERO

From the Department of Medicine, Division of Pulmonary Medicine,Yale University School of Medicine, New Haven, Connecticut 06510, USA

Matthay, R. A., Farmer, W. C., and Odero, D. (1977). Thorax, 32, 539-545. Diagnosticfibreoptic bronchoscopy in the immunocompromised host with pulmonary infiltrates. Nineteenimmunocompromised patients with pulmonary infiltrates underwent diagnostic fibreopticbronchoscopy with transbronchial forceps and brush biopsy. A specific diagnosis was obtainedin 21/25 procedures (10/11 focal lesions and 11/14 diffuse lesions). The most common diagnosiswas infection, and organisms isolated included bacteria, fungi, Pneumocystis carinii, and herpessimplex. A pneumothorax requiring tube drainage occurred in two cases and mild lungparenchymal bleeding was noted in two others. It is concluded that fibreoptic bronchoscopy withforceps and brush biopsy can be performed safely with an excellent diagnostic yield inimmunocompromised hosts with lung lesions. Supplemental oxygen should be administeredduring fibreoptic procedures in these patients and platelet transfusions should be given whenthrombocytopenia is present.

Since its introduction in 1968, flexible fibreopticbronchoscopy has been widely employed in theevaluation and management of patients with pul-monary disorders (Khan et al., 1976). The pro-cedure is easily performed, is associated with fewcomplications, and affords greater visualisation ofthe endobronchial tree than rigid bronchoscopy.Patienits can undergo this procedure with minimaldiscomfort, and biopsy of previously inaccessibleareas of the lung may be accomplished using thestandard biopsy forceps accompanying the flexiblefibreoptic bronchoscope.

Recently, several studies have demonstrated ahigh degree of diagnostic accuracy employingfibreoptic bronchoscopy for determining theaetiology of both focal and diffuse pulmonaryinfiltrates (Levin et al., 1974; Zavala, 1975; Cun-ningham et al., 1977). The safety and diagnosticyield of this procedure applied specifically to im-munocompromised hosts with pulmonary lesionsrequires analysis. These patients are frequently atrisk for lung biopsy procedures, particularly dueto the presence of arterial hypoxaemia, abnormalclotting factors, and thrombocytopenia (Greenmanet al., 1975; Hanson et al., 1976). The purpose ofthis study was to determine whether fibreopticbronchoscopy employing brush and forceps biopsy

is a safe and accurate method for determiningthe aetiology of pulmonary infiltrates in the im-munocompromised patient.

Material and methods

Nineteen consecutive immunocompromisedpatients who underwent fibreoptic bronchoscopyfor diagnosis of pulmonary infiltrates were studiedprospectively. The study population was limited topatients who had established diagnoses ofleukaemia, lymphoma, collagen vascular disease,organ transplantation or other diseases for whichthey had received corticosteroids, cytotoxic drugs,and/or radiation therapy. There were 13 men andsix women with an average age of 44, range 19-75 years. Underlying disease, characteristics ofchest radiographs, arterial blood gas values,platelet count, prothrombin time, partial throm-boplastin time, complications, diagnostic yield, andpatient outcome were analysed.Immunocompromised patients who developed

pulmonary lesions were evaluated with cultures ofsputum, transtracheal aspirates (Ries et al., 1974),blood, urine, and, if indicated, cerebrospinal fluid.If these failed to yield a diagnosis by appropriatestains and cultures fibreoptic bronchoscopy was

539

on April 3, 2022 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.32.5.539 on 1 O

ctober 1977. Dow

nloaded from

http://thorax.bmj.com/

540

performed. Procedures followed were those des-cribed -by Levin et al. (1974) and Ellis (1975).Patients received nothing by mouth for at leastsix hours before the procedure. Premedications in-cluding codeine (30-60 mg) and atropine(0A4-0-6 mg) were given intramuscularly 30minutes before the procedure. The patients'oropharynx and naris were anaesthetised with 4%lidocaine. In addition, viscous lidocaine was usedto anaesthetise and lubricate the naris. TheOlympus BF 5B2 fibreoptic bronchoscope waspassed transnasally (Smiddy et al., 1971). Underdirect vision, the vocal cords were examinedand further anaesthetised with 2% lidocainethrough the bronchoscope's suction and biopsychannel. The trachea was then entered and biopsyperformed. When indicated, to ensure comfortand to facilitate the procedure diazepam (Valium),5-10 mg, was administered intravenously. Topicalanaesthetic (1% lidocaine) was instilled into theairways with careful dose monitoring.

Supplemental oxygen was administered througha single-prong catheter attached to the naris (Pettyet al., 1973). For one patient requiring assistedventilation 'the bronchoscope was passed directlyinto the trachea through an endotracheal tube us-ing an adaptor that provides a tight seal to main-tain adequate -tidal volume and ventilation(Amikam et al., 1972). Patients with plateletcounts less than 50000/mm3 received platelettransfusions (10 platelet packs in most cases) dur-ing and immediately after the procedure to obtainplatelet counts greater than 100 000/mm3.

After complete examination of the tracheo-bronchial tree, including subsegmental bronchi, ifno endotracheal lesions were seen, the broncho-scope was guided directly or by fluoroscopy intothe segment containing the lesion for biopsy.Under fluoroscopic guidance, the standard biopsyforceps that accompanies the Olympus broncho-scope was passed peripherally in a closed positionuntil mild resistance was met. The position of theforceps was carefully checked with fluoroscopy inboth the lateral and anteroposterior projections.The proximity of the forceps to the visceral pleurawas also noted. When satisfactory placement inthe areas of involvement was achieved, the forcepswas retracted 1 to 2 cm and opened, then ad-vanced under fluoroscopic guidance into the lesionor within 1 to 2 cm of the visceral pleura in diffusedisease or until moderate resistance was en-

countered. The forceps was gently closed andslowly retracted while the bronchoscope was

secured in the same position or slowly advancedto ensure accurate repositioning of the bronchialbrush or forceps. The specimen was withdrawn

R. A. Matthay, W. C. Farmer, and D. Odero

through the bronchoscope's biopsy channel.Multiple biopsy specimens were taken from thesame or different areas of involvement. Unilateralbiopsies were done to prevent the possible occur-rence of bilateral pneumothoraces. Bronchialbrushings (Fennessey, 1966; Fry and Manalo-Estrella, 1970; Repsher et al., 1972; Finley et al.,1974) of the same areas were performed and sentfor cytological and microbiological examination.The biopsy specimens were placed in 10% formalinsolution for histological examination. Additionalspecimens were placed in appropriate culturemedia.

Biopsy was performed in the radiology depart-ment using a single plane fluoroscopy unit with animage intensifying screen. Emergency equipmentwith cardiorespiratory arrest or pneumothoraxmanagement was available in all cases. Chestradiographs were taken just before the procedureand immediately afterwards.

Results

Nineteen patients underwent 25 fibreoptic bron-choscopy procedures. Underlying diseases are listedin Table 1. Risk factors for fibreoptic bronchos-copy are indicated in Table 2. Arterial oxygentension among 16 hypoxaemia patients breathingambient air ranged from 34 to 70 (average 55)mmHg. Platelet counts in seven thrombocytopenicpatients ranged from 2400 to 90 000/mm3 (average51 000/mm3). Five patients with platelet countsless than 50000/mm3 received platelet trans-fusions during and immediately after bronchos-

Table 1 Immunocompromised patients (19) andunderlying diseases

Underlying disease No. ofpatients

hlodgkin's disease 6Other lymphoma 2Acute myelogenous leukaemia 4Other leukaemia IRenal transplant 2Bronchogenic carcinoma IMultiple myeloma IHistiocytic medullary reticulosis ISystemic lupus erythematosus I

Table 2 Risk factors for fibreoptic bronchoscopy inimmunocompromised hosts (25 procedures, 19 patients)

Risk factor No. ofprocedures No. ofpatients

Arterial hypoxaemia 18 16Thrombocytopenia 9 7Abnormal partial

thromboplastin time 4 4Abnormal prothrombin time 3 3



j.com/


ctober 1977. Dow

nloaded from


Diagnostic fibreoptic bronchoscopy in inmmunocompromised patients

copy to obtain platelet counts greater than10o OOO/mm3. An abnormal prothrombin time,greater than three seconds beyond control, wasnoted in three patients, and an abnormal partialthromboplastin time, greater than 45 seconds,was identified in four.

DIAGNOSTIC YIELDThe results of fibreoptic bronchoscopy with brushand forceps biopsy are shown in Table 3. Thediagnostic yield of brush biopsy was 63% and oftransbronchial forceps biopsy 68%. The combineddiagnostic yield for both brush and forceps biopsieswas 84% (21/25 procedures). Bronchoscopy resultswere confirmed subsequently by positive bloodcultures, clearing of pulmonary infiltrates withappropriate therapy, open lung biopsy, ornecropsy.Table 4 indicates diagnoses obtained by fibre-

optic bronchoscopy. Secondary neoplasms were allbronchogenic carcinomas. Two patients, aged 29and 41, had previously received radiation therapyto the thorax and intensive chemotherapy forHodgkin's disease. The second tumours in thesetwo patients developed in the previous radiationfield. The third patient had received radiotherapyto the neck only for non-Hodgkin's lymphoma.A specific infection was diagnosed by fibreoptic

bronchoscopy in 14 patients. Table 5 shows thatbacterial infection was identified most commonlyfollowed by fungal organisms. A patient (RM)with histiocytic medullary reticulosis died with aninvasive pulmonary Aspergillus fumigatus infec-tion in spite of amphotericin B therapy (Fig. 1). A

Table 3 Diagnostic yield of fibreoptic bronchoscopyin immunocompromised hosts with pulmonaryinfiltrates (19 patients)

Type ofpulmonary Specific No. of DiagnosticinfiltrMte diagnosis procedures yield %

Diffuse I1 14 79Focal 10 11 91

Total 21 25 84

Table 4 Diagnoses obtained by fibreopticbronchoscopy in immunocompromised hosts withpulmonary infiltrates (25 procedures)

Diagnosis No. ofprocedures

Specific infection 14Secondary neoplasm 3Lung haemorrhage IRadiation pneumonitis ILipoid pneumonitis 1Mucous plug INo specific diagnosis 4

Table 5 Specific pulmonary infections (14) diagnosedin immunocompromised hosts and outcome

No. of InfectionInfection procedures resolved

Bacteria (Klebsiella 5, Nocardia 2,Pseudomonas 1, Staphylococcus aureus 1) 9 7/9

Fungus (Aspergillusfumigatus 1,Aspergillusflavus 1, Candida albicans 1) 3 2/3

Protozoan (Pneumocystis carinii) I 0/1Virus (Herpes simplex) 1 1/1

second patient (MM) with erythroleukaemia de-veloped an invasive pulmonary A. flavus infectionand recovered after amphotericin B and rifampicintherapy (Fig. 2). Candida albicans was identifiedon transbronchial brush and forceps biopsy speci-mens from a patient with acute myelogenousleukaemia. The pulmonary nodular infiltratescleared after treatment with amphotericin B.Pneumocystis carinii was isolated from brush andforceps biopsy specimens of a patient with stageIVB Hodgkin's disease after corticosteroids andcytotoxic chemotherapy had been discontinued.In spite of pentamidine therapy followed bypyramethamine and sulphadiazine therapy thepatient died two months later with respiratoryfailure. At necropsy there was no evidence ofPneumocystis carinii in the lungs. Herpes simplexwas identified on viral cultures of the lung in apatient receiving cytotoxic drugs for Hodgkin'sdisease (stage IVB), and the patient recoveredspontaneously with clearing of the chest radio-graph.No specific diagnosis was obtained by fibreoptic

bronchoscopy in four patients. Subsequently twoof these patients died, and a necropsy revealedradiation pneumonitis in one. No necropsy wasperformed in the other patient. The pulmonaryinfiltrates cleared spontaneously in the thirdpatient with Hodgkin's disease, and in the fourthcase a diagnosis of pulmonary Nocardia infectionwas made on repeat fibreoptic bronchoscopy.

COMPLICATIONSThe morbidity of bronchoscopy with trans-bronchial forceps and brush biopsy was low.Pneumothorax requiring a chest tube occurred asa result of two procedures. Mild, transient hae-moptysis with an increase in alveolar filling on thechest radiograph was noted in two other patients,and in neither instance was blood transfusionnecessary. Both patients with bleeding hadleukaemia and were thrombocytopenic withplatelet counts of 2400/mm3 and 40 000/mm3respectively. In one the partial thromboplastintime was abnormal; however, in both, the pro-

541



j.com/


ctober 1977. Dow

nloaded from



( i

(I)

Fig. 1 Chest radiograph of a 65-year-old man (RM)with histiocytic medullary reticulosis receivingcytotoxic drugs and prednisone: (a) shows a bilateral,predominantly nodular infiltrate; (b) demonstratesthe position of the biopsy forceps in the left upperlung zone during fibreoptic bronchoscopy. Aspergillusfumigatus was isolated on both transbronchial brushand forceps biopsies. Aspergillus precipitins rose froma titre of 1:8 to 1:32.

542



j.com/


ctober 1977. Dow

nloaded from


Diagnostic fibreoptic bronchoscopy in immunocompromised patientsd A$, . ._S] ....,,..0..;....

..i

*,dS ::.'...g_.

..*.e'

(b)

Fig. 2 Chest radiographs of a 52-year-old woman (MM) witherythroleukaemia receiving cytotoxic chemotherapy. Brush andtransbronchial forceps lung biopsies revealed Aspergillus flavus, andA spergillus precipitins rose from a titre of 1:8 to 1:32: (a) shows diffuse,bilateral infiltrates; (b) after four weeks' therapy with amphotericin Band rifampicin reveals considerable clearing of the pulmonary infiltrates.

543



j.com/


ctober 1977. Dow

nloaded from


544

thrombin time was normal. No fatalities wereattributed to fibreoptic bronchoscopy.

OUTCOME

Of the 19 patients who underwent diagnosticfibreoptic bronchoscopy, eight are still alive sevento 16 months later. Three deaths occurred withinone week of the procedure and eight other patientsdied two to 20 (average 10) weeks after bronchos-copy. Resolution of the pulmonary infiltrates withdischarge from hospital occurred in 10 of 14 pul-monary infections (71%) (Table 5).

Discussion

This study confirms that fibreoptic bronchoscopyusing brush and forceps biopsy can be performedin immunocompromised hosts with an excellentdiagnostic yield and with acceptable safety evenwhen abnormal clotting factors, thrombocyto-penia, and arterial hypoxaemia are present. Theimmunocompromised host frequently presentsspecial problems which may increase the risk oflung biopsy procedures (Greenman et al., 1975;Hanson et al., 1976; Cunningham et al., 1977).Arterial hypoxaemia and thrombocytopenia re-

present significant risk factors encountered inimmunocompromised patients undergoing fibre-optic bronchoscopy (Hanson et al., 1976). Sup-plemental oxygen, found to be useful by others(Albertini et al., 1975; Zavala, 1975) to protectagainst severe hypoxaemia during bronchoscopy,appeared to protect hypoxaemic patients in thisstudy. Platelet transfusions administered duringand immediately after bronchoscopy in severelythrombocytopenic patients appeared to preventsevere pulmonary haemorrhage in our study. Thelow incidence of bleeding (8%) in this series maybe related to the fact that only two of 19 patientshad renal failure. It is well known that patientswith renal failure have a tendency to developbleeding (Kazatchkine et al., 1976), and recentlyCunningham et al. (1977) have emphasised thatLiraemic patients in their series had a high inci-dence of pulmonary haemorrhage after trans-bronchial biopsy (45%).

In addition to thrombocytopenia, Greenmanet al. (1975) have indicated that a prolonged pro-

thrombin time predisposes immunocompromisedhosts to pulmonary haemorrhage during lungbiopsy procedures. In the current study only threepatients had abnormal prothrombin times andnone developed pulmonary haemorrhage. Ourresults suggest that with proper precautions, in-cluding platelet transfusions for thrombocytopeniaand continuous supplemental oxygen, diagnostic


fibreoptic bronchoscopy can be performed safelyin immunocompromised patients.The low incidence of pneumothorax observed

in our series may be due in part to the use offluoroscopy to guide biopsy procedures. Severalinvestigators (Levin et al., 1974; Ellis, 1975;Zavala, 1975; Hanson et al., 1976) have indicatedthat fluoroscopic guidance should be used fortransbronchial lung biopsy procedures with thefib,Aeoptic bronchoscope. A blind transbronchiallung biopsy may result in perforation of thevisceral pleura and an increased incidence ofpneumothorax. In addition, the use of fluoroscopyassists in confirming the location of the biopsyinstrument in focal lesions, increasing the chanceof obtaining a diagnosis.The high diagnostic yield obtained by us in

this study of immunocompromised hosts is com-parable to that achieved by others performingtransbronchial lung biopsy in both immunocom-promised and non immunocompromised patients(Andersen and Fontana, 1972; Levin et al., 1974;Ellis, 1975; Zavala, 1975; Hanson et al., 1976;Cunningham et al., 1977). Cunningham et al.(1977) evaluated fibreoptic bronchoscopy specifi-cally in immunosuppressed patients, obtaining a740 0 diagnostic yield (23 of 31 procedures) withforceps transbronchial biopsy and a 28% yield(9 of 31 procedures) with bronchial brush biopsy.In the present study brush biopsy and trans-bronchial forceps biopsy evaluated individuallyprovided similar diagnostic yields of 63 and 68%respectively. However, the diagnostic yield wasincreased to 84% by combining the two pro-cedures. Therefore, we recommend that bothbrush and forceps biopsies be performed inimmunocomprotnised hosts with pulmonaryinfiltrates who are undergoing fibreopticbronchoscopy.

References

Albertini, R. E., Harrell, J. H.. and Moser, K. M.(1975). Management of arterial hypoxemia inducedby fiberoptic bronchoscopy. Chest, 67, 134-136.

Amikam, B.. Landa, J., West, J., and Sackner, M. A.(1972). Bronchofiberscopic observations of thetracheobronchial tree during intubation. AmericanReview of Respiratory Disease, 105, 747-755.

Andersen, H. A., and Fontana, R. S. (1972). Trans-bronchoscopic lung biopsy for diffuse pulmonarydiseases: technique and results in 450 cases. Chest,62, 125-128.

Cutnningham, J. H.. Zavala, D. C.. Corry, R. J., andKeim, L. W. (1977). Trephine air drill, bronchialbrush and fiberoptic transbronchial lung biopsies inimmunosuppressed patients. A merican Review ofRespiratory Disease, 115, 213-220.



j.com/


ctober 1977. Dow

nloaded from


Diagnostic fibreoptic bronchoscopy in immunocompromised patients

Ellis, J. H., Jr. (1975). Transbronchial lung biopsy viathe fiberoptic bronchoscope: experience with 107consecutive cases and comparison with bronchialbrushing. Chest, 68, 524-532.

Fennessey, J. J. (1966). A technique for the selectivecatheterization of segmental bronchi using arterialcatheters. American Journal of Roentgenology, 96,936-943.

Finley, R., Kieff, E., Thomsen, S., Fennessy, J.,Beem, M., Lerner, S., Morello, J. (1974). Bronchialbrushing in the diagnosis of pulmonary disease inpatients at risk for opportunistic infection.American Review of Respiratory Disease, 109, 379-387.

Fry, W. A., and Manalo-Estrella, P. (1970). Bronchialbrushing. Surgery, Gynecology and Obstetrics, 130,67-71.

Greenman, R. L., Goodall, P. T., and King, D. (1975).Lung biopsy in immunocompromised hosts.A merican Journal of Medicine, 59, 488-496.

Hanson, R. R., Zavala, D. C., Rhodes, M. L., Keim,L. W., and Smith, J. D. (1976). Transbronchialbiopsy via flexible fiberoptic bronchoscope: Resultsin 164 patients. American Review of RespiratoryDisease, 114, 67-72.

Kazatchkine, M., Sultan, Y., Caen, J. P., and Bariety,J. (1976). Bleeding in renal failure: A possible cause.British Medical Journal, 2, 612-615.

Khan, M. A., Whitcomb, M. E., and Snider, G. L.(1976). Flexible fiberoptic bronchoscopy. AmericanJournal of Medicine, 61, 151-155.

Levin, D. C., Wicks, A. B., and Ellis, J. H., Jr. (1974).Transbronchial lung biopsy via the fiberopticbronchoscope. American Review of RespiratoryDisease, 110, 4-12.

Petty, T. L., Nett, L. M., and Lakshminarayan, S.(1973). A single nasal prong for continuous oxygentherapy. Respiratory Therapy, 18, 421.

Repsher, L. H., Schroter, G., and Hammond, W. S.(1972). Diagnosis of Pneumocystis carinii pneu-monitis by means of endobronchial biopsy. NewEngland Journal of Medicine, 287, 340-341.

Ries, K., Levison, M. E., and Kaye, D. (1974). Trans-tracheal aspiration in pulmonary infection. Archivesof Internal Medicine, 133, 453-458.

Smiddy, J. F., Ruth, W. E., Kerby, G. R., Renz, L. E.,and Raucher, C. (1971). Flexible fiberoptic broncho-scope (letter). A nnals of Internal Medicine, 75,971.

Zavala, D. C. (1975). Diagnostic fiberoptic bronchos-copy: techniques and results of biopsy in 600patients. Chest, 68, 12-19.

Requests for reprints to: Dr. R. A. Matthay, Divisionof Pulmonary Medicine, Department of Medicine,Yale University School of Medicine, 333 Cedar Street,New Haven, Connecticut 06510, USA.

545



j.com/


ctober 1977. Dow

nloaded from


diagnostic the infiltrates - bmj

Documents