Amiodarone Pneumonitis

Amiodarone Pneumonitis· Bronchoalveolar Lavage Findings in 15 Patients and Review of the Literature Bruno Coudert, M.D.; Fran~ois Bailly, M.D.; Jean NoellDmbard, M.D.; Francis Andre, M.D.; and Philippe Camus, M.D.

Amiodarone (Am) pneumonitis is currently a common and potentially severe adverse reaction, the accurate diagnosis of which remains difficult to establish. Objectives: To determine the contribution of bronchoalveolar lavage (HAL) in the diagnostic workup of patients suspected of having Am pneumonitis. Methods: Diagnosis of Am pneumonitis was established on the basis of (1) development of recent symptoms and pulmonary opacities while receiving the drug, (2) exclusion of other possible causes, and (3) improvement following cessation of Am and/or steroid therapy. (4) Con6rmatory changes were obtained by histopathologic examination in eight cases. HAL was performed in each patient at the time of initial evaluation. Results: Am pneumonitis was diagnosed in 15 consecutive patients between 1985 and 1991. The disease was associated with signi6cant morbidity and mortality. Six patients died; four died of Am pneumonitis. A neutrophilic HAL was found in nine patients (average PMN = 26.6 percent). A mixed pattern (lymphocytic + neutrophilic) was seen in four patients (average: Ly= 19.9 percent; PMN = 11.9 per-

cent). Two patients had a normal BAL. No patient had a lymphocytic pattern. A low CD4 + ICD8 + ratio was seen in two patients. A literature survey indicated 70 cases of Am pneumonitis with detailed information on BAL. The BAL pattern was mixed in 23 (33 percent), neutrophilic in 18 (26 percent), lymphocytic in 15 (21 percent), and normal in 14 (20 percent). No correlation was found between BAL pattern and prognosis. Also, BAL pattern was related neither to daily or total dose of Am nor to duration of treatment with Am. Conclusion: The cellular pro61e of BAL in Am pneumonitis is highly variable, and no cellular pattern of BAL seems to be predictive of a detrimental outcome or of irreversible 6brosis. Aside from excluding other illnesses, and due to its extreme variability, the contribution of BAL differential in the initial workup of patients suspected of having Am pneumonitis is limited. (Chest 1992; 102:1005-12)

kiodarone (Am) is a very effective antiarrythmic, but its use is complicated by numerous adverse effects. In particular, reports of pulmonary toxicity of Am have appeared consistently since the first report of this untoward effect by Rotmensch et aP in 1980. To date, Am pneumonitis emerges as a relatively common and potentially severe illness. 2 We have developed a computerized database (see below) that indicates that more than 400 cases have been reported to date in the literature worldwide. The diagnosis of Am pneumonitis must be established as reliably as possible, because it usually necessitates reduction or even withdrawal of Am therapy. However, such changes may not be devoid of risks, and several fatalities ascribed to recurrence of ventricular dysrythmia have been observed in patients in whom the drug therapy had to be stopped because of Am pneumonitis. 3 Currently, histopathologic study, although not entirely specific, remains the gold standard to establish the diagnosis of Am pneumonitis. 4

However, open lung biopsy is often too aggressive in these fragile patients, and it has been responsible for several deaths. Transbronchiallung biopsy often provides excellent material, but it may suffer occasionally from the limited size of the sample. The yield from thoracoscopic lung biopsy in Am pneumonitis remains to be established. Therefore, it was hoped that bronchoalveolar lavage (BAL) would contribute to the diagnostic workup of patients with Am pneumonitis and obviate the need for lung biopsy. Although early reports on limited groups of patients mentioned increased lymphocyte counts in BAL of patients with Am pneumonitis,s.7 other reports indicated that BAL can display other patterns as well, including a normal one.f~·l0 Therefore, we wish to report our experience with BAL in 15 patients with well-established Am pneumonitis, and review earlier literature on this topic, in order to delineate the possible contribution of BAL to this condition.

*From the Services de Pneumolowe et Reanimation Respiratoire (Drs. Coudert, Lombard, and Camus), Centre Anticancereux Georges Fran~is Leclerc (Dr. Coudert), Hematolo~e (Dr. Bailly), and Cardiolowe (Dr. Andre), Centre Hospitalier Universitaire, Faculte de Medecine et Universite de Bourgogne, Dijon, France. Manuscript received July 8; revision accepted January 21.

PATIENTS AND METHODS

Am = amiodarone; LSD = least signi6cant difference; Ly = lymphocytes; PMN = polymorphonuclear

Inpatients and outpatients were recruited from the Departments of Pulmonary and Cardiovascular Diseases of the Vniversity Hospital of Dijon, Burgundy. Criteria for Am pneumonitis included the following: long-term (>2 months) treatment with Am, the notion of CHEST I 102 I 4 I OCTOBER, 1992

1005

Table I-Explorations Performed in Patients Suspected of Having Amiodarone (Am) Pneumonitis Bronchoalveolar lavage* Patient No.

Bacterial Culture

MT

Malignant cells

Ventilation/Perfusiont Scan

Echocardiography+

Right Heart Catheterization

Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg

Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg

Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg Neg

NO NO NO NO Low probability Low probability Low probability NO NO Low probability NO NO NO NO NO

NO NO NO NC Stable Stable Normal Stable NO Stable Normal NO Stable Stable Stable

NO NO NO Pw: 10mm Hg NO NO NO NO NO NO NO NO NO NO NO

1 2 3 4 5 6

7 8 9

10 11 12 13 14 15

*Bronchoalveolar lavage was processed as follows: bacterial smear and culture, search for Mycobacterium tuberculosis (MT), Coletsos Jensen cultures, cytologic examination for malignant cells. Neg= negative. tVentilation perfusion scan: low probability = ventilation and perfusion scans showed subsegmental defects in the areas of pulmonary opacities seen on chest roentgenograms. NO = not done. :l:Stable = echocardiography at the time of Am pneumonitis was similar to that performed earlier. In particular, there was no appreciable deterioration of left ventricular function. In patient 4, echocardiography was nonconclusive (NC), and a right heart catheterization (SwanGanz) showed normal resting capillary wedge pressure (Pw = 10 mm Hg). a normal chest roentgenogram prior to institution of Am therapy, development of new pulmonary opacities while receiving Am, absence ofexposure to other pneumotoxicants, reasonable exclusion of other causes for the lung opacities (for the latter criterion to be satisfied, two-dimensional [20] echocardiography, Wfc ventilation! perfusion scans, and bacteriologic sampling of sputum or BAL fluid were performed in several patients [Table 1]), and definite improvement or disappearance of the lung opacities following withdrawal of Am therapy and/or administration of steroids. A lung biopsy was

performed in eight patients (open lung biopsy, n =3; thoracoscopic lung biopsy, n = 1; transbronchiallung biopsy, n = 3; autopsy, n = 1) and examined for changes suggestive of Am pneumonitis. ".11 BAL was performed according to established guidelines 12 •1" by instilling four 5O-ml aliquots of prewarmed 0.9 percent saline solution via the fiberoptic bronchoscope (Olympus Corporation, Tokyo, Japan), wedged in the right middle lobe bronchus. Return from the first aliquot was systematically discarded. 12 Smears were made on glass slides by cytocentrifugation at 200 g

Table 2-CUnical Findings, Radiology, Pulmonary Function Tests, and BAL Pattern in 15 Patients with Amiodarone Pneumonitis· Patient No.1 Age, y

1169 'l/12 3/72 4168

5179

6172

7155 8/57

WSI

1
14167

1&'60

Cardiac

Disease*

Daily Dose,t mg

Angina pectoris SVA+VA SVA VA SVA+VA Angina pectoris NOCM VA SVA NOCM+SVA Angina pectoris SVA SVA SVA+VA Angina pectoris

200 400 400 100 400 100 100 400 200 200 200 200 200 200 200

Duration of Treatment, mo 12 84 8 68 3 36 108 29 36 60 36 60 11 36 39

Total Dose, g

Clinical Symptoms*

52 324 70 137 36

C,D,WL,Cy,R D,WL,A,R C,D,WL,A D,WL D,A,Cy,R

365 186

WL,R C,D,WL,Cy,A,R C,D,WL,R D,WL,Cy,A,R F,Cy,R F,R,A D,WL,F,R D,R,WL C,D,R

600

196 160 200 260 44 156 168

R

Chest Pa°2' PaC02, Roentgenogram* mmHg mmHg Mixed I I Mixed I Mixed A Mixed I Mixed I A Mixed Mixed A

69 30 51 64 60 45

29 32 37 21 30 31 39 33

45

42 39

40 56

55 35

34

55 49

38

27 29 37

PFr*

Histopathology*

Outcome

0 M R

Autopsy TBB

Death Recovery§ Recovery Death Recovery§ Recovery Death Death Recovery Recurrence Recovery Recurrence Recovery Recovery Recovery

N

R

R M R R M

OLB TBB TLB OLB OLB

TBB

BAL Sequelae Pattern* R None None R None None R None R None R R None None

Mixed Mixed Mixed Normal Normal PMN PMN PMN PMN PMN PMN PMN PMN PMN Mixed

*SVA = suprnventricular arrhythmia; VA = ventricular arrhythmia, NOCM = nonobstructive cardiomyopathy; C=cough; D=dyspnea; WL =weight loss; A=fatigue; F =fever; Cy =cyanosis; R=rales; A=alveolar opacities; I =interstitial opacities; Mixed = alveolar-interstitial opacities; PIT =pulmonary function tests; N=nonnal; 0 =obstructive; R=restrictive; M=mixed (obstructive +restrictive); TBB = transbronchial biopsy; OLB =open lung biopsy; TLB = thoracoscopic lung biopsy. tIn France, most patients receive amiodarone 5 days a week. *Pattems were determined according to the normal values published by Laviolette;15 lymphocytic >14 percent, neutrophilic >3 percent, mixed Oymphocytic +neutrophilic), and normal. tAfter initial recovef)~ death occurred from ventricular arrhythmia after withdrnwal of amiodarone therapy.

1006

Amiodarone Pneumonitis (Coudert at aJ)

for 10 min (Cytospin, Shandon Corporation, Sewickley, Pa), and stained with May-Griinwald-Giemsa. The differential cell count was done on 300 cells. Four patterns of BAL were defined according to the normal values published by Laviolette l5 : lymphocytic when the percentage of lymphocytes (Ly) was > 14 percent, neutrophilic when neutrophils (PMN) were >3 percent, and mixed when both cell types were increased. Eosinophils were considered normal if <3 percent. Expression of lymphocyte surface markers and CD4 + , CDB +, and null cells were determin~d in some patients (Ortho Diagnostics Kit, Ortho, Raritan, NJ). In eight patients, crosssectional diameters of alveolar macrophages were determined on 50 cells with a micrometer. Results were compared with those obtained in six controls, who were not taking Am, and were free from respiratory disease. Age and smoking history of controls was similar to those of patients with Am pneumonitis. Statistical comparisons were performed using one-way analysis of variance with Fisher LSD (least significant difference) test, as a post hoc procedure. RESULTS

Clinical Data-Chest Roentgenogram-Pulmonary Function Fifteen patients (all male) with Am pneumonitis had their condition diagnosed between 1985 and 1991. Eight (patients 1,3,4,6, 7,8, 9, and 11) were reported earlier as a group in a letter to the editor in French, 16 and one patient (patient 9) was described in the English literature}7 Six patients (patients 2, 3, 5, 7, 10, and 14) were never smokers, three (patients 11, 12, and 15) had stopped smoking more than one year earlier, while the six remaining patients (patients 1, 4, 6, 8, 9, and 13) were current smokers. Clinical data are summarized in Table 2. Cumulated exposure to Am at the time of diagnosis averaged 197 ± 143 g (X± SD) (range, 36 to 600 g), administered over a period of 42 ± 29 months (range, 3 to 108 months). Age ofcontrol patients averaged 61.3 ± 12.8 years; their smoking

history was similar to that of the patients (one current smoker, two ex-smokers, three nonsmokers).

Bronchoalveolar Lavage The number of cells recovered by BAL averaged 532 ± 271 (range, 180 to 1,(00) cells per microliter as compared with 315 ± 73 in control patients (Table 3). A neutrophilic pattern was seen in nine patients, ie, 60 percent (patients 6 through 14). Mean percentage of PMN was 26.6±25.1 (range, 4 to 71 percent). Percentage of Ly was normal in these nine patients (6.3 ± 3.5 percent [range, 0 to 10 percent]). In no patient with neutrophilia was infection a strong diagnostic consideration on clinical grounds, and also because results of bacteriologic investigations were normal (Table 1). Four patients, ie, 27 percent (patients 1 through 3, and 15), had a mixed BAL pattern, with Ly averaging 19.9±2.2 percent (range, 17 to 23 percent) and PMN averaging 11.9±2.5 percent (range, 8 to 14 percent). Eosinophils were increased in seven patients (patients 1, 2, 8, 9, 10, 13, and 14), who all had concomitantly either a neutrophilic or a mixed pattern of BAL. Eosinophils averaged 10.7±4.3 percent of total (range, 4 to 16 percent). Basophils consistently accounted for less than 1 percent of the total. In two patients, ie, 13 percent (patients 4 and 5), BAL differential was entirely normal. No patient had a purely lymphocytic pattern. There was no apparent relationship between smoking status and BAL pattern, either in patients with Am pneumonitis or in control subjects. Cross-sectional diameters of alveolar macrophages, determined in eight patients (patients 1, 2, 4, and 6

Table 3-Bronchoalveolar lAvage Findings in 15 Patients with Amiodarone Pneumonitis Patient No.

Total Cells/JoLl

Ly* Cells/JoLI (%)+

PMN* Cells/JoLI (%)+

Eo* Cells/JoLI (%)+

Mac* Cells/JoLI (%)+

1 2 3 4

800

160 (20)

400 777 440

(8) (14) (12) (3) (2) (13) (12) (22) (71) (4) (17) (73) (24) 40 (4) 123 (14) 19.5±21.5%

(8) 28 (7) 16 (2) 0 (0) 0 (0) 19 (2) 4 (2) 30 (15) 30 (10) 15 (4) 0 (0) 0 (0) 35 (15) 160 (16) 0 (0) 5.4±5.8%

504 (63) 224 (56)

5

6 7 B 9 10 11 12 13 14 15 X±SD

630

940 180 200 300 380 423 400 235 1000 880

532±271

92 (23)

155 (20) 40 (9) 25 (4) 56 (6) 9 (5) 0 (0) 30 (10) 34 (9) 34 (8) 20 (5) 23 (10) 10 (1) 149 (17) 9.7±6.9%

64

56 93 13 13 122 22 44 213 15 72 292 56

64

513 387 592 686 146 124 27 315 309 84 120 790

(66) (88) (94) (73) (81) (62) (9) (83) (73) (21) (51) (79)

607 (69)

Patternt Mixed Mixed Mixed Normal Normal PMN PMN PMN PMN

PMN

PMN PMN PMN PMN Mixed

64.5 ± 22.6%

*Ly = lymphocytes; PMN = neutrophils; Eo = eosinophils; Mac = macrophages. t Patterns were determined according to the normal values published by Laviolette: 15 lymphocytic (Ly) > 14 percent; neutrophilic (PMN) >3 percent; mixed (lymphocytic + neutrophilic); and normal. +Results are expressed as cells per microliter (percentage in parentheses). CHEST I 102 I 4 I OCTOBER, 1992

1007

Table 4-Bronchoalveolar lavage in Amiodarone Pneumonitis: Data from the Literature·: n =70 Author, Year Acar,l983 Brambilla, 1983 Chebat,l983 Dan, 1983 Quyyumi, 1983 Suarez, 1983 Vergnon,l983 Alcoun,l984 Arlet,l984 Darmanata,l984 Venet,l984

N

Patternt

1 1 1 1 1 1 1 1 1 1 1

PMN Ly Ly Mixed Normal Normal Ly Mixed Mixed Mixed Ly Mixed Ly Ly Mixed Ly Normal Normal Normal PMN Mixed PMN Ly Mixed PMN Mixed Ly PMN Ly Ly PMN Ly Normal Ly PMN PMN PMN Mixed Ly Mixed Normal PMN Mixed

4

Breton, 1985 Caubarrere, 1985 Dimitriou, 1985 Ducolone,l985 Adams, 1986 Lok-wan Liu, 1986

Lopez, 1986 Meurice, 1986 Michel, 1986 Moreau, 1986 Scardi,l986 Bensaid, 1987 Bolognesi,1987 Roegel, 1987 Stein, 1987 Xaubet, 1987 Ode, 1988 Alcoun, 1989 Manicardini, 1989 Lavaud, 1990 Nicolet-Chatelain, 1991

1 2 2 2 1

3 4

8 1 1 2

3

2 1 1 1 2 1 1 1

3

2 1 1

1 1 1 1 1 1 2

*Because of the number, articles referred to in this table are not all quoted in the bibliography list. They can be found using the author's name and year of publication. tPatterns were determined according to the normal values published by Laviolette 15 : lymphocytic (Ly > 14 percent), neutrophilic (PMN >3 percent), mixed (lymphocytic + neutrophilic), and normal.

through 10), averaged 22.9±3.6 JLm and 19.9±3.3 JLm. These figures were not different from those in control subjects: 22.0±2.9 JLm and 19.3±2.5 JLm. BAL was repeated after withdrawal of Am therapy in three patients (patients 2, 10, and 13). These patients were receiving oral steroids at the time of the second BAL (since 8, 4, and 8 weeks, respectively). Initially, patient 2 had a mixed BAL pattern, whereas patients 10 and 13 had a neutrophilic pattern. Total number of 1008

cells and percentage of PMN remained elevated in these three patients, despite substantial clinical and roentgenographic improvement. Data from BAL 2 vs BAL 1 regarding each cell type in these three patients were as follows: Ly = 36 vs 23 percent, 2 vs 9 percent, and 6 vs 10 percent; neutrophils = 8 vs 14 percent, 5 vs 4 percent, and 21 vs 24 percent. Lymphocyte surface markers \vere examined in patients 2 and 13. The CD4+/CD8+ ratio \\'as 0.69 and 0.20 at the time of diagnosis, respectively. In patient 2, lymphocyte typing was performed in the second BAL eight weeks after BAL 1, \\,hile the patient was receiving steroids. The CD4 +/CD8 + ratio had increased to 3.6.

Follow-up -

OutCOl1ze

Am therapy was discontinued in all patients, and steroids were given to all but patient 7. Significant improvement was seen in 14 patients. Patient 8 died in the context of respiratory failure and possible ventricular arrhythmia on day 5. Eight patients (patients 2,3,5, 6, 9, 11, 14, and 15) recovered without residual disease. Six patients remained disabled with a restrictive lung function defect (patients 1, 4, 7, 10, 12, and 13). Three of them (patients 1, 4, and 7) ultimately died of respiratory failure, respectively, 6, 2, and 12 months after diagnosis. In two patients (patients 10 and 12), recurrence of pulmonary opacities was seen in previously unaffected areas, respectively, 12 and 1 month following gradual tapering of steroid therapy. In patient 10, pulmonary opacities remained stable, and mutilating interstitial fibrosis was documented ten months after diagnosis by open lung biopsy. Am was substituted for another antiarrhythmic in eight patients (patients 2 through 5, 8, 10, 12, and 14). Despite this change, patients 2 and 5 died suddenly, respectively, 24 and 4 months later, from documented ventricular arrhythmia. Patient 10 sustained supraventricular arrhythmia and pulmonary edema two months after withdrawal of Am therapy. DISCUSSION

The antiarrhythmic, amiodarone, enjoys popularity among cardiologists worldwide. 3 Unfortunately, the drug induces numerous side effects,18 among is which pulmonary toxicity. 2 Interestingly, pulmonary toxicity can be reproduced in laboratory animals. 19 Despite a huge number of reported cases, no reliable predictive test or diagnostic criteria for Am pneumonitis has yet been identified. 2.2o ,21 While an abnormal pretreatment chest roentgenogram or carbon monoxide diffusion test has been regarded as a possible risk factor, 18,22,23 development of Am pneumonitis in a given patient is largely unpredictable. An increase in lymphocytes has been documented Amiodarone Pneumonitis (Coudert et 81)

in the BAL of patients with such drug-induced lung diseases as methotrexate or gold pneumonitis. 24 Likewise, early reports on Am pneumonitis showed lymphocytosis in BAL,2 and it was hoped that alveolar lavage might be helpful in the diagnostic workup of patients suspected of having Am pneumonitis. However, data from the literature coupled to our own observation do not substantiate this vie~ To clarify the contribution of BAL, we reviewed all previously reported cases of Am pneumonitis incorporating detailed information on BAL, and found 70 such cases in 32 publications (Table 4). Using the thresholds outlined above 15 (see under "Patients and Methods" section), a lymphocytic BAL pattern was present in 15 (21 percent) of these 70, a neutrophilic pattern in 18 (26 percent), a mixed pattern in 23 (33 percent), and a normal pattern in 14 (20 percent). Setting the threshold for lymphocytosis at 7 percent, as suggested by some authors,2.5-27 or 21 percent, as proposed by others28 instead of 14 percent, did not affect this distribution dramatically. The results obtained in our patients (Table 3) do not strictly conform to these figures. Instead, the neutrophilic pattern was the dominant pattern in our series with 60 percent, while no patient demonstrated pure lymphocytosis on BAL. In fact, quite different BAL data can be found in authentic cases of AM pneumonitis in the literature as well. For instance, Akoun et a1 29-31 and Venet et al32 described marked BAL lymphocytosis in their patients, whereas others reported that BAL remains normal. 33 -35 These observations suggest that BAL displays a variable pattern in Am pneumonitis, and that alveolitis, as reflected by shifts in cell types in BAL, mayor may not be present at the time of diagnosis of Am pneumonitis. Several explanations may account for these apparent discrepancies among different authors: (1) indeed, any series of patients with Am pneumonitis has a limited number, and the role of chance in these findings cannot be ruled out; (2) the marked topographic heterogeneity of Am pneumonitis, as seen on roentgenograms or CT scan, may also playa role. If one hypothesizes that most patients are lavaged in the right middle lobe, this may bias the results of BAL, because of the heterogenous spatial distribution of pulmonary involvement; (3) temporal changes in the traffic of inflammatory cells may be another explanation, and alveolar lavage may yield different results, depending on the moment it is performed within the time course of the disease. In this connection, evidence has been given in patients with Am pneumonitis that a neutrophilic pattern can proceed to a lymphocytic pattern within a few weeks,36 and the converse has been seen as well. 30 In keeping with this latter hypothesis, animals fed Am long term demonstrate an initial influx of macrophages in BAL, followed by long-

lasting neutrophilic alveolitis, while lymphocytes increase only after week 6. Later on, both lymphocyte and neutrophil counts remain elevated, and this is followed by the development of histopathologic abnormalities. 19 Of note, individual animals can display increases in cell types of different magnitude. For instance, Wilson et a}37 showed that in a batch of rats given the same dosage of Am, BAL neutrophils can be as low as 4 percent and as high as 32 percent. Incidentally, and in contrast to humans in whom BAL lymphocytosis can be as high as 74 percent,38 rats with Am pneumonitis do not exhibit such a high degree of BAL lymphocytosis; indeed, lymphocytes were less than 7 percent in the study of Wilson et al. 19 (4) Another hypothesis stems from the recent observation of Nicolet-Chatelain and her associates35 that administration of Am can be associated with derangements of BAL cell differential count without clinically or roentgenologically overt toxicity. Whether this is related to subclinical alveolitis or is a precocious marker ofAm pulmonary toxicity currently remains uncertain. In a number of patients with Am pneumonitis, suppressor (CD8 +) lymphocytes predominate in BAL, and the CD4 + ICD8 + ratio is less than unity. Similar results have been reported in pneumonitides induced by other compounds,39 but this finding is far from universal, even in patients with pneumonitis induced by a given drug. For instance, in methotrexate lung, Akoun et al39 found a majority ofCD8+ lymphocytes, while White et al24 described CD4 + alveolitis. Information regarding lymphocyte subsets in BAL from patients with Am pneumonitis is available in nine of the 70 cases referred to above. 30·40-42 The CD4+/CDB+ ratio was <1 in eight, and averaged 0.57±0.25. In accordance with these observations, Akoun et al31 recently reported on 22 patients with Am pneumonitis, 16 of whom had lymphocytosis in their BAL; the CD4 +IC D8 + ratio in these patients averaged 0.56±0.41. Of note, in the study of Akoun et al,31 increased relative proportion of CDB+ lymphocytes was also present in patients without significant lymphocytosis in their BAL. A decreased CD4+/CDB+ ratio was seen herein in the two patients in whom lymphocyte typing was performed. In one of them (patient 2), the CD4 +ICDB + ratio increased following discontinuation of Am therapy, an observation already made by others. 43 •44 While the exact significance of the CDB + alveolitis, as well as the possible contribution of other lymphocyte subsets as natural killer cells in controlling influx and activity of suppressor cells45 remain unknown, we believe CDB+ lymphocytosis should be systematically quantitated in Am pneumonitis to delineate its possible diagnostic contribution. It has been established that the likelihood of developing Am pneumonitis is increased in patients receiving elevated daily dosages of Am,21.46.47 and our dataCHEST I 102 I 4 I OCTOBER. 1992

1009

Table 5-Bronchoolveolar lAvage (BAL) Data in 70 CaBes 0/ Amiodarone Pneumonitis (Literature Review): Relationship o/Daily Dose, Duration o/Treatment, Cumulated Dose, and BAL Pattern· BAL Pattern Normal (n= 14) Lymphocytic (n = 15) Neutrophilic (n = 18) Mixed (n = 25)

Daily Dose, mg (range)

Treatment Duration, wk (range)

Total Dose, g (range)

410.0± 165.0t (200-800) 244.7±98.4 (150-500) 315.6±99.5 (150-471) 200.3±64.5 (100-406)

138.8± 167.9 (16-676) 83.3±52.1 (28-180) 124.4 ± 107.8 (16-384) 149.3 ± 114.8 (36-468)

232.4± 180 (28.6-676) 124.16± 162.6 (28-750) 241. 7 ± 201.5 (16-756) 157.3± 118.0 (36-468)

*BAL patterns were determined according to the normal values published by Laviolette. ls See Table 3 and text for explanations. tData are expressed as mean ± SD.

base on 412 cases indicates that the time required to develop pneumonitis is significantly shorter in patients given large daily doses of Am. Reasoning that daily dose or duration oftreatment may influence the degree ofinHammationlalveolitis in Am pneumonitis, we were interested in correlating cellular BAL profile and daily and cumulated dosages of Am in the 70 patients from the literature (Table 5). Patients who had a normal BAL pattern had received significantly greater daily doses of Am than patients who had an abnormal BAL (p
217. Of these 217, 171 (78.8 percent) stabilized or improved, whereas 45 (26.3 percent) remained disabled. Death had occurred by the time of publication in 48, a 22 percent crude mortality rate. Of these 48 deaths, 22 (45.8 percent) were due to Am pneumonitis, whereas the remainder was attributable to other causes, mainly ventricular arrhythmias. 48 Among the six deaths observed in the present series (40 percent mortality rate), four (67 percent) were due to Am pneumonitis. Five ofthe 70 patients from the literature for whom BAL was mentioned had died by the time of publication. Three of them had >3 percent PMN in their BAL. In the present series, three of the four patients who died of Am pneumonitis had elevated neutrophils in their BAL (patient 1 [PMN = 8 percent]; patient 7 [PMN = 12 percent], patient 8 [PMN = 22 percent]). While it cannot be ruled out that BAL neutrophilia indicates increased severity, or is a predictor of death, definitive conclusions cannot be reached for the time being, given the small number of cases available for revie~ Macrophages from rats with phospholipidosis induced by amphiphilics, a class of compounds to which Am belongs, are larger than those from control animals. 49 Therefore, we examined whether a similar observation could be made in patients with Am pneumonitis. However, the diameter of macrophages was similar in patients and in control subjects. While we cannot rule out that the centrifugation procedure (Cytospin) may have eliminated any difference in the size of macrophages between patients and control subjects,50 we doubt that in practice, evaluation of macrophage dimensions will help differentiate toxic from nontoxic patients. On the other hand, BAL macrophages are known to have a foamy cytoplasm in Am pneumonitis2 ; this aspect was not evaluated herein. Nevertheless, foam cells may also be retrieved in subjects exposed to Am without pneumonitis. 35 While we did not find in the present study that cell differential count brings positive information in the diagnostic workup ofthe patient with Am pneumonitis, this by no means indicates that BAL is useless in this Amiodarone Pneumonitis (Coudert et 81)

condition. First, BAL is quite useful to rule out other diseases, mainly infectious processes, as it enables sampling of alveolar fluid for microbiologic cultures. Second, more refined analyses of BAL in Am pneumonitis, which were not performed herein, have shown derangements of phospholipids, which are collectively increased,51 some (eg, phosphatidylcholine) more than others.51 Although the technique is demanding and may not be applicable in routine, it may prove an interesting adjunct in selected patients. Third, in the BAL cells of animals or humans with Am pulmonary toxicity, the amount of Am and its proximal metabolite desethylamiodarone correlates with the amount of phospholipids. 51 .52 The same relationship between the drug and phospholipids has been demonstrated in lung tissue37 but not in BAL fluid. 35 Furthermore, there is increased toxicity in those rats that take up Am more avidly in their lungs. 37 Therefore, the exact relationships between phospholipids, and drug or metabolite content of BAL cells or lung tissue, the degree and type of alveolitis as reflected by BAL differential count, and pulmonary histopathologic study, require additional studies, before the exact contribution of BAL in the diagnosis of Am pneumonitis is delineated. For the time being, however, we believe that results of a BAL differential count should be used neither to prove nor to refute the diagnosis of Am pneumonitis. ACKNOWLEDGMENTS: The authors wish to thank Drs. M. Perrichon, A. Prud'homme, and E Pfitzenmeyer for permission to study and to report details of patients under their care. Mrs. M. Lallemand and R. Trupin gave secretarial assistance. Grant 32* from I'Universite de Bourgogne is ~tefully acknowledged.

REFERENCES 1 Rotmensch H H, Liron M, Tupilski M, Laniado S. Possible association of pneumonitis with amiodarone therapy. Am Heart J 1980; 100:412-13 2 Martin WJ, Rosenow EC. Amiodarone pulmonary toxicity: recognition and pathogenesis: part I and II. Chest 1988; 93:1:1067-75,11:1242-48 3 Rosenbaum MB, Chiale PA, Halpern MS, Nau GJ, Przbylski J, Levi RJ, et aI. Clinical efficacy of amiodarone as an antiarrhythmic agent. Am J Cardiol 1976; 38:934-44 4 Myers JL, Kennedy JI, Plumb VJ. Amiodarone lung: pathologic findings in clinically toxic patients. Hum Patholl987; 18:349-55 5 Chebat J, Caubarrere I. Pneumopathie grave et amiodarone. Therapie 1983; 38:111-12 6 Brambilla C, Chavez L, Piot M, Rigaud D, Paramelle B. Pneumopathie a ramiodarone. Presse Med 1983; 12:1236 7 Vergnon M, Wiesendanger T, Brune J. L'amiodarone: nouvelle etiologie de pneumopathie interstitielle diffuse? Rev Fr Mal Resp 1983; 11:249-50 8 Acar J, Pernes JM, Grimberg D, Jaillon E Fibrose pulmonaire lors d'un traitement par I'amiodarone. Presse Med 1983; 12:47 9 Suarez LD, Poderoso JJ, Elsner B, Bunster AM, Esteva H, Bellotti M. Subacute pneumopathy during amiodarone therapy. Chest 1983; 83:566-67 10 Quyyumi AA, Ormerod L~ Clarke S~ Evans TR, Ward RL. Pulmonary fibrosis: a serious side effect of amiodarone therapy. Eur Heart J 1983; 4:521-24

11 Breton JL, Depierre A, Garnier G. Etude ultrastructuraIe et spectrometrique d'une pneumopathie induite par I'amiodarone. Ann Med Interne 1985; 136:76-80 12 Reynolds HY. BronchoaIveolar lavage. Am Rev Respir Dis 1987; 135:250-63 13 Crystal RG, Reynolds HY, Kalica AR. BronchoaJveolar lavage: the report of an international conference. Chest 1986; 90:12231 14 Chamberlain D~ Braude AC, Rebuck AS. A critical evaluation of bronchoalveolar lavage: criteria for identifying unsatisfactory specimens. Acta Cytoll987; 31:599-605 15 Laviolette M. Lymphocyte fluctuation in bronchoaIveolar lavage fluid in normal volunteers. Thorax 1985; 40:651-56 16 Perrichon M, Coudert B, Foucher ~ Bailly F, Solary E, Jeannin L, et aI. Valeur diagnostique du lavage broncho-alveolaire au cours dues pneumopathies dues a I'amiodarone. Presse Med 1989; 18:542 17 Camus Ph, Lombard IN, Perrichon M, Piard F, Guerin JC, Bejui-Thivolet F, et al. Bronchiolitis obliterans organising pneumonia in patients taking acebutolol or amiodarone. Thorax 1989; 44:711-15 18 Oates]A, Wood AJJ. Drug therapy: amiodarone. N Engl J Med 1987; 316:455-66 19 Wilson BD, Jaworski AJ, Donner ME, Lippmann ML. Amiodarone induced pulmonary toxicity in the rat. Lung 1989; 167:301-11 20 VrobeI TR, Miller PE, Mostow ND, Rakita L. A general overview of amiodarone toxicity: its prevention, detection, and management. Progr Cardiovasc Dis 1989; 31:393-426 21 Marchlinski FE, Gansler TS, Waxman HL, Josephson ME. Amiodarone pulmonary toxicity. Ann Intern Med 1982; 97:83945

22 Ohar JA, Jackson F, Redd RM, Evans GR, Bedrossian CW Usefulness of serial pulmonary function testing as an indicator of amiodarone toxicity. Am J Cardiol 1989; 64: 1322-26 23 Dusman RE, Stanton MS, Miles WM, Klein LS, Zipes D~ Fineberg NS, et aI. Clinical features of amiodarone induced pulmonary toxicity. Circulation 1990; 82:51-9 24 White DA, Rankin JA, Stover DE, Gellene RA, Gupta S. Methotrexate pneumonitis: bronchoaJveolar lavage findings suggest an immunologic disorder. Am Rev Respir Dis 1989; 139:1821 25 Ettensohn DB, Jankowski MJ, Duncan PG, Lalor PA. Bronchoalveolar lavage in the normal volunteer subject, I: technical aspects and intersubject variability; II: safety and results of repeated BAL and use in the assessment of intrasubject variability. Chest 1988; 94:1:275-80; 11:281-85 26 Low RB, Davis GS, Giancola MS. Biochemical analyses of bronchoalveolar lavage fluids of healthy human volunteers smokers and non smokers. Am Rev Respir Dis 1978; 118:863-75 27 Weinberger SE, Kelman JA, Elson NA, Young RC, Reynolds HY, Fulmer JD, et aI. Bronchoalveolar lavage in interstitial lung disease. Ann Intern Med 1978; 89:459-66 28 Reynolds HY, Fulmer JD, KazmierowskiJA, Roberts WC, Frank MM, Crystal RG. Analysis of cellular and protein content of bronchoalveolar lavage Ouid from patients with idiopathic pul-

monary fibrosis and chronic hypersensitivity pneumonitis. J Clin Invest 1977; 59:165-75 29 Akoun GM, Gauthie....Rahman S, Milleron BJ, PerrotJY, Mayaud CM. Amiodarone induced hypersensitivity pneumonitis. Chest 1984; 85:133-35 30 Akoun G, Milleron BJ, Badaro DM, Mayaud CM, Liote HA. Pleural T-Iymphocyte subsets in amiodarone associated pleuropneumonitis. Chest 1989; 95:506-07 31 Akoun GM, Cadranel JL, Blanchette G, Milleron BJ, Mayaud CM. Bronchoalveolar lavage cell data in amiodarone associated pneumonitis: evaluation in 22 patients. Chest 1991; 99:1177-82 CHEST I 102 I 4 I OCTOBER, 1992

1011

32 Venet A, Caubarrere I, Beman G. Five cases ofimmune mediated amiodarone pneumonitis. Lancet 1984; 1:962-63 33 Xaubet A, Roca J, Rodriguez-Roisin R, Herranz J, Marin A, Lomenia F, et al. Bronchoalveolar lavage cellular analysis and gallium lung scan in the assessment of patients with amiodarone induced pneumonitis. Respiration 1987; 52:272-80 34 Adams PC, Gibson GJ, Morley AR, Wright AJ, Corris PA, Reid DS, et al. Amiodarone pulmonary toxicity: clinical and subclinical features. Q J Med 1986; 59:449-71 35 Nicolet-Chatelain G, Prevost MC, Escamilla R, Migueres J. Amiodarone induced pulmonary toxicity. Chest 1991; 99:363-69 36 Israel-Biet D, Venet A, Caubarrere I, Bonan G, Danel C, Chretien J, et al. Bronchoalveolar lavage in amiodarone pneumonitis: cellular abnormalities and their relevance to pathogenesis. Chest 1987; 91:214-21 37 Wilson BD, Clarckson CE, Lippmann ML. Amiodarone pulmonary in8ammation: correlation with drug dose and lung levels ofdrug, metabolite and phospholipid. Am Rev Respir Dis 1991; 143:1110-14 38 Bensaid J, Gueret P, Blanc P, Doumeix Virot P, Pinaud D, et al. Pneumopathie a l'amiodarone: une entite iatrogime souvent meconnue. Ann Cardiol Angeioll987; 36:307-12 39 Akoun GM, Cadranel JL, Milleron BJ, Flammang d'Ortho Mp, Mayaud CM. Bronchoalveolar lavage cell data in 19 patients with drug-associated pneumonitis (except amiodarone). Chest 1991; 99:98-104 40 Akoun GM, Gauthier-Rahman S. Le poumon de l'amiodarone: donnees immunologiques preliminaires. Presse Med 1983; 12:1824 41 Caubarrere I, Bonan G, Venet A, Chebat J, Uzzan D, Gilbert J, et al. Pneumopathies d'hypersensibilite a l'amiodarone et nephropathies associees. Ann Med Interne 1985; 136:311-15 42 Meurice JC, Debacque I, Boita F, Underner M, Bouquet S, Preud'homme JL, et at. Contribution a l'etude des pneumopa-

n,

X. X·· · XVII

AMSTERDAM

1012

THf

43 44 45 46 47 48

49

50 51 52

thies interstitielles au l'lIlrS du traitement par I'amiodarone. Path Bioi 1986; 34: 1074-1lO Scardi S, Ciani F, Pandullo C, Padrini R. Pivotti F. Alveolite ricorrente da amiodarone. G Ital CardioI1986; 16:177-80 Karpel JP, Ligenza C. Aldrich T. Norin J. Amiodarone induced interstitial lung disease: evidence for sensitized T lymphocytes. Am Rev Respir Dis 1988; 137:S347 Karpel JP, Mitsudo S, Norin AJ. Natural killer cell activity in a rat model of amiodarone-indiced iuterstitiallung disease. Chest 1991; 99:230-34 Dusman RE, Stanton MS. Miles WM. Klein LS. Zipes DP, Fineberg NS. et at. Clinical features of amiodarone induced pulmonary toxicity. Circulation 1990; 82:51-9 Mason J'N, and the amiodarone study group. 1(,xicity of amiodarone. Circulation 1985; 72(suppI3):1Il-272 Lombard IN. Pneumopathies de I'amiodamne: analyse de 412 cas de la litterature dont 206 exploitables et de 1I cas personnels. These de Docteur en Medecine de la Facuhe de Medecine de Dijon. Decembre 1990 Reasor MJ, Koshut RA. Castranova V. Biochemical characteristics of rat alveolar macmphages with chlorphentermine induced phospholipidosis: variation with increasing cell size. Exp Mol Patho11979; 31:297-307 Van Meir F. Planimetry of bmnchoalveolar macmphages: importance of preparation and staining techniques. Anal Quant Cytol Histoll99l; 13:261-68 Reasor MJ, Ogle CL. Walker ER. Kacew S. Amiodarone induced phospholipidosis in rat alveolar macrophages. Am Rev Respir Dis 1988; 137:510-18 Martin WJ, Standing JE. Amiodarone pulmonary toxicity: biochemical evidenl'e for a cellular phospholipidosis in the bronchoalveolar lavage of human subjects. J Pbarmacol Exp Ther 1988; 244:774-79

Plan to Attend ACCP's XVII World Congress on Diseases of the Chest

NE:THEALANDS

Amiodarone Pneumonitis (Coudert et at)