clinical features of amiodarone-induced pulmonary...

51

Clinical Features of Amiodarone-InducedPulmonary Toxicity

Raymond E. Dusman, MD, Marshall S. Stanton, MD, William M. Miles, MD,

Lawrence S. Klein, MD, Douglas P. Zipes, MD,

Naomi S. Fineberg, PhD, and James J. Heger, MD

The incidence and clinical predictors of amiodarone pulmonary toxicity were examined in 573patients treated with amiodarone for recurrent ventricular (456 patients) or supraventricular(117 patients) tachyarrhythmias. Amiodarone pulmonary toxicity was diagnosed in 33 of the573 patients (5.8%), based on symptoms and new chest radiographic abnormalities (32 of 33patients) and supported by abnormal pulmonary biopsy (13 of 14 patients), low pulmonarydiffusion capacity (DLCO) (nine of 13 patients), and/or abnormal gallium lung scan (11 of 16patients). Toxicity occurred between 6 days and 60 months of treatment for a cumulative riskof 9.1%, with the highest incidence occurring during the first 12 months (18 of 33 patients).Older patients developed it more frequently (62.7± 1.7 versus 57.4±0.5 years,p=0.018), with nocases diagnosed in patients who started therapy at less than 40 years of age. Gender, underlyingheart disease, arrhythmia, and pretreatment chest radiographic, spirometric, or lung volumeabnormalities did not predict development of amiodarone pulmonary toxicity, whereaspretreatment DLCO was lower in the group developing it (76.0+5.5% versus 90.4±-1.4%,p=0.01). There was a higher mean daily amiodarone maintenance dose in the pulmonarytoxicity group (517±25 versus 409±6 mg, p<0.001) but no difference in loading dose. Nopatient receiving a mean daily maintenance dose less than 305 mg developed pulmonarytoxicity. Patients who developed toxicity had higher plasma desethylamiodarone (2.34±0.18versus 1.92+±0.04 Jg/ml, p=0.009) but not amiodarone concentrations during maintenancetherapy. Death due to pulmonary toxicity occurred in three of 33 patients (9.1%). In conclusion:1) amiodarone pulmonary toxicity occurred in 5.8% of patients and was more common with a

higher amiodarone maintenance dose, advanced age, lower pretreatment DLCO, and higherplasma desethylamiodarone concentrations; and 2) pretreatment chest radiographic, spiromet-ric, and lung volume abnormalities were not predictive for development of amiodaronepulmonary toxicity. (Circulation 1990;82:51-59)

A miodarone hydrochloride, although generallyeffective as an antiarrhythmic drug,'-3 hasboth complex pharmacokinetic properties

and the potential for a wide range of adverse effects.4Perhaps the most serious side effect of amiodaronehas been pulmonary toxicity. Initially reported in1980,5 amiodarone-associated pulmonary toxicity has

From the Krannert Institute of Cardiology, Department ofMedicine, Indiana University School of Medicine, and the Roud-ebush Veterans Administration Medical Center, Indianapolis,Indiana.

Supported in part by the Herman C. Krannert Fund, by grantsHL-06308, HL-07182, HL-28556, and HL-01729 from theNational Heart, Lung, and Blood Institute of the National Insti-tutes of Health, US Public Health Service, and by the AmericanHeart Association, Indiana Affiliate, Inc.Address for correspondence: Douglas P. Zipes, MD, Krannert

Institute of Cardiology, 1001 W. 10th Street, Indianapolis, IN46202.

Received April 26, 1989; revision accepted February 22, 1990.

been reported to occur in 1-17% of treated cases andto have a potentially fatal course.1'6-13 Risk factors,such as underlying pulmonary disease, have beensuggested in some8J'213 but not all9 reports.The present study was undertaken to determine

the incidence and clinical features associated withamiodarone-induced pulmonary toxicity in a largeseries of patients treated at a single medical center.Also examined were the relations of preexistingpulmonary abnormalities, drug doses, and plasmadrug concentrations to the development of amio-darone pulmonary toxicity.

MethodsPatientsThe study population comprised 573 patients who

were treated at the Indiana University Medical Cen-ter and initiated amiodarone therapy between April

by guest on April 26, 2018

http://circ.ahajournals.org/D

ownloaded from

http://circ.ahajournals.org/

52 Circulation Vol 82, No 1, July 1990

TABLE 1. Clinical Characteristics of Patients Treated WithAmiodarone

Patients (n) 573Age (yrs)mean±SD 57.7±+12.6range 5-88

Sex (male/female) 451/122Pretreatment arrhythmiaVT or VF 456SVT 117

Underlying heart diseaseischemic 346cardiomyopathy 94valvular 47primary electrical 62Wolff-Parkinson-White 14congenital 9fibroelastosis 1

VT, nonsustained or sustained ventricular tachycardia; VF,ventricular fibrillation; SVT, supraventricular tachycardia.

1977 and March 1986. All patients receiving amio-darone for more than 3 days during this time spanwere included. The clinical characteristics of thesepatients are outlined in Table 1.

Amiodarone Administration ProtocolAmiodarone therapy was administered in loading

and maintenance dose phases. Initially in our experi-ence with amiodarone, the loading phase was 800mg/day administered for up to 6-8 weeks. Based onclinical and pharmacokinetic data,'4 this loading dosewas later modified to 1,600 mg/day for 1 week, fol-lowed by 800 mg daily for 1 month. During mainte-nance therapy, daily amiodarone doses ranged from50 to 800 mg/day, determined by control of the clinicalarrhythmia and the occurrence of drug-related sideeffects. Plasma concentrations of amiodarone and itsmetabolite, desethylamiodarone, were measured byhigh performance liquid chromatography's every 6-12months during follow-up.

Pulmonary Follow-up ProtocolA pulmonary follow-up protocol was established in

1980, once the potential for pulmonary toxicity wasidentified. Chest radiographs and, when feasible,measurements of spirometry, lung volume, and dif-fusion capacity were obtained pretreatment. Thesepulmonary function measurements included 1-second forced expiratory volume (FEV1), forcedvital capacity (FVC), mid-forced expiratory flow(FEF25-75), the ratio of FEV, to FVC (FEV,IFVC),total lung capacity (TLC) as measured by nitrogenwashout methods, and diffusion capacity (DLCO)as measured by single breath carbon monoxidedilution. All data were recorded as absolute valuesand calculated percent predicted using conven-tional corrections for body surface area, age, andsex of the patient.16 From April 1977 throughMarch 1986, all patients initiated therapy at ourinstitution and were seen in follow-up every 3months for the first 2 years of drug therapy and

every 6 months thereafter in the InvestigationalDrug Clinic by a cardiologist experienced withamiodarone therapy. Chest radiographs wererepeated at 3-6-month intervals in more than onehalf of the patients during the follow-up period, orsooner if symptoms dictated. Pulmonary functiontests were repeated if new symptoms or chestradiographic abnormalities suggested pulmonarytoxicity.

After the release of amiodarone for noninvestiga-tional use in March 1986, some patients opted forcontinued follow-up with their own referral physician.These patients, their physicians, or both were contactedby phone to ascertain pertinent follow-up information.

Other InvestigationsAdditional testing was performed to establish or

confirm the diagnosis of amiodarone pulmonary tox-icity in most cases. Gallium-67 citrate, 2-5 mCi, wasadministered intravenously, and scintigraphic imageswere obtained 48 hours after injection of the radio-isotope. Increased lung uptake was interpreted asconsistent with an inflammatory process.'7 Light andelectron microscopic examination of lung tissueobtained by bronchoscopic or open lung biopsy wasconducted for the following pathologic changes6"13: 1)hyperplasia of type II pneumocytes, 2) alveolar septalthickening, 3) intra-alveolar accumulation of foamymacrophages and pneumocytes, and 4) fibrosis.Bronchoalveolar lavage, more recently described inpatients with suspected amiodarone lung toXicity,18"19was done inconsistently in the present study. In allcases, alternative etiologies such as congestive heartfailure, infection, or malignancy were excluded byright heart catheterization, trials of diuretic therapy,sputum examination, or more detailed radiographicprocedures. Amiodarone therapy was discontinuedwhen the diagnosis of toxicity was made.

Diagnostic CiteriaAmiodarone-induced pulmonary toxicity was a

clinical diagnosis that required two or more of thefollowing criteria: 1) new onset of pulmonary symp-toms such as dyspnea, cough, or pleuritic chest pain;2) new chest radiographic abnormality such as aninterstitial or alveolar infiltrate; 3) a decrease in theDLCO of 20% from the pretreatment value, or ifnone was available, a value less than 80% of pre-dicted; 4) abnormal lung uptake with gallium-67radioisotope; and 5) characteristic histologic changesof lung tissue obtained by bronchoscopic or openlung biopsy.6"3 Exclusion of alternative etiologiessuch as congestive heart failure, infection, or malig-nancy was done in all patients.

StatisticsComparison between patients with and without

amiodarone pulmonary toxicity were made using a ttest for continuous variables. Discrete variables wereanalyzed by a X2 test. Fisher's exact test was used for2x2 tables. Cumulative survival without pulmonary



ownloaded from


Dusman et al Amiodarone-Induced Pulmonary Toxicity 53

10)(0(2 (1 ),,28)

_-60)_-25)

/299)

^341)/37)

/410

/462) OVERALL INCIDENCE: 331573, 6.8S

FIGURE 1. Life table analysis of amiodarone pul-monary toxicity (APT) and duration of therapy in573 patients treated with amiodarone. Numbers inparentheses denote number of patients remainingavailable for analysis.

6 12 18 24 30 36 42 48 54 60 66 72

DURATION OF THERAPY (months)

toxicity was calculated by the Kaplan-Meier method.Ap value of less than 0.05 was considered significant.The duration of the therapy was computed as thelength of time from the start of amiodarone therapyuntil either the end of the study follow-up, death,discontinuation of therapy, loss of follow-up (lastdate for which reliable information was available), orthe development of pulmonary toxicity. A survivalcurve for survival without pulmonary toxicity wascalculated using the Kaplan-Meier product momenttechnique. Failure was development of pulmonarytoxicity. All other endpoints as described above wereconsidered censored data. Figure 1 is the inverse ofthe survival curve or the cumulative incidence ofamiodarone-associated pulmonary toxicity.

ResultsIncidence and Time Course for ToxicityOf the 573 patients who received amiodarone, 33

developed pulmonary toxicity, resulting in an overallincidence of 5.8%. The cumulative incidence ofamiodarone pulmonary toxicity was 9.1% at 121months of therapy using survival analysis (Figure 1).The highest risk for the development of toxicity,corresponding to the steepest portion of the curve,occurred during the first 12 months of therapy. After12 months, the cumulative incidence of toxicity con-tinued to rise steadily, reaching 9.1% at 60 months oftherapy. No cases of amiodarone-induced pulmonarytoxicity developed after 60 months of therapy. Thelongest treated patient was followed for 121 months.The shortest dosing interval in which pulmonarytoxicity occurred was 6 days; this patient had amio-darone discontinued 6 days after beginning therapy(800 mg/day) because of recurrent arrhythmia. Uponpreoperative evaluation 1 month later for left ven-

tricular aneurysmectomy and endocardial resection,he was found to have interstitial infiltrates on chestradiograph, mild dyspnea, and abnormal gallium lunguptake; histologic examination of lung tissueobtained by bronchoscopy and confirmed by openlung biopsy was consistent with pulmonary toxicity.

Before the institution of a specific pulmonaryfollow-up protocol (1977-80), 20 patients discontin-ued amiodarone therapy. None of these patients hadsymptoms or signs suggestive of pulmonary toxicity,and only one patient discontinued the drug forunknown reasons.Three hundred forty-one of the 573 patients were

followed at Indiana University. Ninety-three percentof these subjects had had chest radiographs within 6months of the conclusion of the study. We have nodata among the 232 patients followed outside ofIndiana University regarding completeness of chestradiography follow-up, after they were no longerfollowed by us, beginning March 1986.

Presenting SymptomsThe presenting symptoms of amiodarone-induced

pulmonary toxicity were known in 28 patients. Themajority (71%) had dyspnea as the initial symptom.Other symptoms were cough (25%), fever (21%),nausea (7%), weakness or fatigue (7%), weight loss(4%), and pleuritic chest pain (4%).

Diagnostic TestsThe diagnostic tests used to establish and confirm

the diagnosis of amiodarone pulmonary toxicity areoutlined in Table 2. The diagnosis of pulmonarytoxicity was made solely on the basis of presentingsymptoms and chest radiographic abnormalities in sixof 33 patients. In 25 patients, additional abnormali-ties in either DLCO, gallium scan, histology, or a

TABLE 2. Diagnostic Tests Performed in Patients With Amio-darone Pulmonary Toxicity

Result

Test n Normal Abnormal

Chest radiograph 33 1 32DLCO 13 4 9Gallium scan 16 5 11Histology 14 1* 13

DLCO, diffusion capacity of lung using single breath carbonmonoxide test.

*Inadequate specimen.

10-

9.

8-

7-

6-

5-

4-

3.

2-

I.

ILA.

0W0z

0

z

W

-i

m

0 1 1 5.10,



ownloaded from



combination of these were present to confirm thediagnosis, whereas two patients had pulmonary tox-icity diagnosed only at postmortem examination.New chest radiographic abnormalities occurred in 32of the 33 patients and were most often diffuseinterstitial pulmonary infiltrates, occasionally alveo-lar in distribution, and often involved the upperlobes. Two patients had pleural effusions. Thepatient without a radiographic pulmonary abnormal-ity presented with increased exertional dyspnea after12 months of amiodarone treatment. A 42% declinein DLCO and diffuse gallium lung uptake suggestedthe diagnosis of pulmonary toxicity. DLCO determi-nations were available in 13 patients who presentedwith toxicity and were abnormally low in nine.Restrictive lung function (decreased FVC and TLCand increased FEV1/FVC) occurred in all 15 patientswho had these parameters measured.

Gallium lung scans revealed abnormal radioiso-tope uptake in 11 of 16 patients and generallycorrelated with the abnormalities noted on chestradiograph. Of the five patients with normal galliumscans, two were imaged 2-4 weeks after amiodaronewas discontinued, one had histologic evidence forpulmonary fibrosis at postmortem examination with-out the presence of inflammatory cells, and one hadthe gallium scan done 2 days after amiodaronediscontinuation, with biopsy confirming toxicity. Theremaining patient had superimposed congestiveheart failure, the degree of which was not thoughtsevere enough to solely account for the clinicalpicture, although lung biopsy was not performed.

Fourteen patients had histologic examination oflung tissue, which was obtained by open lung biopsyin three patients, transbronchial biopsy in sevenpatients, and combined open lung and transbronchialbiopsy in one patient. Three patients had histologicexamination postmortem. Thirteen biopsies wereabnormal by the criteria outlined previously. Onepatient had an inadequate specimen obtained bytransbronchial biopsy.

Associated Clinical CharacteristicsClinical and demographic characteristics were

examined for possible association with the occur-rence of amiodarone pulmonary toxicity. Patientswith and without amiodarone pulmonary toxicitywere no different in gender, underlying heart disease,presenting arrhythmia, or concomitant antiarrhyth-mic medications. Patients who developed amio-darone pulmonary toxicity were older at initiation oftherapy (62.7+1.7 years) compared with those with-out pulmonary toxicity (57.4+0.5 years,p=0.018). Nocases of pulmonary toxicity were diagnosed inpatients who started therapy less than or equal to 40years of age (N=57), whereas the incidence of toxic-ity was fairly uniform for each decade of life inpatients beyond 50 years of age (Figure 2).

a.

O.LU0

0

z

AGE (years)

FIGURE 2. Incidence of amiodarone pulmonary toxicity(APT) based on age at which therapy was started.

Pretreatment and Follow-up RadiographsAnalysis of pretreatment chest radiographs re-

vealed no significant difference between patients whodid and did not develop pulmonary toxicity. Duringfollow-up, new chest radiographic abnormalitiesdeveloped in 98 of 571 patients (17%). These newabnormalities were attributed to amiodarone pulmo-nary toxicity in 32 patients, whereas the remainderwere secondary to congestive heart failure (34patients), atelectasis (seven patients), chronicobstructive pulmonary disease (six patients), malig-nancy (four patients), and infection (three patients).Twelve remaining patients had chest radiographicabnormalities of unknown etiology that resolvedspontaneously despite continuation of amiodarone.

Pretreatment Pulmonary Function TestsPretreatment pulmonary function tests of the

groups with and without pulmonary toxicity werecompared. There was no significant difference inmeasured lung volumes or spirometry between thesetwo groups (Table 3). Figure 3 compares the pre-

TABLE 3. Pretreatment Spirometric and Lung Volume Measure-ments in Patients Treated With Amiodarone

Test APT- APT+ p

FVC (1) 3.42±0.04 3.42±0.23 0.99(389) (26)

FEV1/FVC (%) 76.1±0.5 73.1±01.8 0.15

(389) (26)FEF25 -7 (1/sec) 2.19±0.06 1.77±0.25 0.10

(389) (26)TLC (1) 5.71±0.10 5.34±0.32 0.25

(316) (16)Values are mean±SEM; the number in parentheses denotes the

number of patients available for analysis.APT-, amiodarone pulmonary toxicity absent; APT+, amio-

darone pulmonary toxicity present; FVC, forced vital capacity;FEV1/FVC, ratio of 1-second forced expiratory volume to forcedvital capacity; FEF25 75, mid-forced expiratory flow; TLC, totallung capacity.



ownloaded from



30 []APT-| 100 p.01

APT+ 90

CB 25 p.05208s0

E 3t520E0

I0-

0 15- 0

Oramn 050adteecn f rdce

W~~~~~~~~~0

22804mlmnm Hgp002. Weepese

Z 100 30-

20-

I0-

0- 332 200 332

FIGURE 3. Comparison of pretreatment diff-usion capacity

(DLCO) and percent ofpredicted DLCO between the groupswith amiodarone pulmonary toxicity (APT +; strippled bars)and without (APT unfilled bars). Numbers within each bar

represent number ofpatients available for analysis.

treatment DLCO and the percent of predictedDLCO between the two groups. The mean pretreat-

ment DLCO tended to be lower in patients who

subsequently developed toxicity when compared with

those who did not develop toxicity (19.8±e1.7 versus22.8± s0.4ml/min/mm Hg, p=C0.052). When expressedas a percent of the predicted DLCO, however, those

patients who developed toxicity had a significantlylower pretreatment value compared with those with-

out toxicity (76.0±+5.5% versus 90.4±f-1.4%, p=0.01)(Figure 3). The frequency distribution for the per-

cent of predicted DLCO is illustrated in Figure 4.

Thirteen of 20 patients who subsequently developed

pulmonary toxicity had pretreatment percent of pre-

dicted DLCO less than 80%, but the other seven

patients had normal pretreatment values. The sensi-

tivity and specificity of a pretreatment DLCO abnor-

mality for developing pulmonary toxicity were 65%

and 68% respectively. Correspondingly, the positivepredictive value of an abnormal percent of predictedDLCO for developing toxicity was quite low at 11%,

50-

APT-

APT+

z

0

-J 30-

0.

0i 20

although a normal pretreatment value had a negativepredictive value of 97%.

Plasma Drug ConcentrationsPlasma concentrations of amiodarone and its

metabolite, desethylamiodarone, were measuredduring the maintenance period of amiodarone ther-apy (Figure 5). There was no difference in meanplasma amiodarone concentration between thegroups with and without toxicity (3.04+±0.38 versus2.83±0.06 ,xg/ml, p=0.46). However, the meanplasma desethylamiodarone concentrations were sig-nificantly higher in the group with pulmonary toxicitycompared with the group without toxicity (2.34±+0.18versus 1.92±0.04 4g/ml,p=0.009). There was a meanof 3.2 (range, 1-9) determinations of plasma drugand metabolite concentrations done per patient inthe group with amiodarone pulmonary toxicity and amean of 3.9 (range, 1-12) in the group without.

Amiodarone Drug Dose AnalysisAmiodarone dose data were calculated by two meth-

ods. The initial analysis defined the loading period asthe number of days amiodarone was administeredbefore a decrease to a stable dose. There was nosignificant difference between the group with and thegroup without amiodarone pulmonary toxicity withrespect to the actual length of the loading period(mean, 35±12; range, 6-56 days versus mean, 36+16;range, 3-83 days,p =0.47) and the mean daily loadingdose (838±40 versus 848+11 mg, p=0.79). During themaintenance period, however, the patients who devel-oped toxicity received a significantly higher mean dailydose of amiodarone when compared with the groupwithout toxicity (515+25 versus 409±6 mg,p<0.001).The second analysis standardized the loading phase

of amiodarone administration to be the first 30 days ofdrug treatment. During the loading period, there wasno significant difference between the group with andthe group without development of amiodarone pulmo-nary toxicity in the mean daily amiodarone dose(816+35 versus 832±10 mg, p=0.70). During the

FIGURE 4. Frequency distribution ofpretreatmentpercent ofpredicted diffusion capacity (DLCO) anddevelopment of amiodarone pulmonary toxicity(APT +; stippled bars) versus no toxicity (APT -;unfilled bars). Numbers in each bar represent num-ber ofpatients in each group.

% OF PREDICTED DLCO



ownloaded from



3.00,

2.50-

, 2.00

0

.50-

K 1.00-z

0.50-

0-

T3.00-

C@ 2.50-

0o 2.00-

1 1.50-

-OKI 1.00

z' Q50w

353 o0

FIGURE 5. Analysis of plasma concentrations of amio-darone and desethylamiodarone obtained during maintenanceperiod of therapy between groups with amiodarone pulmonarytoxicity (APT +; stippled bars) and without (APT -; unfilledbars). Numbers in each bar represent number of patientsavailable for analysis.

maintenance period, however, the patients who devel-oped toxicity received a significantly higher mean dailydose of amiodarone when compared with the groupwithout toxicity (517+25 versus 409±6 mg, p<0.001)(Figure 6). No patient who received an average dailyamiodarone dose of less than 305 mg during mainte-nance therapy developed pulmonary toxicity. There

was no relation between age and maintenance dose;younger patients did not receive a lower drug dose.The two initial loading dose strategies (1,600 ver-

sus 800 mg) were also analyzed for a possible asso-ciation with the development of amiodarone pulmo-nary toxicity. The proportion of patients starting drugtreatment with either dose was the same for bothgroups with and without toxicity (p=0.97).

Follow-up of Patients with Pulmonary ToxicityThree of the 33 patients with toxicity (9.1%) died

as a direct result of this complication. The timecourse from diagnosis to death ranged from 4 to 30days. Eight additional patients with amiodarone pul-monary toxicity died during follow-up; five deathswere attributed to congestive heart failure and one

death each to sudden cardiac arrest, alcoholic liverdisease, and metastatic breast cancer. Ten patientswere treated with glucocorticoids, of whom one diedof pulmonary toxicity. Two deaths occurred in theremaining 23 patients not treated with steroids. Twopatients diagnosed with amiodarone pulmonary tox-icity had a subsequent rechallenge with the drug.One patient developed progressive pulmonary infil-trates associated with diffuse gallium uptake in thelungs, whereas the second patient did not developsymptoms or signs suggestive of pulmonary toxicitydespite similar drug dosages for both trials.

FIGURE 6. Analysis of mean daily amiodarone dose duringa loadingperiod of30 days and during the maintenanceperiodbetween groups with amiodarone pulmonary toxicity (ATP +;stippled bars) and without (APT -; open bars) groups.

Number in each bar represents number of patients availablefor analysis.

There were 197 total deaths in this patient popu-lation during the follow-up period, of which threewere attributed to pulmonary toxicity as outlinedabove. One hundred forty-nine were due to a cardiacetiology, and 24 were noncardiovascular in origin.Twenty-one patients had unknown causes of deathalthough symptoms suggesting pulmonary toxicitywere not reported antemortem.

DiscussionIn this large series of patients treated with amio-

darone, 33 of 573 patients presented with overtclinical evidence of pulmonary toxicity. Thesepatients usually were symptomatic, had new radio-graphic abnormalities, and had additional diagnosticstudies to help confirm the diagnosis. The diagnosiswas made by considering a constellation of clinicaland laboratory findings, as no single test is pathogno-monic for the disorder.

Incidence and Diagnosis of ToxicityThe reported incidence of amiodarone pulmonary

toxicity in the literature depends on the diagnosticcriteria used. Our incidence of 5.8% is similar to thatpreviously reported6,8'13 and was based on theappearance of new symptomatology, most frequentlydyspnea and cough, and confirmed with additionaldiagnostic testing, including radiographic andnuclear imaging, pulmonary function testing, andhistologic examination. Pulmonary toxicity from amio-darone shares features with other drug-inducedpneumonitis,1920 including the presence of activatedinflammatory cells in the lungs that may be detectedby the accumulation of gallium.17,2' Microscopicpathologic changes noted with lung biopsy likewise

[I0 APT

ESAPT +

900- p =.70

TL. T

APT -

| APT +

p<.00 1

800-

700-

600-

500

400

300

200-

2

wCD0W(00a

0

0

z

w

-c

z

Ul

100'

0-



ownloaded from



resembled other toxic drug effects in the lung andwere not pathognomonic for amiodarone toxicity,6,13but their presence was helpful in supporting thediagnosis. Smith et a110 and Haffajee et al1l' have bothreported a low incidence (1%) of pulmonary toxicityalthough the diagnosis in these studies relied mostlyon radiographic abnormalities without histologic sup-port. This difference in diagnostic criteria and alsorelatively short follow-up (mean, 24 and 9 months,respectively) may be responsible in part for the lowincidence of amiodarone pulmonary toxicity in thesestudies. A higher incidence (17%) of toxicity wasreported by Magro et al,9 who used similar clinicaldiagnostic criteria to those used in the present studybut emphasized symptoms and radiographic abnor-malities without the aid of DLCO abnormalities.Differences in the patient populations studied andcriteria used to make the diagnosis of amiodaronepulmonary toxicity may account for the observeddiscrepancies between our findings and those ofMagro et al.9 Bronchoalveolar lavage, although notapplied systematically in the present study, has beenused to describe the cytologic changes that occurduring amiodarone therapy, but its role in drug-induced toxicity remains to be defined.

Demographic Factors Predisposing to ToxicityWhen demographic parameters were analyzed, a

higher occurrence of pneumonitis was found in theolder patients. There were no cases of amiodaronepulmonary toxicity in those patients who startedtherapy at an age less than 40 years. This latterfinding most likely explains the significant age differ-ence found between our two groups with and withouttoxicity. Other clinical characteristics such as gender,underlying heart disease, and presenting arrhythmiawere not identified as risk factors for development oftoxicity. The highest risk for developing toxicityoccurred during the first 12 months of treatment, andno new cases were diagnosed after 60 months oftherapy. This peak in incidence early in treatmenthas also been observed by others.9

Radiographic AbnornalitiesPretreatment chest radiographic abnormalities did

not identify a subset of patients at risk for developingpulmonary toxicity. New radiographic abnormalitieswere observed in 98 patients, but on further investi-gation, 67% of these were attributable to etiologiesother than amiodarone toxicity. We currently recom-mend surveillance chest radiographs at 3-6-monthintervals during maintenance amiodarone therapy,especially during the first 12 months when the riskappears to be highest. The use of routine chestradiographs during maintenance therapy comple-ments the occurrence of new symptomatology, as thelatter can be subtle and nonspecific, especially inpatients with coexisting congestive heart failure orchronic obstructive lung disease.A weakness of this analysis is our uncertainty

regarding the extent to which chest radiography

follow-up was obtained in patients not followeddirectly by us after March 1986. Though we believethat chest radiographic films were obtained at leastevery 6 months in most of those 232 patients, if theywere not, then some cases of pulmonary toxicitycould have been missed.

Pulmonary Function Testing IncludingDiffusion Capacity

Pretreatment spirometric or lung volume measure-ments were not helpful in identifying a high-riskgroup for developing pulmonary toxicity. An abnor-mally low pretreatment percent of predicted DLCO(i.e., less than 80%) did allow identification ofpatients who were at increased risk for developingtoxicity. The wide range of DLCO values in bothgroups, with and without toxicity, did not allow for aspecific discriminating pretreatment value that waspredictive for the subsequent development of toxic-ity. When the data were analyzed using less than 80%of the predicted DLCO as the criterion for pretreat-ment abnormality, the sensitivity and specificity wereonly 65% and 68%, respectively, and the positivepredictive value 11%. The negative predictive valueof a normal pretreatment percent of predictedDLCO was 97%, however, and identified a group ofpatients at low risk for developing pulmonary toxic-ity. With the onset of pulmonary toxicity, a restrictivephysiologic pattern was seen with a decrease in FVCand TLC and an increase in the FEV1IFVC ratio.The identification of a high-risk subgroup based on

an abnormally low pretreatment DLCO is consistentwith the findings of Kudenchuk et al,12 who found arelative risk of 9.8 for development of pulmonarytoxicity in patients with an abnormal baseline chestradiograph film and DLCO of less than 80% ofpredicted. This contrasts with Magro et al,9 whofound that the DLCO singly or in combination withradiographic abnormalities did not predict develop-ment of pulmonary toxicity in their patients.Based on our present data, it would appear that

amiodarone should not be withheld solely on thebasis of pre-existing lung disease as judged by eitherradiographic or pulmonary function criteria. Withmediocre sensitivity and specificity and poor positivepredictive value, pretreatment DLCO abnormalitiesshould not singly negate amiodarone therapy butrather should identify patients who should be fol-lowed closely for developing toxicity. As alreadyemphasized by Horowitz,22 the effect of pulmonarytoxicity can be devastating, and patients with pre-existing lung disease and decreased reserve tolerateits development poorly.

Amiodarone Dose and Relation to ToxicityThe development of toxicity may be related to

amiodarone dose, as there was a higher average dailydose administered to those patients who subse-quently developed toxicity during the maintenancetherapy. This relation of dose to the development oftoxicity has been found by others.69 Our change to



ownloaded from



higher initial loading doses (1,600 mg/day) was notassociated with an increased risk of toxicity. A dose-dependent effect of the drug on cytotoxicity has beendemonstrated in vitro.23 Total cumulative dose wasoriginally calculated in the first 498 patients of thisseries to correlate this variable with the developmentof pulmonary toxicity. The group of patients devel-oping toxicity had lower cumulative doses adminis-tered because these patients correspondingly had ashorter duration of therapy because of their censor-ing from the study at the time of diagnosis ofpulmonary toxicity. Therefore, our approach of ana-lyzing average daily doses both in the loading andmaintenance phase of amiodarone therapy was usedto circumvent this problem. Although no cases ofamiodarone-induced pulmonary toxicity occurred inthose patients who received less than 305 mg/dayduring maintenance therapy in our patient popula-tion, toxicity has been reported during low-dosemaintenance therapy.24 In the present study, becausethere was no relation between age and drug dose,amiodarone dose was not responsible for thedecreased risk seen in the younger patients.

Plasma Drug Concentrations and Relation to ToxicityMeasurements of plasma drug concentrations dur-

ing therapy revealed significantly higher mean des-ethylamiodarone but not amiodarone concentrationsin the group that developed toxicity. A relationbetween higher amiodarone concentrations anddevelopment of drug side effects such as hepaticdysfunction and neuromuscular and ocular effectshas been shown previously.2526 but pulmonary toxic-ity was also shown to occur at lower concentrations.26It is presently unclear whether desethylamiodaroneplays a role in the development of pulmonary toxicity,other than that plasma concentrations tend to corre-late with amiodarone doses administered duringmaintenance therapy.14

Clinical ImplicationsAmiodarone-induced pulmonary toxicity remains a

significant complication of therapy with this antiar-rhythmic agent and can result in death. Risk may beminimized by administering the lowest effective dailydose of amiodarone during maintenance therapy.The highest risk appears to occur in patients withabnormally low pretreatment pulmonary DLCO val-ues, and this test may identify a higher risk patientfor whom closer surveillance may be warranted. Ahigh index of suspicion is often necessary in estab-lishing the diagnosis of pulmonary toxicity, and mostcases are reversible if detected early.

References1. Heger JJ, Prystowsky EN, Jackman WM, Naccarelli GV,

Warfel KA, Rinkenberger RL, Zipes DP: Amiodarone: Clin-ical efficacy and electrophysiology during long-term therapyfor recurrent ventricular tachycardia or ventricular fibrillation.NEnglJMed 1981;305:539-545

2. Heger JJ, Prystowsky EN. Zipes DP: Clinical efficacy ofamiodarone in treatment of recurrent ventricular tachycardiaand ventricular fibrillation. Am Heart J 1983;106:887-894

3. Wellens HJJ, Brugada P, Abdollah H: Effect of amiodarone inparoxysmal supraventricular tachycardia with or withoutWolff-Parkinson-White syndrome. Am Heart J 1983;106:876-880

4. Zipes DP, Prystowsky EN, Heger JJ: Amiodarone: Electro-physiologic actions, pharmacokinetics and clinical effects. JAm Coll Cardiol 1984;3:1059-1071

5. Rotmensch HH, Liron M, Tupilski M, Laniado S: Possibleassociation of pneumonitis with amiodarone therapy (letter).Am Heart J 1980;100:412-413

6. Marchlinski FE, Gansler TS, Waxman HL, Josephson ME:Amiodarone pulmonary toxicity. Ann Intern Med 1982;97:839-845

7. Mason JW: Toxicity of amiodarone (abstract). Circulation1985;72(suppl III):III-272

8. Porterfield JG, Groshart KD, Golden EB, Iansmith DHS,Dean PJ, Vanderzwaag R: Amiodarone pulmonary toxicity(abstract). Circulation 1985;72(suppl III):III-272

9. Magro SA, Lawrence EC, Wheeler SH, Krafchek J, Lin HT,Wyndham CRC: Amiodarone pulmonary toxicity: Prospectiveevaluation of serial pulmonary function tests. J Am CollCardiol 1988;12:781-788

10. Smith WM, Lubbe WF, Whitlock RM, Mercer J, RutherfordJD, Roche AH: Long-term tolerance of amiodarone treatmentfor cardiac arrhythmias. Am J Cardiol 1986;57:1288-1293

11. Haffajee CI, Love JC, Canada AT, Lesko LJ, Asdourian G,Alpert JS: Clinical pharmacokinetics and efficacy of amio-darone for refractory tachyarrhythmias. Circulation 1983;67:1347-1355

12. Kudenchuk PJ, Pierson DJ, Greene HL, Graham EL, SearsGK, Troubaugh GB: Prospective evaluation of amiodaronepulmonary toxicity. Chest 1984;86:541-548

13. Dean PJ, Groshart KD, Porterfield JG, Iansmith DH, GoldenEB: Amiodarone-associated pulmonary toxicity: A clinical andpathologic study of eleven cases. Am J Clin Pathol 1987;87:7-13

14. Heger JJ, Prystowsky EN, Zipes DP: Relationships betweenamiodarone dosage, drug concentrations and adverse sideeffects. Am Heart J 1983;106:931-935

15. Solow EB, Schwier HS, Heger JJ, Prystowsky EN, Zipes DP:HPLC analysis by amiodarone and its desalkyl metabolite inplasma, red cell hemolysate and saliva (abstract). Clin Chem1983;29:1204

16. Crapo RO, Morris AH: Standardized single breath normalvalues for carbon monoxide diffusing capacity. Am Rev RespirDis 1981;123:185-189

17. van Rooij WJ, van der Meer SC, van Royen EA, van ZandwijkN, Darmanata JI: Pulmonary gallium-67 uptake in amio-darone pneumonitis. J Nucl Med 1984;25:211-213

18. Israel-Biet D, Venet A, Caubarrere I, Bonan G, Danel C,Chretien J, Hance AJ: Bronchoalveolar lavage in amiodaronepneumonitis: Cellular abnormalities and their relevance topathogenesis. Chest 1987;91:214-221

19. Martin WJ, Rosenow EC: Amiodarone pulmonary toxicity:Recognition and pathogenesis (Part 2). Chest 1988;93:1242-1248

20. Rosenow EC: The spectrum of drug-induced pulmonary dis-ease. Ann Intern Med 1972;77:977-991

21. Martin WJ, Rosenow EC: Amiodarone pulmonary toxicity:Recognition and pathogenesis (Part 1). Chest 1988;93:1067-1075

22. Horowitz LN: Detection of amiodarone pulmonary toxicity:To screen or not to screen, that is the question! JAm CollCardiol 1988;12:789-790

23. Martin WJ, Howard DM: Amiodarone-induced lung toxicity:In vitro evidence for the direct toxicity of the drug.Am JPathol1985;120:344-350

24. Pozzi E, Sada E, Luisetti M, DeRose V, Scelsi M: Interstitialpneumopathy and low-dosage amiodarone. Eur J Respir Dis1984;65:620-625



ownloaded from



25. Harris L, McKenna WJ, Rowland E, Holt DW, Storey GCA,Krikler DM: Side effects of long-term amiodarone therapy.Circulation 1983;67:45-51

26. Rotmensch HH, Belhassen B, Swanson BN, Shoshani D,Spielman SR, Greenspon AJ, Greenspan AM, Vlasses PH,Horowitz LN: Steady-state serum amiodarone concentrations:

Relationships with antiarrhythmic efficacy and toxicity. AnnIntem Med 1984;101:462-469

KEY WORDS * pulmonary fibrosis * diffusion capacity .

pneumonitis * drug side effect



ownloaded from


R E Dusman, M S Stanton, W M Miles, L S Klein, D P Zipes, N S Fineberg and J J HegerClinical features of amiodarone-induced pulmonary toxicity.

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1990 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.82.1.51

1990;82:51-59Circulation.

http://circ.ahajournals.org/content/82/1/51the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:



ownloaded from

http://circ.ahajournals.org/content/82/1/51

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


clinical features of amiodarone-induced pulmonary...

Documents