252

y-Globulin Treatment of Acute Myocarditisin the Pediatric Population

Nancy A. Drucker, MD; Steven D. Colan, MD; Alan B. Lewis, MD; Alexa S. Beiser, PhD;David L. Wessel, MD; Masato Takahashi, MD; Annette L. Baker, RN, MSN;

Antonio R. Perez-Atayde, MD; Jane W. Newburger, MD, MPH

Background Myocardial damage in myocarditis is medi-ated, in part, by immunological mechanisms. High-dose intra-venous y-globulin (IVIG) is an immunomodulatory agent thatis beneficial in myocarditis secondary to Kawasaki disease, aswell as in murine myocarditis. Since 1990, the routine man-agement of presumed acute myocarditis at Children's Hospi-tal, Boston, and Children's Hospital, Los Angeles, has in-cluded administration of high-dose IVIG.Methods and Results We treated 21 consecutive children

presenting with presumed acute myocarditis with IVIG, 2g/kg, over 24 hours, in addition to anticongestive therapies. Acomparison group comprised 25 recent historical control pa-tients meeting identical eligibility criteria but not receivingIVIG therapy. Left ventricular function was assessed duringfive time intervals: 0 to 7 days, 1 to 3 weeks, 3 weeks to 3months, 3 to 6 months, and 6 to 12 months. At presentation,the IVIG and non-IVIG groups had comparable left ventric-ular enlargement and poor fractional shortening. Compared

A cute dilated cardiomyopathy can be caused bymany different processes,12 ofwhich myocarditisis among the most perplexing in its pathobiology

and clinical course. Myocardial damage in myocarditismay be mediated by predominantly immunological mech-anisms rather than by the direct effect of viral infectionand replication. However, the efficacy of immunomodu-latory regimens for treatment of viral-induced myocardi-tis remains controversial.3,4

High-dose intravenous y-globulin (IVIG) has beeneffective in the treatment of a variety of immunologi-cally mediated diseases.5-10 Recently, we describedmarked improvement in left ventricular contractilityand myocardial function after IVIG treatment in chil-dren with myocarditis secondary to Kawasaki disease."Further supportive evidence for a beneficial effect ofIVIG in myocarditis comes from a mouse model inwhich polyclonal immunoglobulin has been demon-strated to protect against myocardial damage.12

Received March 5, 1993; revision accepted September 24, 1993.From the Departments of Cardiology and Pathology, Children's

Hospital and the Department of Pediatrics, Harvard MedicalSchool, Boston, Mass (J.W.N., S.D.C., D.L.W., A.L.B., A.R.P.A.);the Department of Pediatrics, Division of Cardiology, Universityof Vermont, Burlington, (N.A.D.); the Division of Pediatric Car-diology, Children's Hospital of Los Angeles, Calif (A.B.L., M.T.);and the Department of Epidemiology and Biostatistics, BostonUniversity School of Public Health, Boston, Mass (A.S.B.).

Reprint requests to Dr Jane W. Newburger, Department ofCardiology, Children's Hospital, 300 Longwood Ave, Boston, MA02115.

with the non-IVIG group, those treated with IVIG had asmaller mean adjusted left ventricular end-diastolic dimensionand higher fractional shortening in the periods from 3 to 6months (P=.008 and P=.033, respectively) and 6 to 12 months(P=.072 and P=.029, respectively). When adjusting for age,biopsy status, intravenous inotropic agents, and angiotensin-converting enzyme inhibitors, patients treated with IVIG weremore likely to achieve normal left ventricular function duringthe first year after presentation (P=.03). By 1 year afterpresentation, the probability of survival tended to be higheramong IVIG-treated patients (.84 versus .60, P=.069). Weobserved no adverse effects of IVIG administration.

Conclusions These data suggest that use of high-doseIVIG for treatment of acute myocarditis is associated withimproved recovery of left ventricular function and with atendency to better survival during the first year after presen-tation. (Circulation. 1994;89:252-257.)Key Words * cardiomyopathy * myocardium * y-globulin

Since 1990, the management of presumed acute my-ocarditis at Children's Hospital, Boston, and Children'sHospital, Los Angeles, has included administration ofhigh-dose IVIG in addition to conventional supportivetherapies for congestive heart failure. The purpose ofour retrospective study was to assess the effect of IVIGtherapy in patients with presumed myocarditis on sur-vival and recovery of left ventricular function.

MethodsSubjectsWe reviewed all consecutive cases of patients with acute

congestive cardiomyopathy associated with left ventricular dys-function verified by echocardiogram who presented to Chil-dren's Hospital, Boston, and Children's Hospital, Los Angeles,between January 1985 and December 1991. All patients pre-senting after 1990 received IVIG; patients first evaluated be-tween 1985 and 1989 served as historical controls. Patients wereselected according to the following criteria: acute (<3 months)onset of congestive heart failure and echocardiographic docu-mentation of diminished left ventricular function. We excludedpatients with evidence of long-standing dilated cardiomyopathyby history or physical examination; documentation of conditionsknown to be associated with acute congestive heart failure,eg, sepsis, metabolic disorder, toxic shock syndrome, HIVinfection, Kawasaki disease, primary arrhythmia, or structuralheart disease; previous treatment with a known cardiotoxicagent; family history of dilated cardiomyopathy; and evidence ofany cause of acute dilated cardiomyopathy other than myocardi-tis by history, physical examination, or myocardial biopsy. Thestudy was approved by each institution's committee on clinicalinvestigation.

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

Drucker et al y-Globulin Treatment of Acute Myocarditis 253

Normal subjects whose data were used to derive z scores ofechocardiographic parameters included 256 children who werefree of known cardiovascular disease, were taking no cardioac-tive medications, and had normal results from physical exami-nations, ECGs, and intracardiac anatomy by two-dimensionalechocardiography.

EchocardiographyData were collected by previously reported methods.1" The

primary echocardiographic outcome measures for this studywere end-diastolic dimension and fractional shortening. Leftventricular end-diastolic dimension was measured at the timeof maximum left ventricular dimension. The left ventricularpercent fractional shortening was calculated as the differencebetween the dimensions at end diastole and end systoledivided by end-diastolic dimension. We assessed recovery ofleft ventricular function during the following five time periodsafter presentation: 0 to 7 days (39 patients), 1 to 3 weeks (30patients), 3 weeks to 3 months (25 patients), 3 to 6 months (19patients), and 6 to 12 months (21 patients). If more than oneechocardiogram was performed in any given time interval, theearliest study was used for analysis. Children less than 2 yearsold usually received sedation with chloral hydrate; however,the physical state of patients during echocardiography (eg,postprandial) was not recorded.

Endomyocardial BiopsyAll endomyocardial biopsies were reviewed by a designated

cardiac pathologist at each center. Of the 39 patients in whoma biopsy was obtained, only two patients had a repeat biopsy.Usually five to seven endomyocardial biopsy specimens wereobtained from each child during catheterization. The samples,generally 1 mm in diameter, were fixed in 10% formaldehydefor conventional light microscopy or immediately fixed in 2.5%glutaraldehyde for electron microscopy. One piece was em-bedded in OTC and kept frozen at -70°F. All diagnosticinterpretations were made after examination of routine histo-logical slides, plastic-embedded sections, and electron micros-copy. A definitive diagnosis of myocarditis was made in thepresence of a lymphocytic infiltrate associated with myocytedegeneration or necrosis. The extent of myocardial damagevaried from a single focus to multiple foci of myocyte degen-eration or necrosis. Patients whose biopsies showed myocytehypertrophy or interstitial fibrosis indicative of chronic diseasewere excluded from the present study. We classified a myocar-dial biopsy as positive if it met the Dallas criteria13 for"borderline" (with increased interstitial lymphocytes) or de-finitive myocarditis. Immunohistochemistry, immunoperoxi-dase staining, and polymerase chain reaction studies on biopsymaterial were not routinely obtained in this retrospectivestudy.

TreatmentIVIG (Immuno AG, Vienna) was administered initially in a

total dose of 2 g/kg over a maximum of 24 hours. In 4 of the 26patients (15.4%) in the IVIG group, an additional dose of 1g/kg was administered 1 week after the initial infusion. Othertherapies for congestive heart failure, including inotropicagents, digoxin, diuretics, and afterload-reducing agents, wereadministered according to the routine practices of each insti-tution. Fifteen of the 26 patients (57%) at Children's Hospital,Boston, and 6 of 20 patients (30%) at Children's Hospital, LosAngeles, received IVIG, reflecting the proportion of patientsat each institution presenting in the period from 1990 to 1991.

Statistical AnalysisPrimary outcomes for this study included survival and recov-

ery of left ventricular function. For continuous variables, weused two-sample t tests to compare the treatment groups at

adjusted for body surface area (end-diastolic dimension) or forage (fractional shortening) using the z score method, in whichthe z score is calculated as (patient value minus populationmean) divided by population standard deviation. We used t teststo compare the mean values of z-score-adjusted fractionalshortening and end-diastolic dimensions to the normative meanof 0. Differences between groups were generally consideredstatistically significant at P=.05; however, to adjust for multiplet testing for differences in end-diastolic dimension and frac-tional shortening using the Bonferroni method, the cutoff forstatistical significance was .01.We compared the y-globulin and conventional therapy

groups with respect to survival and recovery of left ventricularfunction through 1 year after presentation. The differentialfollow-up periods of children in the treatment groups(10.5+±2.1 versus 20.5±3.8 months, respectively; P=.029) pre-cluded valid comparisons beyond this point.The Kaplan-Meier method was used to obtain survival

estimates. We did not distinguish between death and cardiactransplantation. We compared point estimates of survivalbetween groups at two predetermined time points: 6 monthsand 1 year after presentation. We also compared the entiresurvival experience using the log rank test. We used the Coxproportional hazards method to compare the survival experi-ence of the two treatment groups, adjusting for potentialconfounding variables such as age, biopsy status, or treatmentwith intravenous inotropic agents or angiotensin-convertingenzyme (ACE) inhibitors.We defined left ventricular function as "recovered" when

the fractional shortening was within 2 SD of the normal meanfor age. We used an approach identical to that described forsurvival analyses to compare groups with respect to recoveryof left ventricular function. Statistical analyses were performedwith Statistical Analysis Systems (Cary, NC).14

ResultsOf the 46 children with acute myocarditis, 21 (45.7%)

were treated with high-dose IVIG and 25 (54.3%) wererecent historical control patients who did not receiveIVIG therapy. All patients received anticongestive ther-apies. The patient characteristics of the two groups aresummarized in Table 1. Children in the two groups didnot differ significantly in sex distribution, age at presen-tation, time from onset of symptoms to hospital admis-sion, percent with a myocardial biopsy positive for acutemyocarditis, or indexes of illness severity, includinginitial left ventricular end-diastolic dimension, frac-tional shortening, pulmonary capillary wedge pressure,lowest pH, lowest serum bicarbonate, or cardiac index.However, patients receiving IVIG therapy were morelikely to have received treatment with intravenous af-terload-reducing agents (P<.001) and inotropic agents(P<.01) as well as with ACE inhibitors after hospitaldischarge (P=.021).

Five patients (10.9%) were treated with prednisone,of whom three were in the non-IVIG group and two inthe IVIG group. Two patients (both in the non-IVIGgroup) who received prednisone were also treated withcyclosporine. Among the three patients in the non-IVIGgroup, one died, one underwent cardiac transplanta-tion, and one had persistent ventricular dysfunction.Both patients in the IVIG group showed normalizationof left ventricular function.Although survival tended to be better among IVIG-

treated patients, the difference in the overall survivalexperience did not achieve statistical significance in our

baseline and within each study period. We used the X2 test tocompare discrete data. Echocardiographic parameters were

small sample (P>.10; Fig 1). Six months after presen-tation, the probability of survival among IVIG-treated

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

254 Circulation Vol 89, No 1 January 1994

TABLE 1. Comparability at Baseline for ConventionalTherapy With and Without IVIG

IVIG No IVIG PAge <2 y* 16/21 14/25 NS

(76) (56)Male* 10/21 14/25 NS

(48) (56)Positive biopsy* 12/19 8/20 NS

(63) (40)Captopril/enalopriltreatment* 15/17 10/19 .02

(88) (53)Intravenous inotropes* 19/21 13/25 <.01

(90) (52)

Intravenous afterload* 15/21 5/25 <.001

(71) (20)Pulmonary capillary wedgepressure, mm Hgt 14.3+1.94 18.4±3.25 NS

Lowest pHt 7.22±0.37 7.30±0.31 NSLowest serum bicarbonatet 13.5+1.5 17.3+1.17 NS

Cardiac indext 3.10-0.35 3.39+0.32 NS

Time from symptoms toadmission, d 2 4.5 NS(median, interquartilerange) (1-7) (1-17)

IVIG indicates intravenous rglobulin.*Values are no. of patients (%).tValues are mean+SEM.

patients compared with those treated only with conven-tional therapy was .84 versus .69 (P>.10); by 12 months,the probability of survival tended to be higher amongIVIG-treated patients (.84 versus .60, P=.069). Weinvestigated potentially confounding variables (youngage, biopsy positivity, treatment with ACE inhibitors)individually and in combination using Cox proportionalhazards models. The survival estimates were similarafter adjusting for these variables.We explored the effect of IVIG treatment on left

ventricular end-diastolic dimension (Fig 2 and Table 2)and on fractional shortening. End-diastolic dimensionwas z-score adjusted for body surface area, with a higherz score indicating a larger ventricle. At baseline, theend-diastolic dimension of both groups was more di-lated than normal (P<.001); patients who receivedIVIG did not differ significantly from those treated withconventional therapy. Compared with patients who didnot receive IVIG, those treated with IVIG had a smallermean adjusted left ventricular end-diastolic dimensionin the intervals from 3 to 6 months (P=.008) and from6 to 12 months (P=.072) after presentation. Patientstreated with conventional therapy continued to have amean end-diastolic dimension significantly (P<.01)larger than normal in all periods of comparison withinthe first year after presentation, whereas the meanend-diastolic dimension of children treated with IVIGwas not significantly different from normative valuesafter 3 months.

1.0O

A:

Co.0

00

i0.8 F

i------ p NS : r

L -1111 pL____0690.6 [

0.4 F

0.2

0.0

- IVIGNo IVIGp NS

0 3 6 9Time Since Presentation

(months)

12

FlI 1. Kaplan-Meier estimates of survival among patientstreated with intravenous rglobulin (IVIG) or without intravenousrglobulin (No IVIG). Each step represents death or cardiactransplant. The vertical lines represent 95% confidence intervalsaround the point estimates of probability of survival at 6 and 12months.

Fractional shortening was z-score adjusted for age,with lower z scores reflecting poorer ventricular func-tion (Fig 3 and Table 2). At presentation, all patientshad severely depressed left ventricular function. Frac-tional shortening in both groups improved progressivelyover the first year; children treated with IVIG had ahigher fractional shortening than those treated withconventional therapy alone in the time intervals from 3to 6 months (P=.033) and from 6 to 12 months(P=.029). By 6 to 12 months, mean adjusted fractionalshortening in the y-globulin group was not significantlydifferent from normative values (z=-1.01, P>.10),whereas that of children who did not receive IVIGremained significantly depressed (z=-5.51, P<.001).The overall probability of left ventricular function

recovering to normal during the first year after presen-tation tended to be higher among patients who receivedIVIG therapy (P=.06; Fig 4). By 6 months after presen-tation, the probability of recovery of left ventricular

0U)aCVEn.vUUnU,

Q-D

(n

:53

8,

< 1 wk. 1-3 wk. 3 wk.-3 mo. 3-6mo. 6-12 mo.

Time PeriodFiG 2. Plot of left ventricular end-diastolic dimension, z scoreadjusted for body surface area, according to time after presen-tation with presumed acute myocarditis. The solid horizontal lineand shaded area represent the expected population mean±2SD, encompassing the values of 95% of the normal population.Solid circles and vertical lines depict the mean and 95% confi-dence limits (CL) for adjusted end-diastolic dimension in patientstreated with intravenous yglobulin (IVIG) for each time periodsince presentation. Open circles and vertical lines represent themean and 95% CL for the group that did not receive -globulintherapy (No IVIG).

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

Drucker et al y-Globulin Treatment of Acute Myocarditis

TABLE 2. Parameters of Left Ventricular Function by Time After Presentation

Variable IVIG No IVIG PAge at presentation 2.78+0.94 (21) 5.72+1.30 (26) NS

Period 1 FS-Z -10.05±0.54 (19) -8.85+0.44 (20) NS

EDD-Z 2.73t0.68 (18) 4.09±# 0.61 (19)

Period 2 FS-Z -6.58+1.17 (17) -5.79± 1.17 (13) NS

EDD-Z 2.53+0.80 (16) 3.39±t 0.96 (13)

Period 3 FS-Z -4.78+1.01 (14) -7.56± 1.23 (11) NS

EDD-Z 3.01 ÷0.95 (13) 4.48+1.07 (11)

Period 4 FS-Z 3.76± 1.21 (8) -7.46-+ 1.03 (11) .033

EDD-Z 1.83+1.04 (8) 5.96+0.90 (11) .008

Period 5 FS-Z 1.01 +1.56 (8) -5.51 +0.99 (8) .029

EDD-Z 1.75+0.84 (7) 4.43t1.05 (8) .072

IVIG indicates intravenous yglobulin; FS-Z, z score for fractional shortening; and EDD-Z, z score forend-diastolic dimension.Values are mean+SEM (no. of patients).

function among IVIG-treated patients compared withthose in the conventional therapy group was .48 versus.26 (P>.10); by 12 months after presentation, theseprobabilities were 1.00 versus .37 (P<.001).To assess whether the beneficial effect of IVIG on

recovery of left ventricular function could be con-founded by age, biopsy status, or use of ACE inhibitors,we adjusted for these variables individually and incombination using the Cox proportional hazards model.When adjusted for biopsy status alone, IVIG was not asignificant predictor of recovery of left ventricular func-tion. However, when adjusted for age (<2 years or >2years) or age together with biopsy status, recovery ofleft ventricular function in the IVIG group was signifi-cantly better (P=.02 and P=.05, respectively). Similarly,adjustment for long-term treatment with ACE inhibi-tors did not influence the beneficial effect of IVIG onrecovery of left ventricular function (P=.013). Finally,when simultaneously adjusting for age, biopsy status,

0)a

a

ci)0

a

IL

c)(a0

* IVIGa No IVIG

2 mean, 95% CL

-1 if } } p=033 p 029

-8

-1<1 wk. 1-3 wk. 3 wk.-3 mo. 3-6 mo. 6-12 mo.

Time PeriodFIG 3. Plot of fractional shortening, z-score adjusted for age,according to time after presentation with presumed acute myo-carditis. The solid horizontal line and shaded area represent theexpected population mean--2 SD, encompassing the values of95% of the normal population. Solid circles and vertical linesdepict the mean and 95% confidence limits (CL) for adjustedfractional shortening in patients treated with intravenous rglob-ulin (IVIG) for each time period since presentation. Open circlesand vertical lines represent the mean and 95% CL for the groupthat did not receive rglobulin therapy (No IVIG).

intravenous inotropic agents, and ACE inhibitors, IVIGsignificantly influenced the course of recovery of leftventricular function (P=.03).We observed no adverse effects of IVIG administra-

tion. A dose of 2 g/kg required a 40-mL/kg volumeinfusion within 24 hours, but we could not detect anyexacerbation of congestive heart failure in these pa-tients. Our ability to discern worsening congestive heartfailure may have been limited as anticongestive therapyhad already been instituted in all patients. No patientshad hypotension or symptoms suggestive of anaphylaxis.

DiscussionIn this retrospective study of IVIG treatment of

presumed acute myocarditis in children, treatment withIVIG was associated with superior recovery of leftventricular function in multivariate models adjusting forage, biopsy status, and use of inotropic agents and ACEinhibitors. y-Globulin administration was also associ-ated with a trend toward improved survival 1 year afterpresentation. We observed no adverse effects of IVIGadministration.

a 1.00_DOg 0.8

O IL

0.6

a: .4~, ' 046

w>

-04J 4 0.2

0.0120 3 6 9

Time Since Presentation(months)

FIG 4. Kaplan-Meier estimates of recovery of left ventricularfunction among patients treated with intravenous -globulin(IVIG) or without intravenous rglobulin (No IVIG). Each steprepresents recovery of left ventricular function, defined as az-score-adjusted fractional shortening of more than -2. Thevertical lines represent 95% confidence intervals around thepoint estimates of probability of recovery at 6 and 12 months.

255

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

256 Circulation Vol 89, No 1 January 1994

Acute myocarditis may be initiated by viral infection,but subsequent myocardial damage appears to be me-diated by autoimmune mechanisms in addition to directviral infection.15 The use of immunosuppressive agentsfor treatment of acute myocarditis is controversial3,4;both beneficial and deleterious effects have been not-ed.16-'8 Parillo et al16 found an initial improvement inventricular function in patients with dilated cardiomy-opathy who received steroids compared with controlsubjects who received conventional therapy; however,there was no significant difference in ventricular func-tion between the two groups 6 months after presenta-tion. Treatment of children with biopsy-positive myo-carditis with prednisone was found to have a beneficialeffect on outcome.18 Animal studies have almost uni-formly shown that treatment with steroids during theviral replication phase (3 to 5 days after virus injection)enhances myocardial damage,19 whereas steroid treat-ment decreases myocardial necrosis if treatment isdelayed until the phase of lymphocytic infiltration.20-23A murine model supports the premise that IVIG

therapy for myocarditis is beneficial.12 Weller et al12explored whether treatment with mouse polyclonal im-munoglobulin would prevent cardiac inflammation inBalb/c male mice infected intraperitoneally with cox-sackievirus B3. Animals treated with mouse immuno-globulin 2 days before infection showed >50% reduc-tion in myocarditis compared with animals treated withphosphate-buffered saline, human immunoglobulin, ormonoclonal mouse immunoglobulin G to an extraneousantigen. Mouse polyclonal immunoglobulin also mini-mized myocardial damage when given 24 or 48 hoursafter infection.

In clinical studies in humans, IVIG has been shown toaccelerate recovery of myocardial function in myocardi-tis associated with Kawasaki disease." Myocarditis is anearly universal feature of acute Kawasaki disease, andabnormalities of left ventricular systolic function maypersist after resolution of clinical and laboratory in-dexes of systemic inflammation.24 Histological findingsinclude interstitial myocarditis during the acute phaseof the disease25 as well as hypertrophy of myocytes andfibrosis on late biopsy.26 Mean fractional shorteningamong patients treated with aspirin alone reached thenormal range only beyond 1 year after disease onset. Incontrast, mean fractional shortening in IVIG-treatedpatients was normal by 11 to 31 days after diseaseonset."The mechanism by which IVIG may improve myocar-

dial dysfunction in myocarditis is unknown. IVIG couldbe effective by providing specific antibodies to virusesand thus could lead to a more rapid clearing of myocar-dial viral infection. However, in the murine model ofcoxsackievirus myocarditis, the pooled mouse immuno-globulin often contained no neutralizing antibody tocoxsackievirus yet was effective.12 An alternative mech-anism of y-globulin's action may be modulation of theimmune response,27 leading to decreased cardiac in-flammation or to downregulation of proinflammatorycytokines that have direct negative inotropic effects.28-30It appears that the negative inotropic effects result fromthe generation of nitric oxide induced by various cyto-kines.28 In engaging the low-affinity F, y-receptors,IVIG induces synthesis of the interleukin-1 receptorantagonist, and this may result in the downregulation of

nitric oxide synthase.3' We did not test the use of IVIGin patients with known long-standing cardiomyopathy,but its potential mechanisms of action would suggestgreatest usefulness during the acute inflammatory pro-cess. The IVIG product that we used is an intact,unmodified preparation that appeared to cause very fewadverse effects. There are insufficient published data tocompare the efficacy of commercially available IVIGpreparations.These data on the use of y-globulin in acute myo-

carditis should be viewed as preliminary in view ofseveral limitations in our study. Our patient selectionand exclusion criteria attempted to maximize the per-cent of patients with acute myocarditis in the studysample, but discrimination of acute myocarditis fromthe acute presentation of a cardiomyopathy was neces-sarily inexact. Acute and convalescent viral titers werenot systematically obtained in this retrospective study,and repeat endomyocardial biopsies that might haveyielded additional supportive evidence of acute myo-carditis were not performed in children for ethicalreasons. Data on hospital course and subsequent out-come were collected retrospectively and thus are subjectto the problems of incomplete and inconsistent record-ing. In analyzing the change in left ventricular functionover time, not all patients were studied in all timeperiods, allowing for the possibility of selection bias.Patients who received y-globulin were compared withrecent historical controls; although the patient groupsappeared to be comparable at baseline and we adjustedfor known confounders (eg, ACE inhibitors), otheraspects of patient care may have changed over time.Furthermore, the initiating agents for acute myocarditisand subsequent clinical course may differ from year toyear and by geographic location. Finally, the relativelysmall number of patients in our study limits our abilityto assess the extent to which confounding factors maycontribute to the association between y-globulin treat-ment and improved recovery of left ventricular functionas well as the extent to which this association is subjectto effect modification.

In summary, these data suggest that use of high-doseIVIG for treatment of acute myocarditis may be asso-ciated with improved recovery of left ventricular func-tion and with a tendency toward better survival duringthe first year after presentation. Prospective, random-ized, clinical studies are needed to elucidate the effectsof IVIG treatment during the acute stage of myocarditison ultimate outcomes.

AcknowledgmentsThe authors thank Fred S. Rosen, MD, for manuscript

review, and Ms Kathleen M. O'Brien for manuscriptpreparation.

References1. Kereiakes DJ, Parmley WW. Myocarditis and cardiomyopathy.Am

Heart J. 1984;108:1318-1326.2. Colan SD, Spevak PJ, Parness IA, Nadas AS. Cardiomyopathies.

In: Fyler DC, ed. Nadas' Pediatric Cardiology. Philadelphia:Hanley & Belfus, Inc; 1992.

3. O'Connell JB, Robinson JA, Henkin RE, Gunnar RM. Immuno-suppressive therapy in patients with congestive cardiomyopathyand myocardial uptake of gallium-67. Circulation. 1981;64:780-786.

4. Anderson JL, Fowles RE, Unverferth DV, Mason JW. Immuno-suppressive therapy of myocardial inflammatory disease: initial

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

Drucker et al y-Globulin Treatment of Acute Myocarditis 257

experience and future trials to define indications for therapy. EurHeart J. 1987;8(suppl):263-266.

5. Imbach P, Wagner HP, Berchtold W, Gaedicke G, Hirt A, Joller P,Mueller-Eckhardt C, Muller B, Rossi E, Barandun S. Intravenousimmunoglobulin versus oral corticosteroids in acute immunethrombocytopenic purpura in childhood. Lancet. 1985;2:464-468.

6. The National Institute of Child Health and Human DevelopmentIntravenous Immunoglobulin Study Group. Intravenous immuneglobulin for the prevention of bacterial infections in children withsymptomatic human immunodeficiency virus infection. N Engl JMed. 1991;325:73-80.

7. Winston DJ. Intravenous immunoglobulins as therapeutic agents:use in viral infections. Ann Intern Med. 1987;107:371-374.

8. Cooperative Group for the Study of Immunoglobulin in ChronicLymphocytic Leukemia. Intravenous immunoglobulin for the pre-vention of infection in chronic lymphocytic leukemia: a ran-domized, controlled clinical trial. N Engl J Med. 1988;319:902-907.

9. Buckley RH, Schiff RI. The use of intravenous immune globulin inimmunodeficiency diseases. N EnglJ Med. 1991;325:110-117.

10. Newburger JW, Takahashi M, Burns JC, Beiser AS, Duffy CE,Glode MP, Mason WH, Reddy V, Sanders SP. The treatment ofKawasaki syndrome with intravenous gamma globulin. N Engl JMed. 1986;315:341-347.

11. Newburger JW, Sanders SP, Burns JC, Parness IA, Beiser AS,Colan SD. Left ventricular contractility and function in Kawasakisyndrome: effect of intravenous gamma globulin. Circulation. 1989;79:1237-1246.

12. Weller AH, Hall M, Huber SA. Polyclonal immunoglobulin therapyprotects against cardiac damage in experimental coxsackievirus-induced myocarditis. Eur Heart J. 1992;13:115-119.

13. Aretz HT, Billingham ME, Edwards WD, Factor SM, Fallon JT,Fenoglio JJ Jr, Olsen EG, Schoen FJ. Myocarditis, a histopath-ologic definition and classification. Am J Cardiolvasc Pathol 1987;1:3-14.

14. SAS Institute. SAS User's Guide: Statistics. Cary, NC: 1985.15. Kishimoto C, Abelmann WH. In vivo significance of T cells in the

developmental of coxsackievirus B3 myocarditis in mice: immaturebut antigen-specific T cells aggravate cardiac injury. Circ Res. 1990;67:589-598.

16. Parrillo JE, Cunnion RE, Epstein SE, Parker MM, Suffredini AF,Brenner M, Schoer GL, Palmeri ST, Cannon RO, Alling D, WitlesJT, Ferrans V, Rodriquez ER, Faucit AS. A prospective, ran-domized, controlled trial of prednisone for dilated cardiomy-opathy. N Engl J Med. 1989;321:1061-1068.

17. Salvi A, DiLenarda A, Dreas L, Silvestri F, Camerini F. Immuno-suppressive treatment in myocarditis. Int J Cardiol. 1989;22:329-338.

18. Chan KY, Iwahara M, Benson LN, Wilson GJ, Freedom RM.Immunosuppressive therapy in the management of acute myo-carditis in children: a clinical trial. J Am Coll Cardiol. 1991;17:458-460.

19. Tomioka N, Kishimoto C, Matsumori A, Kawai C. Effects ofprednisolone on acute viral myocarditis in mice. JAm Coll Cardiol.1986;7:868-872.

20. O'Connell JB, Reap EA, Robinson JA. The effects of cyclosporineon acute murine coxsackie B3 myocarditis. Circulation. 1986;73:1058.

21. Estrin M, Smith C, Huber S. Coxsackievirus B-3 myocarditis: T-cellautoimmunity to heart antigens is resistant to cyclosporin-Atreatment. Am J Pathol. 1986;125:244.

22. Monrad ES, Matsumori A, Murphy JC, Fox JG, Crumpacker CS,Abelman WH. Therapy with cyclosporine in experimental murinemyocarditis with encephalomyocarditis virus. Circulation. 1986;73:1058-1064.

23. Kishimoto C, Abelmann WH. Absence of effects of cyclosporineon myocardial lymphocyte subsets in coxsackievirus B3 myocarditisin the aviremic stage. Circ Res. 1989;65:934-945.

24. Landing BH, Larson EJ. Pathological features of Kawasaki disease(mucocutaneous lymph node syndrome). Am J Cardiovasc Pathol.1987;1:215-229.

25. Fujiwara H, Hamashima Y. Pathology of the heart in Kawasakidisease. Pediatrics. 1978;61:100-107.

26. Yutani C, Go S, Kamiya T, Hirose 0, Misawa H, Maeda H,Kozuka T, Onishi S. Cardiac biopsy of Kawasaki disease. ArchPathol Lab Med. 1981;105:470-473.

27. Dwyer JM. Manipulating the immune system with immuneglobulin. N Engl J Med. 1992;326:107-116.

28. Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG,Simmons RL. Negative inotropic effects of cytokines on the heartmediated by nitric oxide. Science. 1992;257:387-389.

29. Leung DY, Burns JC, Newburger JW, Geha RS. Reversal of lym-phocyte activation in vivo in Kawasaki syndrome by intravenousgammaglobulin. J Clin Invest. 1987;79:468-472.

30. Hosenpud JD, Campbell SM, Mendelson DJ. Interleukin-1-induced myocardial depression in an isolated beating heart prep-aration. J Heart Transplant. 1989;8:460-464.

31. Arend WP, Smith MF, Janson RW, Joslin FG. IL-1 receptorantogonist and IL-1 beta production in human monocytes areregulated differently. J Immunol. 1991;147:1530-1535.

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from

Perez-Atayde and J W NewburgerN A Drucker, S D Colan, A B Lewis, A S Beiser, D L Wessel, M Takahashi, A L Baker, A R

Gamma-globulin treatment of acute myocarditis in the pediatric population.

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1994 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.89.1.252

1994;89:252-257Circulation. 

http://circ.ahajournals.org/content/89/1/252World Wide Web at:

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

  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 this process is available in the

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

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

by guest on May 23, 2018

http://circ.ahajournals.org/D

ownloaded from