Vd 33 (Suppl), No. 1,1993 Rimd in U.S.A.

Intravenous Immune Globulin for the Prevention of Infections in Children with Symptomatic Human Immunodeficiency Virus Infection1

LYNNE M. MOFENSON AND JOHN MOYE, JR. Pediatric, Adolescent and Maternal AIDS Branch. National Institute of Child Health and Human Developmenr, National Institutes of Health, Bethesda. Maryland 20852

ABSTRACT. Infection with human immunodefw vi- moral arm of the immune system (1). Such Well abnormalities rus (HIV) caw a progressive immunodeficiency, often appear to be multifactorial in origin, with demonstration of both manifested in adults by the development of infections with intrinsic B-cell defects and defects secondary to CD4+ helper relatively me, opportunistic organisms. However, in HIV- dysfunction (1-4). Functional abnormalities of B lymphocytes infected children, recurrent serious infections with common noted in vitro include elevated spontaneous proliferation, in- encapsulated bacteria, as well as recurrent minor bacterial creased numbers of spontaneous plaque-forming cells in the and viral infections, ltmy be early and frequent ntanifesta- peripheral blood, enhanced responsiveness to B-cell growth fac- tions of HIV disease. The use of i.v. immune globulin tors, and refractoriness to normal in vitro signals for B-cell (MG) has been shown to prevent infections in patients activation (1). In vivo manifestations of dysfunction include with primary immnnodefm and in uncontrolled &dies hypergammaglobulinemia, circulating immune complexes, an of HIV-infected children. This article reviews the problem inability to mount an appropriate serologic response after im- of and immunologic predisposition for reemrent infections munization, and an increased frequency of common bacterial in pediatric HIV infection and reports on the use of MG and viral infections. for prophylaxis. Updated analysis fnw the National Insti- Although B- and T-cell dysfunction are seen in both HIV- tute of Child Health pad H~mrnDevelopment MGClin- infected adults and children, the consequences of B-cell dysfunc- ial Trial, a multkente!r, doubbblind, placebcxontrolled tion may be more pronounced in children, particularly those trhl of the safety and efficacy of MGfor the prevention with vertically acqWinfection. In adults, even if response to of infections in children with symptomatic HIV infection, new antigens is diminished, preexisting memory cells provide is presented. Between March 1988 and January 1991,376 some immunity fiom infection with common pathogens. How- children with clinical or immuw1ogic evidence of HIV ever, if defective B-cell function occurs early in life, before the disease who were randomized to receive MGor albmnin development of specific memory cells, recurrent and severe placebo were enrolled in this trial. In children with an entry infection with otherwise common organisms may result. CD4+ count of 200/mm3or higher, MG signifiitly In children with HIV infection, recurrent infection with cod mmeased the time ftee from serious bacterial infections mon bacterial and viral organisms is a frequent and early mani- and significantly decreased the rates of minor bacterial festation of HIV disease. The similarity of the type of infections infections and viral infections. (Pcdiatr Res 33 (Suppl): seen in HIV-infected children to those seen in children with SSO-SSS, 1993) congenital hypogammaglobulinemiaprovided the early rationale for use of IVIG for infection prophylaxis in pediatric HIV Abbreviations infection. This article briefly reviews the problem of and the immunologic pxdkposition for recurrent infections in HIV HIV, human immunodefichcy virus infected children and summarizes the published reports on the MG, intravenous immune globulin use of MG for infection prophylaxis, including updated and NICHD, National Institute of Child Health & Human newly analyzed data from the NICHD IVIG Clinical Trial. Development Although the focus of this article is on the use of IVIG for PCP, Pnwmcytis carinii pneumonia preexposure prophylaxis of infections, other current and poten- ZDV,zidovdh tial uses for MGin pediatric HIV infection include treatment CDC, Centers for Disease Control of HIVassociated thromboqtopenia, treatment of cytomeg- ACE, AIDSClinicll Trial Gmop alovirus and Parvovirus B19 infection, and the potential use of hyperimmune HIV immune globulin for treatment of HIV in fection and prevention of vertical tnmmkion (5,6).

Infection with HN is characterized by a progressive immu- RECURRENT INFECTIONS IN HIV-INFECTED CHILDREN nologic deficiency manifested by depletion of CD4+ helper HIV infection in adults is associated with the development of lymphocytes, resulting in opportunistic infections and a severe infections with relatively rare, opportunistic organisms. In con- deficit in cellular immunity. Additionally, alterations in B-lym- trast, the development of recurrent serious systemic illness caused phocyte function are common, resulting in deficits in the hu- by common encapsulated bacteria is so common in HIV-infected Comspondence:Lynne M. Mofenson, M.D., Amchte Branch Chieffor Clinical children that the federal CDC modified the pediatric AIDS case R-h, wk Addesrmt & Matcroal AIDS Branch, National Institute of definition in 1987 to include recurrent serious bacterial infections Child Health & Human Development, National Mtutm of Health, 6100 Execu- (7). tive Blvd., Room 4B1 I, Rockvilk, MD 20852. as an AIDSindicator disease Although PCP is the most 'Portions of this article were modi6ed from Mofenson LM, Sheonr WT 1992 frequent AIDS-defining illness in children, recurrent serious Use of lVlG in &tric HIV infection. Semh Pediatr I* Dis 31-10. bacterial infections occurred in almost 25% of children with P P

s1

INFANT CHILDREN ADOLESCEN cLY6 1-1-13 YR ++ Yo& Cky ha@Wdinia, Mmlda et al. (15) nated that 79% of LymphoO.nicr - +++ thdr~~pllmonu~proMema~~~~-a4 - ++ +++ uwillgaacff~~oflabo~-~- -- +++ ++ maarh 70% hv& bpital ahmmmm for an acute respl- kWpmamwkkrkmk * ~in#r~raidenScdpatbgsn.lnasmallprospec- tivdytidfor#dIOsltneohoatof27Hnr-hh%~dchildrenand An- -b~ nutrluA.~~~~trota,HIvin&etsb*badsienif- m did uaiaBbcted C0nb.d pasienb (16). NA Rccrnnntminor~~~rratiE%ercrdirrasEkorylrdw dw,aomrnoninHIV-idu%cdtW&~b-&tlrermaPhert l~rr~oftrr- 4 ++ ++ the numbg alfawm- with ++ ?, .*

-. - - --- S82 IbmFEWDN AND MOW3 tive loss of T-bdper mpome to d an* stimuhq thwcfim petients (3 1-39). With the exception of one small comparative appears to be asdatul tlrrith an incns&6 ridr of batted study oomducteci in an older hemophiliac population (39), which infewharaas morc pqrsgive T-bclpsr dysttnction ap= did nat show a benefit of MG on infection prevention, thc pcrrsk>omwhterintbaoowseobHnr~mdtobe studieb~av&tyofb~dbwithMGtherapy,incl~d- aasodatad with a dadine in CI% all number lurd the ouau- ing declslcasGd infictioas and ~tioos,improvements in rmctof gnowtbund wavi#l, a dedist in p24 an* and improvements rresslieasintdevetcd inpedirtricHWiafaeSion,~~QWS~(bybdatedeh~~mdim~vcmcntrrm~ adult~~~can~bepatcnthnyasplriwd. . dflnczioa,~nsponse,aedCD4+cdl~0unt. wa- ofrsMlPtjadQn,~tbs~8geHowewff,in~onofthese~Was~d~to ~co4.+aountnoanrrinyosurl~thcePrtyBcdltheuncQatrOUcdaakuedmostoftheshdies,the~uatc abmrmaW8 and dydmXbd 6 ~cizeinthcstudies,thewidelyvaryingregim~ofMG servtxi in infdihts, tbe-m the Ld; of- patient evaluation fw-up, antigen nsponac in immundaeicrlfy dveHIV- hfhta and the use of same jmduct produced befi 1985 that poten- tially collhbd antibody to HIV. Therefore, the National Insti- IVIGPRO&OP~NSIN tuas of~ealthsponsond two ~~lldomizal,controlled, c in dad I3iV-m-W pmphyiaxisinHIV-~~Onestudy(thcMCHDclinical~toexaminethe&cacyofMGfor~on he Won& for pmp&ykisin m-infwchitdren MG Qinical T*) has completad (40) and the other isbeseduponthec~~tyafthetyptand~uencyof(ACIYi051)~n~o~. infixtionsinHIV-~~tothosesanincMdrcnwitb ~0Wtalhumd hskodoficickicd and the suaxdld ust MCHD MG CLINICAL TRIAL of IVIG to pnnnt infixtiom in thedt congenital disoders (30). In the latw half of the 19808, aevcaal invedpton published MethodF. The NICHD MG trial was a multicenter, double- reports indicating the potential benetit of MGfor the prmntion blind, ~bacontrolldstudy to determine whether MG, in a of infections in HIV-hfected children. As summarized in Table dose of 400 mgFg every 28 d, would reduce the risk of serious 1, then have been six pubWbd case reports, invdw 61 tqackkl infttdion or death in HIV-intkded children. Figun 2 patien4 and three co-k:' repor$, involving 5 i treated povidcs an overview of the protocol design. Nonhemophiliac, Table 1. Summary of studies of ZYIG prophylax@ in HZV-in@& childrm' No. of Typt of Authot (-1 prtiehtr . npOrt Mpdose~ Rme Silvcman el d. (31) -.33 - Case $wk &n; 300 LDHt dccbc w/bemy2* Gupta ef d. (32)' ' 7 Case 1-wk edc&W, 200 supprslor cell function w%svar~2* imwt Wood el d. (33) 1 Case 4oomdLgloading; No btdalidkctiom ~m%ksemy over 19- mo schaaa et al. (34) Imo eadation; 400 rcasfebrileepisodes mgFBmm0 and Moar; LDH Ww, normalized gt0Wth-d- licwmlmt; no- deaths Williams ef al. (35) 2-wk escalation; 200 Less savae infcctiow mgFgenry3-4 ndPced- wk tionIcqgth;* i@ Case Lessirlfktb~lessbog pitlrlmc- *,&mmes in p24 rmtigcn 14 treated ~mg/lysvary2~ Dcasrrakinsspsisand spicodes; 27 untreated P t&de in- arokiacD4+ 00unt;bsaaein

wGm#-X #r bcremd .4 les@mc&LL3ii& ctine loomdksevery=Jk E~survival IVIG PROPHYLAXIS IN HIV-INFECTED CHILDREN S83

NlCHD lVlG Clinical Trial A os/owm - Mrn DLSKIH Phue Ill, Multkonter. Pheobo- cocltroN.4 D0UM.-Blind 6My POPULATION 176 HlV-in- &IWO#I rg.. 1 mo - 12 yrr Non-hom0phili.o Septk Arthritis Symptomatlo - CDC Ckn P-2 w P-1-0 Acute Mastoiditis STRATIFICATIOH w Entry GO4 or ! 200/mm9 w Hi.torl, ot AYW-d.tInm hkctknr (CDC dur P-2-D-llP-2-D-2 not)

3 culturn Clinical symptoms 6 rf % New radiologic abnormality lVlQ Albumin Piacobo ENDPOINTS: ar~omw PNEUYOIU PRIMARY O(I SINUSITIS SH.00E w Roduction in rriouo bacterial infoctionr or death SECONDARY w Roduotion b mkor b.ot.ri.1 lnfoctio~ 1 w Roduotion in rkd ~tionr I 1 W R.dUCti0n kl hoo#wbrkn ACUTE NOT W.UFFIUEY1 EPISODE WTEEPISOaS a TO Fw ptotd IUI~~PCP tl swtl CLA.RCY 2. NlCHD IVIG Chicat Trial design. UL. w luno b10p.l) A r 1 HIV-infected children under 13 y of age with clinical or immu- -yQ< N. I0 .- HIV X*Y 8.r nologic evidence of disease were eligible for entry. Forty- -2=.w - C.-P.I@DM~ ' five parent of children enrolied MI only mild symptoms of ANVJ OR =zz$e%, Hnr disease: 13% were clinically asymptomatic chiEdren with N- Y. .. X-my Urrvllly - AWIHIIII om ah& immune function (CDC pediat& class PlB) and 32% X-my ~~~syosptoms~theai~fiKCDC NOOR --&*Lt- -.-- - CL- I. X-Ray Fig. 3. Classification of serious bacterial infections in NCHD IVIG Clinical Trial. A shows the definition of serious bacterial infections used tlbe besis of entry in the trial; B shows the criteria used for classification of clinically diagnosed pneumonia and sinusitis. BAL, bronchoalveolar lavage. - -

or higher and immunologic abnormalities or HIV-idated symp toms but without prior AIDSdefining opportunistic or bacterial infections. Children were seen monthly for examinstion and inhiom, and information. regarding . dl mmnt-mfections, medications, and hospitabtmm was cwhtdTbe initiation of antiretroviral therapy avd jmphykk for PCP (apccified as vw.1 iwhw. mtig.l cu1tun. .rlba &*cl)u Or m1ibO.y &batlor Or hl8UOW trimetho@/sul&smethoxszok on 3 mmecutive d/wk) was per- mit.at any time after catry into the study, according to the Qkiul w-. CUIUwadmao, ClinM .m-. p~medial?d;redPnddcareasdekmmedby the thftient'~ LO. Ollll, - LO. MI*UI(*... La. OI.1 dU. physician with the conseat of parents or guardians. das- Fig. 4. Clasdication of other infections in NICHD IVIG clinical Definition of oWvmes. Sdous Saious infinfoctons were Trial. sified as either WomWypven or clinically diagnosed (Fig. 3A)- infections Wefedefined -&- a bacterial or oppo*&tic etiology, the episode callyconfunlcdepisodeaot~ti%,bacteremia,~)"cli-cmasey&ap,&, t& septic mk mte of iaUraal W@QY Statisticalanalysis.The times to first seriousbacterial infection acute pneumonia, oracute MtkWW diaen#sBm the MG and pkboarms were compared by nonpimmetric infections defined CpideS of anlpe ~~G~CSS~WS-mUCS anal* me mmM- of itis without defined microbiologic etidoZW tbst :a49b -\ts -ce8 mntfeatment and placebo the &I+- predetermined clinical and radi- C'd@d& M.- @- bytion of times to event were assessed with the log-rank test, sodes of pneumonia awl sinusitis usingtwo-sidedpvaluc~ independent review by two physicians ~~teFor8~~0ndaryanalysesofthee&ctofIW3anminarbacte- laboratory-proven, as acute dinically t%rsa%ot rj?l-d- 'c infections, differences in rates of , .-, d,.. acutewepisode (Fi 38). infedion between the MGand placebo groups per 100 patient- Secoadaryanalysisof'othcr~" yeam mze oompared using hypothesis tests that assume a normal terial infections and viral aad opportunistic distribution. The jackkde procedure was used to estimate stand- an organism was isoM by culture, detected by antigen or ard errors histologic~ordiagnoPadbyspecific~tl&odytestis&tbe ~esulrs.~~-baaed&~~ -. ,- episode waselrrificdr WoqMgyprovea If the eyaluatiug 31, physician repartad a didad Lyndrome felt to k consistent wit&

4 A '&!I Y-"'II-i "@I '-3 V

S84 MOFENSON AND MOYE infected children with an entry m+count of 20()/mm3 or All Chllam rlth EWy CM8- 200/mm! Oroupr la2 ComMnd higher (p= 0.009)but not in children with an entry CD4+ count

of QOO/mm3; additionally, fewer overall bacterial infections, A t acute hospitalizations, and days spent annually in the hospital ' /'"IG were observed in MGrecipients with an entry CD4+ count of 200/mm3. There was no difference in mortality between MG and placebo., however, analysis by cause of death has not yet been completed. \..., ---'-- The following analysis includes data collected through January &...... F 60 - '5..... 15, 1991 (the date of trial termination by the Data Safety and R , Monitoring Board) on 376 enrolled children with 545 patient- 40- Phcmbo ...... years of follow-up. The overall incidence of infections is sum- I a- marized in Table 2. Serious laboratory-proven or clinically di- N agnosed bacterial infections were found in 141 patients (38%). EF 50- Clinically diagnosed serious infections occurred 2.3 times more 0 lo- oma~~~rmlt~aootn frequently than laboratory-proven serious bacterial infections. I Minor bacterial infections were 3 times as frequent as serious "0 ~,SAL~L~,~~IAO~~&O bacterial infections; vid infections occurred at a frequency Dam similar to serious bacterial infections. Opportunistic infections, 6. It-. ID-. tern. although i-uent, oocurred at a rate similar to laboratory- na b 111 r.m r.n r.14 0.m proven bacterial infections. nw*. ,lu r.r r.n O.M r.0 Using events occumng through the date of trial termination, increased significance was observed for MG in prolonging in- P fection-free time. Overall, MG patients experienced a 39% B reduction in serious infections, with 91 infections in 55 IVIG patients compared with 150 infections in 86 placebo patients. in patients with an entry CD4+ count of 200/mm3 or higher, MG significantly prolonged the time free of serious bacterial infection (p = 0.0012) (FG 5A). Analysis confined to laboratory-proven infections similarly revealed a beneficial effect of MGfor chil- F 60 - dren with an entry CD4+ count of 200/mm3 or higher (p = 0.006) (not shown). I In group 1 children with a prior history of AIDS-defining I 50- ...... N / ;i infection who had an entry CD4+ count of 200/mm3 or higher, F so -Placebo there was a significant difference in the time free fiom serious E 0 10- infections between MGand placebo (p = 0.006) (Fig. 5B), with Ovmrall kp rank. P9.0005 the median time to infection over 1 y longer with MG(681 I versus 283 d). However, for group 1 children with an entry s "0 d@,A & A A ,L ,A ,A do "A CD4+ count of <200/mm3 (regardless of prior history of AIDS Dam defining infections), no difference in the time to infection was 0 n 2 m I m. 24 ns. 111s c 3s r.na 1.11 r.m r.41 seen with MG. nw*r b I r.6. r.11 r.n r.11 Children in group 2 (entry CD4+ count of 200/mm3 or higher and no prior history of AIDSdefining infection) were less likely to develop serious infections than children in group 1; however, 100 significantly longer infection-free time with MGwas noted (p = 0.0035) (Fig. 50. A significant effect of MGwas seen only after approximately 11 to 12 mo of therapy. 80 The most frequent laboratory-proven infections were primary 70 - bacderemias (74%) and pneumonia (22%). Similar to other re- M- .-'. :..... ports, the most common bacterial isolate was pneumococcus, I..... z...... accounting for 28% of all isolates. In patients with an entry 60- / i:...... CD4+ count of 200/mm3 or higher, the rate of invasive pneu- 40 - Pkcabo mocddisease per 100 patient-years was 5.7 episodes in pla- 50 - f Table 2. Overall incidence of infections in NZCHD ZVZG clinical E C trial, March 1988 to January 1991 T 101 Om.# lop lmk. P9.0016 I No. pa- No. epi- Infection rate o Type of infection perpatient-year n ;$ ,&&:f& Serious- -r - . - ' "' wprovaor cliai- 141 24 1 cally diagnosed Clinically diagnosed 116 167 0.3 1 Labproven 54 74 0.14 Fig. 5. Probability of remaining free of laboratory-proven and clini- Other infections cally diagnosed serious bacterial infections in study children assigned to Minor bacterial 243 77 1 1.42 IVIG (sdfd line) or placebo (dashed line) in NICHD IVIG Clinical Trial, Viral 156 266 0.49 March 1988-January 199 1. Modified from reference 5. A shows all study Opportunistic 53 90 0.17 children witb entry CD4+ all counts z200/mm3 (p= 0.0012 by the overall log-rank test for the comparison of the study anus); B shows from WT Use *Modified Mofenson LM, Shearer 1992 of WIG in group 1 children with entry CD4+ cell counts 1200/mm3 (p= 0.006); pediatric HNinfection. Semin Pediatr Infect Dis 3:l-10. IVIG PROPHYLAXIS M HIV-INFECTED CHILDREN TtW 3. nrCHD IMG clinical trial March 1988 to January traditionally associated with prevention of infection were de- 1991.- &zt ofZVIG on selected infections in patients with entw creased in IVIG recipients. All five of the reported cases of CD4+ z200/me measles, two of two enteroviral infections, and the single case of hepatitis A occurred in placebo patients. MGt -t There were 55 opportunistic infections in 34 patients with (n = 161) (n = 152) entry CD4+ counts of 200/mm3 or higher. The most frequent Serious bacterial infections infection was with Candida, accounting for 73% of infections, Labproven or clinically diagnosed 26.4# 48.2 followed by PCP (7%); all other etiologies accounted for 5% or Lab-proven: less. There was a similar frequency of opportunistic infections in Streptec~pneumonioc 2.0 5.7 MGand placebo children (Table 3). Otherbactaia 4.0 8.9 Detailed analysis has been done to evaluate the interaction of Clinically diagnosed: MG and concomitant therapy with trimethoprim/sulfameth- Acute pneumonia 10.0 23.9 oxazole administered 3 d/wk for PCP prophylaxis or with ZDV Acute sinusitis 10.4 9.7 (41). As of January 199 1, when the study ended, 50% of study Other infections patients with an entry CD4+ count of 200/mm3 or higher had Minor bacterial 1lS.lg 159.7 received PCP prophylaxis (50% of placebo and 49% of IVIG Otitis 63.5 99.5 patients) and 45% had received ZDV (47% of placebo and 44% Skin/sofi tissue 14.0 23.0 of MGpatients). Time to initiation of either therapy was similar Upper respiratory tract 10.8 15.9 in both groups. Vi 36.01 54.0 For minor bacterial infections and viral infections, risk md-- Herpes simplex virus 7.6 13.7 (MG versus placebo) were used to compare the relative ris ' Opportunistic 10.0 13.3 infection before and after beginning ZDV therapy or PCP 1 Candid spies 6.4 10.6 phylaxis. No sigdlcant differences mre found for either mv. Pneumocystis carinii 1.2 0.4 bacterial infections or viral infections for ZDV (p z 0.15) or *Modified from Mofenson LM, Shearer WT, Moye J, Nugent R, PCP prophylaxis (p r 0.19). Willoughby A, for the MCHD Study Group 1992. Letter to the editor. Both logistic and linear regrexion models were used to evaluate N Engl J Med 326:1636-1637. the interactive effects of MG, PCP prophylaxis, and ZDV t Rate of infection per 100 patient-years. therapy on rates of infection. Regardless of either ZDV therapy # p = 0.002, comparing rate per 100 patient-years, MGvs placebo. or PCP prophylaxis, IVIG reduced the chance of minor bacterial Qp = 0.02, comparing rate per 100 patient-years, MG vs placebo. infection (p = 0.05), the number of minor bacterial infections I p = 0.01, comparing rate per 100 patient-years, MG vs placebo. among those having more than one infection (p= 0.03), and the ccbo (Table number of viral infections among those having more than one patients compared with 20 in MGpatients 3). infection (p = 0.06). There was no significant interaction either MGwas~witha~inpncum~diseasebetween IVIG and ZDV (p = 0.1 1) or between IVIG and PCP regardless of PCP prophylaxis status: of the 18 episodes of disease prophylaxis (p = 0.43) in the rates of minor bacterial infections, pneumocoad in patients with entry CD4+ and there was no significant interaction between IVIG and either counts of 200/md or lugher, five occumd in placebo and two rates in MGpatients who were kviqPCP prophylaxis at least 1 ZDV therapy or PCP prophylaxis in of viral infections (p r 0.17). mo before the event occurred, and eight occurred in placebo and using three in MGpatients who were not receiving prophylaxis before Survival curves, right-censored (only events occurring the event. Although the overall ocmnce of serious infections before initiation of PCP prophylaxis or ZDV) and lefi-cemored with other bacteria was also decreased with MG, infections with (events occumng after initiation of PCP prophylaxis or ZDV) other bacterial species were too few to make meaninm individ- data were used to compare time free of serious bacterial infi.iction ual comparisons. between MGand placebo; in all cases MGwas associated with Acute pneumonia was the most common clinically diagnosed improvement in infection-free time. As discussed previously, serious infection (6674, followed by acute sinusitis. There was a placebo patients experienced episodes of pneumococcal disase 53%decrease in the rate of acute clinically diagnosed pneumonia on or off PCP prophylaxis before an event, and patients rdving per 100 patient-years in IVIG recipients with entry CD4+ counts MGexperienced fewer episodes regardless of PCP prophylaxis of 200/mm3 or higher (Table 3). Statw. Six hundd forty-nine minor bacterial infedions were ob A proportional hazards analysis was performed to look for wedin 205 children with entry CD4+ counts of 200/mm3 or po6sii interactions between MGand either PCP prophylaxis higher. Nearly 60% involved the ear, 13% skin and soft tissue, or ZDV thtrapy in terms of time to first serious bacterial infec- and 10% upper respiratory tract. Sitlyfewer minor bae tion. S~cally,PCP prophylaxis and ZDV therapy were used terial infections per 100 patient-years occumd with MG in as tinmedependent covaiates (22). The resulting analysis indi- patients with entry CD4+ counts of 200/mm3 or higher (p = cated that the previously discussed effect of MG was of the 0.02) (Table 3). IVIG was associated with decreases in rates per same magnitude before and after the use of either PCP prophy- 100 patient-yean of 36% for otitis, 39% for skidsoft tissue laxis a ZDV. ZDV use was assaciated with a nonsigmtlcant infections, and 32% for upper respiratory tract inf' prolongation of time to serious badmid infection, whereas PCP Similarly, significantly fewer viral infections pg 100 patient- prophylaxis was asochted with a dewin time to serious I years were observed with MG in patients with entry CW+ bactmkl infection, prnrumsbly a reflection that children who counts of 200/mm3 or higher (p = 0.01) (Table 3). Two hundred developed infections were more Iikely to be pIaced on prophy- twelve viral infections were diagnosed in 129 -ts in this laxis. group. Of infections with a spedfled etiology, over 50% were due t0herpwSimplexvirw.~were45%f~hapcssim~ DISCUSSION virus infections in MG tecipients. For viral infdoos with a spedled etiology, thm agents for which humoral immunity is Treatment with MGwas safe and well tolerated, with benefit seen across sewed idmious outcome parameters. In MG- and C shows group 2 chiMren (p= 0.0035). The &ted proporti; treated children who entend the trial with a CD4+ count of of chiidren in each study arm dnginfection-free at 6, 12, 18, and 200/mm3 or higher, treatment significantly increased the time 24 mo is indicated beneath eacb g%aph. fne from serious bactcd infections and significantly reduced AND MOYE the number of minor bactd infections and viral infections function of the residual T-helper cell population may be too (42). severe to overcome through provision of antibody alone. Ther- As in other studies, S. pneumoniae was the most common apies that may be succxdd at one level of immune dysfunction identified etiology of serious bacterial infection; MG therapy may not be dcient with increasing levels of dysfunction. was associated with a 62% decrease in pneumococcal disease in patients with entry CD4+ counts of 200/mm3 or higher. PCP ACTG PROTOCOL 05 1 prophylaxis as adminkred in this study and cumntly recom- mended by the Public Health Service (trimethoprim/sulfameth- Although ZDV therapy during the NICHD MGClinical Trial oxazole given 3 d/wk) (22) did not appear to significantly affect did not alter the results, the NICHD trial was not designed to the incidence of pneumococcal disease or alter the benefit of evaluate the efficacy of IVIG in children already receiving ZDV. IVIG. However, the influence of a daily regimen of antibiotic The National Institutes of Health is also sponsoring a double prophylaxis remains to be determined. blind, placebocontrolled trial to evaluate MGin children with Repeated episodes of acute pneumonia cause significant mor- severely symptomatic HIV infection (AIDS or AIDS-related bidity in HIV-infected children. In a recent report, 70% of a complex) receiving ZDV (ACTG protocol 05 1). This trial, which cohort of HIV-infected children fiom New York City required has completed enrollment with 262 patients, closed to accrual hospital admission for an acute respiratory illness with no iden- on August 16, 1990 and is currently ongoing. The sample size tified pathogen, and 30% developed laboratory-proven bgcterial requirements of ACTG 051 assume an infection rate of 25% pneumonia (1 5). In our study, acute pneumonia, the majority after 24 mo and an MGtreatment effect of 60%;lower infection clinically diagnosed, was the most frequently observed serious rates or treatment effects less than 60% may limit the ability of infection. The rate of acute laboratory-proven or clinically diag- the trial to detect a significant treatment difference. The observed nosed pneumonia episodes in placebo-arm patients with entry overall treatment effect of IVIG in the NICHD MG Clinical CD4+ counts of 200/mm3 or higher (26 episodes per 100 patient- Trial at the time of study termination was 39%; however, the years) was more than 6 times greater than that reported for patient population in the NICHD trial was less severely symp healthy preschool children (43). IVIG therapy was associated tomatic than that in ACTG 05 1. with a 53% decrease in acute laboratory-proven bacterial or In some studies, ZDV has been demonstrated to improve T- clinically diagnosed pneumonia episodes per 100 patient-years helper and other immune cell function (48-50). Although the in patients with entry CD4+ counts of 200/mm3 or higher. immunologic correlates of infection risk in HIV-infected children Viral infections were a significant cause of morbidity in this have not been well delineated, it is possible that ZDV, by delaying cohort. MGtherapy was associated with a significant decrease the time to immunologic deterioration, could prolong an infec- in viral infections, particularly herpes simplex virus, in patients tion "risk-fk" period, during which adjunctive immunotherapy with entry CD4+ counts of 200/mm3 or higher. Although cellular may not be necessary. Further evaluation of data from the immunity is thought to play a major role in prevention of NICHD IVIG Clinical Trial, as well as ACE05 1, may better recurrent herpes simplex infection, recent data have suggested delineate the subgroup of HIV-infected children most likely to that antibodydependent cellular cytotoxicity may also play a benefit from MGtherapy. significant role (44,45). It would be anticipated that a protective It is important to note, however, that the effect of ZDV in effect of antibody would therefore depend upon a relatively intact enhancing immunologic and clinical status appears to be tran- cellular immune system, as might be seen in those children with sient (5 1,52). Because the natural history of HIV infection, even CD4+ counts of >2W/mm3. in the treatment era, appears to be one of progressive immuno Limitations of IVIG therapy were noted. In group 2 patients, deficiency, serial combinations of antiretroviral therapy, antibi there was some delay before a treatment effect was seen for otic prophylaxis to prevent PCP, imrnunomodulator therapy serious infections. This may be related to the low incidence of such as MG, and other adjunctive therapies may be necessary events in these patients, who had less advanced HIV disease, and to prevent the eventual development of a downward cycle of the likelihood that infectious complications may not increase in increasingly opportunistic infections in the HIV-infected child. frequency until immunologic attrition occurs, i.e. loss of T-helper recall antigen response (29). In addition, because the half-life of REFERENCES IgG is reduced in hypergammaglobulinemic states (46), attain- I. Lane HC, Fauci AS 1985 Immunologic abnormalities in the acquired immu ment of a month-long protective effect of IVIG may be delayed node6ciency syndrome. Annu Rev Immuad 3477-500 in HIV-infected patients. In patients with primary immunodefi- 2. Am- AI, Schiffinan G, Abrams D, Volberding P, Ziegler J, Conant M ciency receiving monthly IVIG, serum IgG levels show a stepwise 1984 BceU immunodeficiency in acquired immunodeficiency syndrome increase in peak and trough levels after the fmt four to six JAMA 251:1447-1449 3. Pahwa S, FmS, Mena R, Pahwa R I986 F'cdiatric acquired immunodefi infusions before reaching a protective plateau (47). The optimal ciency syndrome: demonstration of B lymphocyte defects in vitro. Diag~ dose and frequency of IVIG administration in pediatric HIV Immunol4:24-30 infection remains to be determined. Further data analysis to 4. Bernstein U,Ochs HD,Wedgwood RI, Rubinstein A 1985 Defective humor& evaluate the timing of infectious events in relation to time of immunity in pediatric acquired immunodeticiency syndrome. J Pediatr 107:352- 357 IVIG infusion is planned. 5. Mofenson LM, Sharer WT 1992 Use of MG in pediatric HIV infecton In children who entered the trial with CD4+ counts of <200/ Semin F'cdiatr Infed Dis 3:l-10 mm3, no significant improvement was noted. The limited num- 6. Yap PL, Todd AAM, WilliPms PE, Hague RA, Mok J, Burns SM, &ettle RI ber of children with entry CD4+ counts of C200/mm3 (50 total 1991 Use of intravenous immu~obulinin acquired immune deficiency patients) the power of this study detect such a syndmw. Cancer 68(suppl)l440-1450 decreases to 7. Centas for Dhm Control 1987 Wonof thc CDC swv&ana case diffetrmce and conclusions regarding the effectiveness of MGin definition fix soquid immunodeficiencysyndrome. MMWR 36(suppl) IS- this population cannot be drawn from this study. However, it is 15s possible that adjunctive immunotherapy with MG is most 8. Oxtoby MJ I991 PaktaUy acquired HIV infection. In: PizLo PA, Wilf' effective in those children with mild-to-moderate immunologic CM (eds) Pediatric AIDS The Challenge of HIV Infection in Infants Children and Adol~tgWilliams & Wilkins, Baltimore, pp 3-21 dysfunction with some preservation of T-helper cell function. In 9. Bemstoin U, BZ, Novick B, Sicklick MJ, Rubimtein A 1985 Bpcterial such children, therapy may compensate for defects in T-helper infecton in the acquired immuaodeficiency syndrome of children. Pediatr recall antigen response and provide protection against a variety Infect Dh J 4472-475 of bacterial and viral pathogens through provision of specific lo. Mnski K, BorLowdry W, Bonk S, Lawrence R, Chandwani S 1988 Fbctmh infections in human immunodeficiency virus-infected children. F'cdiatr In- antibody as well as enhancement of antibodydependentcytotox- fect Dis J 2323-328 icity. As immunologic function continues to deteriorate, the 1 I. Roilida E, Marshell D, Venzon D, Butler K, Hwson R, PiPo RA 1991 combination of CD4+ cell depletion coupled with severe dys- Becterial infections in human immunoddiciency virus type 1-infected chil- IVIG PROPHYLAXIS IN HIV-INFECTED CHILDREN

the impact of central venous cathetersand antire-tmviral agents. Pediatr 37. Calvelli TA, Rubenstein A 1986 Intravenous gamma globulin in infant- lnfeet DisJ 1@813-819 acquired immunodeficiency syndrome. Pediatr Infect Dis J 5:S207-S210 12. Hsu H, Moye J, Ng P, Shea 9, Mofenson L, Kunches L, Tobin S, DeMaria A, 38. Siegal FP, Oleske JM 1986 Management of the acquired immunodeficiency Grdy G, Mclntosb K, Meissner C, Pastemack M, Pelton S, Stechenbeq B, syndrome: is there a role for immune globulin? In: Morel1 A, Hydeggar AE Sullivan J, and the Massachusetts Working Group on Surveillance 1991 (ads) Clinical Use of lntravenous Immunoglobulins. Academic Press, Lon- pneumoooocal bacteremia in children with HIV infection. Program and don, pp 373-384 Abstra*sof the VII International Conference on AIDS, Florence, Italy (abstr 39. Wagner N, Bialek R, Radinger H, Becker M, Brackmann H 199 1 Intravenous W.B. 2062) immunoglobulins in HIV-infected hemophiliac children and adolescents: a 13. Mowsky HS, Gallagher D, Gill FM, Wang WC, Falletta JM, Lande WM, randomized controlled trial over 24 mo. Program and Abstrani of the VII Levy PS, Verter JI, Wethers D, and the Cooperative Study of Sickle Cell International Conference on AIDS, Florence, Italy (abstr W.B. 2067) Disease 1986 Bacteremia in sickle hemoglobinopathis. J Pediatr 109579- 40. NlCHD IVIG Clinical Trial Study Group 1991 Efficacy of intravenous im- 585 munoglobulin for the prophylaxis of serious bacterial infections in sympto- 14. Redd SC, Rutherford GW, Sande MA, Lihn AR, Hadley WK, Facklam RR, matic HIV-infected children. N Engl J Med 32573-80 Spika JS 1990 The role of human immunodeficimcy virus infection in 41. Mofenson LM, Bethel J, Moye J, for the NICHD IVIG Clinical Trial Study pneumdbacteremia in San Francisco residents J Infect Dis 162: 1012- Group 1992 Effect of intravenous immunoglobulin, PCP prophylaxis, and 1017 AZT on prevention of infections in HIV-infected children. Pediatr Res 15. Mmlda J, Pace 9, Bonforte RJ, Kotin NU, Rabinowitz J, Kattan M 1991 31:17OA(abstr 1008) Pulmonary manifestations of HIV infection in children. Pediatr Pulmonol 42. Mofenson LM, Moye J, Bethel J, Hihhorn R, Jordan C, Nugent R, for the 10:231-235 NlCHD lWG Clinical Trial Study Group 1992 Prophylactic intravenous 16. Principi N, Marchisio P, Tornaghi R, Massironi E, Onorato J, Pim P, Libretti immunoglobulin in HIV-infected children with CD4+ counts of 0.20 x lo9/ C 1991 Occurrence of infections in chiJdren infected with human immuno- L or more; effect on viral, opportunistic, and bacterial infections, JAMA deficiency virus. Pediatr Infect Dis J 10:190-193 268:483-488 17. Principi N, Marchisio P, Tornaghi R, Onorato J, Massironi E, Pim P 1991 43. Murphy TF, Hendemn FW, Clyde WA, Collier AM, Denny FW 1981 Pneu- Acute otitis media in human immunodeficiency +infected children. monia: an 1 1-year study in a pediatric practice. Am J Epidemiol 1 13: 12-21 Pediatrics 88:566-571 44. Kohl S 1991 Role of antibodydependent cellularcytotoxicityin defense. against 18. Stiehm ER, Wara DW 1991 Immunology of HIV. In: Piuo PA, Wiert CM herpes simplex infections. Rev Infect Dis 13: 108- 1 14 (eds) Pediatric AIDS: The Challenge of HIV lnfection in Infants, Children 45. Mester JC, Glorioso JC, Rouse BT 1991 Protection against zosteriform spread and Adolescents. Williams & Wilkins, Baltimore, pp 95-1 12 of herpes simplex virus by monoclonal antibodies J Infect Dis 163263-269 19. Leibovitz E, Riguad M, Pdack H, Lawrence R, Chandwani S, Krasinski K, 46. Morell A, Terry MD, Waldmann TA 1970 Metabolic properties of IgG Borkowsky W 1990 Phewnocylis carinii pneumonia in infants infected with subdasses in man. J Clin Invest 49:673-680 the human immunodeficiency virus with more than 450 CD4 T lymphocytes 47. Ochs HD, Fixher SH, Wedgwood RJ, Wara DW, Cowan MJ, Amman AJ, per cubic millimeter. N Engl J Med 323531-533 Saxon A, Budinger MD, ALlred RU, Rousell RH 1984 Comparison of high- 20. Kovacs A, Frederick T, Church J, Eller A, Oxtoby M, Mascola L 1991 CD4 dcse and lowdose intravenous immunoglobulin in patients with primary T-lymphocyte counts and Pneumorystis corjnii pneumonia in pediatric HIV immunodeficiency diseases. Am J Med 76(suppl3A):78-82 infection. JAMA 265: 1698-1 703 48. Roilides E, Venzon D, Pizzo PA, Rubin M 1990 Effects of antiretroviral 2 1. Centers for Disease Control 1991 Guidelines for prophylaxis against Pneume dideoxynucleosides on polymorphonuclear leukocyte function. Antimimb cystiscarnii pneumonia for children infected with human immunodeficiency Agents Chemother 34: 1672- 1677 virus MMWR 40(suad)RR-2: 1-13 49. Roilides E, Clerici M, Butler K, Gress J, Pizzo PA, Shearer GM 1990 Improve- 22. Denny TN,Niven PIS& C,~hadwic~ EG, Krigm L, Amer M, Connor EM, ment of T helper cell function on dideoxyinosine therapy: correlation with Okskc JM, Yogv R 1990 Lymphocyte subsets identified by monoclonal the clinical improvement of HIV-infected children. Programs and Abstracts antibodies in healthy children: ~*~es 2'1: 155A(abstr 916) of the 30th Interscience Conference on Antimicrobial Agents and Chemo- 23. European Collaborative Study 1991 Children born to women with HIV-1 therapy Atlanta, GA (abstr 428) , infection: natural histmy and risk oftmumision. Lancet 337:253-260 50. Rinaldo C, Huang XL, Piazza P, Armstrong J, Rappocciolo G, Pazin G, 24. Klee EB, Jackson DJ, Green SDR, Mokili JLK, Cutting WAM 1991 Immu- McMahon D, Gupta P, Fan Z, Zhang Z, Ho M 1991 Augmentation of noglobulins and 9-2 miaoglobulin in HIV-I infected children in mral Zaire. cellular immune function during the early phase of zidovudine treatment of AIDS 5:1032-1034 AIDS patients. J Infect Dis 164:638-645 25. Pahwa S, Filaig S, Menez R, Pahwa R 1986 Pediatric acquired immunodefi- 51. Yarchoan R 1990 Taking drugs from the test tube to patients (pp 607-612). ciency syndrome demonstration of B-lymphocyte defects in-vitro. Diagn In: Broder S, moderator. Antiretroviral therapy in AIDS (NIH Conference). lmmunol4:24-30 Ann Intern Med 1 13:604-6 18 26. Borkowsky W, Stele U, Grubman S 1987 Antibody responses to bacterial 52. Fischl MA, Richmann DD, Grieco MH,Gottlieb MS, Volberding PA, Laskin toxoids in children infected with human immunodeficiency virus. J Pediatr OL, Leedom JM, Grwpman JE, Mildvan D, Schwley RT 1987 The ef6cacy 1 10563-566 of azidothymidine(AZT) in the treatment of patients with AIDS and AIDS- 27. Shearer GM, Clerici M 1991 Early T-helper cell defects in HIV infection. AlDS related complex: a double-blind, placebo-controlled trial. N Engl J Med 5:245-253 317185-191 28. Lucey DR, Melcher GP, Hendrix CW, Zajac RA, Goetz DW, Butzin CA, Clerici M, Warner RD, Abbadessa S, Hall K, Jaso R, WwIford 9, Miller S, FLOOR Stocks NI. Salinas CM. Wolfe WH. Sheam GM. Boswell RN 1991 Human DISCUSSION immunodeficieoey vihinfecti0n.h the US Force: sermnv&on% D~.schreiber: I want to an alternative expla- clinical staging and assessment of a T-helper functional assay to predict change in CD4+ Tall counts. J Infect Dis 164631437 nation for your consideration, which was not in your article in

29. Roilides~--~-- E. Clerici M. DePalma L. Rubin M. Piuo PA 1991 Helocr T-mu the New Ennland Journal of Medicine. PerhaDs an FC recevtor responses ~~hiklrehinfected wih human immunodeficiency viktype I. defect is ~resentwithin macro~hagesand in &tients with ~DS J Pediatr I 183724-730 and is one mechanism that prebis$ses these dentsto bacterial 30. Mofenson LM, Willoughby A 1991 Passive immmkation. In: Pizzo PA, Wiert CM (eds) Pediatric AIDS. The Meageof HIV Infection in Infants, infection. onewould expect patients who have a high CW+ cell Children and Adolescents ~h& wilkins, pp 633450 population to respond better to IVIG because those patients 3 1. Silvennan BA, Rubinstein A 1985 Serum IactatedehydrogenaseleveLSinadults would be more likely to have bacterial infections than those who and children with AIDS: possible indicator of B cell lymphoproliferative and have low CD4 levels and might have a Tafl-mediated host disease activity; effect of IVIG on enzyme lev~kAm J Med 783728-736 response. Patients who have ad&uak -+ would be using 32. Gupta A, Novick BE, Rubinstein A 1986 Restoration of suppressor Tall functions in children with AlDS following intravenous gamma globulin antibody Fc-dependent mechanisms in clear,ng bacterial infec- treatment. Am J Dis Child 140:143-146 tion and one would suspect that the presentation of antibody 33. wood CC, McNamara IG, Sch- DF, Menill WW, Sbapiro ED 1987 intravenously might overcome their Fc receptor defect. I wodd Revention of pneumococcal bacteremia in a child with ARC. Pediatr Infect have Dredicted that vou would see a decrease in the incidence of D~SJ 6:~-566 infections \n the spectrum of patients that you have 34. Schaad UB, Gianella-Borradori A, Perret B, Imbach P, Mmll A 1988 Intra- batted venous immune globulin in symptomatic paediatric human immunodcfi- Seen, and I would have predicted that the decrease would ciency virus infection. EUI J Pediatr 147:300-303 in the patients who have adequate CD4+ cells because they 35. Williams PE. Hague RA. Yav PL Mok J, Bmtk RP. Couns NA, Eden OB, would be more likely to have bacterial infection that codd be Watson JG 1988 ~nxtmeitof human immunodeficiency virus antibody mrrectable by the pfesentation of antibody through MG.One children with intravenous immunoglobulin. J How Infect 12(suppl D):67- approach would be to demonstrate that those patients who have 73 cOrRCkd 36. Hague RA, Yap PL, Mok JY, Eden 09, Coutts NA, Watson JG, Hargreaves the increased incidence of that by MG are m, Whitelaw JM 1989 intravenous immunoglobulin in HIV infection: the ones with an FC receptor defect. We used that kind of evidence for efficacy of treatment khDis child 64: 1 146- I I 50 rationale in chronic renal failure and reported ow findings, in MOFENSON AND MOYE an article in the New England Journal of Medicine, showing that Dr. Mofenson: Well, as I have said, this is very preliminary the severity of bacterial infection correlated with the severity of data. I saw the data on rate of infection from time of infusion the macrophage Fc receptor defect. only a few days ago. It is just an initial run and we are looking Dr. Mofenson: I wish that we had cells saved; however, we do at a variety of different things. I think that this study did not have serum saved from the children. When the study was de- answer all the questions; I think it has raised even more questions. signed in 1987, the issues around differential loss of recall antigen The optimal dose, the optimal timing, the optimal population were not as well known. It would be very interesting to look at remain to be decided. I am hoping that when the other ACTG recall and alloantigen function in these children, but we just do trial (051) is over we will sit down together, compare the data not have the cells to do that. and the patient populations, and hold a concensus conference to Dr. Steigbigel: Although you observed a decrease in a variety be able to help the clinician know what to do. When people ask of measures of bacterial infection, the time to death did not me now, my answers are as a clinician and not as an NIH staff change. Was there an analysis of what the causes of death were member! I will be happy to do that away from the microphone. and was there a substitution of different causes in one group Dr. Wara: All I hope is that when this consensus conference is versus the other? held you all can figure out a way to infuse IVIG no more Dr. Mofenson: We have not done an analysis by cause of frequently than once a month, because we are not going to be death. While the study was ongoing, we presented a paper at able to handle it otherwise. ICAAC (Interscience Conference on Antimicrobial Agents and Dr. Stiehm: I have never seen a case of varicella on IVIG. I Chemotherapy) about the diagnosis of pneumonia in study chil- am wondering if there were extenuating circumstances; for ex- dren. At that time, a briefblinded review ofdeath for pneumonia ample, could these children have had severe diarrhea and lost all was done. We found that most of the deaths were associated with their gamma globulin if they got the varicella at the end of the opportunistic infection. There were maybe one or two deaths period, or did they have extremely high levels ofgamma globulin, that I remember reviewing in which sepsis or bacterial infection which greatly increases IVIG catabolism? It would be interesting was the primary cause of death. There were a number of deaths to analyze those specific cases. that were due to cardiomyopathy, nephropathy, some that were Dr. Mofenson: I have divided them into primary varicella and due to Candida, meningitis, or CMV (cytomegalovirus) disease, zoster. Some of those primary varicellas may be the phenomenon but very few were related to bacterial infections per se. We need called recurrent varicella. When you think about the use of IVIG to go back and do a blinded, independent review of cause of versus hyperimmune globulin, you don't give IVIG when you death to do that curve over again. are trying to prevent varicella, you give hyperimmune VZIG Dr. Steigbigel: So in that the time to death was not changed, (varicella-zoster immune globulin); but when you want to pre- the benefit in terms of the total patient care would have to rest vent measles, you give immune serum globulin. I think it says a lot on quality-of-life issues. something about the concentration of antibody that may be Dr. Mofenson: Yes. needed to prevent infections in the two groups. After thinking Dr. Steigbigel: Do you have any data on the total hospital about the difference between herpes simplex and varicella and time, including the time that the patient is hooked up to IVIG? about the fact that we give VZIG, I was not completely surprised Dr. Mofenson: The IVIG was administered as outpatient that there were a similar number of cases. It was somewhat less therapy. We do have data on hospital days, and it was reported in the IVIG-treated group but not significantly less. in our original paper. There were significantly fewer hospitali- Dr. Stiehm: VZIG is given intramuscularly and IVIG, of zations in the IVIG-treated patients. We limited ourselves to course, gives about a comparable level of antibody. There have acute care hospitalization and found about 38 fewer hospitali- been studies on that. zations per 100 patient-years in the IVIG group. I would like to Dr. Fischer: You showed a very interesting association between go back and look at cause of hospitalization because I think it the IVIG and the change in the decline of the CD4 counts. You would be even more significant if we looked specifically at may already know the answer to an important question. That is, infection-related hospitalizations. is this a specific effect of IVIG on organisms, thus decreasing Dr. Rubinstein: From 1979 to 1981, we treated children with episodes of infection in these children? Did those who had pediatric AIDS without knowing that they had pediatric AIDS, decreased episodes have less decline in CD4 counts, or did all knowing only that they had a B-cell deficiency. We gave IVIG the children, even those who did not have a decreased number in the same manner as our treatment of agammaglobulinemic of episodes or very few infections, also have a decline in CD4 patients and we found that the once-monthly regimen did not counts? That could be very important in deciding which IVIG give us the same benefit as more frequent regimens. I recall one should be used. child who had three recurrent episodes of pneumococcal sepsis, Dr. Mofenson: The CD4 count data has also just been ana- each one at the end of the period after receiving IVIG. We looked lyzed. The problem is that one has to define which infections to at levels of pneumococcal antibodies in this child 4 wk after the consider in the analysis. Serious bacterial infections occurred in previous gamma globulin infusion, and they were nonexistent. 38% of the children, but minor infections occurred in almost That was why we moved to more frequent infusions, and we still 70% and viral infections occurred in about 38-40%. So if you continue that regimen in those of our children who are not in put them all together, almost everybody had an infection at some your study. Those in your study receive IVIG once monthly, but point. We could try to pull out the IVIG effective group for some those that are on our study are given IVIG every other week. definition of infections and look, but we have not done that yet. The second finding that was obvious in the studies was that Dr. Fischer: One thing that we found, and you have probably we did not see the immunomodulatory effect of IVIG in the seen it too, is that it seems some of the children tend to have a same manner as we do in CVID (common variable immunode- majority of the infections, whereas others will have some infec- ficiency). One of the parameters we looked at was immune tions, but there are really two different groups. That might be complexes. When we gave the IVIG once monthly, as reported worth looking at. in our paper in 1985, the immune complexes went down, only Dr. Mofenson: That is a good idea. to bounce back up after 4 wk to the same level as before. When Dr. Ballow: I have two questions, one for Dr. Mofenson, and we gave IVIG more frequently, we could reduce circulating one for either Dr. Fischer or Dr. Hill. Dr. Mofenson, concerning immune complexes only with a biweekly infusion. My question the metabolism of the IVIG, did you compare children who were to you is, now that you have analyzed the infections and shown hypogammaglobulinemic and those that are hypergammaglob- that the increased infection rate appears at the longest interval ulinemic? I wonder whether that is a variable. after the infusion, would you recommend a more frequent in- Dr. Mofenson: I want to look at that; we have not as yet. The fusion of IVIG to these children? pharmacokinetics of IVIG in HIV-infected children is not pub- lished md we! did not do that study. We do have a without the presence of opsonic anti- and whethex we noad senundwithentysaum, whichis~Oaeofthe~to use the MGeven with penicillin prophylaxis. I uouM~toowgroupis that weIqokutheentrylevels Dr.F~penidh~~bbeeneffictiwfapa &gamma globulin in these children and try Oo kok at time to blswBobaveotherimm~~Iwoukithialrgeni- idkction, but it has not been done yet. ~~~woddbeanal~Youanprobahyaot Dr. Ballow: My second question ooncerns several couurmb gubgtomvera~ofthepneumocacciwiththeMG;theap:aw that wen made about the incidence of pneumdinfectiom ~tobeGapsularserotypesthatyonpPabaMy~'tBrve la my naive way of thinking about it, we can prophylax eas* entibody to. just with penicillin, but 1 want to ask the opsonic investigatm I)r.M~n:Butitwillnotoovavirrd~ whether that is suf'ficient to prevent pneumococcal infketbs Dr.F~Peai~wiUnot.You"a~taboatlMt.