Prophylactic Intravenous Immunoglobulin

Prophylactic

Intravenous Immunoglobulin in HIV-Infected Children With CD4+ Counts of 0.20\m=x\109/L or More Effect

on

Viral, Opportunistic, and Bacterial Infections

M. Mofenson, MD; John Moye, Jr, MD; James Bethel, PhD; Ronald Hirschhorn; Carol Jordan, MSN; Robert Nugent, PhD; for the National Institute of Child Health and Human Development

Lynne

Intravenous

Immunoglobulin Clinical Trial Study Group

Objective.\p=m-\Toevaluate the efficacy of intravenous immunoglobulin (IVIG) for prevention of viral, opportunistic, and minor bacterial infections in children infected with human immunodeficiency virus (HIV). Design.\p=m-\Randomized,double-blind, placebo-controlled, outpatient clinical trial comparing subjects treated with 400 mg of IVIG per kilogram of body weight every 28 days with those given albumin placebo. clinical centers in mainland United States and Puerto Setting.\p=m-\Twenty-eight Rico. Patients.\p=m-\Threehundred seventy-six children infected with human immunodeficiency virus with clinical or immunologic evidence of HIV disease, 313 of whom had entry CD4+ counts of at least 0.20\m=x\109/L (\m=ge\200/mm3). Main Outcome Measures.\p=m-\Theincidence of laboratory-proven and clinically diagnosed viral, opportunistic, and bacterial infections. Main Results.\p=m-\Viralinfections and minor bacterial infections contributed more frequently to morbidity in children with entry CD4+ counts of at least 0.20\m=x\109/L (together over five times as frequent) than did serious bacterial infection, the primary outcome measure of the trial. Opportunistic infections occurred at a similar rate as laboratory-proven serious bacterial infections. In this group of children, IVIG was significantly associated with a decrease in the rate of viral infections and minor bacterial infections per 100 patient-years (36.0 vs 54.0 episodes of viral infection per 100 patient-years, IVIG vs placebo, P=.01; and 115.1 vs 159.7 episodes of minor bacterial infection per 100 patient-years, IVIG vs placebo, P-.02), as well as a decrease in the rate of serious bacterial infections per 100 patient-years (26.4 vs 48.2 episodes per 100 patient-years; P=.002). There was no apparent difference in the rate of opportunistic infections between treatment arms. Conclusions.\p=m-\Beneficialeffect of IVIG was seen across multiple infectious outcome measures, with reductions in serious and minor viral and bacterial infections observed in children with entry CD4+ counts of at least 0.20\m=x\109/L. (JAMA. 1992;268:483-488)

INFECTION with human immunode¬

ficiency virus (HIV) is characterized by progressive immunologie deficiency, From the Pediatric, Adolescent, and Maternal AIDS Branch, Center for Research on Mothers and Children, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Md (Drs Mofenson, Moye, and Nugent); and Westat Incorporated, Rockville, Md (Dr Bethel, Mr Hirschhorn, and Ms

Jordan).

Members of the National Institute of Child Health and Human Development Intravenous Immunoglobulin Clinical Trial Study Group are listed at the end of this article. Reprint requests to Pediatric, Adolescent, and Maternal AIDS Branch, Center for Research on Mothers and Children, National Institute of Child Health and Human Development, National Institutes of Health, 6120 Executive Blvd, Room 450, Bethesda, MD 20892

(Dr Mofenson).

manifested by depletion of CD4+ lym¬ phocytes and resulting in a severe def¬ icit in cellular immunity and the devel¬ opment of opportunistic infections. Al¬ terations in B-lymphocyte function are also observed, with resultant deficits in

the humoral arm of the immune sys¬ tem.14 While B- and T-cell dysfunction is seen in both HIV-infected adults and children, the consequences of B-cell dys¬ function may be more pronounced in chil¬ dren, particularly in those with verti¬ cally acquired infection. In adults, even if response to new antigens is dimin¬ ished, preexisting memory cells provide some measure of immunity from com¬ mon pathogens. However, if defective

Downloaded From: http://jama.jamanetwork.com/ by a National Institutes of Health User on 08/29/2014

B-cell function occurs early in life, prior development of specific memory cells, HIV-related disease may present early as recurrent and severe infection with otherwise common organisms. The frequency of serious systemic ill¬ ness with common encapsulated bacte¬ ria in HIV-infected children is such that the US Centers for Disease Control (CDC) modified the pediatrie acquired to the

immunodeficiency syndrome (AIDS)

case definition in 1987 to include recur¬ rent serious bacterial infections as an AIDS-indicator disease.5 Elevated rates of invasive disease due to Streptococcus

pneumoniae have been reported,69 and recurrent episodes of pneumonia, often without an identified origin, also cause significant morbidity in HIV-infected

children.10 Recurrent minor bacterial in¬ fections such as otitis media are com¬ mon in HIV-infected children,11 and these children may have serious infec¬ tion with common viral organisms.1217 The clinical similarity between the types of bacterial and viral infections observed in HIV-infected children and those seen in children with primary hypogammaglobulinemia formed the ra¬ tionale for studies of intravenous immunoglobulin (IVIG) for infection pro¬ phylaxis in pediatrie HIV infection.18 The National Institute of Child Health and Human Development IVIG Clinical Trial Study Group reported that monthly ad¬ ministration of IVIG significantly pro¬ longed the time free from serious bac¬ terial infections and decreased acute care hospitalizations in HIV-infected children ente'ring treatment with CD4+ lympho¬ cyte counts of at least 0.20xl09/L (200/ mm3).19 This report evaluates the effi¬ cacy of IVIG for prevention of a broader spectrum of infection types in these pa¬ tients over a longer period of follow-up. METHODS The IVIG Clinical Trial

was a ran¬

domized, double-blind, placebo-con¬ trolled outpatient study conducted in 28

clinical centers in the mainland United States and Puerto Rico. The study pro¬ tocol and informed consent forms were reviewed and approved by the institu¬ tional review board at each participat¬ ing center, and written informed con¬ sent was obtained from each child's par¬ ent or legal guardian. Patients random¬ ized to the treatment arm received 400 mg of IVIG (Gamimune N, Cutter Biological, Miles Laboratories Ine, Ber¬ keley, Calif) per kilogram of body weight every 28 days; patients randomized to the placebo arm received 0.1% albumin without preservatives in 10% maltose, visually indistinguishable from the study drug and administered in an identical fashion. Detailed description of the trial design and initial results based on data collected through October 1990 have been reported.19

Eligible patients were asymptomatic but immunologically abnormal, or clin¬ ically symptomatic, HIV-infected nonhemophiliac children younger than 13 years. Forty-five percent of children en¬ rolled had only mild to moderate symp¬

clini¬ cally asymptomatic with abnormal im¬ mune function (CDC pediatrie class P1B),2" and 32% had nonspecific clinical symptoms of HIV meeting criteria for CDC pediatrie class P2A only. Twentytwo percent had a history of AIDS-defining opportunistic or recurrent bacte¬ rial infections at entry (CDC class P2D1 or P2D2). Patients were stratified into two groups on the basis of entry CD4+ count and history of AIDS-defining in¬ fections. Group 1 included children with a history of AIDS-defining opportunis¬ tic or recurrent serious bacterial infec¬ tions, entry CD4+ counts of less than 0.20xl09/L, or both. Group 2 consisted of children without prior AIDS-defining infections and with entry CD4+ counts of at least 0.20 X109/L. Between March 1988 and study ter¬ mination in January 1991, 376 patients were enrolled, 313 with a CD4+ count of at least 0.20X107L and 50 with a CD4+ count of less than 0.20X107L; 13 pa¬ tients lacked an entry CD4+ count. En¬ try CD4+ count was the most important determinant of response to IVIG for se¬ rious bacterial infection prophylaxis in our previous article.19 Because subjects with an entry CD4+ count of at least 0.20 x 107L had a clear response to IVIG, while no efficacy was apparent for those with an entry CD4+ count under 0.20 X109/L, the present article analyzes IVIG efficacy in children with entry CD4+ counts of at least 0.20X 107L from groups 1 and 2 combined. The 13 pa¬ tients (eight placebo and five IVIG) who lacked an entry CD4+ count were ex¬ cluded. Entry characteristics of patients

toms of HIV disease: 13%

were

lacking entry CD4+ counts did not differ from the group as a whole; eight serious bacterial infections (six in placebo and two in IVIG patients), 20 minor bacte¬ rial infections (18 in placebo and two in IVIG patients), and eight viral infec¬ tions (six in placebo and two in IVIG patients) occurred in these children. Children were seen monthly for ex¬

amination and infusion, and information regarding all intercurrent infections (bacterial, viral, and opportunistic) was routinely collected. Prophylaxis against Pneumocystis carinii pneumonia (PCP) on a

3-consecutive-day-per-week regi¬

and use of antiretroviral therapy at any time after study entry was permit¬ ted according to the prevailing medical standard of care as determined by the patient's physician with consent of the men

parent

or

guardian.

Definition of Infectious Outcomes For the analysis of minor bacterial infections, serious and minor viral in¬ fections, and opportunistic infections; if an organism was isolated by culture, detected by antigen or histologie find¬

ings, or diagnosed by specific antibody testing, it was classified as laboratory proven. If the evaluating physician re¬ ported a clinical syndrome thought to be consistent with a bacterial, viral, or op¬ portunistic origin (ie, episodes of otitis media, herpes stomatitis, or oral candidiasis, respectively), the episode was classified as clinically diagnosed.

Serious bacterial infections were clas¬ sified as described previously.19 Clini¬ cally diagnosed serious infections were defined as episodes of acute pneumonia or sinusitis without defined microbio¬ logie origin that met predetermined clin¬ ical and radiologie criteria21; all episodes underwent central, blinded, independent review by two physicians for classifica¬ tion as an acute laboratory-proven, an acute clinically diagnosed, or a "notacute" episode. Bactériologie confirma¬ tion was required for episodes of men¬

ingitis, bacteremia, osteomyelitis, tic

arthritis,

scess

acute

mastoiditis,

or

sep¬ ab¬

of internal organ.

Statistical Analysis The analysis of treatment effect in terms of minor bacterial, viral, and op¬ portunistic infections was complicated by the occurrence of multiple and re¬ curring infections in individuals. There¬ fore, rates of infection were calculated (separately for IVIG and placebo arms) as the number of infections reported for all patients divided by the total number of months of observation for all patients. These rates were then standardized to represent rates per 100 patient-years. The jackknife procedure was used to


Table 1.—Characteristics of Study Children With Entry CD4* Lymphocyte Count ==0.20x109/L According to Treatment Arm Characteristic No. of patients Mean follow-up, mo Mean age at entry, mo Age .10 for comparison of rate per 100 patient-years, iVlG vs placebo, for minor bacterial, serious and minor viral, and opportunistic infections; data not shown). This is similar to what was previously observed for serious bacte¬ rial infection in this patient group.19

Efficacy of IVIG Minor Bacterial Infections.—There were 649 minor bacterial infections ob¬ served in patients with entry CD4+ counts of at least 0.20xl09/L, the vast majority (89%) of which were clinically diagnosed. Fifty-nine percent of minor bacterial infections involved the ear; 13%, skin and soft tissue; 10%, upper respiratory tract; and the remainder of sites accounted for less than 5% each. Of the 577 clinically diagnosed minor bac¬ terial infections, 261 occurred in 93 IVIG patients compared with 316 in 98 pla¬ cebo patients; of the 72 laboratoryproven minor bacterial infections, 27 oc¬ curred in 22 IVIG patients compared with 45 in 30 placebo patients. Significantly fewer minor bacterial in¬ fections per 100 patient-years were ob¬ served in children with entry CD4+ counts of at least 0.20xl09/L receiving IVIG (P=.02; 115.1 compared with 159.7 episodes per 100 patient-years in IVIG and placebo arms, respectively) (Table 2). Intravenous immunoglobulin was as¬ sociated with a 36% decrease in the rate of ear infections per 100 patient-years (63.5 vs 99.5 infections per 100 patientyears for IVIG vs placebo, respectively); a 39% decrease in the rate of skin and soft-tissue infections per 100 patientyears (14.0 vs 23.0 infections per 100 patient-years for IVIG vs placebo, re¬ spectively); and a 32% decrease in the rate of upper respiratory tract infec¬ tions per 100 patient-years (10.8 vs 15.9 infections per 100 patient-years for IVIG vs

placebo, respectively) (Table 3).

There was a similar frequency of grampositive and gram-negative organisms observed in the laboratory-proven infec¬ tions. Intravenous immunoglobulin was associated with a 48% decrease in the rate per 100 patient-years of minor grampositive infections (4.4 compared with 8.4 episodes per 100 patient-years in IVIG and placebo arms, respectively) and 38%

Table 2.—Number and Rate of Infections by Infection Type in Study Patients With Count > 0.20 x 109/L Overall and According to Treatment Arm Overall

Type of Infection

(n=313)*

Immunoglobulin (n=161)f

No.of

Rate per 100 Patient-

No. of

Episodes

years§

Episodes

175

36.7

66

Serious bacterial

Laboratory-

Rate per 100

10.1

proven

Clinically diagnosed

Patient-

No. of

years§ 26.4II

Episodes

Patient-years

109

48.2

6.0

33

14.6


33.6 159.7

10.8

45

19.9

104.3

316

288

115.1H

72

15.1

27

577

121.1

261

212

44.5

90

proven

31

6.5

Clinically diagnosed Opportunistic Laboratory-

55

Viral

76

136.3

proven

Laboratory-

38.0 11.5 2.3

proven


44

36.0#

122

54.0

4.4

20

8.8

31.6

102 30

45.1

10.0

25

Rate per 100

361

649

Laboratory-

( =152)

Placebo

r

20.4

Minor bacterial

Entry CD4* Lymphocyte

13.3

1.6

9.2

21

3.1

8.4

10.2

23

Total of 476.3 patient-years follow-up.

tTotal of 250.2 patient-years follow-up. Total of 226.1 patient-years follow-up. §Totals may not sum due to rounding. ||P=.002 comparing rate per 100 patient-years, immunoglobulin vs placebo. TlP=.02, comparing rate per 100 patient-years, immunoglobulin vs placebo. #P=.01, comparing rate per 100 patient-years, Immunoglobulin vs placebo. Table 3.—Number and Rates of Minor Bacterial Infections by Site and Organism in Entry CD4* Lymphocyte Count == 0.20x109/L According to Treatment Arm

of Minor Bacterial

Infection

(n=161)*

INo. of

Episodes

Patients With

Placebo

Immunoglobulin Type

Study

(n=152)t

Rate per 100

Patient-years^

No. of

Rate per 100

Episodes

Patient-years}:

Site Ear

63.5

159

225

99.5

Skin/soft tissue

35

Upper respiratory Eye Node/gland

27 18

7.2

3.5

Genitourinary Lower respiratory

14

5.6

3.5

3.6

3.5

Gastrointestinal

3.2

3.5

Orodontic Other

0.4

0.4

52 10.8

17

Multiple organisms Gram-positive Streptococcus pyogenes Staphylococcus aureus

15.9 12

Organism

10 8.4

4.4

10 1.2

2.7

0.4

Enterococci

Streptococcus

0.9

viridans

Clostridium difficile

0.4 0.4

Gram-negative

16

Escherichia coli

5.3

7.1

2.8

2.2

Klebsiella

0.4

0.4

Salmonella

0.4

0.9

0.4

0.4

Pseudomonas aeruginosa

species species Campylobacter species Haemophilus influenzae

Total of 250.2 Total of 226.1

Totals may

patient-years follow-up. patient-years follow-up. rounding.

not sum due to


fewer minor

gram-negative

infections

(4.4 vs 7.1 episodes per 100 patient-years

for IVIG vs placebo, respectively). Viral Infections.—Two hundred twelve viral infections were observed in patients with entry CD4+ counts of at least 0.20X109/L. Thirty-three percent involved the respiratory tract, 26% were mucocutaneous, 27% were generalized, and 11% were gastrointestinal. Thirtyone infections (15%) were laboratoryproven, 16 (52%) of which were due to herpes simplex virus. An additional 34 infections with herpes simplex virus were clinically diagnosed. Ofthe 181 clin¬ ically diagnosed viral infections, 79 oc¬ curred in 54 IVIG patients compared with 102 in 64 placebo patients. Of the 31 laboratory-proven viral infections, 11 occurred in eight IVIG patients com¬ pared with 20 in 15 placebo patients. Significantly fewer viral infections per 100 patient-years occurred in children with entry CD4+ counts of at least 0.20X109/L receiving IVIG (P=.01; 36.0 compared with 54.0 episodes per 100 patient-years in IVIG and placebo arms, respectively) (Table 2). Intravenous im¬ munoglobulin was associated with de¬ creases in rates per 100 patient-years of 50% for mucocutaneous infections (8.0 vs 15.9 infections per 100 patient-years for IVIG vs placebo, respectively) and 43% for generalized viral infections (8.8 vs 15.5 infections per 100 patient-years for IVIG vs placebo, respectively) (Table 4). Res¬ piratory tract and gastrointestinal viral infections were only modestly decreased. Forty-five percent fewer laboratoryproven or clinically diagnosed infections due to herpes simplex virus per 100 pa¬ tient-years were observed in IVIG re¬ cipients (7.6 vs 13.7 episodes per 100 patient-years for IVIG vs placebo). All of the reported cases of measles (five), entero viral infections (two), and hepa¬ titis A (one) occurred in placebo patients. The incidence ofinfection with varicellazoster virus was similar between treat¬ ment

arms.

Opportunistic Infections.—Fifty-five opportunistic infections occurred in pa¬

tients with entry CD4+ counts of at least 0.20X109/L. The majority (80%) of in¬ fections were mucocutaneous, followed by respiratory (7%), systemic and gas¬ trointestinal (5% each), and central ner¬ vous system (2%). Eleven infections (20%) were laboratory-proven, four in four IVIG patients and seven in seven placebo patients; there were 21 clini¬ cally diagnosed infections in 14 IVIG patients and 23 episodes in 14 placebo patients. Candida organisms accounted for 73% of laboratory-proven or clini¬

cally diagnosed opportunistic infections, by carinii (7%); each other organism accounted for 5% or less.

followed

Table 4.—Number and Rates of Viral Infections by Site and Organism in Lymphocyte Count £ 0.20 x109/L According to Treatment Arm

Study

Immunoglobulin (n=161)* Type of

Viral Infection

No. of

Placebo

Rate per 100

Patient-yearst

Episodes

Patients With

No. of

Entry

CD4*

(n=152)t Rate per 100

Episodes

Patient-years^

Site

Respiratory tract

36

14.4

15.5

Mucocutaneous

Generalized Gastrointestinal Central nervous system

22

15.9

35

15.5

13

5.7 0.4

0.4

Node/gland Herpes simplex Laboratory-proven Clinically diagnosed Varicella-zoster§ Laboratory-proven Clinically diagnosed

36

19

0.9

Organism

7.6

13.7

2.0

4.9

14

5.6

20

12

4.8

13

5.7

0.4

0.4

4.4

5.3 2.2

Measles

Laboratory-proven Clinically diagnosed Rotavirus

1.2

0.4

Laboratory-proven Clinically diagnosed Respiratory syncytial virus!

0.8

0.4

Cytomegalovirusll

0.4

0.4 0.9 0.9

Enterovirus

Laboratory-proven Clinically diagnosed Hepatitis A|| MolluscumH

1.3

0

"Total of 250.2 patient-years follow-up. -fTotal of 226.1 patient-years follow-up.

JTotals may not sum due to rounding.

§Of those patients treated with immunoglobulin, nine had primary infection (chickenpox) and three had recurrent infection (zoster). Of those patients receiving placebo, 10 had chickenpox and three had recurrent zoster.

IIAII reports laboratory-proven. TIAII reports clinically diagnosed.

There was a similar frequency of op¬ portunistic infections per 100 patientyears in IVIG and placebo patients (Ta¬

ble 2). A modest decrease in the rate of mucocutaneous infection and infection due to Candida species was observed with IVIG (Table 5). While infections with other organisms were too few to make meaningful comparisons between treatment arms, there did not appear to be a trend toward a treatment difference. Serious Bacterial Infections.—With extension of patient follow-up through study termination, IVIG demonstrated a significant effect reducing serious bacte¬ rial infections, as in our prior publication. The rate of laboratory-proven or clini¬ callydiagnosed serious bacterial infection per 100 patient-years decreased by 45% in IVIG recipients (P=.0O2) (Table 2). Twenty-seven percent of serious bac¬ terial infections were laboratory-proven; S pneumoniae was the most frequent isolate (38%). The rate of primary bacteremia was diminished by 68% in IVIG recipients (4.0 episodes per 100 patientyears in IVIG compared with 12.4 epi¬ sodes in placebo patients). Acute pneu-


monia

the most common clinically serious bacterial infection. There was a 58% decrease in the rate of acute clinically diagnosed pneumonia per 100 patient-years in IVIG recipients (12.0 episodes per 100 patient-years in IVIG compared with 25.7 episodes in was

diagnosed

placebo patients).

Concurrent Therapy With Trimethoprim-Sulfamethoxazole for PCP Prophylaxis and Zidovudine.—Detailed analysis has been performed to evaluate the potential confounding effects of con¬

comitant therapy with zidovudine or trimethoprim-sulfamethoxazole, adminis¬ tered 3 days per week for PCP prophy¬ laxis, on the observed efficacy of IVIG in reducing serious and minor bacterial and viral infections.23 As of January 1991, when the study ended, 50% of study patients with entry CD4+ counts of at

least 0.20X 109/L had received PCP pro¬ phylaxis (50% of placebo patients and 49% of IVIG patients) and 45% zidovu¬ dine (47% of placebo and 44% of IVIG patients). Time to initiation of either therapy was similar in both treatment arms. Assessment of relative risk, com-

Table 5.—Number and Rates of Opportunistic Infections by Site and Organism in Entry CD4* Lymphocyte Count sO.20 x109/L According to Treatment Arm

Immunoglobulin (n=161)* Type of Opportunistic Infection

No. of

Rate per 100

Patient-years

Episodes

Study

Placebo

r No. of

Episodes

Patients With

(n=152)t Rate per 100

Patient-years

Site Mucocutaneous

8.0

20

24

10.6

Respiratory Systemic

1.2

0.4

Gastrointestinal

0.4

0.9

Central

nervous

1.3

system

Multiple organisms Candida species Laboratory-proven Clinically diagnosed Pneumocystis cariniit Tinea§ Histoplasma capsulatura Mycobacterium aviumtW Toxoplasma gondii§ Giardia lamblia^

Organism

0.4

0.4

16

6.4

24

10.6

16

6.4

21

9.3

1.3 1.2 1.2

0.4

0.4 0.9 0.4

1

0.4

"Total of 250.2 patient-years follow-up. (Total of 226.1 patient-years follow-up.

reports laboratory-proven. §AII reports clinically diagnosed. [¡Isolates from gastrointestinal tract and

bone

marrow.

paring IVIG and placebo arms, for se¬

rious or minor bacterial and viral infec¬ tion by person-time receiving or not re¬ ceiving zidovudine therapy or PCP pro¬ phylaxis found no significant difference in the risk ratio (IVIG vs placebo) be¬ fore and after receipt of either therapy (Ps.15 for zidovudine and P>.19 for PCP prophylaxis). In addition, propor¬

tional hazards modeling analysis (using a time-dependent covariate analysis) in¬ dicated that neither PCP prophylaxis nor zidovudine therapy accounted inde¬ pendently for the benefit observed with IVIG.23 Linear and logistic regression analyses produced similar results. CONCLUSIONS While recurrent serious bacterial in¬ fections caused significant morbidity in HIV-infected children, minor bacterial and viral infections contributed more fre¬ quently to morbidity (together five times as frequently) than did serious bacterial infections. Beneficial effect of IVIG was noted in children with entry CD4+ counts of at least 0.20xl09/L across multiple in¬ fection parameters, including serious and minor viral and bacterial infections. Viral infections, a measure that would appear not to be confounded by ques¬ tions of antibiotic use, accounted for al¬ most 20% of observed infections in this report. Intravenous immunoglobulin therapy was associated with a signifi¬ cant decrease in overall viral infections in children with entry CD4+ counts of at least 0.20X 109/L, particularly for mucocutaneous and generalized viral infec¬ tions. Viral infections for which augmen-

tation of humoral immunity is preven¬ tive were decreased with IVIG: no cases of measles or hepatitis A were seen in IVIG recipients. While cellular immu¬ nity is thought to play the major role in prevention of recurrence of latent herpes simplex infection, recent data suggest that antibody-dependent cellular cytotoxicity may also play a significant role in protection.24·25 Consistent with a pos¬ sible role for humoral antibody in herpes simplex reactivation, a 45% reductipn in herpes simplex infections was noted in IVIG recipients with entry CD4+ counts of at least 0.20xl09/L. It might be an¬ ticipated that such a protective antibody effect would depend on a relatively in¬ tact cellular immune system, as might be seen in those children with CD4+ counts over 0.20 X107L. Recurrent minor bacterial infections such as otitis media are common in HIVinfected children. In a prospective co¬ hort study of 27 HIV-infected children and matched uninfected controls, the number of children diagnosed with acute otitis media was similar between the HIV-infected and uninfected control groups, but the HIV-infected children had significantly more episodes of oti¬ tis.11 The CD4+ lymphocyte count and CD4+/CD8+ ratio were not significantly different between HIV-infected children with or without otitis episodes, suggest¬ ing that an early, subtle immune deficit, occurring prior to CD4+ cell depletion, may predispose HIV-infected children to recurrent bacterial infections. A sequential loss in T-helper lympho¬ cyte function, occurringprior to CD4+ cell


depletion, has been described in HIVinfected adults and children.26·27 Studies evaluating T-helper cell function by measuring T-cell proliferation and interleukin-2 production in response to different types of antigenic stimuli have demonstrated progressive loss of re¬ sponse, first to recall antigen stimulus, followed by loss of response to alloantigen and finally to mitogen stimulus. These more subtle defects in T-helper function have been correlated with the slope of CD4+ cell decline over time28 and with risk for infections.26 In a study of pediatrie HIV-infected patients, a CD4+ count un¬ der 0.20xl09/L was significantly associ¬ ated with loss of response to all antigenic stimuli, while those children with no de¬ fects were more likely to have CD4+ counts over 0.20x 109/L (P