Immunogenicity and Safety of Meningococcal C Conjugate Vaccine in Children and Adolescents Infected and Uninfected with Human Immunodeficiency Virus in Rio de Janeiro, Brazil

Abstract

Background

We aimed to evaluate the Neisseria meningitidis C conjugated vaccine (MCC) seroconversion and adverse events (AE) in HIV-infected and uninfected children and adolescents in Rio de Janeiro, Brazil.

Methods

HIV-infected or uninfected subjects, 2–18 years old, with CD4+ T-lymphocyte cell (CD4) percentage >15%, without active infection or antibiotic use, were enrolled. All patients were evaluated before and 1–2 months after immunization for seroconversion (defined as ≥4-fold titer increase in human serum bactericidal activity), and for AEs at 20 minutes, 3 and 7 days after immunization. Factors associated with seroconversion among HIV-infected group were studied.

Results

204 subjects were enrolled: 154 HIV-infected and 50 HIV-uninfected. Median age was 12 years and 53% were female. Among the HIV-infected group, 82 (53%) had a history of at least one C clinical category Centers for Diseases Control and Prevention event, and 134 (87%) were using combination antiretroviral therapy (cART). The median nadir CD4 percentage was 13% (0%–47%). 76(37.3%) experienced mild AEs. Seroconversion occurred in 46/154 (30%) of the HIV-infected group, and in 38/50 (76%) of the uninfected group (p<0.01). Factors associated with seroconversion in the HIV-infected group were: Never had a C clinical category event (OR=2.1, 95%CI=1.0–4.4); undetectable viral load at immunization (OR=2.4, 95%CI=1.1–5.2), and higher CD4 nadir/100 cells (OR=1.1, 95%CI=1.0–1.2).

Conclusion

MCC vaccine should be administered to HIV-infected children and adolescents after maximum immunologic and virologic benefit has been achieved with cART. Our data suggest that a single dose of MCC vaccine is insufficient for HIV-infected individuals 2–18 years of age.

INTRODUCTION

In late 2010, the World Health Organization (WHO) estimated that 34 million people were living with HIV worldwide, including 3.4 million children under 15 years of age (1). In Brazil, 16,935 HIV-infected children were reported from 2000 to June 2012 (2). The Brazilian AIDS program was implemented in 1996 and provides universal, free-of-charge outpatient HIV treatment, including antiretroviral drugs, vaccines, and diagnostic and prevention activities. Although the program resulted in decreased HIV related mortality, nearly 20,000 AIDS-related deaths occur annually in Brazil, mainly because of opportunistic infections (2–5).

HIV infection leads to a progressive loss of humoral and cellular immune functions and increases the risk of infection, notably from encapsulated bacteria (6). Studies from developed countries have demonstrated an association between HIV and meningococcal disease (7–10), although some African studies did not corroborate this finding (11–13). The studies that did not find an association were conducted before the implementation of highly active antiretroviral therapy, which could have influenced the results. Recently, Cohen et al. (14) demonstrated a 60-fold higher risk of meningococcal disease among HIV-infected children as compared with HIV-uninfected children in South Africa. In the same study, it was shown that HIV-infected individuals had almost three times higher odds of developing bacteremia than HIV-uninfected individuals. Additionally, Miller et al (15) also found an increased risk of invasive MD in adolescents and adults, in New York.

In Brazil, meningococcal disease (MD) is endemic, with periodic outbreaks (16). Incidence rates have been stable in recent years, with 1.4–2.5 reported cases per 100,000 inhabitants (16). Approximately 40–50% of cases were reported in children under 5 years old, with the highest incidence in infants (17). In some states, a higher incidence rate among adolescents has also been observed (18), similar to reports from European and North American countries (19,20). Since 2000, Brazil has experienced an increase in serogroup C MD, which now accounts for 81.5% of reported cases (17). Meningococcal serogroup C conjugate (MCC) vaccines have been supplied by the public sector for control of outbreaks and for high-risk patients, including HIV-infected children under 13 years old, since 2006. Immunization is given as a single dose at specialized reference centers (21).

MCC vaccine has been shown to be safe and immunogenic in many high-risk populations, with results depending on the degree of immunosuppression (22–24). There are few data about the use of MCC vaccine in HIV-infected children and adolescents (25). The primary aim of this study was to evaluate short-term immunogenicity following administration of a single dose of MCC vaccine in HIV-infected children and adolescents at a reference center in Rio de Janeiro, Brazil, and to compare their response to HIV-uninfected subjects. We also sought to evaluate factors associated with vaccine safety.

MATERIALS AND METHODS

Study design and population

We conducted a prospective cohort study in HIV-infected and uninfected children and adolescents assisted at the Instituto de Puericultura e Pediatria Martagão Gesteira (IPPMG), a tertiary-care hospital of the Universidade Federal do Rio de Janeiro, Brazil. IPPMG is a pediatric reference center for infectious diseases and has been also a reference center for care, research, and training related to pediatric HIV/AIDS since 1989.

HIV infected and uninfected children and adolescents were eligible for inclusion if they were 2–18 years old; had never been immunized with any meningococcal conjugate vaccine; had not received a live vaccine within 4 weeks before study entry; did not plan to receive other vaccines within 2 weeks after entry; had no signs, symptoms or diagnosis of any other immunosuppressive disease (not HIV infection related); had not used systemic immunosuppressive drugs; had not used antibiotics up to 3 weeks before the immunization or immunoglobulin therapy within the last six months; had no history of bleeding disorders or adverse reactions to any vaccine components; and had no psychiatric disorder, including illicit drug or alcohol intoxication at the time of the interview (patient or legal guardian). For female participants with Tanner stage ≥3 and/or history of menarche, a negative pregnancy test was required before immunization. For the HIV-infected group, additional eligibility criteria were: HIV infection, absence of severe or advanced HIV clinical disease at entry (WHO clinical stage 3 or 4), and CD4+ T-lymphocyte cell (CD4) count at or above 350 cell/mm3 and/or 15% at study entry. HIV-uninfected individuals must have had a negative HIV serologic test after 18 months of age, and for those who were sexually active, the same test was repeated at the study entry to confirm that serologic status. The HIV-uninfected group was recruited from a healthy child clinic at IPPMG. Children and adolescents in both groups lived in the same geographic area.

Individuals were enrolled between February 2011 and December 2012 to receive one dose (0.5 mL intramuscular injection) of conjugated meningococcal C oligosaccharide-CRM197, a toxoid of Corynebacterium diphtheria (Chiron/Novartis Vaccines, Siena, Italy). For HIV-infected volunteers, the vaccine was provided by the Brazilian National Ministry of Health; for HIV-non-infected group, it was provided by the investigators, through the research grant.

Demographic, clinical, virologic and immunologic data were obtained from the medical record. Standardized questionnaires involving clinical, laboratory, and socioeconomic data were administered at entry and 1–2 months after immunization.

Investigators observed all volunteers 20 minutes after the immunization. Axillary temperature, local reactions (graded by size in millimeters) and systemic symptoms were recorded. Subjects and their caregivers were trained to measure these parameters and were contacted by telephone at 3 and 7 days after immunization to provide information about these adverse events (AEs). AEs were classified according to a standardized grading system (26).

Sample Size Considerations

Studies in children and adolescents HIV-infected have indicated that immunologic response to other conjugate vaccines was 60% or lower (27, 28); in contrast, healthy children in United Kingdom had an antibody response rate of at least 80% to MCC (29). Accordingly, a sample size of 150 HIV-infected individuals and 50 uninfected individuals was required to detect this difference in immunological response of at least 20% between groups, considering an alpha-error of 0.05, and a beta-error of 0.20.

Immunogenicity Assays

The serum bactericidal activity using human serum as the complement source (hSBA) was determined before immunization and 1–2 months after immunization (30). The positive control for each assay consisted of a pool of post-vaccination human sera with previously determined hSBA titers. The negative control consisted of the complement source in the absence of test serum. Two local strains of serogroup C Neisseria meningitidis were used as target strains: N79/96 (C:2b:P1.10) and N753/00 (C:23:P1.14-6), both provided by Adolfo Lutz Institute Bacteriology Section in São Paulo, Brazil. These strains were from Brazilian patients with invasive MD, with N753/00 representing the current prevalent phenotype in Brazil. The hSBA assays were performed at the Department of Microbiology, Immunology and Parasitology, Universidade do Estado do Rio de Janeiro, Brazil.

Seroprotection was defined as an hSBA ≥1:4 and participants were classified as previously immune if the baseline hSBA was ≥1:4 (31). Seroconversion was defined as a ≥4-fold increase in hSBA titer after vaccination.

Statistical methods

Continuous and categorical demographic characteristics were compared between HIV infected and non-infected individuals using Mann Whitney and Fisher’s exact tests, respectively.

To assess possible short-term predictors of hSBA response among HIV-infected individuals, individuals were categorized in two groups: individuals with a 4-fold increase in hSBA titer after immunization (responders) and those who did not (nonresponders). The two groups were compared for possible risk factors for immunologic response, including age; use of combination antiretroviral therapy (cART); nadir CD4 count and percentage; peak viral load; CD4 count; CD4 percentage; and HIV viral load at immunization. Variables with p-value ≤0.15 were considered for inclusion in a logistic regression multivariate analysis model. Interactions were assessed using the –2 log likelihood ratio test and models with and without interactions were evaluated. Two-tailed tests were used in all analyses.

All analyses were performed by using STATA software (version 2.0; Stata Corp LP, College Station, TX, USA).

Ethical Issues

The study was approved by IPPMG institutional review board (IRB) and the Brazilian Ministry of Health Ethics Commission (CONEP). Participants and/or their parents/guardians provided informed consent before participation; assent was obtained as required by the IPPMG IRB.

RESULTS

Study population

We enrolled 161 HIV-infected and 50 HIV-uninfected. Five subjects in the HIV-infected group were later found to be ineligible: four patients with CD4 counts below the entry threshold and another with prior receipt of MCC vaccine. Two other HIV-infected patients were lost to follow-up, leaving 204 participants for inclusion in the immunogenicity analysis.

At study entry, 53% of participants were female with no differences between HIV-infected or uninfected groups (p-value = 0.12). The median age was 12 years-old (range 2–18): 12.2 for HIV-infected group and 10.3 for HIV-uninfected group (p-value <0.01). There were no differences in median weight (34 kg for HIV-infected vs. 33 kg for HIV-uninfected subjects) or height (141.5 cm vs. 141.1cm). However, there were differences between age-and-gender adjusted z-scores for weight (0.16 in HIV-uninfected group and −0.98 in HIV-infected group, p<0.01) and for height (0.35 in HIV-uninfected group and −1.18 in HIV-infected group, p<0.01). A large proportion of participants in the HIV-infected group (87%) were receiving cART and 53.3% had a history of CDC Class C conditions (Table 1).

Table 1.

Baseline characteristics of HIV-infected subjects by immune response to meningococcal C conjugate vaccine at week 4–8 (N = 154).

Characteristic	Total N = 154	Non-responder N = 108	Responder N = 46	p-value*
Age-median, years (IQR#)	12 (8 to15)	13 (10 to 15)	12 (7 to 15)	0.27
Gender (male)	71 (47%)	54 (50%)	17 (37%)	0.16
Weight-median, kilograms (IQR#)	33 (24 to 48)	36 (26 to 48)	31 (21 to 49)	0.38
Height-median, centimeters (IQR#)	142 (125.5 to 155.5)	141.5 (128 to 154)	142 (118.8 to 156)	0.47
Weight age and gender adjusted z-score - median (IQR#)	−0.88 (−1.81 to −0.19)	−0.89 (−1.82 to −0.24)	−0.80 (−1.48 to −0.04)	0.38
Height age and gender adjusted z-score - median (IQR#)	−1.09 (−1.93 to −0.36)	−1.20 (−2.06 to −0.45)	−0.84 (−1.55 to −0.20)	0.12
Number of individuals living in the same house per individual – median (IQR#)	3. (2 to 5)	3 (2 to 4)	4 (3 to 5)	0.12
History of previous meningococcal polysaccharide C immunization	9 (5.8%)	8 (7.4%)	1 (2.2%)	0.27
Never had a C clinical category event	72 (46.7%)	45 (41.7%)	27 (58.7%)	0.08
On cART& at entry	134 (87%)	93 (86.1%)	41 (89.1%)	0.43
Age at cART& initiation-median, years (IQR#)	5 (2 to 9)	6 (2 to 9)	3 (1 to 9)	0.09
Years in cART& - median, years (IQR#)	6 (4 to 9)	6 (4 to 9)	6 (3 to 8)	0.63
Immunological and virologic variables
Pre-cART&
CD4 count - median (IQR#), cells/mm3	301 (73.5 to 543)	284 (67.8 to 518)	321.5 (221.5 to 840)	0.03
CD4 percentage - median (IQR#)	13 (6 to 17)	12 (5 to 16)	15 (9 to 21)	0.04
CD8 count median (IQR#), cells/mm3	659 (202.5 to 1156)	654.5 (267.8 to 1150.8)	798.5(536.8 to 1225.3)	0.32
CD8 percentage median (IQR#)	40 (29 to 51.5)	44 (32.3 to 53.8)	36.5(27 to 45.5)	0.05
Peak viral load median (IQR#), copies/ml	233,143 (66,252 to 830,000)	221,500 (67,700 to 820,000)	330,000 (119,250 to 1122,750)	0.37
At entry
CD4 count median (IQR#), cells/mm3	686 (527 to 1049)	657.5(526 to 1026.8)	854.5 (594.5 to 1102.5)	0.05
CD4 percentage-median (IQR#)	27 (22.5 to 33)	26.5(23 to 32)	29.5 (26 to 35.5)	0.02
CD8 count median (IQR#), cells/mm3	995 (812.5 to 1397)	1032.5 (852.5 to 1407)	1001 (770 to 1468.5)	0.67
CD8 percentage at entry-median (IQR#)	40 (32 to 46)	41 (36.3 to 47)	38 (32 to 45)	0.06
Undetectable viral load	88 (57.1%)	56 (51.9%)	32 (69.6%)	0.05
Adverse events
Systemic adverse events 7 days post-immunization	32 (20.8%)	27 (25%)	5 (−10.9%)	0.05

^*p value for comparison between the groups, using Mann-Whitney test for continuous variable, and Fisher exact test for categorical variables.

^#IQR = interquartile interval

^&cART = combination antiretroviral therapy

Immunogenicity

hSBA assays with the two local strains showed similar and consistent results. At enrollment, 33/204 (16.2%) of participants had baseline immunity (hSBA titer ≥1:4). The percentage of patients with hSBA titers ≥1:4 pre-immunizaton was higher in the HIV-uninfected group than in the HIV-infected group (30% vs. 11.8%, p-value = 0.02).

Seroprotection at 1–2 months post-immunization was achieved in 104 subjects (51%) and the rate was significantly lower in the HIV-infected group (45.5% vs. 84%, p-value < 0.01). Seroconversion was observed in 84 subjects (41%): 46/154 (30%) in the HIV-infected group versus 38/50 (76%) in the HIV-uninfected group (p-value < 0.01). This difference remained even when subjects with baseline immunity were excluded (43/134 vs. 26/35, 32% vs. 74%, respectively - p-value < 0.01). The geometrical mean bactericidal titers were 13.9 (95% CI=10.2–18.9) overall and were lower in the HIV-infected group (8.3;95%CI=6.1–11.3) than in the HIV-uninfected group (45.2;95%CI=24.6–83.3) (p-value <0.01).

Among HIV-infected patients, there was no association between immune response and demographic characteristics, use of cART at entry, or history of meningococcal polysaccharide vaccination (Table 1).

Predictors of hSBA response in the HIV-infected group

In the multivariable logistic regression model, significantly higher response rates to MCC were found in subjects with Never had a C clinical category event (OR: 2.1; 95% CI:1.0–4.4, P = 0.04), higher CD4 count/100 cells precART (nadir) (OR: 1.1; 95% CI:1.0–1.2, P = 0.04) and viral load undetectable at study entry (OR: 2.4; 95% CI:1.1–5.2, P = 0.02).

Safety

Of 204 participants, 76 (37.3%) experienced any AE; 38 (18.6%) reported immediate AEs, 32 local and 10 systemic reactions (Table 2). AEs were reported in 42.9% and 20% of HIV-infected and HIV-uninfected groups, respectively (p-value < 0.01). All local AEs were mild (grade 1), predominantly pain and tenderness at the injection site. No participants developed a grade 3 or higher AE.

Table 2.

Adverse events to meningococcal C conjugate vaccine immunization within 20 minutes, and at 3 and 7 days (N = 204).

Adverse event*		HIV infected group (n = 154)	HIV-uninfected group (n = 50)	p-value
Immediate - total		30 (19.48%)	8 (16%)	0.78
Local reaction		24 (15.58%)	8 (16%)	0.84
Systemic reactions		10 (6.49%)	0	0.17
Within 3 days - total		42 (27.27%)	3 (6%)	< 0.01
Fever		5 (3.25%)	0	0.45
Local reaction		32 (20.78%)	3 (6%)	0.07
Systemic reactions		16 (10.39%)	0	0.07
Within 7 days - total		15 (9.74%)	1 (2%)	0.09
Fever		2 (1.30%)	0	0.72
Local reaction		4 (2.60%)	1 (2%)	0.84
Systemic reactions		13 (8.44%)	0	0.09
Total		66 (42.86%)	10 (20%)	< 0.01

^*All local reactions were mild (grade 1). No grade ≥ 3 adverse events were reported.

DISCUSSION

In this study, a single dose of MCC was safe but poorly immunogenic in HIV-infected children and adolescents. We observed that only 30% of HIV-infected children and adolescents responded to MCC vaccine as compared to 76% of the HIV-uninfected subjects. In addition, only 45.5% of HIV-infected patients exhibited seroprotection 1–2 months post-immunization.

Our study was consistent with other studies that demonstrated that MCC vaccine was safe and immunogenic in HIV-uninfected individuals (29,32,33), although the seroconversion rate was lower than previously described in healthy children (84–99%)(29,32) and adolescents (85–100%) (25,34). The reasons for this lower response rate among HIV-uninfected population in our study are unknown. However, genetic and immunological differences between populations or pharyngeal carriage rates could have affected the immune response (35). Finally, the use of local meningococcal strains for the hSBA assay could have contributed to the observed differences (35).

Immunologic response to non-live vaccines in HIV-infected populations have generally been lower and less durable as compared with HIV-uninfected populations (28). Conjugate vaccines to encapsulated bacteria such as Haemophilus influenzae type b and Streptococcus pneumoniae have induced lower geometric mean titers in HIV-infected children as compared to exposed, uninfected children (36,37). Although MCC was very effective in other populations with impaired immune response (22–24) and recommended in guidelines for HIV-infected patients (38–41), meningococcal conjugate vaccines have only been recently evaluated in HIV-infected individuals (25,42–44) with only one study evaluating MCC vaccine (25). Data from a study that analyzed immunogenic response to a single dose of quadrivalent A, C W and Y meningococcal polysaccharide conjugate vaccine (MCV4) in HIV-infected and uninfected population also demonstrated lower immunogenic response rates to the serogroup C component in HIV-infected subjects 2–10 years old (43%) (43) and 11–24 years old (52%) (42). This response rate was also lower than that obtained in studies with healthy subjects (45,46) and suggested that the vaccine was somewhat more immunogenic for serogroup C in adolescents than in preadolescents and younger children (42,43). Our study demonstrates that the poor immunity after a single dose of MCV4 vaccine was not due solely to the multivalent characteristic of the vaccine, since we observed similar response to a single dose of MCC.

In a recent Brazilian clinical trial, Bertolini et al also described poor immunogenicity in HIV-infected adolescents to a single dose of MCC (72.1%) (25). Our study differed from that study in that ours had a stricter inclusion criterion for CD4 count (CD4 count >350 for our study vs. 100 for Bertolini et al). We included a younger population and we used human complement for our SBA assay, whereas Berolini et al used rabbit complement, which may be more sensitive, but less specific, for estimating protective immunity than assays that use human complement (47,48). HIV-infected subjects in our study also presented higher CD4 count and viral load at entry and most non-responders subjects had previous CDC classification C, indicating a probable higher immune activated group. Residual immune activation was described even in patients with viral control after cART and may interfere with immunization response (49). We found in another study that, among 36 patients that were include in this study, vaccine response failure was associated with higher frequency of CD4 activation (50).

Our study population had low socioeconomic status and crowded living condition, therefore it could be related with a higher meningococcal pharyngeal colonization rate and possibly baseline immunity. The lower baseline immunity observed in the HIV-infected group might be related to impaired development of natural immunity because of HIV infection. However, HIV non-infected subjects may have been more exposed to natural infection, since they had a higher median number of individuals living in the same household (5 vs. 4 individuals), so they would have been expected to present a higher baseline immunity rate. The role of baseline immunity on the induction of immune response to meningococcal vaccines in HIV-infected patients is still unclear, but studies analyzing immunogenicity in HIV-infected adolescents after a single dose of MCV4 suggested that meningococcal colonization reduces immune response (42) We do not know to what extent the rate of pharyngeal colonization may have affected these results.

Factors independently associated with immune response to MCC were the absence of a CDC clinical C event history, higher CD4 count /100 cells nadir and undetectable viral load at immunization. MCV4 immunogenicity studies in HIV-infected adolescents and young adults also found an association of lower CD4 percentage (<25%), detectable viral load and more advanced HIV disease stage with lower vaccine response at 4 weeks across all 4 serogroups contained in the vaccine (42) as has been described with response to other vaccines in HIV-infected children (27, 28 36,51–53). These results reinforce the importance of early antiretroviral therapy for a better response to immunizations.

AEs in our study were mild and more frequent in the first 3 days after vaccination. The nature and frequency of AEs were similar in both groups, although the overall incidence was higher than those reported in healthy children and adolescents (34). In contrast to the results of the Bertolini study that demonstrated a higher incidence of AEs in the HIV-uninfected group (25), our results showed a higher rate of AE after 3 days post-immunization among HIV-infected subjects. Siberry et al also found a lesser frequency of AE to quadrivalent MCV4 vaccine in the HIV-infected group than those previously reported for healthy subjects (42). The HIV-infected patients enrolled in our study were perinatally infected and therefore 53.3% were clinically classified as CDC C category. As a consequence, they were frequently exposed to healthcare environment and therefore could have been more likely to report AEs. Nevertheless our study demonstrated that MCC vaccine was safe in HIV-infected children, and no severe AE was observed.

Our study has certain limitations. The hallmark of immunity to meningococcal disease was a bactericidal titer in serum of >1:4 measured with human complement. However, this threshold titer may underestimate the extent of protection (31, 48, 54). Even in the absence of serum bactericidal activity, protective activity may be associated with higher-avidity group C anticapsular antibodies (48), which were not measured. There is also the question as to whether non-responders were primed for memory anticapsular responses upon subsequent exposure to the bacteria or a second dose of vaccine. The study also failed to evaluate the role of cART on immunogenicity because the majority of our HIV-infected subjects were on treatment already using this therapy.

Our results suggest that a single dose of MCC vaccine is insufficient for HIV-infected children and adolescents. Further, our data suggest that MCC vaccine should be administered to HIV-infected children after maximum immunologic and virologic benefit has been achieved with cART. In fact, U.S. Advisory Committee on Immunization Practices guidelines recommend MCV4 for all persons 11–12 years old, including those with HIV infection, for which a 2-dose primary series administered 8 weeks apart and a booster dose at 16–18 years are recommended (55). Our study supports the modification to the Brazilian national immunization program that was recently made (56), in which two doses given 8 weeks apart are now recommended for previously unvaccinated HIV-infected toddlers, children and adolescents, with a booster 5 years later.