Immunization against pneumococcal disease in HIV-infected patients: conjugate versus polysaccharide vaccine before or after reconstitution of the immune system (CTN-147)

Background

The introduction of combination antiretroviral therapy (cART) among people living with HIV infection has decreased the incidence of all-cause pneumonia. Pneumococcal pneumonia however still remains more common among HIV-infected than non-HIV-infected individuals, even in subgroups of patients with CD4 counts >500 cells/mm³. Further, the incidence of invasive pneumococcal disease (IPD) is 10 times more frequent among HIV-infected persons than non-HIV infected persons, even in those who were receiving HAART and had near normal CD4 cell counts. ¹

A recent Canadian study found that despite universal access to intensive measures to prevent pneumococcal disease including widespread use of HAART and 23-valent pneumococcal polysaccharide (PPV) immunization, the incidence of IPD remains high in HIV patients. ²

The effectiveness of PPV has been questioned. In immunological studies, PPV has consistently elicited capsule-specific pneumococcal antibodies in HIV-infected individuals, but the magnitude and duration of postimmunization responses in these individuals has often been lower than those seen in immunocompetent hosts. The randomized controlled Ugandan trial reported no protection, potentially harmful effect but in a very unique population. ³ Three out of seven additional studies ⁴ on PPV and all-cause pneumonia showed no significant effect from PPV immunization and four found a protective effect. Study quality and potential confounding factors in these studies were not rigorously assessed. However, based upon these observational studies and expert opinion, current US guidelines recommend immunization of HIV patients with PPV when CD4 counts are above 200 cells/mm³. ⁵ In contrast, the World Health Organization (WHO) advises that PPV may be considered for people with HIV infection in WHO stage 1 or, if CD4 testing is available, with a CD4 count above 500 cells/mm³. ⁶ There are no data supporting either of these recommendations.

Pneumococcal conjugate vaccine has a polysaccharide antigen with an immunogenic carrier which provides a T-cell dependent antigen, stimulating a T-helper-cell response potentially leading to an overall improved immune response.

The aim of our study was to determine whether immunizing HIV-positive patients with pneumococcal vaccine after there has been a reconstitution of the immune system resulted in an improvement in immunogenicity and to determine whether antibody responses to specific pneumococcal serotypes were different in HIV patients given PPV versus pneumococcal heptavalent conjugate vaccine (7PCV).

Methods

This was a multicentre, randomized trial, conducted at 14 institutions across Canada between September 2004 and October 2007. The study population included HIV seropositive volunteers aged 18–65 years, with CD4 cell counts <200 cells/mm³, who were willing to initiate antiretroviral therapy. We excluded individuals who had prior immunization with pneumococcal vaccine, hypersensitivity to components to either vaccine, acute febrile illness at the time of vaccination, documented or suspected pneumococcal infection in the last five years, asplenia, receipt of immune globulin within the last six months or who were pregnant. After signing informed consent, patients were randomly assigned, utilizing a two by two factorial design to receive immediate immunization with either 0.5 mL intramuscularly of PPV (Pneumovax, Merck, Canada) or 0.5 mL intramuscularly of 7PCV (Prevnar, Wyeth) or delay immunization until their immune system had been reconstituted (CD4 count × 2 at 2-month intervals) with antiretroviral therapy. After six months, if the CD4 count had not risen to >200 cells/mm³, patients received the vaccine assigned at randomization. Blood samples for antibody measurements were collected prior to immunization and at one, six and 12 months postimmunization. All participants were assessed for adverse events, by study personnel at clinic visits and by self-report.

Serum was separated within two hours of collection and stored at 280° C before being shipped frozen to the Finnish Institute for Health and Welfare for analyses. Pneumococcal anticapsular immunoglobin (IgG) concentrations were measured for the seven common vaccine serotypes (4, 6B, 9V, 14, 18C, 19F and 23F) with a 22F adsorption step. ⁷ Opsonophagocytic activity (OPA) of antibodies (serotypes 14, 18C, 19F and 23F) were measured by using a multiplexed opsonophagocytic assay. ⁸

Baseline characteristics of the four groups were tabulated. Continuous variables were summarized using medians with interquartile ranges (IQRs). Dichotomous and categorical variables were summarized using counts and percentages. For each serotype, vaccine responders were defined as individuals who had a two-fold increase in IgG concentrations from baseline. This was analysed at one, six and 12 months postimmunization. An ordinal regression model was used to assess the influence of vaccine and delivery strategy on the number of serotypes showing response, adjusting for important baseline factors including age, gender, ethnic group, body mass index and type of antiretroviral drug used. We chose to focus on the number of serotypes showing response because we felt that this was an appropriate summary measure and wished to avoid the problem of multiple outcomes by focusing on each individual serotypes. Linear regression models were also utilized to assess the change in log IgG concentration for each serotype adjusting for preimmunization titre and the factors identified above in the ordinal regression model. The proportion of patients with OPA titre of > 2 in the four groups were compared using Fisher's exact test. SAS software, version 9.2, was utilized for statistical analyses.

The trial was reviewed and approved by institutional ethics boards for all participating institutions. The procedures that were followed were in accordance with the ethical standards of these responsible committees and with the Declaration of Helsinki. Major adverse events, leading to hospitalization, disability or death, were immediately reported to the principal investigator, and the national data monitor.

Results

Seventy-nine patients consented to participate and were randomized into the study.

Sixty-two of the participants were men (78%), 44 patients were Caucasian, with a median age of 41 years (IQR 34–48); median time since HIV diagnosis was 0.3 years (IQR 0.2–2.8), median CD4 count was 60 cells/mm³ (IQR 19–135), median CD8 count was 500 cells/mm³ (IQR 337–772) and median HIV viral load was 5.02 log copies/mL (IQR 4.72–5.49). Forty-two patients were randomized to the immediate immunization arm of the study and 37 to the delayed arm. The median CD4 count prior to immunization in the delayed group (i.e. waiting to immunize until after reconstitution of the immune system) was 182 cells/mm³ (IQR 112–229).

Immunization and sampling was achieved for 74 participants. One patient who had previous pneumococcal pneumonia was inadvertently enrolled into the study. One patient discontinued study due to adverse events from antiretroviral therapy and one patient moved away from the study institution and they were immunized prior to leaving. Two patients could not be contacted after several attempts.

Local reactions (redness, pain and swelling) to vaccination were found to be more common in the immediate vaccination group as compared with the delayed group (P < 0.05; Wilcoxon rank-sum test). No differences in local reactions were observed when comparing the 7PCV versus the PPV (data not shown).

There were no significant differences in the proportional odds and geometric mean concentration of anticapsular IgG for the vaccine serotypes between the individual vaccines at month one, six or 12 (Figure 1). Likewise there were no significant differences between the vaccines when analysing the proportion of patients with OPA titre >2 by serotype (Figure 2). OPEN IN VIEWER

OPEN IN VIEWER

Figure 2

Opsonophagocytic activity (OPA) titre of >2 by vaccine type and by timing of vaccination. Percentage of patients with OPA titre of >2 for each serotype by vaccine type and by timing of vaccine at one and 12 months post-immunization. Solid bars indicate antibody response using pneumococcal polysaccharide vaccine or after immediate immunization; hatched bars indicate antibody response using 7PCV vaccine or after delaying immunization until after reconstitution of the immune system

The proportional odds ratios for delayed versus immediate immunization were 0.54 (95% CI: 0.21–1.40, P = 0.21), 0.341 (95% CI: 0.12–0.96, P = 0.04) and 0.204 (95% CI: 0.07–0.61, P = 0.004) at months one, six and 12 respectively (Figure 1). The difference between immediate and delayed vaccination was consistent across individual serotypes (Figure 3). There was however no difference in OPA responses (OPA scores >2) between the two dosing strategies (Figure 2). OPEN IN VIEWER

There was one report of IPD throughout the 12-month follow-up of the study in a patient who received PPV in the immediate immunization arm.

Discussion

Although some studies have found a decrease in the incidence of pneumococcal disease in HIV-infected patients since the widespread use of cART, it remains one of the common causes of hospital admission in these patients, and the incidence is higher among HIV-infected patients than in age-matched controls. This is especially true in patients with advanced HIV infection. ¹

Untreated HIV-infected patients have reduced antibody responses to PPV, which correlates with falling CD4 cell counts. cART partially restores the immune system by reducing HIV replication and immune activation and increasing CD4 cell counts. However certain immunological abnormalities persist for years after cART initiation, including a low CD4:CD8 ratio, a low naïve-to-memory cell ratio, a low expansion of CD28 effector T-cells and a reduced T-cell repertoire. The effects of CD4 cell count and cART treatment therefore may be important and it has been proposed that delaying pneumococcal vaccine in HIV-infected patients until there has been ‘repair ⁷ of the immune system would favour an improved immunological response to any pneumococcal vaccine. ⁴

Antibodies to the pneumococcal capsule which support killing of Streptococcus pneumonia with complement and phagocytosis have long been considered the primary determinant of protection against the organism. ⁸ In our study we have shown that HIV-infected adults regardless of the vaccine used, produced antibody response more often to pneumococcal capsular polysaccharides after reconstitution of the immune system. Opsonophagocytic assays measure the ability of serum antipneumococcal antibodies to opsonize bacteria and stimulate phagocytosis. The measurement of OPA in vaccine studies is considered informative, especially in case of HIV-infected individuals and the elderly, since their antibodies can have lower functional capacity than those of immunocompetent individuals. ⁸ In our study, we could not show any difference in OPA titres among patients who had their immune system reconstituted versus those who did not. The possible reasons are that all vaccinees are HIV infected and, in general, have low OPA titres and that the number of vaccinees was fairly small. The OPA assay is in general less sensitive than enzyme immunoassay (EIA) and that is why perhaps many sera did not show any OPA activity. For instance, for serotypes 19F and 23, only approximately half of the vaccinees had an OPA titre ≤4 (except in the delayed PCV group).

Studies to date have not adequately defined the efficacy of pneumococcal vaccine in HIV-infected patients; however, immunization with a PPV is still recommended in most guidelines. Reports on the effectiveness of PPV have been inconsistent.^3,4 The authors of the randomized controlled trial ³ concluded that immunization with PPV is ineffective and may be detrimental, but at present there is no good explanation for this unexpected finding. The study was carried out in Uganda inpatients with no access to cART at the time; ³ the conclusions about the efficacy of PPV should be extrapolated to other settings with caution.

In developed countries, most studies have shown that cART with or without immunization has had the most consistent effect on reducing the incidence of pneumonia and IPD.^4,9 7PCV was found to be effective in preventing vaccine serotype-specific IPD in a cohort consisting primarily of HIV-infected Malawian adults (who had recovered from documented IPD); ¹⁰ however, immunological benefit was not seen in other studies.^11,12

In conclusion, we have shown that HIV-infected adults produced a higher immunological response to PPV or 7PCV after the reconstitution of their immune system. This finding supports current recommendations to wait until the CD4 count is greater >200 cells/mm³ before immunization of HIV-infected patients with pneumococcal vaccines. ⁵ This study showed no advantage to recommending immunization of HIV-infected adults with conjugate vaccine over that of polysaccharide vaccine. Newer agents which may result in better protection for our patients including adjuvanted vaccines require further investigation. ¹³