The persisting burden of invasive pneumococcal disease in HIV patients: an observational cohort study

Abstract

Background: The increasing use of highly active antiretroviral therapy (HAART) and pneumococcal immunization along with shifting community exposures may have altered the burden of Streptococcus pneumoniae disease in HIV-infected persons. We describe the burden and risk factors for pneumococcal disease in the modern era of HIV care and evaluate the use of a 23-valent pneumococcal polysaccharide vaccine (PPV-23).

Methods: The incidence of invasive pneumococcal disease (IPD) between January 1st, 2000 and January 1st, 2010 in a regional HIV population in Southern Alberta, Canada was determined by linking comprehensive laboratory and hospital surveillance data. Clinical and epidemiologic data including risk factors for S. pneumoniae, history of pneumococcal immunization, serotypes of infections, and length of any hospitalizations for pneumococcal disease were evaluated with multivariate analysis. CD4 count and viral load at immunization were evaluated with a nested case-control analysis.

Results: In 1946 HIV-patients with 11,099 person-years of follow up, there were 68 distinct episodes of pneumococcal disease occurring in 50 patients. Increased risk was seen if female, age >60, Aboriginal ethnicity, lower education, injection drug use, smoking, nadir CD4 <200/μL, chronic obstructive pulmonary disease, and hepatitis C. Overall, the incidence of IPD was 342/100,000 person-years and was reduced to 187/100,000 within three years of PPV-23 immunization (P < 0.01). Although 78% of patients received PPV-23, 74% of IPD episodes were caused by PPV-23 serotypes. In a case-control analysis, HIV viral load at immunization was significantly predictive of PPV-23 failure, while CD4 count was not. 80% of IPD cases required hospitalization: median length of stay was 7 days (range: 1-71); four patients died.

Conclusions: Despite universal access to intensive measures to prevent pneumococcal disease including the widespread use of HAART and PPV-23 immunization, the incidence of IPD remains high in HIV patients with its associated morbidity and mortality.

Figure 1

Number and type of pneumococcal infections per year. Pneumococcal disease in Southern Alberta HIV patients over 10 years; there is no evidence of a decreasing trend. The peak in 2007 overlaps with a serotype 5 outbreak among Calgary's homeless community [10]; number of active patients at the midpoint of each year.

Figure 2

S. pneumoniae serotypes isolated over 10 years (n = 50). Isolated incidents of S. pneumoniae serotypes from 2000-2009 are described and arranged by coverage of three pneumococcal vaccinations. 40 were isolated from a sterile site only, 4 from bronchoalveolar lavage (BAL) only, and 6 from both a sterile site and BAL. Lines indicate serotypes included in the 7-valent pneumococcal conjugate vaccine (PCV7; n = 5), 13-valent pneumococcal conjugate vaccine (PCV13; n = 23), and the 23-valent pneumococcal polysaccharide vaccine (PPV-23; n = 37). *Serotype 6A is not included in PPV-23.

Figure 3

Incidence of IPD with immunization. Crude incidences with 95% Poisson confidence intervals are shown for PPV-23 serotypes. The incidence of IPD within three years of immunization was significantly lower than the incidence before immunization (P < 0.001). A composite incidence >3 years post-immunization incidence (not shown; mean: 262, 95% CI: 131-469) was also significantly lower than incidence before immunization (P < 0.01).