Only one out of every six vaccinations in HIV-positive individuals against hepatitis B (HBV) produced an antibody response strong enough to protect them from infection, according to data due to appear in the October 1st issue of Clinical Infectious Diseases, and now available online in CID's Electronic Edition. The only factor associated with a protective antibody response was an HIV viral load below 400 copies/ml at the time of vaccination, which contradicts a previous study which found both nadir and current CD4 cell count to be correlated with HBV vaccine response.
UK treatment guidelines recommend that all HIV-positive patients who lack immunity to hepatitis B are vaccinated against HBV, which is many times more infectious than HIV, and is usually transmitted through contact with blood, semen, vaginal fluids or saliva of an HBV-infected person.
Recent studies have found that HBV coinfection significantly increases the risk of death in people with HIV (Rockstroh 2004). A study of patients in the Multicenter AIDS Cohort found that men who were hepatitis B surface antigen-positive were eight times more likely to die of liver-related causes when compared to men with HIV who did not have hepatitis B (Thio 2002).
In order to characterise clinical factors that contribute to HBV vaccine response, investigators from St. Louis and New York in the United States, and Dublin, Ireland, conducted a retrospective cohort study, reviewing 920 medical records at a tertiary care HIV clinic.
Of the 920 records reviewed, 813 patients had complete HBV serologic data. Previous HBV infection was found in 255 (31.4%) and 166 (20.4%) had not received any dose of the HBV vaccine.
The vaccine used at the clinic is the three-dose GlaxoSmithKline (GSK) recombinant hepatitis B vaccine, Engerix-B, given over six months. This is one of two HBV preventative vaccines currently available in the United Kingdom, the other being Aventis Pasteur's HBvaxPRO. In the United States, Merck's Recombivax is the alternative to Engerix-B, although GSK's combination hepatitis A and B preventative vaccine, Twinrix is also approved there. There are no studies comparing the effectiveness of these various vaccines in HIV coinfected individuals.
Of the remaining 392 patients with complete HBV serologic data, 354 completed all three vaccinations. However, 160 did not have follow-up surface antibody tests in order to determine whether the vaccine was, in fact, protective, leaving 194 (just 35% of all eligible vaccine recipients with complete HBV serologic data) to be included in the analysis.
Of the 194 vaccinated individuals who had follow-up testing, 192 (99%) received a 10mg dose for doses 1 and 2, and two (1%) received a 20mg dose. For dose 3, 188 (97%) received a further 10mg dose, five received a 2mg dose, and one received a 40mg dose. The median interval between doses 1 and 3 was seven months (interquartile range [IQR], 6-9 months).
Protective vaccination was defined as hepatitis B surface antibody titre of 10mIU/ml or greater, measured after all three doses had been administered. Just 34 (17.5%) of the 194 reached this level. Last year, the US HIV Outpatients Study (HOPS) found that 19 (37%) out of the 51 who had post-vaccination testing for HBV surface antigen were protected (Tedaldi 2004). However, the criteria in the HOPS paper was merely a positive surface antigen result; without quantification this may have overestimated how many were truly protected.
Univariate analysis found that male gender (p=0.006) and HIV viral load below 400 copies/ml at the time of the first (p
Length of treatment with antiretroviral therapy also correlated with vaccine response: responders received highly active anti-HIV therapy (HAART) for a median of 13.5 months (IQR, 6.1-19.4 months), whereas non-responders received HAART for a median of 3.7 months (IQR, 1.3-15.5; p=0.005).
No significant difference in either the nadir CD4 cell count or in the CD4 cell count at time of vaccination was found, in contrast to the HOPS study which found that the majority of responders had a nadir CD4 cell count above 200 cells/mm3 and higher median CD4 cell counts (584 cells/mm3 vs. 384 cells/mm3, p=0.08) at time of vaccination, as well as an undetectable baseline viral load (63.2% vs. 33.3%, p=0.04) (Tedaldi 2004). In the current study, poor response was seen across the spectrum of CD4 counts.
On multivariate analysis, only an HIV viral load below 400 copies/ml at first dose administration was found to predict success (Odds Ratio [OR], 3.47; 95% CI, 1.5-7.6).
During subsequent 24-month follow-up, no vaccine responder developed HBV infection, compared with 16 (10%) of the non-responders, two of whom developed chronic active hepatitis B and five of whom developed natural immunity.
"Although the overall proportion of patients who developed protection was very poor at 17.5%, responders were truly protected from subsequent infection," write the investigators.
They also comment on the "concordance of missed opportunities" to provide the vaccine to those who need it, and for the low rates of follow-up measurements of hepatitis B surface antibody titre. With success rates this low, they conclude, clinicians should "consider booster vaccinations for patients who do not develop protective antibodies."
Overton ET et al. Undetectable plasma HIV RNA load predicts success after hepatitis B vaccination in HIV-infected persons. Clinical Infectious Diseases 41, electronic edition, 2005.
Rockstroh J et al. Hepatitis B and hepatitis C in the EuroSIDA Cohort: prevalence and effect on mortality, AIDS progression and response to HAART. Eleventh Conference on Retroviruses and Opportunistic Infections San Francisco, abstract 799, 2004.
Tedaldi EM et al. Hepatitis A and B vaccination practices for ambulatory patients infected with HIV. Clinical Infectious Diseases 38: 1483-1489, 2004.
Thio CL et al. Liver disease mortality in HIV-HBV co-infected persons. Ninth Conference on Retroviruses and Opportunistic Infections, Seattle, abstract 656, 2002.