Yellow fever vaccine in HIV-positive people: safe, but not always effective

This article is more than 13 years old. Click here for more recent articles on this topic

A retrospective analysis has found that yellow fever vaccine did not result in any serious adverse events in HIV-positive individuals. However, people with HIV responded less well to the vaccine than their HIV-negative counterparts, and the protective effects of the vaccine wore off more quickly. The study was published in the March 1st issue of Clinical Infectious Diseases. A letter in the online edition of the journal AIDS also warns that yellow fever vaccine may, theoretically, pose risks for people treated with maraviroc (Celsentri).

Yellow fever vaccine (17DV) can effectively prevent this potentially fatal disease; however, there have been concerns that this live-attenuated vaccine cannot be safely given to HIV-positive individuals, whose suppressed immunity might allow the weakened vaccine virus to replicate. Several small studies (such as this) have found yellow fever vaccination to be safe in HIV-positive people with CD4 cell counts above 200 cells/mm3, and US CDC guidelines suggest that the vaccine should be offered to people in that situation.

In this study, researchers identified participants in the Swiss HIV Cohort Study who had reported travel to tropical destinations. A total of 174 had received yellow fever vaccine (17DV): 72 of these were vaccinated while HIV-negative, and 102 after becoming HIV-positive. The protective effect of the vaccination was assessed by neutralisation titre (NT - roughly, the maximum blood plasma dilution ratio at which neutralising antibody reactions can still be measured; ratios of 1:10 or greater are considered protective).


Neutralising titre results were available for 78 HIV-positive vaccine recipients within the first year after vaccination. Within this first year, significantly more HIV-positive participants (13 out of 78, or 17%) had a non-protective response than HIV-negative participants (2 of 66, or 3%). The actual NT scores – a measure of the robustness of protective response – were also lower overall amongst HIV-positive recipients.



A protein on the surface of certain immune system cells, including CD4 cells. CCR5 can act as a co-receptor (a second receptor binding site) for HIV when the virus enters a host cell. A CCR5 inhibitor is an antiretroviral medication that blocks the CCR5 co-receptor and prevents HIV from entering the cell.

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 


A laboratory measurement of the amount, or concentration, of a given component in solution.



The fluid portion of the blood.


Weakened. Attenuated viruses are often used as vaccines because they no longer cause disease but may still stimulate a strong immune response.

The median CD4 cell count of these 78 individuals was 496 cells/mm3, and only six had counts below 200 cells/mm3. There were no statistically significant predictors of reactive versus nonreactive NT in this group, possibly due to the small numbers with lower CD4 cell counts; however, those on successful antiretroviral treatment had a non-significant trend toward protective NT (odds ratio 2.5, 95% confidence interval, 0.7-8.8, p=.17). Other factors that showed trends toward protection were suppressed HIV viral load, higher CD4 cell count, female sex, absence of chronic hepatitis, and older age.

Furthermore, eleven patients who initially had protective responses showed non-protective NTs within five years after vaccination, indicating a more pronounced decline in protective response over time. Over the time period from one to ten years after vaccination, the rate of non-protective response in HIV-positive recipients was 23% – twice that of HIV-negative recipients – although this difference was not statistically significant (p = 0.07).

Another subgroup (four of the 102 HIV-positive patients), with initially non-protective responses, later developed protective NTs after beginning ART and achieving sustained viral suppression – "a phenomenon that," according to the authors, "has never before been described." By contrast, "most" of the patients who lost their protective response were not on ART and had unsuppressed HIV viral load levels at the time of vaccination.


No significant adverse events (hospitalisation or death) within six weeks of vaccination were seen in the 102 HIV-positive individuals who received the 17DV vaccine. There were 48 deaths among the 696 cohort participants who travelled, including three deaths in the 17DV recipients, but these were deemed "clearly not associated" with vaccination.

Due to the size of the group studied, this 0% rate of serious side-effects had a reliable upper 95% confidence limit of 2.9%. In other words, although no serioius adverse events were actually observed, the study cannot rule out the possibility that vaccination-related SAEs could occur at a rate of up to 2.9% in larger groups.


The authors conclude that their findings "reinforce current recommendations about the safety of 17DV in HIV-infected patients… However, a low risk (up to 3%) of [serious side-effects] cannot be excluded." HIV-positive individuals who are not on HIV treatment and who have CD4 cell counts below 350 cells/mm3 "should preferably postpone receipt of 17DV until the plasma HIV RNA level is undetectable, to attain a more vigorous vaccine response." The ten-year period typically recommended for booster vaccinations is also "probably not sufficient for HIV-infected individuals."

Yellow fever vaccine and maraviroc

In a letter to the journal AIDS, Anna Roukens and others from the Leiden University Medical Center note that the CCR5 delta 32 mutation which confers resistance to HIV infection (by down-regulating CCR5 receptor expression on CD4 lymphocytes) has also been associated with increased severity of infection with flaviviruses such as the yellow fever virus. They report that this mutation was recently found in a patient who had a severe adverse reaction to the 17DV vaccine. The authors note that, as maraviroc's mechanism of action is comparable to that of the mutation, this raises an as-yet "purely hypothetical" concern regarding 17DV safety in patients on maraviroc (Celsentri). They suggest that 17DV vaccination be avoided in patients on maraviroc "until further research has ruled out the contribution of a CCR5 mutation in patients with yellow fever vaccine-associated adverse events."


Veit O et al. Immunogenicity and safety of yellow fever vaccination for 102 HIV-infected patients.CID 48:659-66, 2009.

Roukens AH, Visser LG and Kroon FP. A note of caution on yellow fever vaccination during maraviroc treatment: a hypothesis on a potential dangerous interaction. AIDS (online edition), 2009.