Given that severe CD4⁺ T cell deficiency when ART is commenced is a risk factor for an IRIS, prevention of CD4⁺ T cell depletion through early diagnosis and treatment of HIV infection regardless of CD4+ T cell count, as supported by the findings of the START study (33), is the most effective measure for reducing the risk of developing an IRIS. However, ‘late presenters’ with advanced immunodeficiency continue to present and they must be assessed carefully for HIV-related infections that might trigger an IRIS when ART is commenced. Those patients without obvious signs or symptoms of an HIV-related infection should be screened for a subclinical infection. At a minimum, a chest x-ray and assay of serum for cryptococcal antigen, HBV surface antigen and core antibodies, and HCV antibodies should be performed.

Decisions about the optimal timing of ART commencement in patients with an established HIV-related infection should take into consideration the CD4⁺ T cell count and the site of infection. In general, commencement of ART early in the course of therapy for the infection is recommended, particularly when the CD4⁺ T cell count is <50/mL (34, 35).  Several studies examining the timing of ART commencement in the setting of pulmonary TB have also produced results favouring early commencement of ART (36 – 39). 

Special consideration must be given to the timing of ART commencement in patients with an opportunistic infection of the CNS because inflammation resulting from an IRIS in this site carries a particularly high risk of morbidity and mortality. Thus, deferment of ART is recommended in HIV patients with TB meningitis (40) and in the setting of CM, randomised studies suggest ART should be deferred until 5-6 weeks post CM diagnosis (41 – 43). In PML, deferment of ART is usually not an issue because it is the only therapy that may be effective for this condition. There is insufficient evidence to make recommendations about the optimal timing of ART commencement in patients with Toxoplasma encephalitis.

The ART regimen used is not a consideration in preventing the development of an IRIS, though use of an ART regimen that includes nucleos(t)ide analogues with activity against both HIV and HBV in patients with HIV/HBV co-infection will minimise the risk of a hepatitis flare (HBV-IRIS). Integrase strand inhibitors have been reported to be associated with an increased risk of developing an IRIS in observational studies (44, 45) but this was not confirmed in a randomised controlled trial (46). Inclusion of maraviroc in the ART regimen of patients commencing ART with a CD4⁺ T cell count of <100/μL was explored as a means of preventing an IRIS in the CADIRIS study (47) but was ineffective.

Prednisolone therapy during the first 4 weeks of ART decreases the risk of TB-IRIS in HIV patients with CD4⁺ T cell counts <100/mL (48)