Does Antiretroviral Therapy for HIV Reduce the Risk of MS?
Does Antiretroviral Therapy for HIV Reduce the Risk of MS?
In 2011, a case was reported in which a patient with multiple sclerosis (MS) and HIV was treated with antiretroviral therapy (ART)—he did not experience MS-related deficits for over a decade. This prompted researchers in Denmark (Nexø et al) to examine the association between MS and HIV using population-based databases. Although they found a reduced incidence of MS in HIV-positive individuals compared with the general population, their study did not achieve a conventional level of statistical significance. I was recently asked to review the Gold et al study, which has answered the Danish researchers' call for a larger study to investigate whether HIV, or its treatment, has a protective effect on the development of MS.
Similar to the original Danish study, Gold et al used administrative health databases to fulfil their research objectives; in this case, they exclusively used hospital records. A cohort of individuals with a record of HIV between 1999 and 2011 was 'followed up' for subsequent records of MS. A comparison group, which included individuals who had a minor medical or surgical procedure, or injury, during the same time period, was similarly followed for records of MS. Using stratification, the authors were able to control for extraneous factors beyond age and sex, such as socioeconomic status, calendar year and region. Ethnicity is strongly, and differentially, correlated with MS and with HIV. In England, MS is most common among black and minority groups, while most new HIV diagnoses occur in white males. While the authors attempted to take ethnicity into account, they acknowledged that their analysis in this regard was limited because of missing data.
Authors of the case report and data linkage studies believed that it was likely treatment with ART, rather than HIV infection itself, that may have conferred the hypothesised protective effect on the development of MS. It must be considered that in both data linkage studies, the authors were unable to distinguish between those individuals who were treated with ART and those who were not. During the time period of the study, HIV treatment guidelines and clinical practice changed. In the mid-1990s, many favoured early aggressive treatment; however, this approach was criticised in the late 1990s and early 2000s because of concerns regarding side effects and drug resistance. Recently, the pendulum has swung once again in favour of earlier treatment. Without access to more detailed clinical and treatment information on the HIV-positive study population, researchers could not determine whether the individuals in their HIV cohort were in fact exposed to antiretroviral treatment and if so, the duration of their exposure. Given that a substantial number of HIV-positive individuals are diagnosed late in their infection (and therefore are immediately eligible for treatment), the timeframe of the study, and the length of follow-up, a high proportion of individuals in the HIV cohort were likely exposed to HIV treatment; however, without additional information, antiretroviral exposure remains presumptive.
Despite limitations inherent in the nature of the data, the Gold et al study makes a significant contribution to the literature on whether there is an association between HIV and the development of MS. They have answered the call for a larger study and importantly, controlled for confounding to a greater extent than has been done previously. Their replication of previous findings, the magnitude of their effect size and the statistical significance of their results combine to provide further evidence to support the hypothesis that there is an association between HIV (or its treatment) and a reduced risk of MS. However, additional work is required to move beyond hypothesis generation. A logical next step would be to directly examine the association between exposure to ART and the development of MS.
In 2011, a case was reported in which a patient with multiple sclerosis (MS) and HIV was treated with antiretroviral therapy (ART)—he did not experience MS-related deficits for over a decade. This prompted researchers in Denmark (Nexø et al) to examine the association between MS and HIV using population-based databases. Although they found a reduced incidence of MS in HIV-positive individuals compared with the general population, their study did not achieve a conventional level of statistical significance. I was recently asked to review the Gold et al study, which has answered the Danish researchers' call for a larger study to investigate whether HIV, or its treatment, has a protective effect on the development of MS.
Similar to the original Danish study, Gold et al used administrative health databases to fulfil their research objectives; in this case, they exclusively used hospital records. A cohort of individuals with a record of HIV between 1999 and 2011 was 'followed up' for subsequent records of MS. A comparison group, which included individuals who had a minor medical or surgical procedure, or injury, during the same time period, was similarly followed for records of MS. Using stratification, the authors were able to control for extraneous factors beyond age and sex, such as socioeconomic status, calendar year and region. Ethnicity is strongly, and differentially, correlated with MS and with HIV. In England, MS is most common among black and minority groups, while most new HIV diagnoses occur in white males. While the authors attempted to take ethnicity into account, they acknowledged that their analysis in this regard was limited because of missing data.
Authors of the case report and data linkage studies believed that it was likely treatment with ART, rather than HIV infection itself, that may have conferred the hypothesised protective effect on the development of MS. It must be considered that in both data linkage studies, the authors were unable to distinguish between those individuals who were treated with ART and those who were not. During the time period of the study, HIV treatment guidelines and clinical practice changed. In the mid-1990s, many favoured early aggressive treatment; however, this approach was criticised in the late 1990s and early 2000s because of concerns regarding side effects and drug resistance. Recently, the pendulum has swung once again in favour of earlier treatment. Without access to more detailed clinical and treatment information on the HIV-positive study population, researchers could not determine whether the individuals in their HIV cohort were in fact exposed to antiretroviral treatment and if so, the duration of their exposure. Given that a substantial number of HIV-positive individuals are diagnosed late in their infection (and therefore are immediately eligible for treatment), the timeframe of the study, and the length of follow-up, a high proportion of individuals in the HIV cohort were likely exposed to HIV treatment; however, without additional information, antiretroviral exposure remains presumptive.
Despite limitations inherent in the nature of the data, the Gold et al study makes a significant contribution to the literature on whether there is an association between HIV and the development of MS. They have answered the call for a larger study and importantly, controlled for confounding to a greater extent than has been done previously. Their replication of previous findings, the magnitude of their effect size and the statistical significance of their results combine to provide further evidence to support the hypothesis that there is an association between HIV (or its treatment) and a reduced risk of MS. However, additional work is required to move beyond hypothesis generation. A logical next step would be to directly examine the association between exposure to ART and the development of MS.
Source...