Thyroid hormones in HIV-infected patients in the highly active antiretroviral therapy era

Abstract

In a total of 372 determinations of thyroid hormones in HIV-infected patients we found a correlation between free thyroxine and CD4 cell counts, which was maintained after stratification by CD4 cell count and therapeutic groups, and was confirmed by multivariate analysis. Patients with normal free thyroxine had higher CD4 cell counts than those with low free thyroxine. Antiretroviral drugs did not influence free thyroxine levels. Our results support an interrelationship between the thyroid axis and the immune system. The thyroid function of HIV-infected patients during the era of highly active antiretroviral therapy (HAART) is not well known. The only published study in patients treated with HAART [1] found that 8.5% of patients had subclinical hypothyroidism. On the other hand, there is experimental evidence that supports the interrelations between the pituitary–thyroid axis and the immune system [2–6]. This study was conducted to analyse the thyroid hormonal pattern in a large group of HIV-infected patients in good clinical condition, to investigate the possible influence of HAART on thyroid hormones, and to evaluate the factors that could be associated with them, paying special attention to their possible relationship with CD4 cell counts. We studied 202 HIV-infected patients (95% men) without any history of thyroid disease, who underwent a total of 372 determinations of thyroid hormones. The number of determinations per patient ranged from one to four, each separated by a mean interval of 27.3 weeks. All patients were seen at the outpatient clinic, and all were in stable clinical condition in the absence of any acute severe illness or treatment with drugs that could influence thyroid hormones. In addition to free thyroxine and thyroid stimulant hormone (TSH), other clinical, laboratory and therapeutic parameters were recorded. The Spearman's correlation coefficient and Mann–Whitney U test were used for statistical analysis. A stepwise multiple regression analysis was used to identify the independent association of different variables with free thyroxine. A P value less than 0.05 for a two-sided test was considered statistically significant. The mean age of the patients was 37.3 years, the mean CD4 cell count was 453.8 × 106/l, and 63.3% had an undetectable viral load. We found 1.3% of the determinations of free thyroxine to be below and 0% to be above the normal limits. These figures were 0.3% and 3.8%, respectively, for TSH. Biological hypothyroidism, as defined by increased TSH with normal free thyroxine, was found in 3.5% of patients. A significant relationship was observed between free thyroxine and CD4 cell counts (r = 0.3, P < 0.0001), a correlation that was also significant in each of the four sequential determinations (P = 0.0006, P = 0.01, P = 0.03, and P = 0.006, respectively). Despite the very low number of patients with hypothyroidism, as defined by low free thyroxine levels, there were significant differences in the CD4 cell counts between patients with or without hypothyroidism (mean 226.3 versus 456.3 × 106/l, respectively, P = 0.036). Patients were divided into four groups according to the antiretroviral regimens they were receiving at the time of sampling. There was a parallelism between free thyroxine and CD4 cell counts in each therapeutic group (data not shown), which was not caused by the concomitant effect of antiretroviral drugs, because the association persisted in each group when the patients were stratified according to a cut-off level of 500 CD4 cells (Table 1). Finally, a separate analysis of each individual antiretroviral drug did not reveal significant differences in the free thyroxine levels.Table 1: Free thyroxine levels according to specific antiretroviral regimens and CD4 cell counts.A stepwise multiple regression analysis that included all variables studied revealed that only TSH (P < 0.001), and CD4 cell counts (P < 0.001) were independently associated with free thyroxine, whereas free thyroxine (P < 0.001), no history of AIDS (P < 0.001), the absence of infection with hepatotropic viruses (P < 0.001), time since the initiation of HAART (P = 0.002), younger age (P = 0.003), and an undetectable viral load (P = 0.005) were independent predictors of higher CD4 cell counts. In this series, the largest reported to date on the thyroid function of HIV-infected patients, we found that only a small fraction of patients in good clinical condition had biological hypothyroidism, whereas no patient had hyperthyroidism. This is in agreement with the findings of the only other study performed on this topic in the HAART era [1]. Other studies also found a significant correlation between free thyroxine and CD4 cell counts in adults [7] and children [8] with very different clinical stages. However, both were published before the HAART era, and the diversity of the patients’ clinical condition, which clearly affects both parameters, precludes any definitive conclusion. In contrast, our sample, which was very large and homogenous, allowed us to draw reliable conclusions about the genuine relationship between free thyroxine and CD4 cell counts. In this regard, the association found between these parameters is remarkable despite the narrow range of distribution of free thyroxine values; such an association being independent of other variables as evidenced by the multivariate model. Also noteworthy is the substantially lower CD4 cell counts of patients with subnormal levels of free thyroxine compared with those with normal free thyroxine levels. Data from animals and humans support an interaction between the pituitary–thyroid hormones and the immune system [2,9–12]. Our results add further insight into the existence of a functional relationship between these systems in HIV-infected patients. The intimate mechanism of this relationship is unknown, but thyroid dysfunction has been observed in cancer patients after the infusion of IL-2 [13–19], and the infusion of IL-2 in asymptomatic HIV-infected patients resulted in transient increases in TSH, free thyroxine and CD4 cell counts [20]. Considering that IL-2 is the main cytokine that induces T cell activation and differentiation and produces significant increases in CD4 cell counts, it could be hypothesized that the correlation we found between CD4 cell counts and free thyroxine might be mediated, at least partly, by this cytokine. In the only published study to analyse the possible role of antiretroviral drugs [1], patients with biological, subclinical thyroid dysfunction had received a higher cumulative daily dose of stavudine than patients without thyroid dysfunction. We found no differences in free thyroxine values in patients receiving or not receiving this drug or any other antiretroviral drug, and the different antiretroviral drugs were not significantly related to free thyroxine in either the univariate or the multivariate analysis, a not unexpected finding considering that thyroxine is mainly metabolized through peripheral conversion to triiodothyronine and not by hepatic glucuronidation [21]. We conclude that subclinical thyroid dysfunction is uncommonly observed in HIV-infected patients in stable clinical condition, that antiretroviral drugs do not seem to affect the thyroid hormone levels, and that there is a clear association between free thyroxine levels and CD4 cell counts. Our results support the existence of a relationship between the thyroid axis and the immune system in HIV-infected patients, perhaps mediated by cytokines, and add further insights into an area that could have clinical implications.