Background
During HIV-1 infection, serum cytokine levels generally rise with the progression of immunodeficiency and decrease with the initiation of antiretroviral therapy (ART) [
1‐
6]. However, we previously measured the levels of 12 serum cytokines in 35 HIV-infected patients and demonstrated that the level of serum interferon-γ (IFN-γ) exhibits a trend different from those of other cytokines [
7]. Specifically, in a cross-sectional study, we revealed that some patients who were asymptomatic carriers or being treated by ART had a high level of serum IFN-γ. Similarly, in a longitudinal study, we revealed that a high level of serum IFN-γ was sustained in about 30% of the patients after the initiation of ART. Those observations suggested that serum IFN-γ concentration is maintained at a high level in some patients regardless of the state of immunodeficiency or ART. Thus, compared with the concentrations of other cytokines, the level of circulating IFN-γ may have different clinical significance in patients infected with HIV-1 [
8].
It has been reported that initiation of ART reduces the levels of inflammation-associated soluble biomarkers, including serum cytokines such as interleukin-6 (IL-6), although their levels do not return to those observed in non-infected population [
3,
5,
9]. Various factors, such as old age, comorbidities, and death, have been associated with the high level of circulating IL-6 in patients infected with HIV-1 [
9‐
11].
In contrast to the current knowledge about factors influencing IL-6 levels, the parameters associated with the high level of circulating IFN-γ have not been well established. It is unclear whether sustained high level of circulating IFN-γ has any influence on the clinical course of individuals infected with HIV-1. Although it may have an immunostimulatory effect that suppresses viral replication and increases CD4+ cell count, excessive immune activation may result in the reduction of CD4+ cell count and increased likelihood of comorbidity development. In the present study, we examined HIV-1-infected patients who were either naïve to ART, with CD4+ T cell count > 200 cells/μL, or had HIV-1 RNA levels below the detection limit after being on ART for over a year and compared the clinical characteristics between the participants with high and low levels of plasma IFN-γ or IL-6.
Discussion
In the present study, we determined the proportion of HIV-1-infected patients with a high level of plasma IFN-γ (≥5 pg/mL) and examined their clinical characteristics. We demonstrated that the patients with a high plasma IFN-γ level were more likely to be younger, and less likely to have dyslipidemia or to be on a protease inhibitor. The incidence rate of comorbidities, such as diabetes and hypertension, in the high IFN-γ group was equivalent or nominally lower than that of the low IFN-γ group, although significant differences were not observed. Most of these results were not qualitatively similar to the findings reported in patients with high levels of circulating IL-6 in this and previous studies [
10]. Borges et al. performed an analysis of 9864 patients with virological suppression and demonstrated that the high level of plasma IL-6 was associated with multiple factors that affect inflammation, such as old age, use of protease inhibitors, comorbidities, reduced renal function, and others. Some studies have also demonstrated associations between the high level of circulating IL-6 and various clinical outcomes, including death and the onset of acquired immune deficiency syndrome [
9,
11,
17]. Our findings support a possible association between the high level of IFN-γ with endogenous anti-HIV response rather than with inflammation [
8].
Notably, HIV-1-infected patients with a high plasma IFN-γ level had reduced CD4
+ cell count recovery from year 4 on ART and onwards than patients with a low plasma IFN-γ level. Various factors have been shown to be associated with suboptimal CD4
+ cell count recovery, including low nadir/baseline CD4
+ cell count, old age, male gender, prolonged duration between HIV-1 infection and the initiation of ART, hepatitis C virus infection, hepatitis B virus infection, comorbidities, injection drug use, up-regulation of surface markers of lymphocyte activation, and gene polymorphisms of
IL-10 and
IL7RA [
18‐
37]. In addition, many soluble biomarkers have been also examined for the association with suboptimal recovery of CD4
+ cell count [
38,
39]. However, the nadir/baseline CD4
+ cell count may be a confounding factor as it is strongly associated with suboptimal recovery of CD4
+ cell count and influences the circulating level of soluble biomarkers. After controlling for that confounding factor, only a limited number of biomarkers, including IL-6 before ART and interferon-inducible protein-10 after ART, were associated with the suboptimal recovery of CD4
+ cell count [
34,
35]. In the present study, we demonstrated that there was no significant difference in CD4
+ cell counts prior to and 3 years after ART between the low and high IFN-γ groups. Unexpectedly, the proportion of patients who were considered at high risk of reduced CD4
+ cell count recovery, such as older patients and those with comorbidities, was lower in the high IFN-γ group than in the low IFN-γ group. Collectively, these findings suggest that a high level of plasma IFN-γ may be an independent factor associated with lower recovery of CD4
+ cell count.
Our phylogenetic analysis demonstrated that HIV-1 genetic sequence in patients with a high IFN-γ level was distributed among different subtypes in a scattered manner, indicating that gene polymorphisms may be one of the causative factors of high plasma IFN-γ level. IFN-γ + 874 T/A gene polymorphism is located within the first intron and encompasses the binding site for the transcription factor NF-κB. Binding affinity of NF-κB for the + 874 A-allele is lower, and it results in reduced production of IFN-γ. Because it has been reported that individuals with + 874 A-allele may be at a higher risk of tuberculosis [
40], many studies examined the association between this polymorphism and various diseases. In particular, it has been shown that the A-allele was a risk factor for HIV infection [
41], and that HIV-infected patients with the A-allele had a higher risk of HIV-tuberculosis co-infection [
42], lower CD4
+ cell count [
31,
43], and lack of response to ART [
44]. Our findings were not in accordance with previous reports, which demonstrated that patients with the A-allele, associated with lower IFN-γ production, showed a reduction in CD4
+ cell count. These findings suggest that IFN-γ gene polymorphism may not be the primary factor that contributes to the high level of circulating IFN-γ.
In the phylogenetic analysis, we identified four patients that formed a cluster (red boxes in Fig.
4). In addition to the similarities of their genetic sequence, they had common geographical and chronological features, as all four of them were diagnosed in Osaka after 2011. Thus, these patients were likely to have been infected by the novel HIV-1 variant previously described by Mori et al. [
16]. It is still unclear why the patients who were infected with this novel variant had accelerated disease progression. Further investigations of this phenomenon are required because there may be a potential common mechanism underlying the accelerated disease progression and high plasma IFN-γ level.
There were several limitations in our study: it included a relatively small number of patients, especially of those infected with novel HIV-1 variant previously described [
16]. Furthermore, information on some factors that might affect circulating cytokines, such as consumption of tobacco and alcohol [
10], were missing. In addition, the data on CD4
+ cell counts were collected retrospectively, and plasma IFN-γ level was evaluated in a cross-sectional manner. Although we selected the cutoff value using earlier data [
7], evaluation of the cutoff value in this study using the differences of the recovery of CD4
+ cell count after 4 years of ART suggested that 5 pg/mL of IFN-γ concentration was a potential candidate for optimal cutoff value. In addition, our study did not investigate IFN-γ-producing cells. Further studies are required to uncover the biological mechanism contributing to the high level of circulating IFN-γ.