In this study we investigated factors of general immune activation associated with HIV disease progression, and we found significantly lower values in patients receiving prednisolone compared to untreated patients, suggesting that prednisolone may have beneficial effects on immunological correlates of HIV disease progression in otherwise untreated patients. In patients treated with HAART, no additional beneficial effects attributable to prednisolone could be observed. The immune-modulating activity of prednisolone in otherwise untreated patients described in this manuscript is in line with previous observations, reported by Andrieu et al., who observed in an uncontrolled, open-label study an increase in CD4+ T cell counts following administration of 0.3-0.5 mg/kg prednisolone over a period of 12 months . Prednisolone treatment (with a concentration about 4-8 times higher than in our study) was accompanied by a reduction of immune activation. Patients were encouraged to continue 0.3 mg/kg prednisolone as follow up-medication as long as their CD4+ T cell levels remained above baseline levels. Follow-up prednisolone treatment postponed CD4+ T cell decrease (relative to baseline levels) by two years . A few patients experienced relatively mild prednisolone-associated side effects such as face swelling, an increase in body weight and elevated blood pressure.
In another, randomized, placebo-controlled trial, prednisolone was given for 8 weeks with a decreasing dose from 50 mg to 15 mg per day; the main outcome was survival . Elliot and colleagues observed only a trend towards better survival in prednisolone-treated patients compared to HIV controls (21 versus 25 deaths per 100 person years; difference statistically not significant) but a statistically significant increase in the incidence of Kaposi sarcoma (4.2 versus 0 cases per 100 person years). The study population was not being stratified for HHV-8 seropositivity, the causative factor of Kaposi sarcoma, and HHV-8 infection status was not being analyzed. (In another short-term randomized, placebo-controlled trial studying the effects of prednisolone on effusive tuberculous pericaditis in HIV-infected individuals, with a similar dosage and time scale [60 mg starting dose, tapered with 10 mg per week until completion at the sixths week], three cases of Kaposi sarcoma were observed among the 58 study participants and all of them received placebo .) The main difference in the treatment regime towards other studies, including ours, is the relative short duration of prednisolone treatment of only 8 weeks, which may hamper the observation of a therapeutic benefit with the primary endpoint of survival.
Two other randomized, placebo-controlled trials investigated the effects of prednisolone on HIV-infection in the presence of HAART. In a study with a dose of 0.5 mg per kg body weight per day given for 8 weeks, a decrease in CD8+/CD38+ cells was reported with no changes in viral load and CD4+ cell counts . In partial contrast to this, a fixed daily dose of 40 mg prednisolone was given for 8 weeks in another trial, resulting in an increase of CD4+ counts compared to baseline values of 40%, but no changes in HLADR/CD38-positive CD8 cells and naive T cells (CD45RA+/CD62L+) were reported .
We have previously analyzed the CD4+ cell counts of some of our patients in a longitudinal observational study and found that long-term treatment with low-dose prednisolone is safe and - similar to what has been found by Andrieu et al.  - associated with a postponement of CD4+ T cell loss by about 2 years relative to baseline values [29, 30]. Our present study may offer a mechanistic explanation for the T cell preserving activity of low-dose prednisolone, i.e. by reducing general immune activation. Unfortunately, the previously described CD4+ T cell preserving activity of prednisolone cannot be directly addressed in this study (although a trend towards higher counts in prednisolone-treated patients was visible), as this would require a longitudinal analysis and probably also a more homogenous patient sample with respect to time to infection and therapy initiation.
The strong association between general immune activation and HIV disease progression is also seen in studies investigating the effects of HAART on immune activation. Similar to what we found in our study, HAART does not only suppress virus replication but also reduces immune activation [18–20]. In some patients - the so-called virological nonresponders - HAART fails to suppress virus replication but effectively restores immune activation and CD4 T cell counts . On the other hand, immunological nonresponders, who do not regenerate helper T cell counts despite HAART-mediated suppression of virus replication, show enhanced immune activation . Massanella et al. reported in a cross sectional study that patients with limited CD4 T-cell repopulation under HAART exhibited high T-cell activation rates (CD38+, HLA-DR+), suggesting that immunomodulating strategies should be envisaged to treat discordant patients .
General immune activation not only correlates with progression of immunodeficiency, but also with neurological complications of HIV infection. HIV patients with cognitive impairment or brain atrophy exhibit more elevated sCD14 levels than neurologically asymptomatic HIV-positive controls . This link between monocyte activation and the neurological complications of HIV infection fits well into the overall picture in which microglia activation has been identified as a correlate with immunodeficiency virus-mediated neuropathy [41, 42].
Due to the non-prospective and non-randomized design of our study, our findings should be interpreted with caution, as potential yet undiscovered confounding mechanisms may have biased the result. Moreover, the low sample size of our study is relatively low and differences in individual viral loads (although not statistically significant between untreated and prednisolone-treated patients) are possible limitations. As we found significantly lower immune activation in patients treated with prednisolone compared to untreated patients, the critical reader might object that this difference might have been caused by an unintended enrichment of natural slow-progressors in the group of prednisolone-treated patients over time, who would lower the average immune activation in this group due to intrinsically lower immune activation in these patients. (The argument in this case would be that patients who progressed under prednisolone sooner or later would have left this group to become treated with HAART. Hence, over time, patients with an intrinsic slower disease progression would have become enriched in this group. At a closer look, this mechanism would, however, also apply for patients in the untreated group - compared to which prednisolone-treated patients exhibited significantly lower activation.) We therefore checked whether there is any correlation in prednisolone-treated patients between immune activation (CD8/CD38, sCD14, LBP, suPAR) and the duration of prednisolone medication or the (known) time of infection and found no correlation whatsoever for any of the markers (P values of correlations ranged between 0.2748 and 0.9399, data not shown). These findings suggest that a prednisolone-independent enrichment over time of slow-progressing patients in this group has not taken place, otherwise one should have expected to observe lower levels of immune activation in patients with longer treatment times.
As for the absence of detectable effects of prednisolone in HAART-treated patients in almost all analyzed parameters in this study, this may well be attributed to the heterogeneity of the patients (differences in CD4 counts, time to infection), differences in the composition of individual antiretroviral regimens that may have different effects on immune activation or to the low dose of prednisolone used. Further studies specifically addressing the effects of prednisolone in the presence of HAART looking at more homogenous patient groups are therefore needed. Studying the effects of immune modulators in the presence of HAART (in contrast to immune modulators preceding HAART) is of particular interest, since there is residual immune activation in HAART-treated patients associated with accelerated disease progression .
In countries with lesser access to (and therefore later onset of) HAART, prednisolone medication preceding antiretroviral treatment might be considered as a supplement regimen in HIV therapy, provided that prednisolone-mediated immune modulation is safe and translates into a clinical benefit for the patients. In order to address this question, we are currently performing a randomized, double-blinded, placebo-controlled clinical trial to assess the effects of low-dose prednisolone on immune activation and HIV disease progression in otherwise treatment-naïve patients, recruiting 326 patients in Tanzania (trial name: "ProCort1"; registry: http://ClinicalTrials.gov; registration number: NCT01299948). Recruitment is now completed and first results will be available soon.