Distinct T-helper cell responses to Staphylococcus aureus bacteremia reflect immunologic comorbidities and correlate with mortality

Abstract

Background: The dysregulated host immune response that defines sepsis varies as a function of both the immune status of the host and the distinct nature of the pathogen. The degree to which immunocompromising comorbidities or immunosuppressive medications affect the immune response to infection is poorly understood because these patients are often excluded from studies about septic immunity. The objectives of this study were to determine the immune response to a single pathogen (Staphylococcus aureus) among a diverse case mix of patients and to determine whether comorbidities affect immune and clinical outcomes.

Methods: Blood samples were drawn from 95 adult inpatients at multiple time points after the first positive S. aureus blood culture. Cox proportional hazards modeling was used to determine the associations between admission neutrophil counts, admission lymphocyte counts, cytokine levels, and 90-day mortality. A nested case-control flow cytometric analysis was conducted to determine T-helper type 1 (Th1), Th2, Th17, and regulatory T-cell (Treg) subsets among a subgroup of 28 patients. In a secondary analysis, we categorized patients as either having immunocompromising disorders (human immunodeficiency virus and hematologic malignancies), receiving immunosuppressive medications, or being not immunocompromised.

Results: Higher neutrophil-to-lymphocyte count ratios and higher Th17 cytokine responses relative to Th1 cytokine responses early after infection were independently associated with mortality and did not depend on the immune state of the patient (HR 1.93, 95% CI 1.17-3.17, p = 0.01; and HR 1.13, 95% CI 1.01-1.27, p = 0.03, respectively). On the basis of flow cytometric analysis of CD4 T-helper subsets, an increasing Th17/Treg response over the course of the infection was most strongly associated with increased mortality (HR 4.41, 95% CI 1.69-11.5, p < 0.01). This type of immune response was most common among patients who were not immunocompromised. In contrast, among immunocompromised patients who died, a decreasing Th1/Treg response was most common.

Conclusions: The association of both increased Th17 responses and increased neutrophil counts relative to lymphocyte counts with mortality suggests that an overwhelming inflammatory response is detrimental. However, the differential responses of patients according to immune state suggest that immune status is an important clinical indicator that should be accounted for in the management of septic patients, as well as in the development of novel immunomodulatory therapies.

Keywords: Helper T cells; Sepsis; Staphylococcus aureus.

Fig. 1

Association between neutrophil and lymphocyte counts on the first day of Staphylococcus aureus bacteremia. Patients were categorized as non-survivors (red) or survivors (blue). Linear regression lines were generated for both survivors and non-survivors

Fig. 2

Associations between cytokine profile scores. Patients were categorized as non-survivors (red) or survivors (blue). Linear regression lines were generated for both survivors and non-survivors. Only values measured at time point 1 are displayed. a T-helper type 17 cell (Th17) vs. Th1 scores. b Th17 vs. Th2 scores. c Th1 vs. Th2 scores

Fig. 3

Neutrophil and lymphocyte counts for patients grouped by presence or absence of immunocompromising medical conditions or medications. a Neutrophil counts. b Lymphocyte counts. c Neutrophil-to-lymphocyte count ratio. Solid lines represent mean values. **p < 0.01, ***p < 0.001 by analysis of variance with Tukey’s posttest for multiple comparisons

Fig. 4

Cytokine profile scores for patients grouped by presence or absence of immunocompromising medical conditions or medications. Only values measured at time point 1 are displayed. a T-helper type 17 cell (Th17) score. b Th1 score. c Th2 score. d Th17 score/Th1 score. e Th17 score/Th2 score. f Th1 score/Th2 score. Solid lines represent mean values. p values were all > 0.05 by analysis of variance with Tukey’s posttest for multiple comparisons

Fig. 5

Trajectories of T-cell subsets. Average trends in immune markers were determined using linear mixed models. Patients who were not immunocompromised are represented by solid lines. Patients who were receiving immunosuppressive medications are represented by dashed lines. The p value is a test for differences in slopes between groups. a T-helper type 17 cells (Th17). b Th1 cells. c Th2 cells. d Regulatory T cells (Treg)

Fig. 6

Trajectories of T-helper type 17 cell (Th17)/regulatory T cell (Treg), Th1/Treg, and Th2/Treg ratios. Average trends in immune markers were determined using linear mixed models. Patients who were not immunocompromised are represented by solid lines. Patients who were receiving immunosuppressive medications are represented by dashed lines. The p value is a test for differences in slopes between groups. a, b, c All 28 patients. d, e, f Only the 14 patients who died. g, h, i Only the 14 patients who survived