Changes in Host Immune-Endocrine Relationships during Tuberculosis Treatment in Patients with Cured and Failed Treatment Outcomes

Abstract

A bidirectional communication between the immune and endocrine systems exists and facilitates optimum responses in the host during infections. This is in part achieved through changes in secretion patterns of hypothalamic hormones induced by inflammatory cytokines. The aim of this study was to elucidate the immune-endocrine alterations during tuberculosis (TB) treatment in patients with cured and failed TB treatment outcomes. Blood samples were collected from 27 cured and 10 failed patients and hormone as well as cytokine concentrations quantified at baseline, week 4, and month 6 of TB treatment. Hormone profiles of the two treatment outcome groups were different from each other prior to as well as during TB treatment. Treatment response effects were observed for cortisol, estradiol, T3, T4 ghrelin, leptin, amylin, adiponectin, and dehydroepiandrosterone (DHEA). Trends suggest that T4, amylin, and DHEA concentrations were different between treatment outcomes, although these did not reach statistical significance. Relationships between endocrine and inflammatory markers and the biological pathways involved differed between cured and failed treatment patients. These results highlight the complex interaction between the endocrine and immune system during active TB disease and throughout treatment and suggest that endocrine markers in conjunction with inflammatory markers may be useful in predicting unfavorable treatment outcomes.

Keywords: biomarkers; cytokines; endocrine; immune; metabolic hormones; steroid hormones; tuberculosis.

Figure 1

Differential expression of plasma hormone concentrations before (baseline), during (week 4), and after tuberculosis treatment (month 6). Hormone concentrations, measured by Luminex (cortisol, T3, T4, ghrelin, leptin, and estradiol) and ELISA [dehydroepiandrosterone (DHEA), amylin (Tot), adiponectin, and growth hormone], were clustered according to group and time point (cured group n = 25 and failed group n = 10). Qlucore Omics explorer software was used to generate the heat maps around the mean hormone concentrations of the log-transformed data. Bright yellow indicates +2 (fourfold) upregulation from the mean (black) and bright blue indicates a −2 downregulation from the mean.

Figure 2

Changes in circulating hormone concentrations during the course of tuberculosis (TB) treatment. Plasma from TB patients (males and females) who were cured (n = 25) and who failed (n = 10) TB treatment were collected before (baseline) and during TB treatment (week 4 and month 6). Plasma collected was subjected to hormone measurement by Luminex analysis [cortisol (A), T3 (D), T4 (E), ghrelin (F), leptin (G), and estradiol (B)] and ELISA [dehydroepiandrosterone (DHEA) (J), amylin (Tot) (H), adiponectin (I), and growth hormone (C)]. Data were analyzed by a linear mixed-effects model of hormone concentrations by treatment outcome and time point as fixed effects and patient ID as random effect. Results are presented as means of the log-transformed data with SD. A p-value of < 0.05 was regarded as significantly different. *p-value < 0.05; **p-value < 0.01; ***p-value < 0.001.

Figure 3

Correlations between plasma hormone concentrations of tuberculosis patients who were cured and who failed treatment at baseline, week 4, and month 6. Hormone concentrations were measured by Luminex analysis (cortisol, T3, T4, ghrelin, leptin, and estradiol) and ELISA [dehydroepiandrosterone (DHEA), amylin (Tot), adiponectin, and growth hormone] (cured: n = 25; failed: n = 10). Average Pearson correlation coefficients and p-values of 50 imputed datasets are presented. Correlation coefficients are expressed as percentages in the colored squares (bottom-left section) and the p-values in the white squares (top-right section). The colorimetric scale represents the correlation coefficient where dark blue squares indicate strong positive correlations (+1) and dark red strong negative correlations (−1).

Figure 3

Correlations between plasma hormone concentrations of tuberculosis patients who were cured and who failed treatment at baseline, week 4, and month 6. Hormone concentrations were measured by Luminex analysis (cortisol, T3, T4, ghrelin, leptin, and estradiol) and ELISA [dehydroepiandrosterone (DHEA), amylin (Tot), adiponectin, and growth hormone] (cured: n = 25; failed: n = 10). Average Pearson correlation coefficients and p-values of 50 imputed datasets are presented. Correlation coefficients are expressed as percentages in the colored squares (bottom-left section) and the p-values in the white squares (top-right section). The colorimetric scale represents the correlation coefficient where dark blue squares indicate strong positive correlations (+1) and dark red strong negative correlations (−1).

Figure 4

Bi-plots depicting plasma hormone and serum inflammatory marker interaction. Interactions between hormones and inflammatory markers were determined before tuberculosis treatment (baseline) in patients who were cured (n = 25) and those who failed (n = 10) treatment. The principal component analysis (PCA) was done using the prcomp function in R and the bi-plot generated using the factoextra package.

Figure 4

Bi-plots depicting plasma hormone and serum inflammatory marker interaction. Interactions between hormones and inflammatory markers were determined before tuberculosis treatment (baseline) in patients who were cured (n = 25) and those who failed (n = 10) treatment. The principal component analysis (PCA) was done using the prcomp function in R and the bi-plot generated using the factoextra package.