Leaky Gut Syndrome Is Associated with Endotoxemia and Serum (1→3)-β-D-Glucan in Severe Dengue Infection

Abstract

The hallmark of severe dengue infection is the increased vascular permeability and hemodynamic alteration that might be associated with an intestinal permeability defect. However, the mechanisms underlying the gastrointestinal-related symptoms of dengue are not well characterized. A prospective observational study was conducted on patients with dengue who were categorized according to: (i) febrile versus critical phase and (ii) hospitalized patients with versus without the warning signs to evaluate the gut barrier using lactulose-to-mannitol excretion ratio (LEMR). Serum endotoxins, (1→3)-β-D-glucan (BG), and inflammatory parameters were measured. A total of 48 and 38 patients were enrolled in febrile illness and critical phase, respectively, while 22 and 64 patients presented with or without the warning signs, respectively. At enrollment, a positive LEMR test was found in 20 patients (91%) with warning signs, regardless of phase of infection. Likewise, serum endotoxins and BG, the indirect biomarkers for leaky gut, prominently increased in patients who developed severe dengue when compared with the non-severe dengue (endotoxins, 399.1 versus 143.4 pg/mL (p < 0.0001); BG, 123 versus 73.8 pg/mL (p = 0.016)). Modest impaired intestinal permeability occurred in dengue patients, particularly those with warning signs, and were associated with endotoxemia and elevated BG. Thus, leaky gut syndrome might be associated with severity of dengue infection.

Keywords: dengue; endotoxins; intestinal permeability; lactulose-to-mannitol excretion ratio; leaky gut; β-D-glucan.

Figure 1

The schematic diagram represents (A) study flow diagram and (B) timeline of the study design.

Figure 2

Urinary excretion of mannitol and lactulose from participants among three groups (healthy control subjects (n = 5) and dengue patients with (n = 22) or without warning signs (n = 64)). (A) The percentages (%) of urinary mannitol and lactulose excretion demonstrated significantly higher lactulose excretion in dengue patients with warning signs than in those without warning signs, but there was no significant difference in mannitol excretion among the groups. (B) Comparisons of the lactulose-to-mannitol excretion ratio (LMER) demonstrated by the violin plot again demonstrated a significantly higher LMER in dengue patients with warning signs than in dengue patients without warning signs (20 from 22 versus 11 from 64 participants). The transverse dotted line demonstrates the median and quartiles. A dot refers to an individual value. The light blue area indicates the 90th percentile of the upper normal limits (0.1187) as a reference and serves as the cutoff value for a positive LMER test. (**** p < 0.0001 and ns stands for nonsignificant.)

Figure 3

Ranges of measured lipopolysaccharide-binding protein (LBP), (1→3)-β-D-glucan (BG), cytokines, and lipopolysaccharides (LPS). The detection range of tested serum samples on 1 (t = 0) and 7 (t = 7) of admission from patients who developed severe dengue (red, n = 27) and nonsevere dengue (green, n = 59) in comparison with serum samples from healthy control subjects (black, n = 5) is demonstrated. Heavy bars indicate medians, and error bars represent 95% CIs. Each value is indicated by a dot. Log transforms with nonparametric statistics were performed as skewed data. (A) demonstrates a tendency of higher LBP in patients who underwent severe dengue infection compared with those with nonsevere dengue and healthy control subjects at enrollment (significantly decreased LBP in severe dengue patients on day 7). (B) demonstrates significantly increased BG levels in severe dengue patients compared with nonsevere dengue patients. (C,D) demonstrate significantly higher levels of interleukin (IL)-6, IL-8, IL-1β, and TNF-α in severe dengue patients than in nonsevere dengue patients and healthy control subjects. (E) demonstrates significantly increased circulating LPS in severe dengue infection at enrollment before antibiotic treatment. (F) demonstrates that most severe dengue patients had a positive lactulose-to-mannitol excretion ratio (LMER) corresponding to the presence of warning signs at enrollment. (**** p < 0.0001, *** p < 0.001, ** p < 0.01, and * p < 0.05).

Figure 3

Ranges of measured lipopolysaccharide-binding protein (LBP), (1→3)-β-D-glucan (BG), cytokines, and lipopolysaccharides (LPS). The detection range of tested serum samples on 1 (t = 0) and 7 (t = 7) of admission from patients who developed severe dengue (red, n = 27) and nonsevere dengue (green, n = 59) in comparison with serum samples from healthy control subjects (black, n = 5) is demonstrated. Heavy bars indicate medians, and error bars represent 95% CIs. Each value is indicated by a dot. Log transforms with nonparametric statistics were performed as skewed data. (A) demonstrates a tendency of higher LBP in patients who underwent severe dengue infection compared with those with nonsevere dengue and healthy control subjects at enrollment (significantly decreased LBP in severe dengue patients on day 7). (B) demonstrates significantly increased BG levels in severe dengue patients compared with nonsevere dengue patients. (C,D) demonstrate significantly higher levels of interleukin (IL)-6, IL-8, IL-1β, and TNF-α in severe dengue patients than in nonsevere dengue patients and healthy control subjects. (E) demonstrates significantly increased circulating LPS in severe dengue infection at enrollment before antibiotic treatment. (F) demonstrates that most severe dengue patients had a positive lactulose-to-mannitol excretion ratio (LMER) corresponding to the presence of warning signs at enrollment. (**** p < 0.0001, *** p < 0.001, ** p < 0.01, and * p < 0.05).

Figure 4

The relationship between the lactulose-to-mannitol excretion ratio (LMER) and serum lipopolysaccharides (LPS) and (1→3)-β-D-glucan (BG). (A) Increased LMER demonstrates a better correlation with serum LPS than (B) BG. The ROC (receiver operating characteristic) curves for (C) the severe dengue prediction of BG, LMER, and LPS and for (D) the severe dengue infection confirmation from BG, LMER, and LPS are shown.