Rifaximin Improves Thrombocytopenia in Alcoholic Cirrhosis
Rifaximin Improves Thrombocytopenia in Alcoholic Cirrhosis
The baseline clinical and laboratory characteristics of the three study groups are shown in Table 1. Thrombocytopenia was mild in 53 and 60% of treated and untreated patients respectively. Anaemia and leukopenia were present in 17 (74%) and 6 (26%) patients with thrombocytopenia. None of the patients without thrombocytopenia had anaemia or leukopenia. Patients with thrombocytopenia had significantly higher prevalence of ascites, varices and encephalopathy compared with those without thrombocytopenia. Serum levels of bilirubin and albumin, prothrombin time and Child-Pugh score were also significantly higher in patients with thrombocytopenia than in those without thrombocytopenia. Platelet counts were inversely and significantly correlated with Child-Pugh score (r = 0.625; P = 0.001) and spleen size (r = −0.455; P = 0.02) in patients with thrombocytopenia. No significant correlations were found between platelet counts and Child-Pugh score or spleen size in patients with normal platelet counts (data not shown). No side effects were reported by the use of rifaximin.
Relation to Platelet Counts Patients with thrombocytopenia had significantly higher baseline mean endotoxin and IL-6 levels compared with those without thrombocytopenia (2.76 ± 0.69 vs. 0.64 ± 0.09 EU/ml; P < 0.001 and 24.26 ± 3.38 vs. 2.66 ± 0.74 pg/ml; P = 0.001 respectively) (Fig. 1). Mean IL-1 and TNF-α levels in patients with thrombocytopenia were 4.9 ± 1.9 and 6.2 ± 2 pg/ml respectively; only two patients without thrombocytopenia had detectable levels of IL-1 and TNF-α. Platelet counts were inversely and significantly correlated with the levels of endotoxin and IL-6 in patients with thrombocytopenia (r = −0.589; P = 0.003 and r = −0.464; P = 0.02 respectively) (Fig. 2), but not in those without thrombocytopenia. Inverse correlations close to statistical significance were noted between platelet counts and the levels of IL-1 and TNF-α in patients with thrombocytopenia (r = −0.402; P = 0.08 and r = −0.415; P = 0.07 respectively).
(Enlarge Image)
Figure 1.
Plasma endotoxin levels (A) and serum interleukin-6 levels (B) in cirrhotic patients with (n = 23) and without (n = 10) thrombocytopenia. The markers represent the mean value. The whiskers represent standard error of the mean.
(Enlarge Image)
Figure 2.
Correlation between platelet counts and plasma endotoxin levels (A) and between platelet counts and serum interleukin-6 levels (B) in cirrhotic patients with thrombocytopenia (n = 23).
Relation to Liver Disease Severity Endotoxin levels were significantly correlated with Child-Pugh score in patients with thrombocytopenia (r = 0.685; P = 0.0003); significant correlations were also found between IL-1, IL-6 and TNF-α levels and the Child-Pugh score (r = −0.476; P = 0.01, r = −0.512; P = 0.01, and r = −0.438; P = 0.03 respectively) in these patients.
Mean platelet count increased significantly after the administration of rifaximin (83 100 ± 9700 vs. 99 600 ± 11 200/μl; P = 0.006), whereas it remained unaltered in untreated patients (85 200 ± 9200 vs. 91 300 ± 10 100/μl) (Fig. 3A). In total, platelet counts increased in 11 patients treated with rifaximin (84.6%) (Fig. 3). The median increase of platelet counts was 16.3% (−10.4 − 89.8) at the end of this study. Platelet counts increased by 24.4% (−10.4 − 89.8) in patients with moderate/severe thrombocytopenia and by 15.6% (−4.5 − 27.7) in patients with mild thrombocytopenia, but the difference did not reach statistical significance. Mean endotoxin levels fell significantly after rifaximin treatment (1.28 ± 0.41 vs. 2.54 ± 0.86 EU/ml; P = 0.005), but were not changed in untreated patients (Fig. 4A). In total, endotoxin levels decreased in 12 patients treated with rifaximin (92.3%) (Fig. 4B). Rifaximin induced significant decreases in mean IL-1, IL-6 and TNF-α levels (Table 2; no significant changes in endotoxin and cytokine levels were noted in untreated patients (data not shown). The changes in platelet counts were significantly correlated with the changes in endotoxin levels (r = −0.573; P = 0.04). Correlations close to significance were noted between platelet count changes and changes in TNF-α (r = −0.554; P = 0.05) and IL-6 levels (r = −0.495; P = 0.07) (Fig. 5).
(Enlarge Image)
Figure 3.
(A) Mean platelet count at baseline and after 4 weeks in cirrhotic patients treated with rifaximin (n = 13) and in those who received no treatment (n = 10). The markers represent the mean value. The whiskers represent standard error of the mean, (B) individual changes of platelet counts after rifaximin treatment.
(Enlarge Image)
Figure 4.
(A) Plasma endotoxin levels at baseline and after 4 weeks in cirrhotic patients treated with rifaximin (n = 13) and in those who received no treatment (n = 10). The markers represent the mean value. The whiskers represent standard error of the mean, (B) individual changes of endotoxin levels after rifaximin treatment.
(Enlarge Image)
Figure 5.
Correlation between changes in platelet counts and changes in plasma levels of endotoxin (A) and serum levels of tumour necrosis factor-α (B) and interleukin-6 (C) in cirrhotic patients with thrombocytopenia after rifaximin treatment (n = 13).
Rifaximin increased Hb levels in patients with anaemia (n = 9) and WBC counts in patients with leukopenia (n = 4); both changes reached closed to statistical significance (10.87 ± 0.62 vs. 10.55 ± 0.67 g/dl; P = 0.05 and 3170 ± 390 vs. 2610 ± 320/μl; P = 0.08 respectively). Hb levels and WBC counts did not change in patients without anaemia or leukopenia treated with rifaximin. Hb levels and WBC counts did not significantly change in untreated patients with or without anaemia or without leukopenia (only two patients had leukopenia in this group).
Results
Clinical and Laboratory Characteristics
The baseline clinical and laboratory characteristics of the three study groups are shown in Table 1. Thrombocytopenia was mild in 53 and 60% of treated and untreated patients respectively. Anaemia and leukopenia were present in 17 (74%) and 6 (26%) patients with thrombocytopenia. None of the patients without thrombocytopenia had anaemia or leukopenia. Patients with thrombocytopenia had significantly higher prevalence of ascites, varices and encephalopathy compared with those without thrombocytopenia. Serum levels of bilirubin and albumin, prothrombin time and Child-Pugh score were also significantly higher in patients with thrombocytopenia than in those without thrombocytopenia. Platelet counts were inversely and significantly correlated with Child-Pugh score (r = 0.625; P = 0.001) and spleen size (r = −0.455; P = 0.02) in patients with thrombocytopenia. No significant correlations were found between platelet counts and Child-Pugh score or spleen size in patients with normal platelet counts (data not shown). No side effects were reported by the use of rifaximin.
Baseline Endotoxin and Cytokine Measurements
Relation to Platelet Counts Patients with thrombocytopenia had significantly higher baseline mean endotoxin and IL-6 levels compared with those without thrombocytopenia (2.76 ± 0.69 vs. 0.64 ± 0.09 EU/ml; P < 0.001 and 24.26 ± 3.38 vs. 2.66 ± 0.74 pg/ml; P = 0.001 respectively) (Fig. 1). Mean IL-1 and TNF-α levels in patients with thrombocytopenia were 4.9 ± 1.9 and 6.2 ± 2 pg/ml respectively; only two patients without thrombocytopenia had detectable levels of IL-1 and TNF-α. Platelet counts were inversely and significantly correlated with the levels of endotoxin and IL-6 in patients with thrombocytopenia (r = −0.589; P = 0.003 and r = −0.464; P = 0.02 respectively) (Fig. 2), but not in those without thrombocytopenia. Inverse correlations close to statistical significance were noted between platelet counts and the levels of IL-1 and TNF-α in patients with thrombocytopenia (r = −0.402; P = 0.08 and r = −0.415; P = 0.07 respectively).
(Enlarge Image)
Figure 1.
Plasma endotoxin levels (A) and serum interleukin-6 levels (B) in cirrhotic patients with (n = 23) and without (n = 10) thrombocytopenia. The markers represent the mean value. The whiskers represent standard error of the mean.
(Enlarge Image)
Figure 2.
Correlation between platelet counts and plasma endotoxin levels (A) and between platelet counts and serum interleukin-6 levels (B) in cirrhotic patients with thrombocytopenia (n = 23).
Relation to Liver Disease Severity Endotoxin levels were significantly correlated with Child-Pugh score in patients with thrombocytopenia (r = 0.685; P = 0.0003); significant correlations were also found between IL-1, IL-6 and TNF-α levels and the Child-Pugh score (r = −0.476; P = 0.01, r = −0.512; P = 0.01, and r = −0.438; P = 0.03 respectively) in these patients.
Effects of Rifaximin on Thrombocytopenia and Endotoxin and Cytokine Levels
Mean platelet count increased significantly after the administration of rifaximin (83 100 ± 9700 vs. 99 600 ± 11 200/μl; P = 0.006), whereas it remained unaltered in untreated patients (85 200 ± 9200 vs. 91 300 ± 10 100/μl) (Fig. 3A). In total, platelet counts increased in 11 patients treated with rifaximin (84.6%) (Fig. 3). The median increase of platelet counts was 16.3% (−10.4 − 89.8) at the end of this study. Platelet counts increased by 24.4% (−10.4 − 89.8) in patients with moderate/severe thrombocytopenia and by 15.6% (−4.5 − 27.7) in patients with mild thrombocytopenia, but the difference did not reach statistical significance. Mean endotoxin levels fell significantly after rifaximin treatment (1.28 ± 0.41 vs. 2.54 ± 0.86 EU/ml; P = 0.005), but were not changed in untreated patients (Fig. 4A). In total, endotoxin levels decreased in 12 patients treated with rifaximin (92.3%) (Fig. 4B). Rifaximin induced significant decreases in mean IL-1, IL-6 and TNF-α levels (Table 2; no significant changes in endotoxin and cytokine levels were noted in untreated patients (data not shown). The changes in platelet counts were significantly correlated with the changes in endotoxin levels (r = −0.573; P = 0.04). Correlations close to significance were noted between platelet count changes and changes in TNF-α (r = −0.554; P = 0.05) and IL-6 levels (r = −0.495; P = 0.07) (Fig. 5).
(Enlarge Image)
Figure 3.
(A) Mean platelet count at baseline and after 4 weeks in cirrhotic patients treated with rifaximin (n = 13) and in those who received no treatment (n = 10). The markers represent the mean value. The whiskers represent standard error of the mean, (B) individual changes of platelet counts after rifaximin treatment.
(Enlarge Image)
Figure 4.
(A) Plasma endotoxin levels at baseline and after 4 weeks in cirrhotic patients treated with rifaximin (n = 13) and in those who received no treatment (n = 10). The markers represent the mean value. The whiskers represent standard error of the mean, (B) individual changes of endotoxin levels after rifaximin treatment.
(Enlarge Image)
Figure 5.
Correlation between changes in platelet counts and changes in plasma levels of endotoxin (A) and serum levels of tumour necrosis factor-α (B) and interleukin-6 (C) in cirrhotic patients with thrombocytopenia after rifaximin treatment (n = 13).
Effects of Rifaximin on Co-existing Anaemia and Leukopenia
Rifaximin increased Hb levels in patients with anaemia (n = 9) and WBC counts in patients with leukopenia (n = 4); both changes reached closed to statistical significance (10.87 ± 0.62 vs. 10.55 ± 0.67 g/dl; P = 0.05 and 3170 ± 390 vs. 2610 ± 320/μl; P = 0.08 respectively). Hb levels and WBC counts did not change in patients without anaemia or leukopenia treated with rifaximin. Hb levels and WBC counts did not significantly change in untreated patients with or without anaemia or without leukopenia (only two patients had leukopenia in this group).
Source...