Dialysis reduces portal pressure in patients with chronic hepatitis C

Abstract

The purpose of this study was to characterize changes in hepatic venous pressures in patients with chronic hepatitis C. The histology and laboratory data from patients with chronic hepatitis C who underwent a transjugular liver biopsy (TJLB) and hepatic venous pressure gradient measurement were analyzed. Portal hypertension was defined as hepatic venous pressure gradient > or =6 mm Hg. A single pathologist masked to hepatic venous pressure gradient scored liver sections for inflammation and fibrosis. The patients with high-grade inflammation (relative risk [RR] 2.82, P = 0.027, multivariate analysis) and late-stage fibrosis (RR 2.81, P = 0.022) were more likely to have a hepatic venous pressure gradient > or =6 mm Hg, while the patients on dialysis (RR 0.32, P = 0.01) were less likely to have a hepatic venous pressure gradient > or =6 mm Hg. The patients on dialysis (n = 58) had an elevated serum blood urea nitrogen and creatinine when compared with those who were not (n = 75) (47.6 +/- 3.3 and 7.98 +/- 0.4 vs. 25.9 +/- 2.0 and 1.66 +/- 0.22 mg/dL, respectively; P < 0.001). While the hepatic venous pressure gradient increased with the rising levels of liver fibrosis in the latter group (P < 0.01), it did not change in the patients on dialysis (P = 0.41). The median hepatic venous pressure gradient was especially low in late-stage fibrosis patients on dialysis when compared with the latter group (5 vs. 10 mm Hg, P = 0.017). In patients on dialysis, serum transaminases were low across all levels of fibrosis. Twenty-three of the 92 patients with early fibrosis had a hepatic venous pressure gradient > or =6 mm Hg. In patients with chronic hepatitis C, concomitant TJLB and hepatic venous pressure gradient measurement identify those who have early fibrosis and portal hypertension. Long-term hemodialysis may reduce portal pressure in these patients.

FIG. 1

Representative photomicrographs of fibrosis stages and grades of inflammation are shown. (A) Photomicrographs of trichromestained liver section with increasing stages of fibrosis (1-4) are shown. Fibrous bands are stained blue (white arrows). Expansion of fibrous septae indicates late-stage fibrosis. (B) Photomicrographs of hematoxylin and eosin-stained liver section with increasing grades of inflammation (1-4) are shown. Infiltration of hepatic parenchyma with mononuclear inflammatory cells (white arrows) indicates high-grade inflammation (see Patients and Methods section for details).

FIG. 2

Study design.

FIG. 3

Effect of fibrosis on hepatic venous pressures in patients with chronic hepatitis C. Stage of fibrosis is shown on the x-axis. WHVP (A) and HVPG (B) are shown on the y-axis. Box plots show the median HVPG, 25th to 75th percentile box, and complete range of measurements. There was significant association between stage of fibrosis and hepatic venous pressures. WHVP and the HVPG increased with the progression of fibrosis (Kruskal–Wallis test, P < 0.01 for both; n indicates the number of patients in each group).

FIG. 4

Effect of inflammation on hepatic venous pressures in patients with chronic hepatitis C. Grade of inflammation is shown on the x-axis. WHVP (A) and HVPG (B) are shown on the y-axis. Box plots show the median WHVP and HVPG, 25th to 75th percentile box, and complete range of measurements. Because the number of patients with an inflammation grade of 0 (n = 2) and 4 (n = 1) were too low to calculate the median inflammatory scores, they were grouped with the patients with inflammation grades 1 and 3, respectively. The WHVP and HVPG increased with severity of inflammation (Kruskal–Wallis test, P = 0.03 and P = 0.0002, respectively; n indicates the number of patients in each group).

FIG. 5

Effect of dialysis on HVPG across stages of fibrosis in patients with chronic hepatitis C. Stage of fibrosis is shown on the x-axis and HVPG is shown on the y-axis. Box plots show median HVPG, 25th to 75th percentile box, and complete range of measurements. (A) With progressive increase in fibrosis, the HVPG increased in the patients with normal renal function (solid bars, Kruskal–Wallis test, P < 0.01). (B) In the patients on dialysis, there was no difference in the HVPG across all levels of fibrosis (hatched bars, Kruskal–Wallis test, P = 0.4; n indicates the number of patients in each group).

FIG. 6

Effect of dialysis on HVPG across grades of inflammation in patients with chronic hepatitis C. Grade of inflammation is shown on the x-axis and HVPG is shown on the y-axis. Box plots show the median HVPG, 25th to 75th percentile box, and complete range of measurements. Because the number of patients with an inflammation grade of 0 (n = 2) and 4 (n = 1) were too low to calculate the median inflammatory scores, they were grouped with the patients with inflammation grades of 1 and 3, respectively. (A) With a progressive increase in inflammation, the HVPG increased in patients with normal renal function (solid bars, Kruskal–Wallis test, P = 0.01; n indicates the number of patients in each group). (B) No such change was observed in patients on dialysis (hatched bars, Kruskal–Wallis test, P = 0.09; n indicates the number of patients in each group).

FIG. 7

Impact of dialysis on HVPG among patients with early- and late-stage fibrosis as well as low- and high-grade inflammation. Stages of fibrosis (A) and grade of inflammation (B) are shown on the x-axis and HVPG is shown on the y-axis. Box plots show the median HVPG, 25th to 75th percentile box, and complete range of measurements. As shown in panel A, there is no difference in the HVPG among the patients with stage 0–2 fibrosis. However, the patients on dialysis (hatched bars) with stage 3–4 fibrosis have a significantly lower HVPG when compared with rest of the cohort (solid bars). As shown in panel B, a similar trend was observed in patients with high-grade inflammation. However, it was not statistically significant (n indicates the number of patients in each group).