bs_bs_banner
bs_bs_banner
doi:10.1111/jgh.13006
H E PAT O L O G Y
Relationship and interaction between serum sodium concentration and portal hemodynamics in patients with cirrhosis Hitoshi Maruyama, Takayuki Kondo, Soichiro Kiyono, Tadashi Sekimoto, Masanori Takahashi, Hidehiro Okugawa and Osamu Yokosuka Department of Gastroenterology and Nephrology, Chiba University Graduate School of Medicine, Chiba, Japan
Key words ascites, cirrhosis, hyponatremia, portal hypertension. Accepted for publication 14 April 2015. Correspondence Dr Hitoshi Maruyama, Department of Gastroenterology and Nephrology, Chiba University Graduate School of Medicine, 1-8-1, Inohana, Chuou-ku, Chiba 260-8670, Japan. Email: [email protected] The authors state they have no conflicts of interest.
Abstract Background and Aim: To examine the relationship between hyponatremia and portal hemodynamics and their effect on the prognosis of cirrhosis. Methods: Portal hemodynamic parameters measured by Doppler ultrasound and serum sodium concentrations were examined in 153 cirrhosis patients (mean age 62.2 ± 12.0 years; median observation period, 34.1 m). Results: Study participants included 16 patients with hyponatremia (Na < 135 mEq/L), who showed a significantly greater frequency of possessing a splenorenal shunt (SRS; P = 0.0068), and 137 patients without hyponatremia. Serum sodium concentrations were significantly lower in patients with SRS than in those without (P = 0.0193). An increased prothrombin time-international normalized ratio was a significant predictive factor for developing hyponatremia a year later (8/96; Hazard ratio 14.415; P = 0.028). The cumulative survival rate was significantly lower in patients with hyponatremia (46.7% at 1 and 3 years) than in those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001). The cumulative survival rate was significantly lower in patients who had developed hyponatremia after 1 year (100% at 1 year, 62.5% at 3 years) than those who had not (100% at 1 year, 89.0% at 3 years; P < 0.001). The cumulative survival rate was significantly worse in patients with both hyponatremia and SRS (20% at 1 year). Conclusions: There was a close linkage between the serum sodium concentration and portal hemodynamic abnormality, presence of SRS, and their interaction may negatively influence the prognoses in cirrhosis.
Introduction Cirrhosis is the most advanced stage of chronic liver disease, with varying degrees of hepatic dysfunction and impaired portal hemodynamics. Unlike when in an asymptomatic, compensated condition, decompensated patients suffer from serious complications which influence the disease’s prognosis as well as the patient’s quality of life.1 Nonetheless, predicting long-term outcomes based on clinical findings may be extremely useful for the practical management of patients with cirrhosis. Hyponatremia is frequently observed in patients with cirrhosis, showing a relationship with various clinical manifestations and a poor prognosis.2–4 A pathophysiology of hyponatremia is based on systemic hemodynamic abnormalities caused by impaired renal water retention.5–7 Since the systemic circulation may be affected by portal hypertension due to the increased intrahepatic resistance, the hyperdynamic state and the development of collateral vessels,1,8 this suggests a close link between portal hemodynamics and hyponatremia.
As a minimally invasive procedure, Doppler ultrasound (US) allows the real-time observation of blood flow under physiological conditions. Investigators have reported the efficacy of this technique for evaluating portal hemodynamics in cirrhosis patients.9,10 However, the relationship between portal parameters determined by Doppler US and hyponatremia, and their combined effect on the prognosis of cirrhosis has not been fully studied. Therefore, the aim of this study was to examine the relationship between hyponatremia and portal hemodynamics, and their interaction on the prognosis of cirrhosis.
Methods Patients. We conducted a subgroup analysis using medical records collected from patients with cirrhosis, who had participated in a prospective study regarding the influence of portal hemodynamics on clinical outcomes in cirrhosis. Subjects were recruited from among patients who underwent both upper
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1635
Hyponatremia and portal hemodynamics
H Maruyama et al.
gastrointestinal endoscopy and Doppler US for the assessment of portal hypertension, between the period November 2007 to November 2012. Cirrhosis was diagnosed by evaluating liver biopsy or blood samples, combined with findings from US, contrast-enhanced computed tomography (CT), or magnetic resonance imaging taken during routine testing for hepatocellular carcinoma (HCC). Informed written consent was obtained from all patients. The study was approved by the institutional review board and was considered to have an appropriate design for publication purposes. The work undertaken conforms to the provisions of the Declaration of Helsinki (as revised in Tokyo, 2004). Patients who were excluded from the study included those (i) with HCC detected as hypervascular nodules in the arterial phase with washout in the portal venous or late phase on contrastenhanced CT/magnetic resonance imaging11 or a treatment history of HCC; (ii) using vasoactive drugs, such as β-blockers, which are not licensed for the treatment of portal hypertension in Japan; (iii) receiving antiviral therapy prior to, or during, the study period; (iv) receiving interventional procedures such as a transjugular intrahepatic portosystemic or peritoneo-venous shunt prior to, or during, the study period; and (v) aged below 18 or over 90 years. Hyponatremia was defined as a serum sodium concentration lower than 135 mEq/L. Gastroesophageal varices were classified according to the general rules of the Japan Research Society for Portal Hypertension:12 small, medium, or large. A variceal appearance was assessed by endoscopic findings taken during the 6 months before/after a Doppler US study. Hepatic encephalopathy (HE) was assessed using a West Haven grading system,13 with grade II or above classified as overt HE. The degree of ascites was assessed based on clinical and US findings in reference to guidelines:14,15 these were labeled subclinical when ascites were only detectable by US examination, and clinical when ascites caused abdominal distension. The latter was classified as overt ascites whether or not the patients was under diuretic therapy. Any change in the degree of ascites was monitored by a US and physical examination at least twice per year. Patients with ascites were treated according to established guidelines,14,15 using diuretics with a salt-restricted diet and occasional paracentesis. The study used model for end-stage liver disease (MELD) score16 and MELD-Na score3 as clinical parameters. Subjects who participated in this study were followed up, at least every 3 months in an outpatient clinic, to determine their physical status and to collect blood samples. The observation period was defined as the time between the initial US examination and the date of the last hospital visit, for treatment of collateral vessels, liver transplantation, or patient death.
US examination. B-mode and Doppler US were performed with patients in the supine position after fasting for 4 hours or more, using an SSA-770A or 790A (Toshiba; Tokyo, Japan) with a 3.75 MHz convex probe. Spleen size (mm2) was calculated by multiplying the distance from the splenic hilum to the caudal polar angle, measured by two intersecting lines.17 The upper limit of normal used in the study was 2000 mm2. Portal blood flow measurement was performed using a pulsed Doppler method, with the sampling point at a width corresponding to the diameter of the vessel and at an angle less than 60 degrees 1636
between the US beam and the vessel. The mean velocity (cm/s) was calculated by automatic tracing of the wave spectrum, for approximately one cardiac cycle, by integration of the time-based average value of the spectrum in the portal trunk. The mean flow volume (mL/min) was calculated by multiplying the mean flow velocity for 1 second by the cross-section of the vessel, and then multiplying the calculated value by 60 s. When collateral vessel was detected, the diameter was measured by B-mode US and the flow direction was assessed by color Doppler. The data used for analysis were average values, calculated using measurements taken between two to four times. All US examinations were performed by qualified personnel, HM or MT, each with more than eight years’ of experience in this technique.
Statistical analysis. Continuous variables were compared using Student’s t-test or Mann–Whitney’s U-test. Categorical variables were compared using the chi-square test. The cumulative survival rate was calculated using the Kaplan–Meier method and the difference was compared with the log-rank test. Multivariate analysis was performed using logistic regression analysis. Cox regression analysis was used to evaluate risk factors and results were presented as a hazard ratio (HR) with a 95% confidence interval (CI). Statistical significance was defined as P < 0.05. Statistical analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC, USA).
Results Patient characteristics. There were 153 patients (age: mean ± standard deviation, 62.2 ± 12.0 years; range, 20–89; male 79, female 74); 63 were classed as Child–Pugh A, 66 as Child– Pugh B, and 24 as Child–Pugh C (Table 1). At baseline, 16 patients were affected by hyponatremia, with 137 patients unaffected; the frequency of ascites was significantly higher, and serum albumin levels were significantly lower, respectively, in the former (12/16, 3.0 ± 0.6 g/dL) than in the latter (47/137, P = 0.002; 3.3 ± 0.5 g/ dL, P = 0.023). Similarly, the Child–Pugh and the MELD scores, respectively, were significantly higher in the former (8.8 ± 2.3; 13.8 ± 5.2) than in the latter (7.2 ± 1.9, P = 0.002; 11.2 ± 3.8, P = 0.015). As for portal hemodynamic parameters, the presence of a splenorenal shunt (SRS) was significantly more frequent in the hyponatremia group (5/16, 31.3%) than in the nonhyponatremia group (12/137, 8.8%; P = 0.0068). Serum sodium concentrations were significantly lower in patients with an SRS (135.9 ± 4.8 mEq/L, 126–142) than in those without an SRS (138.9 ± 3.2 mEq/L, 125–146; P = 0.0193). The median interval time between US and endoscopic examinations was 0.2 months (0–10.6). In 103 patients with serum sodium concentration between 125 and 140 mEq/L, the clinical data were analyzed using the MELD-Na score instead of the MELD score. The MELD-Na score was significantly higher in patients with hyponatremia (19.6 ± 4.3, 13–27) than in those without (13.3 ± 4.5, 7–27; P < 0.001). In addition, the MELD-Na score was significantly higher in patients with SRS (16.9 ± 5.9, 9–27) than in those without (13.9 ± 4.7, 7–27, P = 0.036). Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642
© 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
H Maruyama et al.
Table 1
Hyponatremia and portal hemodynamics
Patient characteristics at baseline
Number of subjects Age (years) (mean ± SD [range]) Sex (male/female) Etiology (virus/alcohol/PBC/NASH/AIH/others) Ascites (−/+) Gastroesophageal varices (−/+) History of variceal treatment (−/+) History of variceal bleeding (−/+) Hepatic encephalopathy (−/+) Child–Pugh score (mean ± SD [range]) Model for end-stage liver disease score (mean ± SD [range]) Blood test Sodium (mmol/L) (mean ± SD [range]) Potassium (mmol/L) (mean ± SD [range]) Chlorine (mmol/L) (mean ± SD [range]) Bilirubin (mg/dL) (mean ± SD [range]) Albumin (g/dL) (mean ± SD [range]) Prothrombin time-INR (mean ± SD [range]) Platelet count (104/μL) (mean ± SD [range]) Portal hemodynamics Diameter of portal trunk (mm) (mean ± SD [range]) Mean velocity in the portal trunk (cm/s) (mean ± SD [range]) Mean flow volume in the portal trunk (ml/min) (mean ± SD [range]) Left gastric vein (hepatofugal) Short gastric vein (hepatofugal) Splenorenal shunts (hepatofugal) Paraumbilical vein (hepatofugal) Inferior mesenteric vein (hepatofugal) Diameter of collateral vessels > 11.2 mm (median) Spleen (cm2) (mean ± SD [range]) Portal vein thrombosis (−/+)
Non-hyponatremia (Na ≥ 135)
Hyponatremia (Na < 135)
137 62.3 ± 12.1 (40–89) 71/66 45/32/14/10/8/28 90/47 39/98 104/33 91/46 133/4 7.2 ± 1.9 (5–13) 11.2 ± 3.8 (6–26)
16 61.4 ± 10.9 (42–79) 8/8 7/3/3/0/1/2 4/12 4/12 14/2 12/4 14/2 8.8 ± 2.3 (6–13) 13.8 ± 5.2 (7–24)
139.5 ± 2.5 (135–146) 4.0 ± 0.4 (3.0–5.6) 107.3 ± 3.5 (98–119) 1.9 ± 2.3 (0.3–20.2) 3.3 ± 0.5 (1.8–4.7) 1.30 ± 0.25 (0.94–2.18) 9.8 ± 5.7 (1.7–43.9)
131.4 ± 3.0 (125–134) 4.1 ± 0.6 (3.3–5.4) 98.9 ± 4.3 (91–104) 3.2 ± 3.9 (0.5–16.9) 3.0 ± 0.6 (1.9–4.0) 1.29 ± 0.28 (0.98–1.88) 10.0 ± 5.8 (1.8–24.4)
11.4 ± 2.4 (4.9–18.3) 13.1 ± 3.0 (7.1–21.8) 856.8 ± 382.6 (130–2165) 81 (59.1%) 27 (19.7%) 12 (8.8%) 31 (22.6%) 10 (7.3%) 10 (7.3%) 29.7 ± 12.8 (10.3–69.9) 128/9
10.8 ± 1.9 (6.4–14.6) 11.7 ± 3.3 (6.3–15.6) 696.7 ± 316.5 (320–1430) 11 (68.8%) 4 (25.0%) 5 (31.3%) 5 (31.3%) 0 (0%) 1 (6.3%) 29.8 ± 18.5 (12.9–90.5) 15/1
P values
0.659 0.890 0.668 0.002 0.770 0.296 0.489 0.062 0.002 0.015 < 0.001 0.377 < 0.001 0.209 0.023 0.841 0.852 0.289 0.102 0.121 0.457 0.618 0.007 0.442 0.264 0.878 0.978 0.961
−, absence; +, presence. Hepatic encephalopathy −, grade 0∼I; +, grade II∼IV (West-Haven grading system). AIH, autoimmune hepatitis; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrhosis; SD, standard deviation.
Changes in sodium concentrations during a 1-year follow-up. Ninety-six of 137 patients without hyponatremenia at baseline were properly followed up for more than 1 year in our department; 8 of 96 patients developed hyponatremia, 1 year after the enrollment. A prothrombin time-international normalized ratio (PT-INR) was the significant baseline factor for predicting the transition from non-hyponatremia to hyponatremia (HR: 14.415; 95% CI: 1.336–155.487; P = 0.028). However, baseline MELD-Na score was not a significant factor for developing hyponatremia after 1-year follow-up. Clinical findings after the 1-year follow-up were compared between patients who had developed hyponatremia and those who had not (Table 2): the MELD score and the PT-INR were significantly higher in the former (13.1 ± 4.0; 1.47 ± 0.26) than in the latter (10.4 ± 2.9, P = 0.018; 1.25 ± 0.23, P = 0.017).
Prognosis. Thirty-eight patients died during the study period: 31 of hepatic failure, two of unknown cause, one of gastrointesti-
nal perforation, one of sepsis, one of lung cancer, one of oral cavity cancer, and one of rupture of an abdominal aortic aneurysm. In addition, nine patients received a liver transplantation. Cumulative overall survival rates were 87.3% at 1 year, 73.7% at 3 years, and 60.9% at 5 years. There were two significant predictive factors for a poor prognosis: the presence of ascites (HR 4.191; 95% CI 2.084–8.428; P < 0.001) and a low sodium concentration (HR 0.856; 95% CI 0.743–0.986; P = 0.031; Table 3). Multivariate analysis showed that the presence of ascites was the only significant factor for poor prognosis in the analysis of 103 patients using the MELD-Na (HR 3.329, 95% CI, 1.212–9.143; P = 0.02). Multivariate analysis revealed the presence of ascites was the only significant factor for mortality in patients without hyponatremia (Table 4). The cumulative survival rate was significantly lower in patients with ascites at baseline (70.3% at 1 year; 43.8% at 3 years; 34.0% at 5 years) than in those without (97.8% at 1 year; 92.5% at 3 years; 77.5% at 5 years; P < 0.001). The cumulative survival rate was significantly lower in patients with hyponatremia at baseline (46.7% at both 1 and 3 years) than
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1637
Hyponatremia and portal hemodynamics
Table 2
H Maruyama et al.
Comparison of clinical findings between patients who developed hyponatremia and those who did not over a 1-year period
Number of subjects Age (years) (mean ± SD [range]) Sex (male/female) Etiology (virus/alcohol/PBC/NASH/AIH/others) Ascites (−/+) Gastroesophageal varices (−/+) History of variceal treatment (−/+) History of variceal bleeding (−/+) Hepatic encephalopathy (−/+) Child–Pugh score (mean ± SD [range]) Model for end-stage liver disease score (mean ± SD [range]) Blood test Potassium (mmol/L) (mean ± SD [range]) Chlorine (mmol/L) (mean ± SD [range]) Bilirubin (mg/dL) (mean ± SD [range]) Albumin (g/dL) (mean ± SD [range]) Prothrombin time-INR (mean ± SD [range]) Platelet count (104/μL) (mean ± SD [range]) Portal hemodynamics Diameter of portal trunk (mm) (mean ± SD [range]) Mean velocity in the portal trunk (cm/s) (mean ± SD [range]) Mean flow volume in the portal trunk (ml/min) (mean ± SD [range]) Left gastric vein (hepatofugal) Short gastric vein (hepatofugal) Splenorenal shunts (hepatofugal) Paraumbilical vein (hepatofugal) Inferior mesenteric vein (hepatofugal) Diameter of collateral vessels > 11.2 mm (median value) Spleen (cm2) (mean ± SD [range]) Portal vein thrombosis (−/+)
Non-hyponatremia (Na ≥ 135)
Hyponatremia (Na < 135)
P values
88 61.6 ± 12.6 (20–81) 51/37 30/22/6/6/6/18 61/27 23/65 70/18 60/28 86/2 6.9 ± 1.6 (5–12) 10.4 ± 2.9 (7–19)
8 58.8 ± 13.3 (38–72) 3/5 2/1/2/0/1/2 3/5 2/6 5/3 5/3 8/0 8.8 ± 2.6 (6–13) 13.1 ± 4.0 (9–19)
0.539 0.264 0.478 0.068 0.944 0.264 0.742 0.667 0.080 0.018
4.0 ± 0.4 (3.0–5.1) 107.5 ± 3.3 (100–115) 1.5 ± 0.8 (0.3–4.5) 3.4 ± 0.5 (1.8–4.7) 1.25 ± 0.23 (0.95–2.14) 9.4 ± 4.3 (2.7–26.1)
3.9 ± 0.3 (3.4–4.4) 105.9 ± 5.4 (98–115) 2.7 ± 2.8 (0.7–9.1) 3.1 ± 0.5 (2.4–3.9) 1.47 ± 0.26 (1.16–1.90) 14.8 ± 13.6 (4.4–43.9)
0.511 0.431 0.248 0.130 0.017 0.300
11.4 ± 2.3 (4.9–18.3) 13.3 ± 3.2 (7.1–21.8) 840.4 ± 380.6 (130–2165) 50 (56.8%) 19 (21.6%) 9 (10.2%) 17 (19.3%) 5 (5.7%) 8 (9.1%) 29.3 ± 12.3 (10.3–57.2) 82/6
11.4 ± 3.9 (5.3–16.7) 11.7 ± 2.2 (9.5–14.8) 947.5 ± 307.0 (530–1230) 6 (75.0%) 1 (12.5%) 1 (12.5%) 3 (37.5%) 1 (12.5%) 2 (25.0%) 36.7 ± 14.0 (20.2–56.2) 7/1
0.954 0.242 0.503 0.318 0.544 0.840 0.225 0.446 0.158 0.115 0.554
−, absence; +, presence. Hepatic encephalopathy −, grade 0∼I; +, grade II∼IV (West-Haven grading system). AIH, autoimmune hepatitis; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrhosis; SD, standard deviation. Table 3
Cox regression analysis for the predictors of mortality Univariate
Ascites (−/+) Child–Pugh score MELD score Sodium (mmol/L) Chlorine (mmol/L) Bilirubin (mg/dL) Albumin (g/dL) Prothrombin time-INR
Multivariate
HR (95% CI)
P values
HR (95% CI)
P values
6.113 (3.248–11.506) 1.621 (1.398–1.880) 1.153 (1.078–1.233) 0.792 (0.726–0.865) 0.877 (0.824–0.932) 1.224 (1.138–1.316) 0.369 (0.227–0.600) 3.937 (1.331–11.649)
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.013
4.191 (2.084–8.428) — — 0.856 (0.743–0.986) — — — —
< 0.001 — — 0.031 — — — —
−, absence; +, presence. Child–-Pugh score was excluded from the multivariate analysis against the possibility of co-linearity. CI, confidence interval; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
in those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001; Fig. 1). In addition, the cumulative survival rate was significantly lower in patients who had developed hyponatremia 1 year after the enrollment (n = 8; 100% at 1 year; 62.5% at 3 years) than in those without (n = 88; 100% at 1 year; 89.0% at 3 years; P < 0.001; 1638
Fig. 2). The median observation period was 34.1 months (0.3–86.0). When the prognosis of cirrhosis was assessed with respect to the sodium concentration combined with the presence/absence of an SRS, the cumulative survival rate was significantly worse in
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
0.229 1.002 (0.998–1.007)
Discussion
0.431 0.156 0.303 0.354 0.667 0.517 1.070 0.935 1.087 0.729 1.554 1.407
5.143 (2.657–9.958)
1.640 (1.384–19.444)
1.186 0.886 1.273 0.397 5.857 2.630
Child–Pugh score
MELD score Chlorine (mmol/L) Bilirubin (mg/dL) Albumin (g/dL) Prothrombin time-INR Inferior mesenteric vein (hepatofugal)
(1.096–1.282) (0.805–0.974) (1.157–1.401) (0.234–0.673) (1.800–19.052) (1.022–6.773)
< 0.001 0.013 < 0.001 0.001 < 0.001 0.045
(0.904–1.266) (0.853–1.026) (0.928–1.273) (0.373–1.423) (0.209–11.560) (0.502–3.944)
— — < 0.001
0.001 3.621 (1.686–7.775) < 0.001
P values HR (95% CI) P values
Multivariate Univariate
HR (95% CI)
Non-hyponatremia (Na ≥ 135)
−, absence; +, presence. Child–Pugh score was excluded from the multivariate analysis against the possibility of co-linearity. CI, confidence interval; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
1.003 (1.001–1.006)
0.013
0.605 1.258 (0.527–3.001)
patients with both hyponatremia and the presence of an SRS (20% at 1 year) than in others (91.8% at 1 year, 76.3% at 3 years, and 63.1% at 5 years in patients with neither hyponatremia nor an SRS, P < 0.0001; 77.4% at 1 year, 72.6% at 3 years, and 58.1% at 5 years in patients with hyponatremia or an SRS, P = 0.014; Fig. 3).
Diameter of portal trunk (mm) Mean flow volume in the portal trunk (mL/min)
1.875 (1.052–3.341)
0.033
P values Multivariate
HR (95% CI) P values HR (95% CI)
Univariate
Hyponatremia and portal hemodynamics
Ascites (−/+)
Table 4
Cox regression analysis for the predictors of mortality in patients with and without hyponatremia
Hyponatremia (Na < 135)
H Maruyama et al.
Our study reveals the strong interrelationship between hyponatremia and the simultaneous presence of a portal hemodynamic abnormality, that is, an SRS; furthermore, cirrhosis patients with both showed a much poorer prognosis, similar to a negative synergistic effect. Previous studies have also suggested a potential link between the presence of an SRS and poor liver function, as exemplified by diseased liver in the decompensation stage,18,19 thus supporting our data. Although the precise mechanisms involved in the joint damaging effect of hyponatremia and portal hemodynamic dysfunction remain undefined, care should be taken with cirrhosis patients presenting with both hyponatremia and an SRS in clinical practice. Many kinds of markers exist to predict the long-term outcome of cirrhosis, including HCC, as well as Child–Pugh and MELD scores.20–23 Needless to say, the presence of esophageal varices and/or ascites are significant factors for a poor prognosis. A recent study reported that the 1-year mortality in compensated patients was 3.4% (95% CI: 1.8–5.9) in patients without varices (n = 53), and 7.3% (95% CI: 4.9–10.4) in those with varices (n = 67).24 In decompensated patients, the 1-year mortality was 21.9% (95% CI: 18.3–26.0) in patients with ascites (n = 194) and 18.1% (95% CI: 14.5–22.3) in those with variceal bleeding (n = 129). A more recent study has shown that the 5-year mortality was 1.5% in compensated patients without esophagogastric varices (n = 202), 10% in compensated patients with varices (n = 216), 20% in those with bleeding and without other complications (n = 75), 30% in those with a first, non-bleeding decompensating event (n = 206), and 88% in those with a second decompensating event (n = 213), respectively (P < 0.0001 for all).25 In regard to these data, the presence of hyponatremia, as demonstrated by our study (cumulative survival rate, 46.7% at 1 and 3 years), should be considered a sign of serious condition that becomes critical when combined with an SRS (cumulative survival rate, 20% at 1 year). It is reported that the MELD-Na was superior to the MELD to predict mortality, being an important predictor of survival among candidates for liver transplantation.3 The MELD-Na offers additional advantages over the MELD in the prognostic accuracy in patients with acute decompensated hepatitis,26 and postoperative 90-day mortality in cirrhosis after surgical procedures.27 The present study detected that the MELD-Na score showed a significant difference according to the presence/absence of SRS, indicating a close linkage with the portal hemodynamic abnormality. Our study may also be the first to report the effect of timerelated changes in sodium concentration, over a 1-year interval, on the prognosis of cirrhosis. As expected, the transition from nonhyponatremia to hyponatremia suggests a worsening condition, with the PT-INR a significant predictive factor for this change. Our results clearly demonstrate the importance of such parameters as sodium concentration and PT-INR as useful biomarkers in the management of cirrhosis patients.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1639
Hyponatremia and portal hemodynamics
H Maruyama et al.
Figure 1 Cumulative survival rate was significantly lower in patients with hyponatremia at baseline (46.7% at 1 year and 3 years) than those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001). Solid line, patients without hyponatremia; dashed line, patients with hyponatremia. Median observation period, 34.1 months (0.3–86.0).
Figure 2 Cumulative survival rate was significantly lower in patients who had developed hyponatremia 1 year after the enrollment (n = 8; 100% at 1 year, 62.5% at 3 years) than those without (n = 88; 100% at 1 year, 89.0% at 3 years; P < 0.001). Solid line, patients without development of hyponatremia; dashed line, patients with development of hyponatremia.
Hyponatremia may be the most common electrolyte abnormality seen in daily medical practice and is generally defined as a serum sodium concentration lower than 136 mEq/L.6,28 However, there is another argument that the definition of hyponatremia should be set at a sodium concentration < 130 mEq/L in patients with cirrhosis.7,29 The frequency of hyponatremia varies considerably, even with the same definition: 20.8%–49.4% for Na < 135 mEq/L,3,30 15–30% for Na 130–135 mEq/L,31 and 7–30% for Na < 130 mEq/L.5,31 These data strongly suggest a population dependency in the distribution of impaired electrolytes; patient background including etiology, severity of liver disease, and the studied lesion may affect the lower limit of sodium concentration. In addition, as it is reported that even mild, chronic 1640
hyponatremia may induce long-term adverse effects,32 the definition of hyponatremia should be based on the study design. Increased production of arginine vasopression, caused by a nonosmotic hypersecretion related to circulatory dysfunction, is the main reason for hyponatremia in patients with advanced cirrhosis.2 In particular, excessive, solute-free water retention is a key pathophysiological feature of impaired water balance, resulting in hyponatremia as well as ascites, the so-called dilutional hyponatremia.2,33 In fact, a higher frequency of ascites in patients with hyponatremia was detected in the present study and the presence of ascites was a significant factor for poor prognosis even in the analysis using the MELD-Na score instead of the MELD score; however, the presence of ascites is no longer a significant factor of
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
H Maruyama et al.
Hyponatremia and portal hemodynamics
Figure 3 Cumulative survival rate was worse in patients with both hyponatremia and the presence of an SRS (20% at 1 year) than the others (91.8% at 1 year, 76.3% at 3 years, and 63.1% at 5 years in patients with both non-hyponatremia and absence of an SRS, P < 0.0001; 77.4% at 1 year, 72.6% at 3 years and 58.1% at 5 years in patients with hyponatremia or the presence of an SRS, P = 0.014). Solid line, patients with neither hyponatremia nor an SRS; dashed line, patients with hyponatremia or the presence of an SRS; dashed-dotted line, patients with both hyponatremia and the presence of an SRS. SRS, splenorenal shunt.
mortality in patients with hyponatremia, suggesting a common pathophysiology between hyponatremia and ascites. There were some limitations to our study. Firstly, the number of patients was relatively small, and there may also be a potential bias in the etiology of liver disease since the study was undertaken in Asia. In addition, as we did not collect data relating to the hepatic venous pressure gradient, the influence of serum sodium concentration on portal pressure remains unclear. More importantly, since this was not a prospective study, further validation is needed with a larger patient population. In conclusion, the present study clearly demonstrates that there is a close association between serum sodium concentration and portal hemodynamic abnormality, that is, the presence of an SRS. The interaction between hyponatremia and the presence of an SRS suggests a deleterious influence on a prognosis of cirrhosis. These factors require strict monitoring during the daily care of cirrhosis patients so as to provide more prudent patient management.
References 1 Tsochatzis EA, Bosch J, Burroughs AK. Liver cirrhosis. Lancet 2014; 383: 1749–61. 2 Gines P, Guevara M. Hyponatremia in cirrhosis: pathogenesis, clinical significance, and management. Hepatology 2008; 48: 1002–10. 3 Kim WR, Biggins SW, Kremers WK et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N. Engl. J. Med. 2008; 359: 1018–26. 4 Ahluwalia V, Heuman DM, Feldman G et al. Correction of hyponatraemia improves cognition, quality of life, and brain oedema in cirrhosis. J. Hepatol. 2015; 62: 75–82. 5 Gines P, Berl T, Bernardi M et al. Hyponatremia in cirrhosis: from pathogenesis to treatment. Hepatology 1998; 28: 851–64. 6 Adrogue‘ HJ, Madias NE. Hyponatremia. N. Engl. J. Med. 2000; 342: 1581–9. 7 Gianotti RJ, Cardenas A. Hyponatraemia and cirrhosis. Gastroenterol. Rep. (Oxf) 2014; 2: 21–6.
8 Sanyal AJ, Bosch J, Blei A, Arroyo V. Portal hypertension and its complications. Gastroenterology 2008; 134: 1715–28. 9 Bolondi L, Li Bassi S, Gaiani S, Barbara L. Doppler flowmetry in portal hypertension. J. Gastroenterol. Hepatol. 1990; 5: 459–67. 10 Baik SK. Haemodynamic evaluation by Doppler ultrasonography in patients with portal hypertension: a review. Liver Int. 2010; 30: 1403–13. 11 Bruix J, Sherman M. American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53: 1020–2. 12 The Japan Society for Portal Hypertension. The General Rules for Study of Portal Hypertension, 3rd edn. Tokyo: Kanehara, 2013. 13 Ferenci P, Lockwood A, Mullen K, Tarter R, Weissenborn K, Blei AT. Hepatic encephalopathy—definition, nomenclature, diagnosis, and quantification: final report of the working party at the 11th World Congresses of Gastroenterology, Vienna, 1998. Hepatology 2002; 35: 716–21. 14 Moore KP, Aithal GP. Guidelines on the management of ascites in cirrhosis. Gut 2006 ; 55 (Suppl. 6): vi1–12. 15 Runyon BA. Introduction to the revised American Association for the Study of Liver Diseases Practice Guideline management of adult patients with ascites due to cirrhosis 2012. Hepatology 2013; 57: 1651–3. 16 Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31: 864–71. 17 Maruyama H, Kamezaki H, Kondo T, Sekimoto T, Takahashi M, Yokosuka O. Sonographic and clinical features of collateral vessels at the splenic hilum in cirrhosis. Clin. Radiol. 2014; 69: e140–5. 18 Ohnishi K, Sato S, Saito M et al. Clinical and portal hemodynamic features in cirrhotic patients having a large spontaneous splenorenal and/or gastrorenal shunt. Am. J. Gastroenterol. 1986; 81: 450–5. 19 Tarantino G, Citro V, Conca P et al. What are the implications of the spontaneous spleno-renal shunts in liver cirrhosis? BMC Gastroenterol. 2009; 9: 89. 20 Kamath PS, Wiesner RH, Malinchoc M et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33: 464–70.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1641
Hyponatremia and portal hemodynamics
H Maruyama et al.
21 Said A, Williams J, Holden J et al. Model for end stage liver disease score predicts mortality across a broad spectrum of liver disease. J. Hepatol. 2004; 40: 897–903. 22 D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J. Hepatol. 2006; 44: 217–31. 23 Bruno S, Zuin M, Crosignani A et al. Predicting mortality risk in patients with compensated HCV-induced cirrhosis: a long-term prospective study. Am. J. Gastroenterol. 2009; 104: 1147–58. 24 Zipprich A, Garcia-Tsao G, Rogowski S, Fleig WE, Seufferlein T, Dollinger MM. Prognostic indicators of survival in patients with compensated and decompensated cirrhosis. Liver Int. 2012; 32: 1407–14. 25 D’Amico G, Pasta L, Morabito A et al. Competing risks and prognostic stages of cirrhosis: a 25-year inception cohort study of 494 patients. Aliment. Pharmacol. Ther. 2014; 39: 1180–93. 26 Hsu CY, Lin HC, Huang YH et al. Comparison of the model for end-stage liver disease (MELD), MELD-Na and MELD-Na for outcome prediction in patients with acute decompensated hepatitis. Dig. Liver Dis. 2010; 42: 137–42. 27 Cho HC, Jung HY, Sinn DH et al. Mortality after surgery in patients with liver cirrhosis: comparison of Child–Turcotte–Pugh, MELD and MELDNa score. Eur. J. Gastroenterol. Hepatol. 2011; 23: 51–9.
1642
28 Upadhyay A, Jaber BL, Madias NE. Incidence and prevalence of hyponatremia. Am. J. Med. 2006; 119: 30–5. 29 European Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J. Hepatol. 2010; 53: 397–417. 30 Gaglio P, Marfo K, Chiodo IIIJ. Hyponatremia in cirrhosis and end-stage liver disease: treatment with the vasopressin V2-receptor antagonist tolvaptan. Dig. Dis. Sci. 2012; 57: 2774–85. 31 Corona G, Giuliani C, Parenti G et al. Moderate hyponatremia is associated with increased risk of mortality: evidence from a meta-analysis. PLoS One 2013; 18: e80451. 32 Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild chronic hyponatremia is associated with falls, unsteadiness and attention deficits. Am. J. Med. 2006; 119: 71.e1-8. 33 Angeli P, Wong F, Watson H, Ginès P. CAPPS investigators. Hyponatremia in cirrhosis: results of a patient population survey. Hepatology 2006; 44: 1535–42.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
bs_bs_banner
doi:10.1111/jgh.13006
H E PAT O L O G Y
Relationship and interaction between serum sodium concentration and portal hemodynamics in patients with cirrhosis Hitoshi Maruyama, Takayuki Kondo, Soichiro Kiyono, Tadashi Sekimoto, Masanori Takahashi, Hidehiro Okugawa and Osamu Yokosuka Department of Gastroenterology and Nephrology, Chiba University Graduate School of Medicine, Chiba, Japan
Key words ascites, cirrhosis, hyponatremia, portal hypertension. Accepted for publication 14 April 2015. Correspondence Dr Hitoshi Maruyama, Department of Gastroenterology and Nephrology, Chiba University Graduate School of Medicine, 1-8-1, Inohana, Chuou-ku, Chiba 260-8670, Japan. Email: [email protected] The authors state they have no conflicts of interest.
Abstract Background and Aim: To examine the relationship between hyponatremia and portal hemodynamics and their effect on the prognosis of cirrhosis. Methods: Portal hemodynamic parameters measured by Doppler ultrasound and serum sodium concentrations were examined in 153 cirrhosis patients (mean age 62.2 ± 12.0 years; median observation period, 34.1 m). Results: Study participants included 16 patients with hyponatremia (Na < 135 mEq/L), who showed a significantly greater frequency of possessing a splenorenal shunt (SRS; P = 0.0068), and 137 patients without hyponatremia. Serum sodium concentrations were significantly lower in patients with SRS than in those without (P = 0.0193). An increased prothrombin time-international normalized ratio was a significant predictive factor for developing hyponatremia a year later (8/96; Hazard ratio 14.415; P = 0.028). The cumulative survival rate was significantly lower in patients with hyponatremia (46.7% at 1 and 3 years) than in those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001). The cumulative survival rate was significantly lower in patients who had developed hyponatremia after 1 year (100% at 1 year, 62.5% at 3 years) than those who had not (100% at 1 year, 89.0% at 3 years; P < 0.001). The cumulative survival rate was significantly worse in patients with both hyponatremia and SRS (20% at 1 year). Conclusions: There was a close linkage between the serum sodium concentration and portal hemodynamic abnormality, presence of SRS, and their interaction may negatively influence the prognoses in cirrhosis.
Introduction Cirrhosis is the most advanced stage of chronic liver disease, with varying degrees of hepatic dysfunction and impaired portal hemodynamics. Unlike when in an asymptomatic, compensated condition, decompensated patients suffer from serious complications which influence the disease’s prognosis as well as the patient’s quality of life.1 Nonetheless, predicting long-term outcomes based on clinical findings may be extremely useful for the practical management of patients with cirrhosis. Hyponatremia is frequently observed in patients with cirrhosis, showing a relationship with various clinical manifestations and a poor prognosis.2–4 A pathophysiology of hyponatremia is based on systemic hemodynamic abnormalities caused by impaired renal water retention.5–7 Since the systemic circulation may be affected by portal hypertension due to the increased intrahepatic resistance, the hyperdynamic state and the development of collateral vessels,1,8 this suggests a close link between portal hemodynamics and hyponatremia.
As a minimally invasive procedure, Doppler ultrasound (US) allows the real-time observation of blood flow under physiological conditions. Investigators have reported the efficacy of this technique for evaluating portal hemodynamics in cirrhosis patients.9,10 However, the relationship between portal parameters determined by Doppler US and hyponatremia, and their combined effect on the prognosis of cirrhosis has not been fully studied. Therefore, the aim of this study was to examine the relationship between hyponatremia and portal hemodynamics, and their interaction on the prognosis of cirrhosis.
Methods Patients. We conducted a subgroup analysis using medical records collected from patients with cirrhosis, who had participated in a prospective study regarding the influence of portal hemodynamics on clinical outcomes in cirrhosis. Subjects were recruited from among patients who underwent both upper
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1635
Hyponatremia and portal hemodynamics
H Maruyama et al.
gastrointestinal endoscopy and Doppler US for the assessment of portal hypertension, between the period November 2007 to November 2012. Cirrhosis was diagnosed by evaluating liver biopsy or blood samples, combined with findings from US, contrast-enhanced computed tomography (CT), or magnetic resonance imaging taken during routine testing for hepatocellular carcinoma (HCC). Informed written consent was obtained from all patients. The study was approved by the institutional review board and was considered to have an appropriate design for publication purposes. The work undertaken conforms to the provisions of the Declaration of Helsinki (as revised in Tokyo, 2004). Patients who were excluded from the study included those (i) with HCC detected as hypervascular nodules in the arterial phase with washout in the portal venous or late phase on contrastenhanced CT/magnetic resonance imaging11 or a treatment history of HCC; (ii) using vasoactive drugs, such as β-blockers, which are not licensed for the treatment of portal hypertension in Japan; (iii) receiving antiviral therapy prior to, or during, the study period; (iv) receiving interventional procedures such as a transjugular intrahepatic portosystemic or peritoneo-venous shunt prior to, or during, the study period; and (v) aged below 18 or over 90 years. Hyponatremia was defined as a serum sodium concentration lower than 135 mEq/L. Gastroesophageal varices were classified according to the general rules of the Japan Research Society for Portal Hypertension:12 small, medium, or large. A variceal appearance was assessed by endoscopic findings taken during the 6 months before/after a Doppler US study. Hepatic encephalopathy (HE) was assessed using a West Haven grading system,13 with grade II or above classified as overt HE. The degree of ascites was assessed based on clinical and US findings in reference to guidelines:14,15 these were labeled subclinical when ascites were only detectable by US examination, and clinical when ascites caused abdominal distension. The latter was classified as overt ascites whether or not the patients was under diuretic therapy. Any change in the degree of ascites was monitored by a US and physical examination at least twice per year. Patients with ascites were treated according to established guidelines,14,15 using diuretics with a salt-restricted diet and occasional paracentesis. The study used model for end-stage liver disease (MELD) score16 and MELD-Na score3 as clinical parameters. Subjects who participated in this study were followed up, at least every 3 months in an outpatient clinic, to determine their physical status and to collect blood samples. The observation period was defined as the time between the initial US examination and the date of the last hospital visit, for treatment of collateral vessels, liver transplantation, or patient death.
US examination. B-mode and Doppler US were performed with patients in the supine position after fasting for 4 hours or more, using an SSA-770A or 790A (Toshiba; Tokyo, Japan) with a 3.75 MHz convex probe. Spleen size (mm2) was calculated by multiplying the distance from the splenic hilum to the caudal polar angle, measured by two intersecting lines.17 The upper limit of normal used in the study was 2000 mm2. Portal blood flow measurement was performed using a pulsed Doppler method, with the sampling point at a width corresponding to the diameter of the vessel and at an angle less than 60 degrees 1636
between the US beam and the vessel. The mean velocity (cm/s) was calculated by automatic tracing of the wave spectrum, for approximately one cardiac cycle, by integration of the time-based average value of the spectrum in the portal trunk. The mean flow volume (mL/min) was calculated by multiplying the mean flow velocity for 1 second by the cross-section of the vessel, and then multiplying the calculated value by 60 s. When collateral vessel was detected, the diameter was measured by B-mode US and the flow direction was assessed by color Doppler. The data used for analysis were average values, calculated using measurements taken between two to four times. All US examinations were performed by qualified personnel, HM or MT, each with more than eight years’ of experience in this technique.
Statistical analysis. Continuous variables were compared using Student’s t-test or Mann–Whitney’s U-test. Categorical variables were compared using the chi-square test. The cumulative survival rate was calculated using the Kaplan–Meier method and the difference was compared with the log-rank test. Multivariate analysis was performed using logistic regression analysis. Cox regression analysis was used to evaluate risk factors and results were presented as a hazard ratio (HR) with a 95% confidence interval (CI). Statistical significance was defined as P < 0.05. Statistical analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC, USA).
Results Patient characteristics. There were 153 patients (age: mean ± standard deviation, 62.2 ± 12.0 years; range, 20–89; male 79, female 74); 63 were classed as Child–Pugh A, 66 as Child– Pugh B, and 24 as Child–Pugh C (Table 1). At baseline, 16 patients were affected by hyponatremia, with 137 patients unaffected; the frequency of ascites was significantly higher, and serum albumin levels were significantly lower, respectively, in the former (12/16, 3.0 ± 0.6 g/dL) than in the latter (47/137, P = 0.002; 3.3 ± 0.5 g/ dL, P = 0.023). Similarly, the Child–Pugh and the MELD scores, respectively, were significantly higher in the former (8.8 ± 2.3; 13.8 ± 5.2) than in the latter (7.2 ± 1.9, P = 0.002; 11.2 ± 3.8, P = 0.015). As for portal hemodynamic parameters, the presence of a splenorenal shunt (SRS) was significantly more frequent in the hyponatremia group (5/16, 31.3%) than in the nonhyponatremia group (12/137, 8.8%; P = 0.0068). Serum sodium concentrations were significantly lower in patients with an SRS (135.9 ± 4.8 mEq/L, 126–142) than in those without an SRS (138.9 ± 3.2 mEq/L, 125–146; P = 0.0193). The median interval time between US and endoscopic examinations was 0.2 months (0–10.6). In 103 patients with serum sodium concentration between 125 and 140 mEq/L, the clinical data were analyzed using the MELD-Na score instead of the MELD score. The MELD-Na score was significantly higher in patients with hyponatremia (19.6 ± 4.3, 13–27) than in those without (13.3 ± 4.5, 7–27; P < 0.001). In addition, the MELD-Na score was significantly higher in patients with SRS (16.9 ± 5.9, 9–27) than in those without (13.9 ± 4.7, 7–27, P = 0.036). Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642
© 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
H Maruyama et al.
Table 1
Hyponatremia and portal hemodynamics
Patient characteristics at baseline
Number of subjects Age (years) (mean ± SD [range]) Sex (male/female) Etiology (virus/alcohol/PBC/NASH/AIH/others) Ascites (−/+) Gastroesophageal varices (−/+) History of variceal treatment (−/+) History of variceal bleeding (−/+) Hepatic encephalopathy (−/+) Child–Pugh score (mean ± SD [range]) Model for end-stage liver disease score (mean ± SD [range]) Blood test Sodium (mmol/L) (mean ± SD [range]) Potassium (mmol/L) (mean ± SD [range]) Chlorine (mmol/L) (mean ± SD [range]) Bilirubin (mg/dL) (mean ± SD [range]) Albumin (g/dL) (mean ± SD [range]) Prothrombin time-INR (mean ± SD [range]) Platelet count (104/μL) (mean ± SD [range]) Portal hemodynamics Diameter of portal trunk (mm) (mean ± SD [range]) Mean velocity in the portal trunk (cm/s) (mean ± SD [range]) Mean flow volume in the portal trunk (ml/min) (mean ± SD [range]) Left gastric vein (hepatofugal) Short gastric vein (hepatofugal) Splenorenal shunts (hepatofugal) Paraumbilical vein (hepatofugal) Inferior mesenteric vein (hepatofugal) Diameter of collateral vessels > 11.2 mm (median) Spleen (cm2) (mean ± SD [range]) Portal vein thrombosis (−/+)
Non-hyponatremia (Na ≥ 135)
Hyponatremia (Na < 135)
137 62.3 ± 12.1 (40–89) 71/66 45/32/14/10/8/28 90/47 39/98 104/33 91/46 133/4 7.2 ± 1.9 (5–13) 11.2 ± 3.8 (6–26)
16 61.4 ± 10.9 (42–79) 8/8 7/3/3/0/1/2 4/12 4/12 14/2 12/4 14/2 8.8 ± 2.3 (6–13) 13.8 ± 5.2 (7–24)
139.5 ± 2.5 (135–146) 4.0 ± 0.4 (3.0–5.6) 107.3 ± 3.5 (98–119) 1.9 ± 2.3 (0.3–20.2) 3.3 ± 0.5 (1.8–4.7) 1.30 ± 0.25 (0.94–2.18) 9.8 ± 5.7 (1.7–43.9)
131.4 ± 3.0 (125–134) 4.1 ± 0.6 (3.3–5.4) 98.9 ± 4.3 (91–104) 3.2 ± 3.9 (0.5–16.9) 3.0 ± 0.6 (1.9–4.0) 1.29 ± 0.28 (0.98–1.88) 10.0 ± 5.8 (1.8–24.4)
11.4 ± 2.4 (4.9–18.3) 13.1 ± 3.0 (7.1–21.8) 856.8 ± 382.6 (130–2165) 81 (59.1%) 27 (19.7%) 12 (8.8%) 31 (22.6%) 10 (7.3%) 10 (7.3%) 29.7 ± 12.8 (10.3–69.9) 128/9
10.8 ± 1.9 (6.4–14.6) 11.7 ± 3.3 (6.3–15.6) 696.7 ± 316.5 (320–1430) 11 (68.8%) 4 (25.0%) 5 (31.3%) 5 (31.3%) 0 (0%) 1 (6.3%) 29.8 ± 18.5 (12.9–90.5) 15/1
P values
0.659 0.890 0.668 0.002 0.770 0.296 0.489 0.062 0.002 0.015 < 0.001 0.377 < 0.001 0.209 0.023 0.841 0.852 0.289 0.102 0.121 0.457 0.618 0.007 0.442 0.264 0.878 0.978 0.961
−, absence; +, presence. Hepatic encephalopathy −, grade 0∼I; +, grade II∼IV (West-Haven grading system). AIH, autoimmune hepatitis; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrhosis; SD, standard deviation.
Changes in sodium concentrations during a 1-year follow-up. Ninety-six of 137 patients without hyponatremenia at baseline were properly followed up for more than 1 year in our department; 8 of 96 patients developed hyponatremia, 1 year after the enrollment. A prothrombin time-international normalized ratio (PT-INR) was the significant baseline factor for predicting the transition from non-hyponatremia to hyponatremia (HR: 14.415; 95% CI: 1.336–155.487; P = 0.028). However, baseline MELD-Na score was not a significant factor for developing hyponatremia after 1-year follow-up. Clinical findings after the 1-year follow-up were compared between patients who had developed hyponatremia and those who had not (Table 2): the MELD score and the PT-INR were significantly higher in the former (13.1 ± 4.0; 1.47 ± 0.26) than in the latter (10.4 ± 2.9, P = 0.018; 1.25 ± 0.23, P = 0.017).
Prognosis. Thirty-eight patients died during the study period: 31 of hepatic failure, two of unknown cause, one of gastrointesti-
nal perforation, one of sepsis, one of lung cancer, one of oral cavity cancer, and one of rupture of an abdominal aortic aneurysm. In addition, nine patients received a liver transplantation. Cumulative overall survival rates were 87.3% at 1 year, 73.7% at 3 years, and 60.9% at 5 years. There were two significant predictive factors for a poor prognosis: the presence of ascites (HR 4.191; 95% CI 2.084–8.428; P < 0.001) and a low sodium concentration (HR 0.856; 95% CI 0.743–0.986; P = 0.031; Table 3). Multivariate analysis showed that the presence of ascites was the only significant factor for poor prognosis in the analysis of 103 patients using the MELD-Na (HR 3.329, 95% CI, 1.212–9.143; P = 0.02). Multivariate analysis revealed the presence of ascites was the only significant factor for mortality in patients without hyponatremia (Table 4). The cumulative survival rate was significantly lower in patients with ascites at baseline (70.3% at 1 year; 43.8% at 3 years; 34.0% at 5 years) than in those without (97.8% at 1 year; 92.5% at 3 years; 77.5% at 5 years; P < 0.001). The cumulative survival rate was significantly lower in patients with hyponatremia at baseline (46.7% at both 1 and 3 years) than
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1637
Hyponatremia and portal hemodynamics
Table 2
H Maruyama et al.
Comparison of clinical findings between patients who developed hyponatremia and those who did not over a 1-year period
Number of subjects Age (years) (mean ± SD [range]) Sex (male/female) Etiology (virus/alcohol/PBC/NASH/AIH/others) Ascites (−/+) Gastroesophageal varices (−/+) History of variceal treatment (−/+) History of variceal bleeding (−/+) Hepatic encephalopathy (−/+) Child–Pugh score (mean ± SD [range]) Model for end-stage liver disease score (mean ± SD [range]) Blood test Potassium (mmol/L) (mean ± SD [range]) Chlorine (mmol/L) (mean ± SD [range]) Bilirubin (mg/dL) (mean ± SD [range]) Albumin (g/dL) (mean ± SD [range]) Prothrombin time-INR (mean ± SD [range]) Platelet count (104/μL) (mean ± SD [range]) Portal hemodynamics Diameter of portal trunk (mm) (mean ± SD [range]) Mean velocity in the portal trunk (cm/s) (mean ± SD [range]) Mean flow volume in the portal trunk (ml/min) (mean ± SD [range]) Left gastric vein (hepatofugal) Short gastric vein (hepatofugal) Splenorenal shunts (hepatofugal) Paraumbilical vein (hepatofugal) Inferior mesenteric vein (hepatofugal) Diameter of collateral vessels > 11.2 mm (median value) Spleen (cm2) (mean ± SD [range]) Portal vein thrombosis (−/+)
Non-hyponatremia (Na ≥ 135)
Hyponatremia (Na < 135)
P values
88 61.6 ± 12.6 (20–81) 51/37 30/22/6/6/6/18 61/27 23/65 70/18 60/28 86/2 6.9 ± 1.6 (5–12) 10.4 ± 2.9 (7–19)
8 58.8 ± 13.3 (38–72) 3/5 2/1/2/0/1/2 3/5 2/6 5/3 5/3 8/0 8.8 ± 2.6 (6–13) 13.1 ± 4.0 (9–19)
0.539 0.264 0.478 0.068 0.944 0.264 0.742 0.667 0.080 0.018
4.0 ± 0.4 (3.0–5.1) 107.5 ± 3.3 (100–115) 1.5 ± 0.8 (0.3–4.5) 3.4 ± 0.5 (1.8–4.7) 1.25 ± 0.23 (0.95–2.14) 9.4 ± 4.3 (2.7–26.1)
3.9 ± 0.3 (3.4–4.4) 105.9 ± 5.4 (98–115) 2.7 ± 2.8 (0.7–9.1) 3.1 ± 0.5 (2.4–3.9) 1.47 ± 0.26 (1.16–1.90) 14.8 ± 13.6 (4.4–43.9)
0.511 0.431 0.248 0.130 0.017 0.300
11.4 ± 2.3 (4.9–18.3) 13.3 ± 3.2 (7.1–21.8) 840.4 ± 380.6 (130–2165) 50 (56.8%) 19 (21.6%) 9 (10.2%) 17 (19.3%) 5 (5.7%) 8 (9.1%) 29.3 ± 12.3 (10.3–57.2) 82/6
11.4 ± 3.9 (5.3–16.7) 11.7 ± 2.2 (9.5–14.8) 947.5 ± 307.0 (530–1230) 6 (75.0%) 1 (12.5%) 1 (12.5%) 3 (37.5%) 1 (12.5%) 2 (25.0%) 36.7 ± 14.0 (20.2–56.2) 7/1
0.954 0.242 0.503 0.318 0.544 0.840 0.225 0.446 0.158 0.115 0.554
−, absence; +, presence. Hepatic encephalopathy −, grade 0∼I; +, grade II∼IV (West-Haven grading system). AIH, autoimmune hepatitis; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrhosis; SD, standard deviation. Table 3
Cox regression analysis for the predictors of mortality Univariate
Ascites (−/+) Child–Pugh score MELD score Sodium (mmol/L) Chlorine (mmol/L) Bilirubin (mg/dL) Albumin (g/dL) Prothrombin time-INR
Multivariate
HR (95% CI)
P values
HR (95% CI)
P values
6.113 (3.248–11.506) 1.621 (1.398–1.880) 1.153 (1.078–1.233) 0.792 (0.726–0.865) 0.877 (0.824–0.932) 1.224 (1.138–1.316) 0.369 (0.227–0.600) 3.937 (1.331–11.649)
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.013
4.191 (2.084–8.428) — — 0.856 (0.743–0.986) — — — —
< 0.001 — — 0.031 — — — —
−, absence; +, presence. Child–-Pugh score was excluded from the multivariate analysis against the possibility of co-linearity. CI, confidence interval; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
in those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001; Fig. 1). In addition, the cumulative survival rate was significantly lower in patients who had developed hyponatremia 1 year after the enrollment (n = 8; 100% at 1 year; 62.5% at 3 years) than in those without (n = 88; 100% at 1 year; 89.0% at 3 years; P < 0.001; 1638
Fig. 2). The median observation period was 34.1 months (0.3–86.0). When the prognosis of cirrhosis was assessed with respect to the sodium concentration combined with the presence/absence of an SRS, the cumulative survival rate was significantly worse in
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
0.229 1.002 (0.998–1.007)
Discussion
0.431 0.156 0.303 0.354 0.667 0.517 1.070 0.935 1.087 0.729 1.554 1.407
5.143 (2.657–9.958)
1.640 (1.384–19.444)
1.186 0.886 1.273 0.397 5.857 2.630
Child–Pugh score
MELD score Chlorine (mmol/L) Bilirubin (mg/dL) Albumin (g/dL) Prothrombin time-INR Inferior mesenteric vein (hepatofugal)
(1.096–1.282) (0.805–0.974) (1.157–1.401) (0.234–0.673) (1.800–19.052) (1.022–6.773)
< 0.001 0.013 < 0.001 0.001 < 0.001 0.045
(0.904–1.266) (0.853–1.026) (0.928–1.273) (0.373–1.423) (0.209–11.560) (0.502–3.944)
— — < 0.001
0.001 3.621 (1.686–7.775) < 0.001
P values HR (95% CI) P values
Multivariate Univariate
HR (95% CI)
Non-hyponatremia (Na ≥ 135)
−, absence; +, presence. Child–Pugh score was excluded from the multivariate analysis against the possibility of co-linearity. CI, confidence interval; HR, hazard ratio; INR, international normalized ratio; MELD, model for end-stage liver disease.
1.003 (1.001–1.006)
0.013
0.605 1.258 (0.527–3.001)
patients with both hyponatremia and the presence of an SRS (20% at 1 year) than in others (91.8% at 1 year, 76.3% at 3 years, and 63.1% at 5 years in patients with neither hyponatremia nor an SRS, P < 0.0001; 77.4% at 1 year, 72.6% at 3 years, and 58.1% at 5 years in patients with hyponatremia or an SRS, P = 0.014; Fig. 3).
Diameter of portal trunk (mm) Mean flow volume in the portal trunk (mL/min)
1.875 (1.052–3.341)
0.033
P values Multivariate
HR (95% CI) P values HR (95% CI)
Univariate
Hyponatremia and portal hemodynamics
Ascites (−/+)
Table 4
Cox regression analysis for the predictors of mortality in patients with and without hyponatremia
Hyponatremia (Na < 135)
H Maruyama et al.
Our study reveals the strong interrelationship between hyponatremia and the simultaneous presence of a portal hemodynamic abnormality, that is, an SRS; furthermore, cirrhosis patients with both showed a much poorer prognosis, similar to a negative synergistic effect. Previous studies have also suggested a potential link between the presence of an SRS and poor liver function, as exemplified by diseased liver in the decompensation stage,18,19 thus supporting our data. Although the precise mechanisms involved in the joint damaging effect of hyponatremia and portal hemodynamic dysfunction remain undefined, care should be taken with cirrhosis patients presenting with both hyponatremia and an SRS in clinical practice. Many kinds of markers exist to predict the long-term outcome of cirrhosis, including HCC, as well as Child–Pugh and MELD scores.20–23 Needless to say, the presence of esophageal varices and/or ascites are significant factors for a poor prognosis. A recent study reported that the 1-year mortality in compensated patients was 3.4% (95% CI: 1.8–5.9) in patients without varices (n = 53), and 7.3% (95% CI: 4.9–10.4) in those with varices (n = 67).24 In decompensated patients, the 1-year mortality was 21.9% (95% CI: 18.3–26.0) in patients with ascites (n = 194) and 18.1% (95% CI: 14.5–22.3) in those with variceal bleeding (n = 129). A more recent study has shown that the 5-year mortality was 1.5% in compensated patients without esophagogastric varices (n = 202), 10% in compensated patients with varices (n = 216), 20% in those with bleeding and without other complications (n = 75), 30% in those with a first, non-bleeding decompensating event (n = 206), and 88% in those with a second decompensating event (n = 213), respectively (P < 0.0001 for all).25 In regard to these data, the presence of hyponatremia, as demonstrated by our study (cumulative survival rate, 46.7% at 1 and 3 years), should be considered a sign of serious condition that becomes critical when combined with an SRS (cumulative survival rate, 20% at 1 year). It is reported that the MELD-Na was superior to the MELD to predict mortality, being an important predictor of survival among candidates for liver transplantation.3 The MELD-Na offers additional advantages over the MELD in the prognostic accuracy in patients with acute decompensated hepatitis,26 and postoperative 90-day mortality in cirrhosis after surgical procedures.27 The present study detected that the MELD-Na score showed a significant difference according to the presence/absence of SRS, indicating a close linkage with the portal hemodynamic abnormality. Our study may also be the first to report the effect of timerelated changes in sodium concentration, over a 1-year interval, on the prognosis of cirrhosis. As expected, the transition from nonhyponatremia to hyponatremia suggests a worsening condition, with the PT-INR a significant predictive factor for this change. Our results clearly demonstrate the importance of such parameters as sodium concentration and PT-INR as useful biomarkers in the management of cirrhosis patients.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1639
Hyponatremia and portal hemodynamics
H Maruyama et al.
Figure 1 Cumulative survival rate was significantly lower in patients with hyponatremia at baseline (46.7% at 1 year and 3 years) than those without (91.8% at 1 year, 76.8% at 3 years; P < 0.001). Solid line, patients without hyponatremia; dashed line, patients with hyponatremia. Median observation period, 34.1 months (0.3–86.0).
Figure 2 Cumulative survival rate was significantly lower in patients who had developed hyponatremia 1 year after the enrollment (n = 8; 100% at 1 year, 62.5% at 3 years) than those without (n = 88; 100% at 1 year, 89.0% at 3 years; P < 0.001). Solid line, patients without development of hyponatremia; dashed line, patients with development of hyponatremia.
Hyponatremia may be the most common electrolyte abnormality seen in daily medical practice and is generally defined as a serum sodium concentration lower than 136 mEq/L.6,28 However, there is another argument that the definition of hyponatremia should be set at a sodium concentration < 130 mEq/L in patients with cirrhosis.7,29 The frequency of hyponatremia varies considerably, even with the same definition: 20.8%–49.4% for Na < 135 mEq/L,3,30 15–30% for Na 130–135 mEq/L,31 and 7–30% for Na < 130 mEq/L.5,31 These data strongly suggest a population dependency in the distribution of impaired electrolytes; patient background including etiology, severity of liver disease, and the studied lesion may affect the lower limit of sodium concentration. In addition, as it is reported that even mild, chronic 1640
hyponatremia may induce long-term adverse effects,32 the definition of hyponatremia should be based on the study design. Increased production of arginine vasopression, caused by a nonosmotic hypersecretion related to circulatory dysfunction, is the main reason for hyponatremia in patients with advanced cirrhosis.2 In particular, excessive, solute-free water retention is a key pathophysiological feature of impaired water balance, resulting in hyponatremia as well as ascites, the so-called dilutional hyponatremia.2,33 In fact, a higher frequency of ascites in patients with hyponatremia was detected in the present study and the presence of ascites was a significant factor for poor prognosis even in the analysis using the MELD-Na score instead of the MELD score; however, the presence of ascites is no longer a significant factor of
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
H Maruyama et al.
Hyponatremia and portal hemodynamics
Figure 3 Cumulative survival rate was worse in patients with both hyponatremia and the presence of an SRS (20% at 1 year) than the others (91.8% at 1 year, 76.3% at 3 years, and 63.1% at 5 years in patients with both non-hyponatremia and absence of an SRS, P < 0.0001; 77.4% at 1 year, 72.6% at 3 years and 58.1% at 5 years in patients with hyponatremia or the presence of an SRS, P = 0.014). Solid line, patients with neither hyponatremia nor an SRS; dashed line, patients with hyponatremia or the presence of an SRS; dashed-dotted line, patients with both hyponatremia and the presence of an SRS. SRS, splenorenal shunt.
mortality in patients with hyponatremia, suggesting a common pathophysiology between hyponatremia and ascites. There were some limitations to our study. Firstly, the number of patients was relatively small, and there may also be a potential bias in the etiology of liver disease since the study was undertaken in Asia. In addition, as we did not collect data relating to the hepatic venous pressure gradient, the influence of serum sodium concentration on portal pressure remains unclear. More importantly, since this was not a prospective study, further validation is needed with a larger patient population. In conclusion, the present study clearly demonstrates that there is a close association between serum sodium concentration and portal hemodynamic abnormality, that is, the presence of an SRS. The interaction between hyponatremia and the presence of an SRS suggests a deleterious influence on a prognosis of cirrhosis. These factors require strict monitoring during the daily care of cirrhosis patients so as to provide more prudent patient management.
References 1 Tsochatzis EA, Bosch J, Burroughs AK. Liver cirrhosis. Lancet 2014; 383: 1749–61. 2 Gines P, Guevara M. Hyponatremia in cirrhosis: pathogenesis, clinical significance, and management. Hepatology 2008; 48: 1002–10. 3 Kim WR, Biggins SW, Kremers WK et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N. Engl. J. Med. 2008; 359: 1018–26. 4 Ahluwalia V, Heuman DM, Feldman G et al. Correction of hyponatraemia improves cognition, quality of life, and brain oedema in cirrhosis. J. Hepatol. 2015; 62: 75–82. 5 Gines P, Berl T, Bernardi M et al. Hyponatremia in cirrhosis: from pathogenesis to treatment. Hepatology 1998; 28: 851–64. 6 Adrogue‘ HJ, Madias NE. Hyponatremia. N. Engl. J. Med. 2000; 342: 1581–9. 7 Gianotti RJ, Cardenas A. Hyponatraemia and cirrhosis. Gastroenterol. Rep. (Oxf) 2014; 2: 21–6.
8 Sanyal AJ, Bosch J, Blei A, Arroyo V. Portal hypertension and its complications. Gastroenterology 2008; 134: 1715–28. 9 Bolondi L, Li Bassi S, Gaiani S, Barbara L. Doppler flowmetry in portal hypertension. J. Gastroenterol. Hepatol. 1990; 5: 459–67. 10 Baik SK. Haemodynamic evaluation by Doppler ultrasonography in patients with portal hypertension: a review. Liver Int. 2010; 30: 1403–13. 11 Bruix J, Sherman M. American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53: 1020–2. 12 The Japan Society for Portal Hypertension. The General Rules for Study of Portal Hypertension, 3rd edn. Tokyo: Kanehara, 2013. 13 Ferenci P, Lockwood A, Mullen K, Tarter R, Weissenborn K, Blei AT. Hepatic encephalopathy—definition, nomenclature, diagnosis, and quantification: final report of the working party at the 11th World Congresses of Gastroenterology, Vienna, 1998. Hepatology 2002; 35: 716–21. 14 Moore KP, Aithal GP. Guidelines on the management of ascites in cirrhosis. Gut 2006 ; 55 (Suppl. 6): vi1–12. 15 Runyon BA. Introduction to the revised American Association for the Study of Liver Diseases Practice Guideline management of adult patients with ascites due to cirrhosis 2012. Hepatology 2013; 57: 1651–3. 16 Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31: 864–71. 17 Maruyama H, Kamezaki H, Kondo T, Sekimoto T, Takahashi M, Yokosuka O. Sonographic and clinical features of collateral vessels at the splenic hilum in cirrhosis. Clin. Radiol. 2014; 69: e140–5. 18 Ohnishi K, Sato S, Saito M et al. Clinical and portal hemodynamic features in cirrhotic patients having a large spontaneous splenorenal and/or gastrorenal shunt. Am. J. Gastroenterol. 1986; 81: 450–5. 19 Tarantino G, Citro V, Conca P et al. What are the implications of the spontaneous spleno-renal shunts in liver cirrhosis? BMC Gastroenterol. 2009; 9: 89. 20 Kamath PS, Wiesner RH, Malinchoc M et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33: 464–70.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1641
Hyponatremia and portal hemodynamics
H Maruyama et al.
21 Said A, Williams J, Holden J et al. Model for end stage liver disease score predicts mortality across a broad spectrum of liver disease. J. Hepatol. 2004; 40: 897–903. 22 D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J. Hepatol. 2006; 44: 217–31. 23 Bruno S, Zuin M, Crosignani A et al. Predicting mortality risk in patients with compensated HCV-induced cirrhosis: a long-term prospective study. Am. J. Gastroenterol. 2009; 104: 1147–58. 24 Zipprich A, Garcia-Tsao G, Rogowski S, Fleig WE, Seufferlein T, Dollinger MM. Prognostic indicators of survival in patients with compensated and decompensated cirrhosis. Liver Int. 2012; 32: 1407–14. 25 D’Amico G, Pasta L, Morabito A et al. Competing risks and prognostic stages of cirrhosis: a 25-year inception cohort study of 494 patients. Aliment. Pharmacol. Ther. 2014; 39: 1180–93. 26 Hsu CY, Lin HC, Huang YH et al. Comparison of the model for end-stage liver disease (MELD), MELD-Na and MELD-Na for outcome prediction in patients with acute decompensated hepatitis. Dig. Liver Dis. 2010; 42: 137–42. 27 Cho HC, Jung HY, Sinn DH et al. Mortality after surgery in patients with liver cirrhosis: comparison of Child–Turcotte–Pugh, MELD and MELDNa score. Eur. J. Gastroenterol. Hepatol. 2011; 23: 51–9.
1642
28 Upadhyay A, Jaber BL, Madias NE. Incidence and prevalence of hyponatremia. Am. J. Med. 2006; 119: 30–5. 29 European Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J. Hepatol. 2010; 53: 397–417. 30 Gaglio P, Marfo K, Chiodo IIIJ. Hyponatremia in cirrhosis and end-stage liver disease: treatment with the vasopressin V2-receptor antagonist tolvaptan. Dig. Dis. Sci. 2012; 57: 2774–85. 31 Corona G, Giuliani C, Parenti G et al. Moderate hyponatremia is associated with increased risk of mortality: evidence from a meta-analysis. PLoS One 2013; 18: e80451. 32 Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild chronic hyponatremia is associated with falls, unsteadiness and attention deficits. Am. J. Med. 2006; 119: 71.e1-8. 33 Angeli P, Wong F, Watson H, Ginès P. CAPPS investigators. Hyponatremia in cirrhosis: results of a patient population survey. Hepatology 2006; 44: 1535–42.
Journal of Gastroenterology and Hepatology 30 (2015) 1635–1642 © 2015 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
