Natural history and outcomes of patients with liver cirrhosis complicated by hepatic hydrothorax

Table 1 Baseline characteristics of patients with cirrhosis during the index hospital admission for hepatic hydrothorax (n = 84)

Characteristic	Statistic
Age, mean ± SD, yr	58.3 ± 11.5
Male, n (%)	46 (54.8)
Diabetes mellitus, n (%)	23 (27.4)
Ischaemic heart disease, n (%)	7 (8.3)
Heart failure, n (%)	4 (4.8)
Chronic obstructive lung disease, n (%)	5 (6.0)
Hepatocellular carcinoma, n (%)	23 (27.4)
Other cancer, n (%)	3 (3.6)
Chronic kidney disease, n (%)	11 (13.1)
Current smoker, n (%)	12 (14.3)
Current alcohol drinker, n (%)	9 (10.7)
Aetiology of liver disease1, n (%)
Hepatitis C virus	25 (29.8)
Non-alcoholic fatty liver	17 (20.2)
Alcoholic liver disease	13 (15.5)
Hepatitis B virus	9 (10.7)
Primary biliary cirrhosis/sclerosing cholangitis	7 (8.3)
Cryptogenic	7 (8.3)
Autoimmune hepatitis	6 (7.1)
Budd-Chiari syndrome	3 (3.6)
Metabolic/others2	7 (8.3)
MELD score, mean ± SD	26.8 ± 7.1
Platelet count, median (IQR), per nL	103 (65-132)
Platelet count < 50/nL, n (%)	9 (10.7)
Anticoagulation, n (%)	6 (7.1)
Previous ascitic taps, n (%)	69 (82.1)
Beta-blockers, n (%)	11 (13.1)
Lactulose, n (%)	34 (40.5)
Rifaximin, n (%)	14 (16.7)
SBP prophylaxis, n (%)	32 (38.1)

¹Categories are not mutually exclusive as > 1 aetiology may be attributed to each patient.

²Hemochromatosis, α1-antitrypsin deficiency, sarcoidosis.

MELD: Model for end-stage liver disease; SBP: Spontaneous bacterial peritonitis; IQR: Interquartile range.

Table 2 Clinical features of hepatic hydrothorax, management, and complications

Characteristic	Index admission (n = 84)	Readmissions (n = 83)
NYHA severity of dyspnoea, n (%)
Class I	13 (15.5)	4 (4.3)
Class II	23 (27.4)	37 (39.4)
Class III	28 (33.3)	22 (23.4)
Class IV	20 (23.8)	31 (33.0)
Distribution of hydrothorax, n (%)
Unilateral, left	13 (15.5)	10 (12.1)
Unilateral, right	68 (81.0)	72 (86.8)
Bilateral	3 (3.6)	1 (1.2)
Admission hepatic encephalopathy, n (%)	12 (14.3)	14 (16.9)
De novo hepatic encephalopathy, n (%)	11 (13.1)	10 (12.0)
Thoracocentesis performed, n (%)	53 (63.1)	62 (74.7)
Pneumothorax, n (% of procedures)	6 (11.3)	11 (17.7)
Haematoma, n (% of procedures)	2 (3.8)	1 (1.6)
Pre-procedure platelet infusion, n (%)	11 (13.3)	6 (7.2)
Pre-procedure blood products1, n (%)	18 (21.7)	12 (14.5)
Pleural fluid drained, mean ± SD, litres	2.8 ± 1.7	3.0 ± 1.5
Indwelling intercostal catheter, n (%)	10 (11.9)	8 (9.6)
Infection, n (%)	17 (20.2)	14 (16.9)
Acute kidney injury, n (%)	30 (35.7)	28 (33.7)
Concurrent ascites drainage, n (%)	23 (27.1)	24 (28.9)
Volume drained, mean ± SD, litres	6.0 ± 3.1	5.5 ± 2.1
TIPS performed, n (%)	2 (2.4)	8 (9.6)
Furosemide, n (%)
None	27 (32.1)	9 (10.8)
20 to < 100 mg daily	52 (61.9)	60 (72.3)
100 to 240 mg daily	5 (5.6)	14 (16.9)
Spironolactone, n (%)
None	27 (32.1)	18 (21.7)
25 to < 100 mg daily	15 (17.9)	22 (26.5)
100 to 300 mg daily	42 (50.0)	43 (51.8)

¹One or more of fresh frozen plasma, cryoprecipitate, or prothrombin complex concentrate.

NYHA: New York Heart Association; TIPS: Transjugular intrahepatic portosystemic shunt.

Table 3 Univariable Cox regression on patient survival to 12 mo

Enduring variables	Overall1, 12-mo survival	Transplant-free, 12-mo survival
	HR (95%CI)	HR (95%CI)
Age, per 5-yr	1.13 (0.94-1.35)	0.95 (0.85-1.07)
Male sex	2.05 (0.84-4.99)	1.68 (0.94-3.00)
Diabetes	0.56 (0.25-1.50)	0.39 (0.19-0.81)
Ischaemic heart disease	2.00 (0.68-5.92)	0.80 (0.29-2.23)
Heart failure	0.86 (0.12-6.38)	0.35 (0.05-2.55)
Chronic obstructive lung disease	NE	1.02 (0.32-3.30)
Chronic kidney disease	1.13 (0.38-3.32)	0.82 (0.37-1.83)
Hepatocellular carcinoma	2.11 (0.92-4.81)	1.55 (0.87-2.76)
Hepatitis B	3.36 (1.32-8.53)	1.68 (0.75-3.74)
Hepatitis C	0.81 (0.32-2.05)	1.28 (0.71-2.28)
Alcoholic liver disease	0.76 (0.22-2.55)	0.94 (0.44-2.01)
Non-alcoholic fatty liver	0.51 (0.15-1.73)	0.43 (0.18-1.02)
Current smoking	5.01 (2.15-11.7)	2.27 (1.13-4.56)
Current alcohol intake	0.68 (0.42-1.13)	0.93 (0.68-1.26)
Time varying variables
MELD score, per 5-points	1.28 (0.93-1.77)	1.82 (1.41-2.34)
NYHA functional class	1.14 (0.74-1.75)	1.09 (0.81-1.46)
TIPS procedure	0.37 (0.05-2.83)	0.13 (0.02-0.96)
Acute kidney injury	2.57 (1.13-5.88)	2.36 (1.35-4.11)
Thoracocentesis	0.51 (0.23-1.16)	1.28 (0.70-2.34)
Pneumothorax	2.37 (0.80-7.00)	2.78 (1.39-5.56)
Intercostal catheter inserted	1.56 (0.53-4.62)	1.85 (0.90-3.82)
Admission hepatic encephalopathy
None	1.00 (reference)	1.00 (reference)
Mild	0.75 (0.17-3.25)	1.21 (0.54-2.70)
Severe	11.4 (3.28-39.6)	7.93 (2.36-26.6)
Encephalopathy during admission2	2.30 (0.84-6.32)	3.20 (1.65-6.24)
Any encephalopathy	2.30 (0.97-5.45)	2.21 (1.22-4.03)

¹Censored at time of liver transplantation.

²New-onset during admission not present on admission, all episodes were mild.

HR: Hazard ratio; CI: Confidence interval; MELD: Model for end-stage liver disease; NYHA: New York Heart Association; TIPS: Transjugular intrahepatic portosystemic shunt; NE: Not estimable.

Table 4 Multivariable Cox regression on patient survival to 12 mo

Model and variables	AHR (95%CI)	P value	Bootstrap CI	P value
Overall survival1
Age, per 5-yr increase	1.30 (1.01-1.67)	0.043	0.91-1.85	0.15
MELD score, per 5 points	1.53 (1.06-1.67)	0.025	0.89-2.66	0.13
Current smoking	8.65 (3.43-21.9)	< 0.001	2.54-29.5	0.001
Acute kidney injury	2.91 (1.21-6.97)	0.017	1.04-8.12	0.042
Transplant-free survival
Hepatic encephalopathy	2.00 (1.08-3.67)	0.030	0.93-4.28	0.078
MELD score, per 5 points	1.78 (1.38-2.29)	< 0.001	1.31-2.44	< 0.001
Current smoking	3.11 (1.49-6.52)	0.003	1.16-8.30	0.024
Acute kidney injury	2.18 (1.20-3.96)	0.011	1.10-4.31	0.026

¹Censored for liver transplantation.

MELD: Model for end-stage liver disease; AHR: Adjusted hazard ratio; CI: Confidence interval.