1. Role of Non-selective Beta-blocker In Portal Hypertension
목요세미나
Role of Non-selective Beta-blocker In Portal Hypertension
R3 박영수/Pf. 최광현

3. Introduction
EA Tsochatzis, J Bosch, AK Burroughs. Liver cirrhosis. Lancet, 383 (2014), pp

4. Introduction
EA Tsochatzis, J Bosch, AK Burroughs. Liver cirrhosis. Lancet, 383 (2014), pp
R De Franchis. Expanding consensus in portal hypertension. Report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension. Revising consensus in portal hypertension. J Hepatol, 63 (2015), pp

5. Introduction
C Villanueva, A Albillos, J Genescà, et al. Development of hyperdynamic circulation and response to β-Blockers in compensated cirrhosis with portal hypertension.
Hepatology, 63 (2016), pp
RJ Groszmann, G Garcia-Tsao, J Bosch, et al. Beta-blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med, 353 (2005), pp

6. Aim
In patients with compensated cirrhosis and CSPH, to assess whether long-term treatment with β-blockers might prevent disease progression to clinical decompensation or death

7. Methods: Participants
8 hospital in Spain
Enrollment: Jan 18, 2010 ~ July 31, 2013
Inclusion criteria
Age: 18~80 years
Cirrhosis : biopsy, compatible clinical, biochemical, ultrasonographic findings
No previous decompensation of cirrhosis
No high risk esophageal varices (no varices or small varices without red signs)
Absence of ascites by ultrasonography
Exclusion criteria
Previous decompensation
Absence of CSPH
Portal thrombosis
HCC
Baseline bilirubin > 3 mg/dL
PLT < 30x 103
PT INR > 2.7, PT % < 30
Renal failure : Cr > 2 mg/dL
Comorbidity with life expectancy < 12 months
Contraindication or hypersensitivity of β-blockers
Previous treatment with β-blockers or nitrates
Anticoagulant treatment
Active antiviral for hepatitis C
Pregnancy or lactation

8. Methods: Study design HR > 55/min, SBP > 90 mmHg Up to 3 weeks
Start with 40mg bid
up to 160 mg bid
Start with 6.25mg/day
up to 25mg/day

9. Methods: Outcomes Primary outcomes Secondary outcomes
Decompensation (Ascites, portal hypertension-related GI bleeding, overt hepatic encephalopathy) of cirrhosis or death
Secondary outcomes
Development of each complication (ascites, GI bleeding, overt hepatic encephalopathy)
Development of SBP and other bacterial infections
Development of varices and of high-risk varices
Changes in hepatic dysfunction (Child-Pugh score and MELD score)
Development of HCC
Adverse events
survival
GI bleeding : any episode of hematemesis or melena, evaluated by endoscopy
Variceal bleeding : Baveno criteria
Ascites : compatible signs on examination and confirmed by US or paracentesis
Refractory ascites, HRS, SBP, overt hepatic encephalopathy : diagnosed according to EASL guideline
Encephalopathy : defined according to West Haven criteria, grade II 초과
Varices : Large- not flattened by insufflation, small – flattened by insufflation
High-risk varices : Large varices and small varices with red wale marks or occurring in patients with Child-Pugh class C

10. Methods: Procedure Baseline Annually Every 6 months Blood sample HVPG
Ultrasonography
Jan 18, 2010 ~ July 31, 2013
Planned until Oct 31, 2016
Newly licensed direct-acting antivirals(DAA) for hepatitis C
 study finished in June, 2015
Baseline
Month 1
Month 3
Every
6 months
Each visit
Endpoints
Adverse events
HR (by a study nurse)
Pill count (adequate: ≥ 70%, poor: < 30%)
Annually
Hemodynamic study (HVPG, Cardiopulmonary pressure, cardiac output)
Upper endoscopy

11. Statistical analysis Sample size: Intention-to-treat strategy
In untreated compensated cirrhosis with CSPH, 2-year decompensation risk 25%, Absolute risk-reduction 15%,
Two-tailed test (α 0.05, β 0.2), 5% loss of f/u
 105 patients required in each group
Intention-to-treat strategy
Categorical variables : Fisher’s exact test
Continuous variables : Student’s t test
Skewed or ordinal data : Wilcoxon rank-sum test
Primary and secondary outcomes : time-to-event variables, stratum(acute response to BB). HR and 95% CI
non-liver related deaths: Competing-risk analysis
Probabilities : Cumulative incidence functions, Gray’s test
Comparison of the tow study groups : Cox model
Survival function : Kaplan-Meier method
Event rates of endpoints : Stratified log-rank test
Subgroup analysis : liver function, cause of cirrhosis, presence of varices, baseline HVPG

12. Results: Study profile

13. Results: Baseline and procedural characteristics
Placebo group
(n=101)
β-blockers
group (n=100)
Baseline characteristics
Sex
Male
64 (63%)
59 (59%)
Female
37 (37%)
41 (41%)
Age (years)
59 (11)
60 (10)
Cause of cirrhosis
Alcohol
14 (14%)
19 (19%)
Hepatitis C virus
59 (58%)
54 (54%)
Alcohol and hepatitis C virus
8 (8%)
9 (9%)
NASH
5 (5%)
Others
12 (12%)
13 (13%)
Diabetes
21 (21%)
22 (22%)
Dyslipidaemia
15 (15%)
Arterial hypertension
34 (34%)
45 (45%)
Child-Pugh class A
81 (80%)
80 (80%) B
20 (20%) C
Child-Pugh score
5·8 (0·9)
5·7 (0·9)
Model for end-stage liver disease score
6·8 (0·3)
6·6 (0·3)
Oesophageal varices*
None
43 (43%)
44 (44%)
Small
58 (57%)
56 (56%)
Gastric varices†
1 (1%)
2 (2%)
Portal-systemic collaterals by ultrasound‡
11 (11%)
18 (18%)
Splenomegaly§
67 (66%)
Liver stiffness, kPa¶
30·4 (16)
28·7 (13)
Weight, kg
76 (16)
76 (15)
BMI, kg/m2
27 (5)
27 (4)
Placebo group
(n=101)
β-blockers
group (n=100)
Procedural characteristics
Duration of follow-up (months)
Mean
37 (16)
36 (16)
Median (IQR)
37 (27–47)
37 (26–47)
Lost to follow-up‖
4 (4%)
9 (9%)
Abstinence from alcohol**
88 (87%)
82 (82%)
Development of portal thrombosis
5 (5%)
3 (3%)
Liver transplantation††
1 (1%)
Randomised to propranolol (or identical tablets of placebo)
Number of patients
68 (67%)
67 (76%)
Dose (mg/day)
95 (81)
95 (76)
80 (40–90)
80 (40–120)
Withdrawal‡‡ 2 6
Randomised to carvedilol (or identical tablets of placebo)
33 (33%)
20 (6)
19 (7)
18·8 (18·8–25)
18·8 (12·5–25)
Withdrawal§§ 4
Table 1. Baseline and procedural characteristics
Values are mean (SD) or n (%), unless otherwise stated. NASH=non-alcoholic steatohepatitis. BMI=body-mass index.
*None denotes absence of oesophageal varices on endoscopy and small varices denotes varices that were flattened by insufflation.
†Patients with oesophageal and fundal varices (gastro-oesophageal varices type 2 according to Sarin's classification).
‡Among patients with portal-systemic collaterals, small oesophageal varices were present in 13 (72%) of 18 patients in the placebo group and in five (45%) of 11 in the β blockers group.
§Spleen diameter >12 cm on ultrasound.
¶Liver stiffness by transient elastography.
‖For patients lost to follow-up, mean follow-up time in the placebo group was 18 months (IQR 14–24) and in the β-blockers group was 18 months (IQR 9–20). Ten (77%) of the 13 patients lost to follow-up had history of active or previous alcohol intake versus 78 (41%) of 188 who were not lost to follow-up; there were no other significant differences between patients lost and not lost to follow-up.
**In the placebo group 17 (77%) of the 22 patients with alcoholic cause of cirrhosis (14 alcohol only and eight alcohol plus hepatitis C virus) were abstinent, as compared with 15 (54%) of the 28 patients with alcoholic cause of cirrhosis in the β-blockers group (19 alcohol, nine alcohol plus hepatitis C virus).
††In the placebo group one patient received an orthoptic liver transplant at month 39 from inclusion and in the β-blockers group three patients received an orthoptic liver transplant between month 27 and 48 from inclusion. Previous decompensation of cirrhosis occurred in the four patients who received a liver transplantation (all had ascites and three of them also had encephalopathy). Two of them (one in the placebo group) also developed a hepatocellular carcinoma.
‡‡Withdrawal of placebo or propranolol (due to side-effects or non-compliance) occurred in two patients of the placebo group after a median period of 22 months (IQR 18–34) and in six patients of the β blockers group after a median period of 26 months (15–30).
§§Withdrawal of placebo or carvedilol (due to side-effects or non-compliance) occurred in four patients of the placebo group after a median period of 24 months (IQR 19–37) and in two patients of the β blockers group after a median period of 28 months (21–33).

14. Results: Primary outcomes
Table 2. Baseline and follow-up haemodynamic variables and liver and renal function
Values are mean (95% CI). A second haemodynamic study at 1 year of follow-up was performed in 156 patients (86%) of the 181 remaining in the study. A third haemodynamic study at 2 years was performed in 86 patients (61%) of the 141 remaining in the study. A fourth study at 3 years was performed in 47 patçients (52%) of the 90 remaining in the study.
*p values from the mixed model for repeated measurements analysis for the terms treatment effect, treatment-by-time interaction, time effect, and randomisation stratum. p values for the baseline measurement term were always <0·001.
†At each time period, haemodynamic studies and laboratory controls were performed in patients who remained in the study and accepted to repeat the procedure. In the β-blockers-group, 94 patients remained in the study at 1 year of follow-up, 72 at 2 years, and 42 at 3 years of follow-up. In the placebo group, 87 patients remained in the study at 1 year of follow-up, 69 at 2 years, and 48 at 3 years of follow-up.
‡Liver function was assessed at each time period by Child-Pugh score and model for end-stage liver disease score. During follow-up, these scores were available in patients remaining in the study, not in those with previous decompensation or death.
Risk adjustment
HR 0.47 (95% CI )
p value=0.025

15. Results: Secondary outcomes

16. Results: Treatment group

17. Results : Haemo-dynamic variables and liver and renal function
Baseline
12 months follow-up
24 months follow-up
36 months follow-up
p values*
Patients†
β blockers
100/100
78/88
44/69
22/40 ..
Placebo
101/101
78/87
42/69
25/46
Hepatic venous pressure gradient
Absolute values, mm Hg
ptreat<0·001
14·5 (14 to 15)
12·8 (12 to 14)
13·0 (12 to 14)
12·9 (12 to 14)
14·8 (14 to 16)
15·0 (14 to 16)
14·9 (14 to 16)
14·6 (13 to 16)
Change from baseline, %
−12 (−15 to −8)
−10 (−12 to − 5)
−10 (−15 to −4)
1·5 (−2 to 5)
1·2 (−3 to 6)
−0·9 (−6 to 5)
Cardiac output
Absolute values, L/min
ptreat=0·049
6·0 (5·6 to 6·5)
5·3 (4·8 to 5·7)
5·4 (4·9 to 5·9)
5·5 (4·9 to 6·1)
5·9 (5·5 to 6·4)
5·7 (5·1 to 6·1)
5·9 (5·3 to 6·5)
−12 (−15 to −7)
−8 (−13 to −2)
−7 (−13 to −1)
0·6 (−3 to 4)
−5 (−11 to −1)
−0·3 (−6 to 6)
Cardiac index
Absolute values, L/min per m2
ptreat=0·010
3·4 (3·1 to 3·6)
2·9 (2·7 to 3·2)
2·9 (2·7 to 3·3)
3·0 (2·7 to 3·4)
3·2 (3.0 to 3.4)
3·5 (3·2 to 3·8)
3·2 (2·8 to 3·5)
3·2 (2·9 to 3·6)
−15 (−21 to −7)
−11 (−19 to −3)
−10 (−20 to 0)
11 (4 to 18)
−3 (−14 to 4)
−2 (−12 to 8)
Mean arterial pressure
ptreat=0·007
98 (95 to 100)
92 (89 to 95)
92 (89 to 96)
95 (91 to 99)
96 (93 to 98)
91 (88 to 94)
92 (88 to 95)
93 (88 to 97)
−5 (−7 to 2)
−5 (−8 to −2)
−2 (−6 to 2)
−3 (−5 to −1)
−2 (−5 to 1)
−2 (−5 to 2)
Heart rate
Absolute values, beats per min
ptreat=<0·001
74 (72 to 76)
61 (59 to 64)
62 (59 to 65)
60 (57 to 64)
73 (71 to 75)
72 (69 to 74)
69 (65 to 72)
71 (67 to 74)
−16 (−19 to 14)
−16 (−18 to −12)
−17 (−21 to −13)
−1 (−3 to 1)
−5 (−7 to −2)
−3 (−6 to 1)
Pulmonary artery pressure
ptreat=0·81
16·5 (15 to 18)
17·3 (16 to 19)
18·6 (17 to 20)
21·6 (19 to 24)
17·0 (15 to 18)
18·7 (17 to 21)
17·0 (15 to 19)
7·5 (−1 to 16)
18 (5 to 29)
19 (17 to 21)
5 (−3 to 13)
18 (5 to 31)
4 (−11 to 19)
Table 2. Baseline and follow-up haemodynamic variables and liver and renal function
Values are mean (95% CI). A second haemodynamic study at 1 year of follow-up was performed in 156 patients (86%) of the 181 remaining in the study. A third haemodynamic study at 2 years was performed in 86 patients (61%) of the 141 remaining in the study. A fourth study at 3 years was performed in 47 patçients (52%) of the 90 remaining in the study.
*p values from the mixed model for repeated measurements analysis for the terms treatment effect, treatment-by-time interaction, time effect, and randomisation stratum. p values for the baseline measurement term were always <0·001.
†At each time period, haemodynamic studies and laboratory controls were performed in patients who remained in the study and accepted to repeat the procedure. In the β-blockers-group, 94 patients remained in the study at 1 year of follow-up, 72 at 2 years, and 42 at 3 years of follow-up. In the placebo group, 87 patients remained in the study at 1 year of follow-up, 69 at 2 years, and 48 at 3 years of follow-up.
‡Liver function was assessed at each time period by Child-Pugh score and model for end-stage liver disease score. During follow-up, these scores were available in patients remaining in the study, not in those with previous decompensation or death.

18. Results : Haemo-dynamic variables and liver and renal function
Baseline
12 months follow-up
24 months follow-up
36 months follow-up
p values*
Pulmonary wedge pressure
Absolute values, mm Hg ..
ptreat=0·37
β blockers
8·8 (8 to 10)
10·3 (9 to 11)
10·9 (9 to 12)
13·4 (11 to 15)
Placebo
9·7 (9 to 11)
10·0 (9 to 11)
10·6 (9 to 12)
10·1 (8 to 12)
Change from baseline, %
35 (21 to 53)
66 (39 to 93)
81 (46 to 117)
13 (−7 to 32)
29 (1 to 57)
22 (−11 to 56)
Right atrial pressure
ptreat=0·94
5·6 (5 to 6)
6·1 (5 to 7)
7·0 (6 to 8)
8·2 (7 to 10)
6·4 (6 to 7)
6·5 (5 to 7)
6·8 (6 to 8)
6·6 (5 to 8)
45 (12 to 78)
87 (45 to 129)
103 (49 to 158)
24 (−8 to 55)
65 (20 to 110)
64 (21 to 117)
Serum creatinine
Absolute values, mg/100 mL
ptreat=0·35
0·8 (0·5 to 1·1)
0·9 (0·6 to 1·4)
1·1 (1·0 to 1·5)
0·9 (0·5 to 1·1)
0·8 (0·4 to 1·2)
0·2 (−7 to 8)
0·5 (−4 to 10)
4 (−7 to 12)
−4 (−16 to 7)
−4 (−17 to 8)
−3 (−14 to 8)
Weight
Absolute values, kg
ptreat=0·072
76 (73 to 79)
77 (75 to 79)
77 (74 to 81)
78 (75 to 81)
76 (73 to 80)
77 (74 to 80)
2·3 (1 to 4)
1·7 (0·2 to 3)
2·9 (1 to 4)
1 (−0·4 to 2)
1 (−0·6 to 2)
0·5 (−1·3 to 2)
Child-Pugh score
Absolute values
ptreat=0·21
5·7 (5·5 to 5·9)
5·9 (5·7 to 6·1)
5·8 (5·4 to 5·9)
5·8 (5·5 to 5·9)
5·9 (5·6 to 6·2)
6 (3 to 10)
2 (−2 to 5)
3 (−2 to 6)
2 (−1 to 5)
−1 (−3 to 2)
3 (−1 to 7)
Model for end-stage liver disease score‡
ptreat=0·085
6·6 (5·9 to 7·3)
7·2 (6·5 to 7·9)
7·2 (6·4 to 8·1)
7·5 (6·6 to 8·5)
6·8 (6·1 to 7·5)
6·1 (5·3 to 6·8)
6·5 (5·6 to 7·3)
7·8 (6·7 to 9·8)
26 (−8 to 51)
19 (−18 to 55)
30 (−13 to 66)
−11 (−27 to 42)
24 (−15 to 64)
68 (20 to 116)
Table 2. Baseline and follow-up haemodynamic variables and liver and renal function
Values are mean (95% CI). A second haemodynamic study at 1 year of follow-up was performed in 156 patients (86%) of the 181 remaining in the study. A third haemodynamic study at 2 years was performed in 86 patients (61%) of the 141 remaining in the study. A fourth study at 3 years was performed in 47 patçients (52%) of the 90 remaining in the study.
*p values from the mixed model for repeated measurements analysis for the terms treatment effect, treatment-by-time interaction, time effect, and randomisation stratum. p values for the baseline measurement term were always <0·001.
†At each time period, haemodynamic studies and laboratory controls were performed in patients who remained in the study and accepted to repeat the procedure. In the β-blockers-group, 94 patients remained in the study at 1 year of follow-up, 72 at 2 years, and 42 at 3 years of follow-up. In the placebo group, 87 patients remained in the study at 1 year of follow-up, 69 at 2 years, and 48 at 3 years of follow-up.
‡Liver function was assessed at each time period by Child-Pugh score and model for end-stage liver disease score. During follow-up, these scores were available in patients remaining in the study, not in those with previous decompensation or death.

19. Results: HVPG response in treatment group
40/78 (51%)
28/78 (36%)

20. Results: Change in haemodynamic variables

21. Results: Long-term outcomes
Placebo group
β-blocker group
Decompensation 24 12
Death from decompensated cirrhosis 9 4
Placebo group
(n=101)
β-blockers
group (n=100)
Risk (95% CI)*
p value†
Decompensation or death
Overall‡
27 (27%)
16 (16%)
0·51 (0·26–0·97)
0·0412
Secondary outcomes
Ascites
20 (20%)
9 (9%)
0·42 (0·19–0·92)
0·030
Gastrointestinal bleeding
3 (3%)
4 (4%)
1·52 (0·34–6·82)
0·61
Overt hepatic encephalopathy
5 (5%)
0·92 (0·40–2·21)
0·98
Death from any cause
11 (11%)
8 (8%)
0·54 (0·20–1·48)
0·23
Varices
56 (56%)
58 (58%)
1·15 (0·65–2·02)
0·72
High-risk varices§
25 (25%)
0·60 (0·30–1·21)
0·15
Spontaneous bacterial peritonitis
2 (2%)
0·49 (0·10–2·70)
0·40
Other bacterial infections¶
19 (19%)
15 (15%)
0·81 (0·41–1·59)
0·54
Hepatorenal syndrome
1 (1%)
0·99 (0·06–15·96)
0·96
Hepatocellular carcinoma
17 (17%)
13 (13%)
0·76 (0·37–1·54)
0·43
EVL β-blocker group vs Placebo group :
11/16 (69%) vs 18/25 (72%)
The primary outcome, excluding bleeding
β-blocker group vs Placebo group :
14 pts (14%) vs 27 pts (27%)
(HR 0.42, 95% CI ; p=0.012)
Table 3. Long-term outcomes
Percentages are crude incidences of events occurring at any time during the follow-up. *
Values indicate the hazard ratio of an outcome in the β-blockers group as compared with the placebo group. †
Comparison of cumulative incidences by competing-risk analysis (differences assessed by Gray's test). ‡
The absolute reduction in the incidence of the primary outcome was of 11% (95% CI 0–22).
Among patients with high-risk varices, oesophageal variceal ligation to prevent bleeding was performed in 18 (72%) of 25 patients in the placebo group versus 11 (69%) of 16 in the non-selective β-blockers group.
Including spontaneous bacterial peritonitis, and other documented bacterial infections during follow-up. doses

22. Results: Causes of death
6 died without decompensation
4 of them was liver-related

23. Results:
“Baseline and follow-up HVPG values in the overall series of patients, comparing those developing the primary end-point vs those who did not”
HVPG (% change vs baseline) 10 5 -5
-10
Patients with decompensation and/or death
Patients surviving without decompensation
M M M M36
(-10% [1%] vs 3% [2%], p=0.001)
(-9% [2%] vs 1% [4%], p=0.020)

24. Results: HVPG decrease at 1-year group

25. Results : Adverse events
Placebo
group
(n=101)
β blockers
(n=100)
Overall
88 (87%)
84 (84%)
Probably related to treatment*
30 (30%)
39 (39%)
Very probably related to treatment†
15 (15%)
16 (16%)
Total number of adverse events‡ 45 57
Major adverse events§
2 (2%)
4 (4%)
Syncope 1 2
Bradyarrhythmia
Heart failure
Acute coronary syndrome
Minor adverse events¶
28 (28%)
36 (36%)
Weakness 17 23
Shortness of breath 4
Dizziness or hypotension 9
Bradycardia 3
Headache
Impotence
Gastrointestinal symptoms
Other 6 7
*Adverse events that were subjectively considered by the investigator to be probably related to drug treatment.
†Adverse events that were subjectively considered by the investigator to be very probably related to drug treatment.
‡Some patients had more than one type of adverse event probably or very probably related to treatment: in the placebo group, 101 patients had 45 adverse events and in the β-blockers group 100 patients had 57 adverse events.
§Adverse effects were considered severe if the health or safety of the patient was endangered.
¶In the placebo group, 28 patients had 43 minor adverse events. In the β-blockers group, 36 patients had 52 minor adverse events and, in this group, two patients had both minor and major complications.

26. Discussion
In patients with compensated cirrhosis and CSPH, long-term treatment with NSBB improves decompensation-free survival, mainly by decreasing the incidence of ascites.
Competing-risk analysis
Considering non-liver related death as a competing event
Adequate β blockade: significant reduction in heart rate and cardiac index  improvement in HVPG

27. Discussion Current study vs Timolol study
Patient selection : HVPG ≥ 10 mmHg, varix(-) or small varix without red signs vs HVPG ≥ 6 mmHg, varix (-)
In cirrhosis with HVPG < 10 mmHg, the hyperdynamic syndrome is still underdeveloped and β-blockers induce a much lower effect on portal pressure than when CSPH develops
Carvedilol has a greater HVPG decreasing effect than propranolol
Intrinsic vasodilator activity (anti-α-adrenergic activity and enhanced release of NO)
Slightly improvement in outcomes and better adherence to therapy with carvedilol than with propranolol
All patients could have been treated upfront with carvedilol, thus omitting the need to assess HVPG response

28. Discussion
Treatment with β blockers was particularly successful in patients with small varices.
Patients with varices have a more developed hyperdynamic circulation and higher portal pressure than those without varices
Current guidelines; prophylactic treatment with β blockers or oesophageal variceal ligation to prevent bleeding once high-risk varices develop.
Use of β blockers in patients with small varices ; β blockers might prevent the progression of small to large varices
The benefit of β blockers was mainly apparent after the first 24 months of follow-up, progressively increasing thereafter.

29. Discussion HCV cirrhosis with active infection
 new DAA were introduced during the last year of study
 DAA can improve liver fibrosis and portal hypertension and can prevent decompensation
 study termination before planned
Despite of sustained virological response, a large portion of patients with cirrhosis and CSPH still maintain CSPH and are at risk of decompensation, even if HVPG improves.  β blockers might still be needed after virological response

30. Discussion: limitation
The results cannot be generalized to all patients with cirrhosis
Exclusion of high-risk varices
Using HVPG
Several non-invasive tools (Elastography, imaging techniques by detecting varices or portal-systemic collaterals)
Open-label titration period or heart-rate changes might affect study blindness
A study nurse measured vital signs to keep investigators unaware
Small size
Sample size assumption was reached

32. Introduction
Oesophageal variceal haemorrhage is one of the major complications of portal hypertension
The most important predictor of haemorrhage is the size of varices
Therefore, it seems logical to delay the growth of small oesophageal varices to reduce the incidence of acute bleed and overall mortality to large size by reducing portal pressure.
NSBBs reduce portal pressure used for primary prophylaxis against variceal haemorrhage in cirrhotic patients with large oesophageal varices

33. Introduction
Cumulative probability of (A) remaining free of growth of esophageal varices and (B) survival.
Sarin SK , Mishra SR , Sharma P , et al . Early primary prophylaxis with beta-blockers does not prevent the growth of small esophageal varices in cirrhosis: a randomized controlled trial. Hepatol Int 2013;7:248–56.
Merkel C , Marin R , Angeli P , et al. , Gruppo Triveneto per l'Ipertensione Portale. A placebo-controlled clinical trial of nadolol in the prophylaxis of growth of small esophageal varices in cirrhosis. Gastroenterology 2004;127:476–84.

34. Aim
To compare carvedilol with placebo for early primary prophylaxis to prevent the progression of small oesophageal varices in cirrhotic patients

35. Methods: Participants
Patients with cirrhosis presenting to the Department of Hepatology, Institute of Liver and Biliary Sciences (ILBS), from November 2010 to December 2012
1:1 ratio randomised, placebo-controlled clinical trial
Inclusion criteria
Cirrhotics with small oesophageal varices (≤ 5mm in diameter) without history of previous variceal bleeding
Cirrhosis was diagnosed on clinical, radiological or laboratory parameters and/or liver biopsy.
Exclusion criteria
Age <18 yrs or > 75 yrs
Non-cirrhotic portal hypertension
Child-Turcotte-Pugh score > 12
Refractory ascites
Acute kidney injury
History of prior EVL or sclerotherapy
History of surgery for portal hypertension
Significant cardiopulmonary comorbidity
Uncontrolled diabetes
Peripheral vascular disease
Presence of any malignancy that significantly affects survival
Evidence of alcoholic hepatitis or active alcohol abuse with last intake ≤ 1month or refusal to participate in the study

36. Methods: Endoscopic assessment
Size: Small (< 5mm) vs Large (> 5mm)
Measured against the gap between the jaws of the open endoscopic biopsy forceps (5mm)
Grade I and II based on the presence of varices in one or both phases of respiration and with full air insufflation or not
The interobserver variability was determined
In case of any disagreement, the recorded image were retrieved and reviewed by a third endoscopist

37. Methods: Randomisation and Treatments
Randomisation : independent statistician, block randomization method, single-blind
Carvedilol group
Start dose of 3.125mg twice daily
At 1 week, increased to 6.25mg twice daily, maximum dose of 12.5mg twice daily
SBP > 100 mmHg and HR ≥ 55
Placebo group
Start with one tablet twice daily
Gradually increased, finally four, twice a day in the next 3 weeks

38. Methods: Follow-up Clinical visit : 1 week, 6, 12, 18 and 24 months
During each follow-up, clinical examination, CBC, kidney and liver function tests were done
Upper GI endoscopy (UGIE) at baseline and at 6-9 monthly intervals
Every patient must complete at least three endoscopic assessments during the follow-up
All endoscopist were kept unware of the treatment arm
HVPG at baseline, at 1 year
By introducing a 7 Fr. Swan-Ganz catheter via the transfemoral approach into a major hepatic vein
The occluded position was checked by the absence of reflux after the infection of 1-2ml of contrast medium and appearance of a sinusoidogram
A mean of three readings was taken
Compliance assessed through pill count and periodic telephonic conversations every fortnightly

39. Outcomes Primary outcomes
non-progression from small to large oesophageal varices at 2 years
Secondary outcomes
Reduction in HVPG at 1 year
HVPG non-responders : HVPG reduction of <10% compared with baseline
HVPG responders : HVPG reduction of ≥20% compared with baseline
HVPG partial responders : HVPG reduction of 10%≤ < 20% compared with baseline
Survival at 24 months
Treatment-related adverse events

40. Methods: After primary outcome
Patients developing large oesophageal varices were subjected to EVL and underwent repeat UGIE every 3 weeks to assess for EVL after the initial EVL session, until variceal eradication was achieved.
They were followed up with 3-6 monthly once surveillance to see for variceal recurrence.
Bleeding from oesophageal varices was to be treated by a combination of medical and endoscopic treatment

41. Statistical analysis Sample size calculation
Assumed non-pregression to large varices in carvedilol group (80%) and in placebo (50%)
Α 5%, power 90% -> 114 cases
Further 20% attrition rate, 70 cases in each arm
Baseline parametric data : the proportion, mean±SD, median with IQR
The intention-to-treat and per-protocol analyses
The difference in the groups : χ2 test, unpaired t test or Mann-Whitney test
Comparing the change before and after the intervention : Paired t test
The cumulative probability of development of large varices and overall survival : Kaplan-Meier plots, log- rank test
The agreement between the two endoscopists was seen by к statistics
The change in the HR and mean arterial pressure measurements between the two treatment groups at different time intervals : repeated measures followed by post hoc comparison by least square method

42. Results: Study design
The agreement between the two endoscopist : 0.83 (к statistics)

43. Results: Baseline characteristics
Intention-to-treat analysis
Per-protocol analysis
Parameters
Carvedilol group (N=70)
Placebo group (N=70)
p Value
Carvedilol group (N=66)
Placebo group (N=67)
Age (in years)
48.8±10.3
48.8±10.5
0.99
48.91±10.30
49.18±10.57
0.88
Male:female
60:10
59:11
0.81
56:10
56:11
0.84
Ascites, n (%)
6 (8.6)
11 (15.7)
0.19
6 (9.09)
10 (14.93)
0.30
Total bilirubin (mg/dL)*
2.4±1.9
1.8 (1.1–3.05)
3.2±4.9
1.7 (1.25–3.2)
2.5±1.9
1.73 (1.1–3.05)
3.3±5.1
1.7 (1.23–3.2)
0.77†
Serum albumin (mg/dL)
3.33±0.68
3.24±0.66
0.39
3.32±0.69
3.25±0.66
0.56
Serum creatinine (mg/dL)
0.77±0.29
0.73±0.25
0.32
HVPG (mm Hg)
14.49±4.31
14.97±4.76
0.52
14.32±4.34
15.0±4.8
CTP
6.85±1.73
6.96±1.83
0.72
6.87±1.74
6.89±1.78
0.94
MELD
9.39±2.58
9.86±2.79
0.35
9.52±2.61
9.79±2.72
0.59
Aetiology of cirrhosis
 Hepatitis B
5 (7.1%)
15 (21.4%)
0.02
 Hepatitis C
7 (10.0%)
8 (11.4%)
0.78
 Ethanol
18 (25.7%)
0.55
 Cryptogenic
34 (48.6%)
28 (40.0%)
0.31
 Others
9 (12.9%)
1 (1.4%)
0.008
*Median with IQR along with mean±SD.
†By Mann–Whitney U test.
CTP, Child-Turcotte-Pugh score; HVPG, hepatic venous pressure gradient; MELD, model for end stage liver disease; S Alb, serum albumin; S Creat, serum creatinine; T Bil, total bilirubin.
HBV LC, tenofovir 300mg or entecavir 0.5mg orally once daily if HBV DNA ≥ 2000 IU/mL
HCV LC, PEG-IFN α-2a/2b and weight-based ribavirin for 24/48 weeks

44. Results: Primary outcomes
Intention-to-treat analysis
Per-protocol analysis
Parameters
Carvedilol
group (N=70)
Placebo
p Value
group (N=66)
group (N=67)
The cumulative probability of non-progression to large varices (%)
79.4
61.4
78.5
59.8
The mean time to non-progression to large varices (months)
20.8
( )
18.7
( )
0.04
20.7
( )
18.5
( )
0.03
Free from progression to large varices
Carvedilol vs placebo
56(80%) vs 45(64.3%), p=0.04
Time (months) 6 12 24
Gastric varices (n=13)
Carvedilol 2 1
Placebo 4

45. Results: Overall survival
The cumulative probability of survival at 24 months
96.3% in carvedilol group vs 85.8% in placebo group, p=0.034
Deaths
Carvedilol group(2) : renal cell carcinoma(1), urosepsis(1)
Placebo group(8) : intracranial haemorrhage(1), pneumonia followed by septic shock(3), acute coronary syndrome(2), renal failure(2)
No liver-related morbidity or mortality

46. Results: Reduction in HVPG
HVPG at baseline
HVPG at 1 year
p value
Carvedilol (n=52)
14.5±4.3
13.4±5.8
0.07
Placebo (n=48)
14.66±4.7
13.9±5.1
0.29
The overall mean percentage
Carvedilol vs Placebo
(- 8.64%) vs (+ 0.33%), p=0.22

47. Results: Adverse events
In carvedilol group
Non-compliance (3)
Lightheadeness (2) : spontaneously resolved
Bradycardia with hypotension(1) : the drug had to be discontinued
In placebo group
Non-compliance (1)
Follow-up loss (2)

48. Discussion
Portal hypertension is the single most important predictor of complications in cirrhosis of liver
Staging classification of cirrhosis (Baveno IV)
Mortality : 3.4% per year in stage 2 vs 20% per year in stage 3
 Early primary prophylaxis is important
Previous study (propranolol, nadolol) didn’t carefully studies the HVPG response with the progression of varices
de Franchis R . Evolving consensus in portal hypertension. Report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005;43:167–76.

49. Discussion
Carvedilol delays the progression from small to large varices
The dual effect of carvedilol on α1 and β receptors
HVPG reduction : Carvedilol > propranolol
A modest reduction of portal pressure by 8-10% maintained over a long period of time may have a significant effect
Need for search for newer treatment options to reduce portal pressure and hepatic fibrosis
Terutroban significantly reduced portal pressure in CCl4- cirrhotic rats by action at thromboxane-A2/prostaglandin- endoperoxide recetors on the hepatic vascular endothelium leading to amelioration of fibrosis
Overall better survival at 2 years was noted by use of carvedilol; none of them were liver related
Bañares R , Moitinho E , Matilla A , et al . Randomized comparison of long-term carvedilol and propranolol administration in the treatment of portal hypertension in cirrhosis. Hepatology 2002;36:1367–73.
Rosado E , Rodríguez-Vilarrupla A , Gracia-Sancho J , et al . Terutroban, a TP-receptor antagonist, reduces portal pressure in cirrhotic rats. Hepatology 2013;58:1424–35.

50. Discussion: Strength & limitation
Homogenous set of patients was included
Nearly 70% of the patients had repeat HVPG performed at 1 year
One of the largest in early primary prophylaxis study
This study could have been further strengthened if HVPG could have been repeated after 2 years
Need for reliable and validated assessment by non-invasive methods like transient elastography

51. Critical appraisal 주논문
Invasiveness : hemodynamic study including HVPG in enrollment
Open-label titration period : breaking the blindness
부논문
HVPG reduction in Carvedilol group vs placebo group
p value >0.05 -> not significantly different
Overall survival in Carvedilol group vs placebo group
None was liver-related -> does carvedilol really improve the overall survival?

52. Critical appraisal
New guideline for early primary prophylaxis of decompensation of cirrhosis, including variceal hemorrhage, ascites, etc…
Patient selection : CSPH, small varices, non-alcoholic …
Drug selection : Carvedilol, propranolol, new drugs …
Predictive factor : HVPG reduction, non-invasive methods(transient elastography) …
G Garcia-Tsao, JG Abraldes, A Berzigotti, J Bosch. Portal hypertensive bleeding in cirrhosis: Risk stratification, diagnosis and management: 2016 practice guidance by the American Association for the Study of Liver Diseases. Hepatology, 65 (2017), pp
β blockers : Carvedilol > Propranolol> …
New drugs