Portal hypertension, varices and bleeding

Pathogenesis of portal hypertension

25% hepatic flow comes from portal vein from the GIT (75% direct via hepatic artery)
portal hypertension= incr pressure in portal vein
- pressure=flow x resistance
  - resistance= 70% fixed + 30% dynamic (i.e. amenable to drug manipulation)
- 2 theories for cause of incr pr:
  - "forward theory"- incr splanchnic flow overcomes massive porto-systmeic flow
  - "backward theory" - incr resistance to outflow
get venous and arterial stealing in hyperdynamic circulation (i.e. incr BP and HR)
- venous steal
  - incr intrahepatic pressure causes decrease in portal vein flow ("backed up")
  - blood flows from portal vein straight into hepatic vein (bypass the liver) via portosystemic collaterals
  - results in hepatic ischaemia
- arterial steal
  - increased flow to splanchnic arteries
  - vasodilation and blood pooling in gut
  - decreased flow to peripheries, head and neck, kidneys (renal underperfusion)

1 of 6

present in 30-40% of pt w/ cirrhosis
- in 60% with decompensated cirrhosis
risk of bleeding is 30%
- rebleed irsk up to 70%
- 1 yr mortality following a bleed up to 40%
risk factors for variceal bleeding
- appearance
  - large size, red signs, active bleeding
- wall tension
- portal pressure
- alcohol intake
- HCC
- MELD/SOFA/APACHE score
  - MELD= model for end-stage liver disease
    - looks at bilirubin, creatinine, INR and dialysis

2 of 6

3 of 6

volume expansion
- aim to preserve tissue perfusion
- restrictive transfusion (to get pt Hb to 7-8) in haemodynamcacilly pts
- little evidence for use of blood products
- must correct hypovolaemia and promote tissue oxygenation (tissue ox= SaO2 x CO x Hb)
antibiotics prophylaxis
- reduce infection (65%), rebleeding (47%) and mortality (21%)
Terlipressin
- tri-glycl vasopressin analogue
- 2mg= 21% decr in portal pr for 4 hrs
- mecahnism of action:
  - V1 R - splanchnic vasoconstriction
  - V2 R - reduce renal free water clearance
  - incr SVRI, incr CV (incr renal perfusion)
  - reduces renin
- 39% reduced mortality

4 of 6

VBL
- first line for acute bleed (prev. sclerotherapy)
- multibanders rountinely used
- good initial control but rebleed in 50%
- best when combined w/ vasoactice drugs
  - incr control, decr rebleed but NO survival benefit
ballon tamponade
- good initial haemostasis
- controls oesophageal and gastric varcieal bleeding
- high rebleed rate and local complications (ischaemia)
  - dangerous if inexperienced user
TIPSS (transjugular intrahepatic portosystemic stent-shunt)
- new and promising
- artificial stent b/w portal vein and hepatic vein
- better than VBL/BT for rebleed but no effect on total mortality
- excellent haemostasis as salvage therapy but high mortality

5 of 6

drug therapy
- NSBB reduces bleeding and mortality
- nitrates (isosorbite mononitrate/ISMN) effective in pts who don't respond to NSBB
  - little benefit of adding VBL
- NSBB + ISMN as effective as VBL
endoscopic tehrapy
- sclerotherapy is effective but assoc w/ a/e
- VBL- earlier eradication ,better outcomes, less a/e
- sclerotherapy and VBL combination not recommended bc lots of a/e

6 of 6