FK506 is implicated in
producing veno-occlusive lesions and associated centrilobular necrosis.
Histology in relation to rejection within a 1 year period has been previously
assessed, but long term assessment has not been undertaken.
This study
assessed biopsies and biochemistry of patients transplanted in Birmingham as part of the original European
Multicentre Study(1). The age and sex distribution and reason for
transplantation were similar in both groups. The largest proportion of patients
in both groups were transplanted for primary biliary cirrhosis (FK: 52.5%, CyA:
41.5%).
Protocol biopsies were performed at 1 week, 1, 2 and 3 years and 3-5
yearly thereafter. Most patients also had a time zero (reperfusion) biopsy.
Additional biopsies were taken when clinically indicated. 40 and 41 patients
started in the FK and CyA arms respectively, 4 from each were withdrawn prior to
a 1 week protocol biopsy, a further 9 were withdrawn from the FK arm and 12 from
the CyA arm prior to the 1 year biopsy and a further 4 FK and 9 CyA were
withdrawn after the 1st year. The median time to withdrawal in the 1st year
(excluding failures before 1 week) was FK: 64 days (14-125 days) and CyA: 27.5
days (13-186 days). The median time to withdrawal after the first year was FK: 4
years (1.2-5.8 years) and CyA: 4.4 years (1.8-7.7 years). There was no
significant difference between the two groups in either the time to withdrawal
or the number of biopsies taken at any time point.
Biopsies were
assessed semiquantitatively or quantitatively, blinded to treatment group, for
portal tract infiltrate, bile duct inflammation, portal vein inflammation,
interface hepatitis, percentage of portal tracts lacking bile ducts and hepatic
artery branches, bile ductular proliferation, hepatic vein inflammation,
perivenular (zone 3) inflammation, centrilobular necrosis, centrilobular
congestion/haemorrhage, micro- and macro- vesicular steatosis, lobular
inflammation, hepatocyte apoptosis, ballooning, cholestasis, parenchymal and
portal granulomas, zone 3 fibrosis, periportal fibrosis, perisinusoidal fibrosis
and hepatic veno-occlusive lesions, Time zero biopsies were also assessed for
neutrophil aggregates. The AST, alkaline phosphatase and bilirubin levels at the
time of biopsy were also analysed. Differences between the groups were assessed
using non-parametric statistics and statistical significance was considered at
p<0.05
RESULTS:
TIME ZERO:
There was no significant difference
between the two groups in features of a preservation-reperfusion injury which
was mild in both groups with a median peak day 1 or 2 AST of FK: 625 (95-3095)
and CyA: 554 (152-8350).
LESS THAN 1 YEAR:
Day 7
protocols: The CyA group had significantly more inflammation of the portal
tracts, bile ducts and portal veins, interface hepatitis, centrilobular
necrosis, centrilobular congestion/haemorrhage and hepatocyte apoptosis. These
changes were associated with significantly higher AST and bilirubin
levels.
All biopsies in first month (excluding time zero):
Centrilobular congestion/haemorrhage, AST and bilirubin remained
significantly higher in the CyA group, whereas the other histological features
were no longer significantly different. Cholestasis was significantly greater in
the CyA group. Two CyA grafts failed from rejection and 1 FK graft.
This
indicates there is more severe rejection in day 7 protocols in CyA, however over
the first month, the rejection tends to even out.
All biopsies 1 Month to
1 Year
FK had significantly more bile duct loss (median of 16.7% v 1.1%)
and zone 3 fibrosis. 4 grafts were lost in the FK group from rejection related
causes, 2 from chronic rejection. 2 grafts were lost in the Cya group from
rejection related causes, one from chronic rejection. There is no significant
difference between the two groups for graft failure / death.
MORE
THAN 1 YEAR
All biopsies > 1 year
There was
significantly more centrilobular inflammation in the FK group and significantly
more cholestasis in the CyA group. There was 1 graft failure in the CyA group
and possibly another (died in Egypt of unknown cause) and no graft failures in
the FK group after 1 year.
GRANULOMAS AT ANY
TIME
There were significantly more patients with at least one biopsy
showing a granuloma (in either portal or parenchymal locations) in the FK group
(37.9% V 6.7%). If patients transplanted for PBC were excluded this no longer
reached statistical significance (FK:16.7% v CyA: 5.6%). In patients
transplanted for PBC the significant difference remained (FK: 53% v CyA:
9.1%).
CENTRILOBULAR NECROSIS (CLN)
With the
exception of day 7 protocol biopsies, when CyA had more severe CLN, there was no
significant difference between the severity of CLN in biopsies between the two
groups.
There were 17 patients that never had a CLN score > 1, only 1
patient developed graft failure, which was due to recurrent hepatitis B. There
were no graft failures due to rejection. There were 18 patients that developed
severe bridging (grade 3) CLN, one patient was lost to follow-up. 5/17 (29.4%)
grafts failed: 3 due to chronic rejection, 1 due to veno-occlusive disease and 1
related to persistent CLN and the development of chronic hepatitis. This is
statistically significant.
There was a correlation between the severity of
centrilobular necrosis occurring before 1 year and the degree of zone 3 fibrosis
after 1 year. Mostly, the fibrosis was delicate and bridged between hepatic vein
branches at its worst. In some cases it appeared to be reversible and usually
appeared to have no impact on graft function or clinically detected portal
hypertension.
VENO-OCCLUSIVE LESIONS
There was no
significant difference between the two groups at any time
point.
CONCLUSIONS
FK506 is not
associated with an increase in CLN or veno-occlusive lesions. CLN is associated
with the development of zone 3 fibrosis, which alone does not appear to effect
the graft function. Bridging CLN is associated with a higher rate of graft loss
than grafts with no or minimal CLN. FK506 may cause a granulomatous hepatitis.
The increased centrilobular inflammation in long-term FK506 patients may be due
to the absence of dual treatment with azathioprine used in the long-term CyA
patients. CyA is associated with histological cholestasis.
1.
Randomised trial comparing tacrolimus (FK506) and cyclosporin in prevention of
liver allograft rejection. European FK506 Multicentre Liver Study Group. Lancet
1994; 344(8920):423-428.