Transcript Document

The European experience
with protocol biopsies
Ian Roberts
Oxford, UK
Oxford Pathology
Should protocol biopsies be part
of routine patient monitoring?
What is the evidence from the European experience of
protocol biopsies?
How does this differ from the American, Asian and
Australian experience?
How relevant is published data from elsewhere in the
world to European practice?
Should protocol biopsies be part
of routine patient monitoring?
Protocol biopsies = allograft biopsies performed at pre-determined
time points, irrespective of graft dysfunction.
Functions: graft selection, donor disease, baseline for posttransplant pathology, subclinical rejection, early diagnosis and
quantification of chronic allograft nephropathy.
Risk
Benefit
morbidity associated with
allograft biopsy
information that can be acted on
to improve graft outcome
Impact of disease in the donor
kidney: implantation biopsies
Biopsies of donor kidneys:
210 transplants, 129 no GS, 81 showed 1-60% GS
No GS
>20% GS
GFR at 1 year
62
36
5 year graft survival
80%
35%
Escofet et al. Transplantation 2003;75:344-6
Impact of disease in the donor
kidney: implantation biopsies
Baseline for interpretation of later biopsy changes.
Interpretation of sub-optimal function.
Choice of immunosuppressive regimen.
What level of risk is associated
with protocol biopsies?
Retrospective audit of 2127 protocol biopsies audited for
complications
No deaths
One graft loss
3 episodes of haemorrhage requiring intervention.
3 further patients requiring transfusion.
2 episodes of peritonitis
Furness et al. Transplantation 2003;76:969-73
What determines the potential
benefit of protocol biopsies?
For protocol biopsies performed to diagnose subclinical
rejection, the likelihood of benefit to an individual patient
is a product of:
Frequency of subclinical rejection at the time of protocol
biopsy in the patient population
X
The increased graft longevity to be gained from treating
the subclinical rejection
Incidence of subclinical rejection
Reference
Number of patients
Time of biopsy
Subclinical AR (%)
Borderline (%)
Rush et al, 95
25
1 month
3 months
6 months
20
16
12
44
28
8
Seron et al, 97
98
3 months
4
12
Rush et al, 98
35
1 month
3 months
6 months
43
27
15
11
21
29
Jain et al, 00
78
3 months
1
11
Jain et al, 00
50
1 week
4
8
Nankivell et al, 01
102
3 months
29
49
Shapiro et al, 01
28
1 week
25
21
Gloor et al, 02
114
3 months
3
11
Roberts et al, 04
115
1 week
1 month
13
8
12
16
Impact of immunosuppression on
subclinical rejection
119 SPK recipients
Subclinical rejection 60.% at 1 month, 45.7% at 3 months
No significant impact of immunosuppression on subclinical
rejection at 1 month.
Protocol biopsies 3-12 months:
odds ratio
Tacrolimus-MMF
1
CsA(S)-azathioprine
21.8
CsA(N)-azathioprine
22.1
CsA(N)-MMF
11
Nankivell et al Transplantation 2004;78: 242-49
Impact of immunosuppression on
subclinical rejection
Case-control study 98 patients
Tacrolimus-MMF-prednisolone vs CsA-MMF-prednisolone
4 months protocol biopsies scored using Banff 97:
Acute score
Chronic score
Tacrolimus
0.61 ± 1.01
1.57 ± 1.97
CsA
1.26 ± 1.45
1.51 ± 1.59
p=0.0115
p=ns
Relative risk of acute inflammation with Tacrolimus 0.3
(CI 0.11-0.84)
Moreso et al Transplantation 2004;78: 1064-8
Impact of immunosuppression on
subclinical rejection
Comparison of subclinical rejection in two eras:
Group 1 (1992-5) received CsA, azathioprine and
prednisolone
Group 2 (2001-3) stratified at transplantation according to
immunological ris:
High risk: basiliximab induction, tacrolimus, MMF & pred
Medium risk: CsA, MMF & pred
Low risk: CsA, azathioprine & pred
Roberts et al ERA/EDTA 2005
Impact of immunosuppression on
subclinical rejection
Roberts et al ERA/EDTA 2005
The incidence of early SCR is less
than it appears
SCR = histological changes meeting recognised criteria for
acute rejection in patients with good, stable graft function.
but this also applies to pre-clinical and post-clinical rejections.
True SCR = immune-mediated graft injury, not associated with
acute graft dysfunction if untreated.
The incidence of early SCR is less
than it appears
Day 7 protocol biopsies
Acute rejection
8 treated immediately
Borderline changes
4 treated immediately
10 (13%)
1 of the 2 untreated patients
developed CR in <3 days
9 (12%)
3 of the 5 untreated patients
developed CR in <3 days
Roberts et al Transplantation 2004;77:1194-98
The incidence of early SCR is less
than it appears
Day 28 protocol biopsy
10/13 patients who had subclinical borderline changes had
been treated for acute rejection in the previous 3 weeks
Roberts et al. Transplantation 2004;77:1194-98
Protocol biopsies 10 days after treatment of clinical acute
rejection.
Over 25% of patients with full clinical response to antirejection therapy have persistent histological signs of
rejection.
Mazzucchi et al. Transplantation 1999;67:430-34
How common is subclinical
antibody-mediated rejection?
Day 7 protocol biopsies in 37 living donor recipients
C4d positivity in 11 (30%)
9/11 showed clinical rejection
6/11 showed histological rejection
2/37 (6%) subclinical C4d positivity
Sund et al. Transplantation 2003;75:1204-13
How common is subclinical
antibody-mediated rejection?
48 day 7 protocol biopsies (Cad/LRD 41/7)
C4d positivity in 6 (13%)
5/6 showed clinical and histological rejection
all 5 with rejection had post-transplant DRA
One patient C4d positive, no histological rejection, no
circulating alloantibodies, stable graft function (ie.
doesn’t meet Banff criteria for humoral rejection) never treated for rejection, sCr 104 micromol/L at 3
months
Koo et al. Transplantation 2004;78:398-403
How common is subclinical
antibody-mediated rejection?
Multicentre study, 551 protocol biopsies
Diffuse C4d positivity in 2%
Focal C4d positivity in 2.4%
Inter-centre variation reflecting variable numbers of presensitised patients.
Subclinical C4d positivity had no significant impact on
graft outcome.
Mengel et al. Am J Transplant 2005;5:1050-56
How common is subclinical
antibody-mediated rejection?
Does subclinical C4d positivity equate to subclinical
antibody-mediated rejection?
C4d positivity not due to antibody deposition (false
positive).
Accommodation?
Does early SCR contribute to
chronic graft injury?
120 recipients of kidney-pancreas grafts
Protocol biopsies at 1 & 2 weeks, 1, 3, 6 & 12 months,
annually thereafter. 70% of patients suffered clinical or
SCR/borderline changes in first 3 months.
Patients with acute SCR in first 3 months showed higher
grades of CAN in subsequent biopsies than did patients with
no SCR (2.48 ±0.25 vs 0.32 ±0.09).
But: 3 month risk of SCR was increased by a previous
episode of severe rejection (hazard ratio 2.46), no data on
treatment of SCR.
Nankivell et al. N Engl J Med 2003;349:2326-33
Does early SCR contribute to
chronic graft injury?
Day 28 protocol biopsies:
19 patients with subclinical rejection or borderline changes.
4 treated immediately, 3 for clinical rejection within one week.
12 never subsequently received treatment for rejection:
1 late graft failure at 6 years post-transplantation
2 patients died with functioning grafts (sCr 159 & 168)
9 had functioning grafts at 6 years
median sCr 119 umol/l (range 90-185)
Roberts et al Transplantation 2004;77:1194-98
Does treatment of SCR improve
graft outcome?
72 patients randomised to protocol biopsies 1, 2, 3, 6 and
12 months vs 6 and 12 months only.
In the biopsy group, 80% of 1 month biopsies showed
either acute rejection or borderline changes.
Sub-clinical rejection treated.
Early biopsy group had less chronic damage in 6 month
biopsies and lower serum creatinine at 2 years than
controls.
Rush et al. J Am Soc Nephrol 1998;9:2129-34
Protocol biopsies and the early
diagnosis of CAN
Rationale:
Renal scarring on histology is an earlier/more
sensitive marker of chronic damage than is serum
creatinine/GFR - by the time renal function deteriorates, it’s
too late.
Protocol biopsies and the early
diagnosis of CAN
Assumptions:
1. Early histological evidence of chronic damage predicts
chronic graft failure.
2. Intervention in patients with good renal function but CAN
on biopsy can prevent/limit progression.
Problems:
1. Sampling error: subcapsular scar, vessel disease
2. Stable chronic damage
How common is early CAN?
41 patients (10 HLA identical LRD, 31 cadaveric donor kidneys).
Protocol biopsies at 3 months and 2 years.
No CAN in recipients of HLA identical kidneys.
Cadaveric kidneys:
3 months
2 years
CAN
25%
50%
grade I
7
10
grade II
1
5
CAN correlates with donor age
Legendre et al. Transplantation 1998;65:1506-09
Is sampling error a problem in
CAN diagnosis?
310 protocol biopsies from 155 patients.
Protocol biopsies at 4 & 14 months, follow-up 5-12 years.
graft survival
No CAN in either biopsy 54 (35%)
100%
CAN in 2nd but not 1st 39 (25%)
81.6%
CAN in 1st but not 2nd 19 (12%)
82.6%
CAN in both
69.4%
43 (28%)
Assuming CAN does not regress, 25% of biopsies are misclassified,
greater than increase in CAN from 4-14 months.
Seron et al. Kidney Int 2002;61:727-33
Does early CAN predict chronic
graft failure?
621 protocol biopsies in two multicentre MMF trials
Protocol biopsies at baseline, 1 & 3 years.
1 year: sCr (>1.5 mg/100ml) in 20%, CADI score (>2) in 60%
Mean CADI score:
baseline
1 year
3 years
1.3 ± 1.1
3.3 ± 1.8
4.1 ± 2.2
CADI score 1 year
<2
2-3.9
>4
Graft failure 3 years
0
4.6%
16.7%
CADI score predicts graft survival even if normal renal function
Yilmaz et al. J Am Soc Nephrol 2003;14:773-9
Temporal relationship between
histological CAN & graft function
51 consecutive cadaveric transplant recipients
Protocol biopsies at 0, 3, 6, 12 months
3 months
6 months
12 months
CAN
4%
12%
49%
Cr Cl
56 ± 2 mL/min
24 months
56 ± 2 mL/min
Baboolal et al. Kidney Int 2002;61:686-96
Does early CAN predict chronic
graft failure?
57 living donor transplants
Protocol biopsy at day 0 (n=57), 1 week (n=43) & 1 year (n=33)
Poor correlation between arteriosclerosis in day 0 and 1 week
biopsies (sampling error).
Glomerulosclerosis, interstitial fibrosis and tubular atrophy, but not
arteriosclerosis, increased from baseline to 1 year.
Extent of chronic tubulointerstitial damage in 1 year biopsies, but not
baseline biopsies, correlated with sCr at 1 & 3 years.
Best predictor of renal function at 3 years was sCr at 1 year.
Sund et al. Nephrol Dial Transplant 1999;14:2445-54
Temporal relationship between
histological CAN & graft function
258 patients
Protocol biopsies at 6, 12 and 26 weeks
In multivariate analysis, CAN in biopsy 3 (n=70) predicted by:
CAN in biopsy 2
Lower GFR at biopsy 2
Nephrocalcinosis
Acute rejection
Renal functional decline preceded morphological changes of CAN
Schwarz et al 2005;67:341-48
Does type of CAN predict chronic
graft failure?
3 month protocol biopsies in 280 patients
All stable renal function, sCr <300 micromol/L.
Number
10 year graft survival
No CAN
174
95%
CAN (cv0)
87
82%
CAN (cv1+)
21
41%
Seron et al. Transplantation 2000;69:1849-55
Can we do better than traditional
histology?
Computerised image analysis with Sirius Red stain
Protocol biopsies in 68 patients at 6 months
Volume of interstitial fibrosis correlated better with time to graft failure
(r=0.64) than did Banff chronic and acute sum scores (r=0.28 & 0.35).
Multivariate model including all of the above was highly correlated
with time to graft failure (r=0.7, p=<0.0001).
Grimm et al. J Am Soc Nephrol 2003;14:1662-8
Can we do better than traditional
histology?
TGF-beta
Protocol biopsies in 40 patients at day 0, 1 wk, 6 mnths
IF for active TGF-beta & semiquantitative scanning laser confocal
microscopy.
active TGF-beta from day 0 to 1 wk to 6 mnths (p<0.0001)
DGF  TGF-beta, but no association with acute rejection or IS drug
(CsA vs Tac).
No correlation between TGF-beta expression at any timepoint and
isotope GFR at 12 months.
Jain et al. Transplant Int 2002;15:630-4
Can we do better than traditional
histology?
Other techniques with potential:
computerised morphometry combined with IH for:
type III collagen
tenascin
smooth muscle actin
Can we do better than traditional
histology?
Computerised image analysis (Sirius red) vs Banff
scoring
Protocol biopsies in 33 patients at day 0, 1 year
Sirius red (Vint) correlates with Banff chronic score r=0.439, p=0.0003
1 year biopsies:
Banff chronic score correlated with 8-10 year sCr, p=0.01, and with
late graft loss, p=0.0445
Sirius red (Vint) showed similar trend but did not reach statistical
significance
Sund et al. Nephrol Dial Transplant 2004;19:2838-45
Can we do better than traditional
histology?
Other techniques with potential:
quantification of mast cells
Roberts et al. J Clin Pathol 2000;53:858-62
Can we do better than traditional
histology?
Other techniques with potential:
quantification of mast cells
Biopsy diagnosis
Acute rejection, subsequently
progressing to chronic
Acute rejection, not
progressing to chronic
Chronic rejection
Mast cells/mm2 of cortex
Median (interquar tile range)
3.0 (1.9-5.1)
p value (Mann-Whitney U)
2.7 (1.0-5.3)
acute vs chronic rejection
<0.0001
27.1 (20.7-29.8)
Acute CsA toxicity
2.0 (1.3-2.5)
Chronic CsA toxicity
10.6 (7.2-21.3)
acute vs chronic CsA
toxicity <0.0001
Roberts et al. J Clin Pathol 2000;53:858-62
CAN as a predictor of graft
outcome
CAN alone is not a useful label.
What matters is:
1. Progression of CAN - need at least 2 biopsies
2. Type of CAN - cv lesions most significant
3. Severity of CAN - quantitative image analysis
better than Banff scoring
4. Immune activity
Does intervention in subclinical
CAN make a difference?
Assumption 2:
Intervention in patients with good, stable renal
function but CAN on biopsy can prevent/limit
progression.
Evidence?
Would you change immunosuppression in a patient
with good stable graft function on the basis of
some minor chronic changes in the biopsy?
Should protocol biopsies be part
of routine patient management?
Baseline biopsies: yes - useful in the interpretation of
subsequent pathology.
Post-transplantation: maybe, but must be clear on why they
are being performed and have a protocol for acting on the
results of the biopsies.
Subclinical rejection: overall benefit depends on the frequency
of SCR in the unit. The benefit of treating SCR remains
unproven.
Early CAN: a good early surrogate marker for graft survival in
clinical trials. Relevance to management of individual
patients unclear.