A brief review on KDPI and its utility in assessement of kidneys

1. Kidney Donor Profile Index (KDPI)
Evolving idea or false solace
Goutham Kumar, MD
Consultant
Abdominal Transplant Surgery
Manipal Hospitals

2.  Introduction to the concept of KDPI
 Calculation of KDPI
 Parameters considered in KDPI and their significance
 Intended uses
 KDPI in Kidney Allocation
 Limitations of KDPI
Synopsis

3. Introduction to KDPI

4. Transplant decision

5. The quality of organ at present is objectively determined by
Expanded Criteria Donor (ECD)
Age >60
Age >50 with at least two of the following conditions:
I.Hypertension
II.Serum creatinine > 1.5 mg/dl
III.Cerebrovascular accident
Issues:
Donor age – main criteria in assessment of quality of an organ
Low predictive value because of binary (yes/no) characteristics
ECD stigma – High discard rate (44% in 2010) possibly due to ECD stigma
Why KDPI over ECD?

6. Why KDPI?
KDPI illuminates the fact that not all ECDs are alike
-Some ECD kidneys have reasonably good estimated quality
-Some SCD kidneys actually have lower estimated quality than some ECDs

7. That’s Why KDPI
KDPI
Proposed in order to improve risk stratification of kidneys so that clinicians
can more easily determine if a kidney offer is right for his or her candidate.
The “kidney donor profile index” (KDPI) is a score that informs clinicians about
the potential longevity of a kidney relative to other kidneys.
The calculation of KDPI includes 10 donor factors and is a more precise
measure of donor quality as compared to ECD/SCD dichotomy (only 4 donor
characteristics)
continuous PERCENTILE “score” instead of a binary (yes/no) indicator

8. Calculation of KDPI

9.  Calculation of KDRI (Kidney Donor Risk Index)
 Mapping of KDRI to KDPI - Relative risk scale to cumulative percentile scale
Steps involved in the calculation of KDPI

10. • University of Michigan - Ann Arbor
• Multivariable Cox proportional hazards regression model - to estimate the
association between the donor/recipient/transplant factors and graft survival
( Stratified by recipient age, diabetes status and transplant center)
• 69440 - first time, kidney-only, adult transplant
• 10 donor factors and 4 transplant factors were identified with significant
hazard ratio - to model KDRI
• In calculation of KDPI, Donor only KDRI (KDRI-RAO) is used as transplant
factors/recipient characteristics are different for each case.

14. 170cm
80kg

15. Correlation between KDRI and KDPI

16. Summary of KDPI calculation

17. Donor factor
Reference donor Donor 1 Donor 2 Donor 3 Donor 4 Donor 5 Donor 6 Donor 7 Donor 8 Donor 9 Donor 10
Age (yr) 40 30 40 40 40 40 40 40 40 40 40
Race Non-Black Non-Black Black Non-Black Non-Black Non-Black Non-Black Non-Black Non-Black Non-Black Non-Black
Hypertensive No No No No No Yes No No No No No
Diabetic No No No No No No Yes No No No No
Serum creatinine 1 1 1 1 1 1 1 1.5 1 1 1
Cause of death Non-CVA Non-CVA Non-CVA Non-CVA Non-CVA Non-CVA Non-CVA Non-CVA CVA Non-CVA Non-CVA
Height(cm) 170 170 170 160 170 170 170 170 170 170 170
Weight(kg) 80 80 80 80 70 80 80 80 80 80 80
HCV status Negative Negative Negative Negative Negative Negative Negative Negative Negative positive Negative
DCD
No No No No No No No No No No Yes
KDRI-median 0.82 0.72 0.98 0.86 0.85 0.93 0.93 0.91 0.89 1.04 0.93
KDPI 31 18 48 35 34 43 43 41 39 54 44
How do the Individual donor characteristic affects KDPI?

18. Rules of thumb
 Ideal donor according to KDPI – Age 19, 80kg, 6’3’’, non-CVA, DBD with no
comorbidities
 Age increase of 1yr increases KDPI by 1%, every 10yr increase by 12%
 Black donor - Higher KDPI by 17%, literally means a 23 yo black donor would be
similar to 40 yo non-black donor
 Height – Every 2cm increase in height reduces 1% in KDPI, and vice versa.
 Weight – Every 10kg decrease in weight from reference donor weight increases 4%
in KDPI. No change in increase in weight
 Hypertension & Diabetes – 12% higher KDPI each
 Serum creatinine – 10% Higher KDPI if serum creatinine is increased from 1 to
1.5, then 1% for every 1mg/dL increase in Cr
 HCV positivity increases KDPI by 23%
 DCD donor increases KDPI by 13%

19. Can we use KDRI/KDPI for other organs?
 For Liver and pancreas - Separate organ risk indices LDRI and PDRI respectively are developed,
But does KDRI correlate well with LDRI/PDRI with similar discriminatory power.
 On the other hand, KDRI was also shown to have very modest discriminatory power (c=0.54) for
heart transplant outcomes and very little association with lung transplant outcomes (c=0.52)
Stewart DE et al. Is the Kidney Donor Risk Index (KDRI) a Useful Predictor of Graft
Survival for Non-Renal Organs? ATC Abstract #464.

20. Intended uses of KDPI

21. KDPI as an analytical tool on DonorNet® - some
intended uses
1. Evaluation of deceased donor kidney
2. Differentiate - Lower versus higher quality ECD kidneys
3. Determination of need to implant both the kidneys or not
4. May help OPO predict which donors/organs may be more
less difficult to place.
5. In future, a part of allocation algorithm

22. Evaluation of deceased donor kidney

23. Evaluation of deceased donor kidney

24. Correlation between KDPI and Graft survival

25. Differentiate - Lower versus higher quality ECD kidneys

27. KDPI in organ allocation

28. Problems with Current allocation system
 High discard rates
 Access variability due to geography and biology
 Longevity Mismatch in graft/patient survival
 waiting time has become the primary driver of kidney allocation
 Histocompatibility components have diminished
 This overreliance led to a system that does not accomplish any goal other than
transplanting the candidate waiting the longest
 Fails to acknowledge different needs for different candidates (e.g., speed over
quality)

29. Increasing need and high discard rates for cadaveric kidneys

30.  Optimal utilization of procured cadaveric kidneys – Reduce discard rates
 Promote graft survival for those at highest risk of re-transplant.
 Minimize loss of potential graft function through better longevity matching
 Improve efficiency and utilization by providing better information about kidney
offers
 Provide comprehensive data to guide transplant decision making
 Reduce differences in access variability due to geography and biology
Ideal allocation system

31.  Candidate age,
 Length of time on dialysis,
 Any prior organ transplant, and
 Diabetes status.
Estimated Post-Transplant Survival (EPTS)

32. Proposed Allocation system - Model

35. New system forecasted to result in: 8,380 additional life years gained annually
Expected Outcomes

36. New proposed system – Higher transplant rates in younger population
Expected Outcomes

37. Limitations of KDPI

38. Limitations - KDPI
1. Is statistics and data analysis for KDPI sound enough?
The use of risk factors as prognostic tool for the purpose of prospective
individual risk stratification often yields disappointing results.
Ware JH. The limitations of risk factors as prognostic tools. N Engl J Med 2006;355: 2615-7.

39.  The predictive power of the KDPI is only moderate (c=0.62), but when
highest and lowest quintiles were compared then c-statistic improved
to 0.78. Better in extremes and ‘muddy’ in middle
 Though there is a definite difference between the kidneys with KDPI of
10 and 90, there may not be any difference in KDPI of 40 and 60.
Different KDPI means different interpretation

40. Biopsy and trauma not included in KDPI
KDPI does not include all donor factors potentially associated with kidney
graft outcomes.
Biopsy
No information about damage, trauma, or abnormalities in kidneys
KDPI was developed using graft outcomes from strictly adult transplant recipients; It can
only be used in pediatric donors but not recipients.
.

41. May reduce live-donor kidney transplants?
 Potential reduction in live donor kidney transplants in younger patients (53%
in age 18-34, 41% in 35-49)
 Overdependence on less available live donors for older patients (33% in 50-
64, 28% in 65 & older
.
Hippen BE, Thistlethwaite JR Jr, Ross LF. Risk, prognosis, and unintended consequences in
kidney allocation N Engl J Med. 2011

42. KDPI based allocation system – disadvantageous for
certain groups
 Patients with short term diabetes with similar outcomes as non-diabetics
 Elderly recipients

43. Conclusions in this topic are still a
matter of debate.

44. Thank you