Kidney Week

Abstract: SA-PO033

The Combination of Area Under Curve of Estimated Glomerular Filtration Rate for 2 Years and Annual Rate Change of Estimated Glomerular Filtration Rate Predicts Long-Term Outcome in Kidney Transplants

Session Information

• Transplantation: Clinical Outcomes
  October 27, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Transplantation

• 1802 Transplantation: Clinical

Authors

• Lee, Eun jeong, Samsung Medical Center, Seoul, Korea (the Republic of)

Eun jeong Lee,

• Baeg, Song in, Samsung Medical Center, Seoul, Korea (the Republic of)

Song in Baeg,

• Boo, Hyo jin, Samsung Medical Center, Seoul, Korea (the Republic of)

Hyo jin Boo,

• Kim, Minjung, Samsung Medical Center, Seoul, Korea (the Republic of)

Minjung Kim,

• Jang, Hye Ryoun, Samsung Medical Center, Seoul, Korea (the Republic of)

Hye Ryoun Jang,

• Lee, Jung eun, Samsung Medical Center, Seoul, Korea (the Republic of)

Jung eun Lee,

• Kim, Dae Joong, Samsung Medical Center, Seoul, Korea (the Republic of)

Dae Joong Kim,

• Kim, Yoon-Goo, Samsung Medical Center, Seoul, Korea (the Republic of)

Yoon-Goo Kim,

• Oh, Ha Young, Samsung Medical Center, Seoul, Korea (the Republic of)

Ha Young Oh,

• Huh, Wooseong, Samsung Medical Center, Seoul, Korea (the Republic of)

Wooseong Huh,

Background

Improvement of short-term outcomes in kidney transplant (KT) has required clinical trials to evaluate long-term hard outcomes for validation of new therapies. However, because of time and cost, it is very difficult to conduct a clinical trial using hard outcomes in KT. To solve the problem, the use of surrogate marker should be considered. We examined the possibility of the combination of area under curve of estimated glomerular filtration rate for 2 years (AUCeGFR_2yrs) and % change in estimated glomerular filtration rate (eGFR) between years 1 and 2 after KT as a surrogate marker for long-term graft failure.

Methods

We studied 1423 kidney transplants performed from 1996 to 2013 at Samsung Medical Center, Seoul, Korea, including 202 graft losses (time to graft failure, median 8.4 years [5.4 - 12.4]) and 54 deaths (time to death, median 9.1 years [5.7 - 13.2]). Combination of AUCeGFR_2yrs (>1300 ml/min^.month vs < 1300 ml/min^.month) and % change in eGFR (> 2% vs < 2%) was assessed to determine risk of graft failure using Cox proportional hazard analysis.

Results

The combination was significantly associated with graft failure (p < 0.0001). Patients with AUCeGFR_2yrs < 1300 ml/min^.month and % change in eGFR < 2% formed 16.7% of all patients and showed higher graft failure risk (hazard ratio [HR], 3.36; 95% confidence interval [95% CI], 2.52 to 4.48). The Harrell C-index of the combination was 0.65 (95% CI, 0.60 to 0.69), and was internally validated via 5-fold cross-validation (average Harrell C-index, 0.64; 95% CI, 0.60 to 0.68). We also evaluated a known surrogate marker, > 30% decline in eGFR between years 1 and 3 after KT. The incidence of > 30% decline in eGFR was 6.9% of patients. HR of graft failure was 7.18 (95% CI, 5.22 to 9.89) and Harrell C-index was 0.65 (95% CI, 0.63 to 0.66).

Conclusion

We conclude the proposed combination might be useful as a surrogate outcome in KT trials in that it requires shorter surveillance period (2 years) than the known surrogate marker (3 years) while having comparable predictability. External validations should be conducted.