Review
Copyright ©The Author(s) 2020.
World J Transplant. Sep 18, 2020; 10(9): 230-255
Published online Sep 18, 2020. doi: 10.5500/wjt.v10.i9.230
Table 1 Summary of novel biomarker studies of chemokines associated with immunologic outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Rabant et al[11], 2015244UrineuCXCL9, uCXCL10RejectionCXCL9/10a correlated with ti+mvi (i+t; g + ptc) CXCL10: Cra diagnosed TCMR and ABMR (AUC > 0.75); CXCL10: Cr + DSAa improved the diagnosis of ABMR (AUC = 0.83)
Hricik et al[12], 201521UrineCXCL9RejectionuCXCL9 predicts AR by a median of 15 d before clinical detection
Faddoul et al[13], 2018184Urine and plasmaIFN-γ ELISpot; CXCL9ACRCXCL9 predictive of ACR; IFN-γ predictive of 1 year ↓eGFR; neither predicted 5-yr outcomes
Xu et al[14], 201887PlasmaCirculating fractalkine, IFN-γ and IP-10ARFractalkine on day 0, IP-10 at +7 and IFN-γ on +7 had the highest AUC (0.866) for predicting AR in 1 mo (sensitivity 86.8%; specificity 89.8%)
Tefik et al[15], 201965 (9 rejection, 56 stable)PlasmaIL-2, IL-8RejectionIL-2b and IL-8c predict AR; IL-2b and IL-8d levels correlated with ↓ 3 mo eGFR in the AR group
de Holanda et al[16], 201873PlasmasCD30Rejection; Graft survivalPlasma CD30 at +7, +14 associated w AR (P = 0.036). No difference in 5 yr graft survival
Table 2 Summary of micro-ribonucleic acid-related novel biomarker studies associated with immunologic outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Matz et al[17], 2016160PlasmamiR-15B, miR-103A, miR-106ATCMRmiR-15Ba,b, miR-103Aa,b and miR-106Aa,b discriminated patients with stable graft function from patients with TCMR and UTI
Matz et al[18], 2018111PlasmamiR-223-3p; miR-424-3p; miR-145-5p; miR-15b-5pABMR, TCMR, IFTAmiR-223-3p, miR-424-3p and miR-145-5p distinguished TCMR and ABMR from stable graft function; mir-145-5P decreased in IFTA (AUC 0.891) compared to stable graft function
Millán et al[19], 201780UrinemiR-142-3p, miR-210-3p and miR-155-5p, CXCL10Rejection↑miR-142-3p, ↑miR-155-5p, ↑CXCL10 + ↓miR-210-3p (AUC = 0.875) and CXCL10 (AUC = 0.865) discriminate rejectors and nonrejectors (sensitivity 85%, 84% and specificity 86% and 80% respectively)
Halloran et al[20], 2017519Allograft biopsyMolecular Microscope® Diagnostic System (MMDx™)/microRNATCMR, ABMRAgreement between MMDx™ and histology = 77% for TCMR, 77% for ABMR, and 76% for no rejection with blinding to histology, HLA. MMDx™c agreed with clinical judgment (87%) more than histology (80%)
Ledeganck et al[21], 201911 studiesAllograft biopsymicroRNATCMR, ABMR, cABMR↑miR-142, miR-155, miR-223 and ↓miR-125, miR-30, miR-204 predict TCMR, ABMR, cABMR
Lorenzen et al[23], 201593UrinelcRNA; RP11-354P17.15-001 (L328)TCMRRP11-354P17.15-001d (L328) was associated with acute TCMR (AUC = 0.76) sensitivity 49%, specificity 95%; L328 can detect subclinical TCMR
Table 3 Summary of leukocyte subclass related biomarkers associated with immunologic outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Luque et al[25], 2019175Plasmadonor reactive memory B cells (mBC)ABMRFor-cause bx: mBC in 100% ABMR/DSA+ and most cABMR, +/- DSA [24/30 (80%) and 21/29 (72.4%)]. Protocol bx: mBC > dnDSA was observed at 6 and 24 mo (8.8% vs 7.7% and 15.5% vs 11.1%) and identified pts with ongoing subABMR (AUC = 0.917, 0.809)
Gorbacheva et al[26], 2016PlasmamCD4RejectionMurine models with sensitized mCD4 T cells had SCr > 1 mg/dL (1.7 ± 0.6 mg/dL by 6–8 d post-transplant) and developed graft failure. At rejection, these recipients had DSA and ↑ frequencies of donor–reactive T cells producing IFN-γ compared with controls
Yazdani et al[27], 201995PlasmaNK gene expression model -> NK cellsRejectionNK cells predict ABMRavs no rejection (AUC = 0.98); ABMRbvs TCMR (AUC = 0.91) as well as histology: 22/24 biopsies with mvi (g + ptc) had ↑ NK levels (AUC = 0.89) Moreover, activated NK cells had the best predictive capability of graft failure at 1-2 yr (AUC = 0.74). NK cell infiltrationd predicted graft failure independent of histology
Cortes-Cerisuelo et al[28], 201723PlasmaCD28+CD4+RejectionCD28+CD4+ T cell frequency is associated with rejection on belatacept based IS
Table 4 Summary of gene expression related biomarkers associated with immunologic outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Friedewald et al[10], 2019308PlasmaBlood based biomarker/gene expression profileSubclinical acute rejectionGEP AR biomarker predicted sc-AR (sensitivity 64%, specificity 87%, PPV = 61%, NPV = 88%)
Zhang et al[29], 2019113PlasmaTRExRejection at 3 mo, Graft failureTREx predicts sc-AR at 3 mo in 113 KTRs (AUC = 0.830; NPV = 0.98, PPV = 0.79)
Crespo et al[30], 201775PlasmakSORT™ + ELISpotSubclinical rejectionkSORT™ + ELISpot predict sc-ARa, sc-TCMRa and sc-ABMRa (AUC > 0.85)
First et al[32], 2019192; 45PlasmaTruGraf® GEPSurveillance of patients with stable allograft functionIn 87.5% of the cases, investigators’ clinical decisions were influenced by TruGraf® results. In 45 patients TruGraf® supported 87% of clinical decisions with 93% of investigators stating they would use TruGraf® in subsequent patient care
Sigdel et al[33], 2019150 KTRs (43 stable, 45 AR, 19 borderline AR, 43 BKVN)UrineCommon rejection module (11 genes)Rejection10/11 genes were elevated in AR when compared to stable graft function. Psmb9 and CXCL10 could classify AR versus stable graft function as accurately as the 11-gene model (sensitivity = 93.6%, specificity = 97.6%); uCRM score differentiate AR from stable graft function (AUC = 0.9886)
Table 5 Summary of donor-derived cell-free deoxyribonucleic acid biomarkers associated with immunologic outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Oellerich et al[35], 2019189Plasmadd-cfDNARejectionIn pts with BPR, dd-cfDNA(cp/mL) was 3.3x and dd-cfDNA(%) 2.0x higher (82 cp/mL; 0.57%) than in stable pts w/o rejection (25 cp/mL; 0.29%). dd-cfDNA abs number > dd-cfDNA % (AUC = 0.73). OR = 7.31 for dd-cfDNA (cp/mL)
Stites et al[36], 202079 KTRs with TCMR 1A/borderline rejectionPlasmadd-cfDNAeGFR, dnDSA, Future rejection↑dd-cfDNA predict adverse outcomes: Among patients with ↑dd-cfDNAa, eGFR ↓ by 8.5% vs 0% in ↓dd-cfDNA pts. dnDSA seen in 40% (17/42) vs2.7%b and future or persistent rejection occurred in 9 of 42 ptsa (21.4% vs 0%)
Bloom et al[38], 2017102Plasmadd-cfDNARejectionDistinguished any rejection from non-rejection along with ABMR from non-ABMR
Huang et al[40], 201963Plasmadd-cfDNAABMRdd-cfDNA discriminated ABMRc [median 1.35%; interquartile range (IQR): 1.10%-1.90%] from no rejection (median 0.38%, IQR: 0.26%-1.10%). dd-cfDNA did not distinguish TCMR from no rejection
Whitlam et al[41], 201961Plasmadd-cfDNAaABMR cABMRgd-cfDNA and fraction were predictive of aAMR (AUC = 0.92, 0.85) and composite dx of ABMR (AUC = 0.91, 0.89). gd-cfDNA w/ modest sensitivity (0.90; 0.85) and specificity (0.88, 0.79) for aAMR and ABMR
Table 6 Summary of biomarkers associated with graft quality
Ref.nSampleBiomarkersOutcomeStudy conclusion
Parikh et al[46], 2016671PerfusateNGAL, L-FABP6 mo eGFREach doubling of perfusate NGAL and L-FABP were independently associated with ↓6-month eGFR (1.7mL/min per 1.73m2; 1.48mL/min per 1.73m2 )
Moser et al[47], 201741PerfusateMMP-2, LDH, NGALBiomarker levelsMMP-2a,b, LDHa,b, and NGALa,b were found in highest perfusate concentrations in DCD kidneys, followed by DBD and living donor allografts
Hamaoui et al[48], 201710PerfusatePerfusate lactatePerfusion10 DCD porcine kidneys perfused via HMP with modified AL solutionc had significantly ↓ perfusion lactate levels (3.1 vs 4.1 mmol/L) during reperfusion than those in UW solution
van Smaalen et al[49], 2017390PerfusateExtracellular histone concentration1 yr graft survival(extracellular histone) was associated w/ 1 year graft failure (HR = 1.386) 1 year graft survival was ↑ for the ↓ extracellular histone groupd (83% vs 71%) , maintained up to 5 yearse (76% vs 65%)
Weissenbacher et al[50], 201911PerfusateNGAL, KIM-1Kidney quality↑ perfusate NGAL level was found in the lowest quality kidney. In the perfused kidneys w/o urine recirculation, NGAL and KIM-1 ↓ over time. Small sample size; NGAL/ KIM-1 not predictive of kidney quality
Hosgood et al[53], 201756UrineNGAL, endothelin-1Kidney quality per EVKP score↑ levels of NGAL and ET-1 were associated with ↑ EVKP scoref (P < 0.05)
Table 7 Summary of biomarkers associated with delayed graft function
Ref.nSampleBiomarkersOutcomeStudy conclusion
Parikh et al[46], 2016671PerfusateNGAL, IL-18, L-FABPDGFBase (NGAL) was significantly ↑ in allografts with DGFa. This was also observed in post values of IL-18a and base/post perfusate L-FABP levelsb. These biomarkers did not significantly correlate with DGF development on multivariate adjustment
van Smaalen et al[49], 2017390PerfusateExtracellular histone concentrationDGFExtracellular histone concentration was significantly ↑ in the DGF group (median 0.70 µg/mL (IQR 0.4 to 0.98) compared to grafts that functioned immediatelyc (median, 0.42 (IQR 0.07 to 0.78). Interestingly there was no significant difference in extracellular histone concentration in grafts with primary non-function vs DGF
van Balkom et al[57], 201740PerfusateLeptin, GM-CSF, periostin, plasminogen activator inhibitor-1, osteopontinDGF5 perfusate proteins/158 tested predicted DGF. Leptin and GM-CSF -> most predictive. Validation with 40 kidneys found that leptin, GM-CSF + BMI predict DGF (AUC = 0.89 (95%CI: 0.74 to 1.00), which performed better than KDRI and DGF risk calculator (AUC 0.55, 0.59)
Roest et al[58], 201948PerfusatemicroRNA mir-505-3pDGFIn 8 DCD and DBD donors, ↑ levels of perfusate microRNA mir-505-3p correlated with DGFb (OR 1.12). This was confirmed via validation of 40 allografts, of which 20 developed DGFb. Interestingly, this predictive capability held true solely for DCD allograftsc
Truche et al[59], 201941Urine and PlasmauNGAL, uNAG, LDH, UCrDGFDGF -UNGAL, UNAG AUC 1, 0.96 (0.84-1.0) , urinary tubular injury biomarker-to-creatinine ratio, and LDH AUC = 1 and 0.92 (95%CI: 0.73 to 1.0)
Pianta et al[60], 201581UrineUrinary clusterin, IL-18, KIM-1, NGALDGFUrinary clusterin predicted DGF at 4 h (AUC = 0.72 (95%CI: 0.57 to 0.97), as did IL-18 , KIM-1 and NGAL; eGFR at 90 d was inversely correlated with urinary clusterin at 12 hb (Pearson r = −0.26, and 7 db (Pearson r = −0.25)
Reese et al[61], 20161304UrineMicroalbumin, NGAL, KIM-1, IL-18, L-FABPAKI, DGF, 6-month eGFRMicroalbumin, NGAL, KIM-1, IL-18, L-FABP from deceased donors at procurement; predictive of AKI; NGAL associated with DGF (RR = 1.21 (95%CI: 1.02 to 1.43), NGAL and L-FABP associated with lower 6 mo eGFR
Nielsen et al[62], 2019225Plasma and urinepNGAL, uNGAL uL-FABP, urine cystatin C, urine YLK-40DGF, 1 yr mGFR/eGFRpNGAL 1 d after tx -> associated with DGF. Did not correlate to 12-mo eGFR; no relation w L-FABP, cystatin C, and YLK-40
Koo et al[63], 201694UrineMicroalbumin, NGAL, KIM-1, IL-18, L-FABPDGF, 1 yr graft functionNGAL predicts AKI; NGAL + L-FABP predicts DGF (AUC 0.758, 0.704); NGAL + L-FABP + Cr better than DGF calculator and KDPI. L-FABP predictive of 1 yr graft functionb
Li et al[64], 20191036Urine and plasmauNGAL, pNGALDGFComposite AUC for 24 hours uNGAL was 0.91 (95%CI: 0.89 to 0.94) and the overall DOR for 24 hours uNGAL was 24.17; sensitivity 0.88, specificity 0.81. The composite AUC for 24 hours pNGAL was 0.95 (95%CI: 0.93 to 0.97) with an overall DOR for 24 hours pNGAL = 43.11 with sensitivity 0.91 and specificity 0.86
Bank et al[65], 201974 (DCD KTRs)UrineUrinary TIMP-2DGFTIMP-2/mOsm on day-1 and day-10 identified patients with DGF (AUC = 0.91) and prolonged DGF (AUC = 0.80); Consecutive TIMP-2/mOsm values showed a ↓ in TIMP-2/mOsm before an ↑estimated glomerular filtration rate, predicting resolution of fDGF
McGuinness et al[66], 201694hsa-miR-217; hsa-miR-125bDGFmiRNA + donor age + type donation predicted DGF in 83% of cases (61% sensitivity, 91% specificity)
Ledeganck et al[21], 201911 studiesAllograft biopsymicroRNADGFUpregulation of miR-21-3P and miR-182-5p associated with DGF
Table 8 Summary of biomarkers associated with cardiovascular events and cardiovascular mortality
Ref.nSampleBiomarkersOutcomeStudy conclusion
Foster et al[68], 2017508Urine and plasmaCystatin C, B2M, CrCV events, Mortality, Kidney failureHR eGFRcys and HR eGFRB2M < 30 vs 60+ were 2.02a (95%CI: 1.09 to 3.76) and 2.56b (95%CI: 1.35 to 4.88) for CV events; 3.92c (95%CI: 2.11 to 7.31) and 4.09b (95%CI: 2.21 to 7.54) for mortality; and 9.49c (95%CI: 4.28 to 21.00) and 15.53b (95%CI 6.99 to 34.51) for kidney failure
Bansal et al[69], 20161027UrineuNGAL, uKIM-1, IL-18, L-FABP, UCrCV events, Graft failure, mortalityEach ↑ log in uNGAL/Cr associated with a 24% ↑ risk of CV events (aHR = 1.24 (95%CI: 1.06 to 1.45), graft failure (1.40; 1.16 to 1.68), and risk of death (1.44; 1.26 to 1.65). uKIM-1/Cr and IL-18/Cr associated with higher risk of death (1.29; 1.03 to 1.61 and 1.25; 1.04 to 1.49 per log increase)
Park et al[70], 20171184 (300 CVD, 371 death, 513 random sub-cohort)Urineurine alpha 1 microglobulin [A1M], monocyte chemoattractant protein-1 [MCP-1], procollagen type I [PINP] and type III [PIIINP] N-terminal amino peptide)CV events, Death↑uA1M (HR per doubling of biomarker = 1.40 (95%CI: 1.21 to 1.62), MCP-1 [HR 1.18 (1.03 to 1.36)], and PINP [HR = 1.13 (1.03 to 1.23)]were associated with CVD events and death (HR per doubling α1m = 1.51 (95%CI: 1.32 to 1.72); MCP-1 = 1.31 (1.13 to 1.51); PINP = 1.11 (1.03 to 1.20)
Devine et al[71], 2020367PlasmaST2CV events, CV mortality, All-cause mortality↑ ST2 was associated with CV events (aHR = 1.31 (95% CI: 1.00 to 1.73); significantly for CV mortalityd (aHR = 1.61; (95%CI: 1.07 to 2.41; P = 0.022), The addition of ST2, to risk prediction models for CV mortality/events failed to improve their predictive accuracy
Yepes- Calderón et al[72], 2020604PlasmaMalondialdehydeCV mortalityDuring a follow-up period, 110 KTRs died, with 40% CV death. MDA was significantly associated with the risk for CV mortality. The association between MDA concentration and the risk for CV mortality was stronger in KTRs with ↓ eGFR [HR 2.09 (95%CI: 1.45-3.00) per 1-SD increment]
Table 9 Summary of biomarkers associated with infectious outcomes
Ref.nSampleBiomarkersOutcomeStudy conclusion
Fernández-Ruiz et al[73], 2017100PlasmasCD30Bacterial infectionsCD30 correlates to bacterial infection at baselinea and 1 moa, 3 moa, and 6 moa after KT. Patients with sCD30 ≥ 13.5 ng/mL had lower 12-mo bacterial infection-free survivalb (35.0% vs 80.0%) Baseline sCD30 levels ≥ 13.5 ng/mL is a risk factor for infectionc (HR: 4.65; 2.05-10.53)
Sadeghi et al[74], 201670PlasmaIL-23CMV infectionPatients with post-KT CMV disease (n = 13; 150 ± 106 d post-KT range 41–363 d) had higher pre-KT IL-23d (8.6 ± 4.4 vs 8.0 ± 17) and IL-23/Cr ratiosd than patients w/o CMV disease post-KT (n = 57). Pre-KT IL-23 plasma level of > 7 pg/mL is a risk factor for post-KT CMV infection/reactivation and symptomatic infectione (RR = 4.50, 95%CI: 1.23 to 16.52) ROC curve analysis post-KT CMV disease showed a sensitivity of 69% and a specificity of 67%
Leone et al[75], 2019189Plasma94ins/delE37delATTG NFKB1 polymorphismCMV infection65% of CMV infections occurred in ins/ins group. Survival free from CMV was 54.7% for ins/ins group and 79.4% for del carriers one-year post-KT. A multivariate regression for del carriers showed a ↓ risk of CMV infectionf and recurrence for ins/ins KTRsg (HR = 0.224, 0.307)
Kim et al[77], 2017385UrineUrine microRNA bkv-miR-B1-5p and bkv-miR-B1-3pBKVN↑ bkv-miR-B1-5p and bkv-miR-B1-3p in KTRs w biopsy proven BKVN distinguished them from disease free recipients (AUC = 0.989, 0.985). Only 13 KTRs with BKVN
Abend et al[78], 2017116PlasmaDonor BK virus antibody, recipient BK virus antibodyPost-transplant BK viremiaDonor BK virus antibody seropositivity correlated to post-transplant BK viremiah (OR = 5.0; 95%CI: 1.9-12.7). The authors did not examine for BKVN however
Ho et al[79], 2018107UrineCXCL10BKVN↑CXCL10 correlated with t+ii (uCXCL10/creatinine, 1.23 ng/mmol vs 0.46 ng/mmol; AUC = 0.69) and mvi, specifically ptci (uCXCL10/creatinine, 1.72 ng/mmol vs 0.46 ng/mmol; AUC = 0.69) compared to normal histology. Urinary CXCL10i corresponded with BKV, but not CMV viremia. These urine CXCL10 findings were confirmed in the independent validation set
Table 10 Summary of biomarkers associated with post-transplant malignancy
Ref.nSampleBiomarkersOutcomeStudy conclusion
Hope et al[83], 201582 (56 KTRs +malignancy, 26 KTRs - malignancy)PlasmaLDH; IFN-γ; ELISpotPost- transplant malignancyLow NK cell function -> HR 2.1 (0.97-5.00) metastatic Ca, Ca-related death, septic death
Pontrelli et al[84], 2019156: 93 KTRs, 34 controls + malignancy, 29 healthy subjectsPlasmaIL-27Post-transplant malignancyIL-27 plasma levels were able to discriminate patients with post-transplant neoplasia with a specificity of 80% and a sensitivity of 81%
Garnier et al[85], 201989PlasmaCD4+CD45RCPost-transplant malignancyKTRs with a low CD4+CD45RChigh population (< 51.9%) carried a 3.7 fold risk of cancera (HR = 3.71 (95%CI: 1.24 to 11.1), CD4+CD45high population having a 20-fold higher risk of rejectionb (HR = 21.7 (95%CI: 2.67-176.2)
Table 11 Summary of biomarkers associated with post-transplant diabetes mellitus
Ref.nSampleBiomarkersOutcomeStudy conclusion
Heldal et al[87], 2018852Plasma20 biomarkersPTDM6/20 biomarkers associated with PTDM; significant include soluble TNF type 1a Pentraxin 3a macrophage migration inhibitory factora and endothelial protein C receptorb
Yepes-Calderón et al[88], 2019516PlasmaMalondialdehydePTDMMDA was inversely associated with PTDM, independent of immunosuppressive therapy, transplant-specific covariates, lifestyle, inflammation, and metabolism parametersa (HR, 0.55; 95%CI, 0.36-0.83 per 1-SD increase)
Table 12 Summary of biomarkers associated with graft survival and/or patient survival
Ref.nSampleBiomarkersDysfunctionStudy conclusion
de Holanda et al[16], 201873PlasmasCD30Rejection; Graft survivalsCD30 at +7, +14 associated with ARa. No difference in 5 yr graft survival
Koo et al[63], 201694Urinemicroalbumin, NGAL, KIM-1, IL-18, L-FABPDGF, slow graft function , 1 yr graft functionNGAL predicts AKI; NGAL + L-FABP predicts DGF, slow graft function (AUC 0.758, 0.704); NGAL + L-FABP + Cr better than DGF calculator and KDPI. L-FABP predictive of 1 yr graft functionb
Foster et al[68], 2017508Urine and plasmaCystatin C, B2M, CrCV events, Mortality, Kidney failureHR eGFRcys and HR eGFRB2M < 30 vs 60+ were 2.02c (1.09-3.76) and 2.56d (1.35-4.88) for CV events; 3.92e (2.11-7.31) and 4.09d (2.21-7.54) for mortality; and 9.49e (4.28-21.00) and 15.53d (6.99-34.51) for kidney failure
Bansal et al[69], 20161027UrineuNGAL, KIM-1, IL-18, L-FABP, UcrCV events, Graft failure, mortalityEach ↑ log in uNGAL/Cr associated with a 24% ↑ risk of CV events (aHR 1.24; 1.06 to 1.45), graft failure (1.40; 1.16 to 1.68), and risk of death (1.44; 1.26 to 1.65). uKIM-1/Cr and IL-18/Cr associated with higher risk of death (1.29; 1.03 to 1.61 and 1.25;1.04 to 1.49 per log increase)
O’Connell et al[89], 2016204BiopsyGene set of 13 genesIFTA, Graft loss at 2/3 yrGene set prediction > clinicopathologic variables (AUC 0.967 > AUC 0.706, AUC 0.806) for IFTA; predicted graft loss at 2 and 3 years (AUC 0.842, 0.844), validated in 2 public datasets
Ix et al[90], 2017748UrineUrine A1M, MCP-1, procollagen type III and type I amino-terminal amino pro-peptideGraft failureIn adjusted models, ↑ concentrations of urine A1M (HR per doubling, 1.73; 1.43-2.08) and MCP-1 (HR per doubling, 1.60; 1.32-1.93) were associated with allograft failure. With the adjustment, urine A1M (HR per doubling, 1.76; 95%CI: 1.27-2.44)] and MCP-1 levels (HR per doubling, 1.49; 95%CI: 1.17-1.89) remained associated with allograft failure
Heylen et al[92], 2018154BiopsyDNA methylation1-yr graft function↑ methylation risk scoref at transplant predicted chronic injury at 1 yr (OR 45; 98 to 499; P < 0.001; AUC 0.919) vs standard baseline clinical risk factors, including age, donor criteria, donor last SCr, CIT, anastomosis time, HLA mismatches (combined AUC 0.743) sensitivity, specificity, and PPV, NPV values of MRS-based ROC curves were 90%, 90%, 95%, and 82%
Park et al[70], 20171184 (300 CVD, 371 death, 513 random sub-cohort)UrineUrine A1M MCP-1, PINP and PIIINPCV events, MortalityIn adjusted models, higher urine AlM (HR per doubling of biomarker = 1.40 (95%CI: 1.21 to 1.62), MCP-1 [HR = 1.18 (1.03 to 1.36)], and PINP [HR = 1.13 (95%CI: 1.03 to 1.23) were associated with CVD events. These three markers were also associated with death (HR per doubling A1M = 1.51 (95%CI: 1.32 to 1.72); MCP-1 = 1.31 (1.13 to 1.51); PINP = 1.11 (95%CI: 1.03 to 1.20)
Smedbråten et al[91], 2017382PlasmaCL-L1, CL-K1CV mortality, Graft survival, Patient survival↑CL-L1 (≥ 376 ng/mL) and ↑CL-K1 (≥ 304 ng/mL) levels at transplantation were associated with mortality in multivariate Cox analysesg [HR = 1.50 (95%CI: 1.09 to 2.07) and HR = 1.43 (95%CI: 1.02 to 1.99)] ↑CL-K1 levels were associated with CV mortality. No association between measured biomarkers and death-censored graft loss was found
San Segundo et al[93], 2019133PlasmaAbs number peripheral blood Treg cellsDeath-censored graft survival↑ Treg cells 1 yr post-KTh showed better DCGL (5-yr survival, 92.5% vs 81.4%). 1-yr Treg cellsh showed a ROC AUC of 63.1% (95%CI: 52.9 to 73.2) for predicting DCGL. After multivariate Cox regression analysis, an ↑ number of peripheral blood Treg cellsh was protective factor for DCGL (HR = 0.961 (95%CI: 0.924 to 0.998), irrespective of 1-yr proteinuria and renal function