Title: Each additional hour of cold ischemia time significantly increases the risk of graft failure and mortality following renal transplantation
Abstract: Although cold ischemia time has been widely studied in renal transplantation area, there is no consensus on its precise relationship with the transplantation outcomes. To study this, we sampled data from 3839 adult recipients of a first heart-beating deceased donor kidney transplanted between 2000 and 2011 within the French observational multicentric prospective DIVAT cohort. A Cox model was used to assess the relationship between cold ischemia time and death-censored graft survival or patient survival by using piecewise log-linear function. There was a significant proportional increase in the risk of graft failure for each additional hour of cold ischemia time (hazard ratio, 1.013). As an example, a patient who received a kidney with a cold ischemia time of 30h presented a risk of graft failure near 40% higher than a patient with a cold ischemia time of 6h. Moreover, we found that the risk of death also proportionally increased for each additional hour of cold ischemia time (hazard ratio, 1.018). Thus, every additional hour of cold ischemia time must be taken into account in order to increase graft and patient survival. These findings are of practical clinical interest, as cold ischemia time is among one of the main modifiable pre-transplantation risk factors that can be minimized by improved management of the peri-transplantation period. Although cold ischemia time has been widely studied in renal transplantation area, there is no consensus on its precise relationship with the transplantation outcomes. To study this, we sampled data from 3839 adult recipients of a first heart-beating deceased donor kidney transplanted between 2000 and 2011 within the French observational multicentric prospective DIVAT cohort. A Cox model was used to assess the relationship between cold ischemia time and death-censored graft survival or patient survival by using piecewise log-linear function. There was a significant proportional increase in the risk of graft failure for each additional hour of cold ischemia time (hazard ratio, 1.013). As an example, a patient who received a kidney with a cold ischemia time of 30h presented a risk of graft failure near 40% higher than a patient with a cold ischemia time of 6h. Moreover, we found that the risk of death also proportionally increased for each additional hour of cold ischemia time (hazard ratio, 1.018). Thus, every additional hour of cold ischemia time must be taken into account in order to increase graft and patient survival. These findings are of practical clinical interest, as cold ischemia time is among one of the main modifiable pre-transplantation risk factors that can be minimized by improved management of the peri-transplantation period. During the past 10 years in renal transplantation, the widespread use of mycophenolic acid, tacrolimus, and novel induction therapies has played a major role in decreasing the incidence of acute rejection episodes.1.Gonwa T. Johnson C. Ahsan N. et al.Randomized trial of tacrolimus+mycophenolate mofetil or azathioprine versus cyclosporine+mycophenolate mofetil after cadaveric kidney transplantation: results at three years. [Miscellaneous Article].Transplant. 2003; 75: 2048-2053Crossref PubMed Scopus (132) Google Scholar,2.Knight S.R. Russell N.K. Barcena L. et al.Mycophenolate mofetil decreases acute rejection and may improve graft survival in renal transplant recipients when compared with azathioprine: a systematic review.Transplantation. 2009; 87: 785-794Crossref PubMed Scopus (135) Google Scholar,3.Ekberg H. Bernasconi C. Tedesco-Silva H. et al.Calcineurin inhibitor minimization in the Symphony study: observational results 3 years after transplantation.Am J Transplant. 2009; 9: 1876-1885Crossref PubMed Scopus (261) Google Scholar However, mid-term kidney graft and patient survival have not improved as much as expected. This could be because of the increased age of both recipients and donors and is associated with the higher frequency of expanded criteria donors (ECDs).4.Metzger R.A. Delmonico F.L. Feng S. et al.Expanded criteria donors for kidney transplantation.Am J Transplant. 2003; 3: 114-125Crossref PubMed Scopus (516) Google Scholar Another endeavor should be achieved to reduce delayed graft function (DGF) risk.5.Yarlagadda S.G. Klein C.L. Jani A. Long-term renal outcomes after delayed graft function.Adv Chronic Kidney Dis. 2008; 15: 248-256Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar DGF is well known to influence mid-term graft outcomes, and it increases hospitalization duration and the frequency of concomitant acute rejection.6.Jayaram D. Kommareddi M. Sung R.S. et al.Delayed graft function requiring more than one-time dialysis treatment is associated with inferior clinical outcomes.Clin Transplant. 2012; 26: E536-E543Crossref PubMed Scopus (26) Google Scholar,7.Giral-Classe M. Hourmant M. Cantarovich D. et al.Delayed graft function of more than six days strongly decreases long-term survival of transplanted kidneys.Kidney Int. 1998; 54: 972-978Abstract Full Text Full Text PDF PubMed Scopus (205) Google Scholar The incidence and severity of DGF have remained stable but varies from 25 to 50% among deceased donor kidneys.8.Tapiawala S.N. Tinckam K.J. Cardella C.J. et al.Delayed graft function and the risk for death with a functioning graft.J Am Soc Nephrol. 2010; 21: 153-161Crossref PubMed Scopus (143) Google Scholar,9.Yarlagadda S.G. Coca S.G. Formica R.N. et al.Association between delayed graft function and allograft and patient survival: a systematic review and meta-analysis.Nephrol Dial Transplant. 2009; 24: 1039-1047Crossref PubMed Scopus (538) Google Scholar DGF is the consequence of well-described risk factors,10.Perico N. Cattaneo D. Sayegh M.H. et al.Delayed graft function in kidney transplantation.Lancet. 2004; 364: 1814-1827Abstract Full Text Full Text PDF PubMed Scopus (759) Google Scholar,11.Irish W.D. Ilsley J.N. Schnitzler M.A. et al.A risk prediction model for delayed graft function in the current era of deceased donor renal transplantation.Am J Transplant. 2010; 10: 2279-2286Crossref PubMed Scopus (275) Google Scholar among which cold ischemia time (CIT) seems to be one of the main explicative variables.12.Opelz G. Döhler B. Multicenter analysis of kidney preservation.Transplantation. 2007; 83: 247-253Crossref PubMed Scopus (166) Google Scholar,13.Chapal M. Le Borgne F. Legendre C. et al.The DGFS: a useful scoring system for the prediction and management of delayed graft function following kidney transplantation from heart beating deceased donors.Kidney Int. 2014; (e-pub ahead of print, 4 June 2014; doi)https://doi.org/10.1038/ki.2014.188Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar CIT acts at least in part through pathophysiological pathways that induce ischemia reperfusion injuries.10.Perico N. Cattaneo D. Sayegh M.H. et al.Delayed graft function in kidney transplantation.Lancet. 2004; 364: 1814-1827Abstract Full Text Full Text PDF PubMed Scopus (759) Google Scholar,14.Mikhalski D. Wissing K.M. Ghisdal L. et al.Cold ischemia is a major determinant of acute rejection and renal graft survival in the modern era of immunosuppression.Transplantation. 2008; 85: S3-S9Crossref PubMed Scopus (143) Google Scholar To improve mid-term outcomes, it could be preferable to optimize the transplantation organization with the aim of shortening CIT as most as possible and preventing ischemia injury through other strategies such as machine perfusion, before treating lesions already established in the graft. Even though CIT is a well-known risk factor among the renal transplantation community, its precise etiological role on mid-term graft outcomes is still under debate as illustrated by the wide heterogeneity of results observed in the literature. On one hand, some authors have shown that CIT was not significantly associated with graft survival among transplanted patients.15.Kayler L.K. Srinivas T.R. Schold J.D. Influence of CIT-induced DGF on kidney transplant outcomes.Am J Transplant. 2011; 11: 2657-2664Crossref PubMed Scopus (111) Google Scholar, 16.Kayler L.K. Magliocca J. Zendejas I. et al.Impact of cold ischemia time on graft survival among ECD transplant recipients: a paired kidney analysis.Am J Transplant. 2011; 11: 2647-2656Crossref PubMed Scopus (117) Google Scholar, 17.Johnson R.J. Fuggle S.V. O’Neill J. et al.Factors influencing outcome after deceased heart beating donor kidney transplantation in the United Kingdom: an evidence base for a new national kidney allocation policy.Transplantation. 2010; 89: 379-386Crossref PubMed Scopus (51) Google Scholar, 18.Summers D.M. Johnson R.J. Hudson A. et al.Effect of donor age and cold storage time on outcome in recipients of kidneys donated after circulatory death in the UK: a cohort study.Lancet. 2013; 381: 727-734Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar On the other hand, numerous other studies established that CIT represents a major risk factor of graft survival.12.Opelz G. Döhler B. Multicenter analysis of kidney preservation.Transplantation. 2007; 83: 247-253Crossref PubMed Scopus (166) Google Scholar,14.Mikhalski D. Wissing K.M. Ghisdal L. et al.Cold ischemia is a major determinant of acute rejection and renal graft survival in the modern era of immunosuppression.Transplantation. 2008; 85: S3-S9Crossref PubMed Scopus (143) Google Scholar,19.Ojo A.O. Wolfe R.A. Held P.J. et al.Delayed graft function: risk factors and implications for renal allograft survival.Transplant. 1997; 63: 968-974Crossref PubMed Scopus (839) Google Scholar, 20.Salahudeen A.K. Haider N. May W. Cold ischemia and the reduced long-term survival of cadaveric renal allografts.Kidney Int. 2004; 65: 713-718Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar, 21.Quiroga I. McShane P. Koo D.D.H. et al.Major effects of delayed graft function and cold ischaemia time on renal allograft survival.Nephrol Dial Transplant. 2006; 21: 1689-1696Crossref PubMed Scopus (265) Google Scholar, 22.Morris P.J. Johnson R.J. Fuggle S.V. et al.Analysis of factors that affect outcome of primary cadaveric renal transplantation in the UK.Lancet. 1999; 354: 1147-1152Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar, 23.Van der Vliet J.A. Warlé M.C. Cheung C.L.S. et al.Influence of prolonged cold ischemia in renal transplantation.Clin Transplant. 2011; 25: E612-E616Crossref PubMed Scopus (52) Google Scholar, 24.Hernández D. Estupiñán S. Pérez G. et al.Impact of cold ischemia time on renal allograft outcome using kidneys from young donors.Transpl Int. 2008; 21: 955-962Crossref PubMed Scopus (26) Google Scholar Nevertheless, there is no consensus whether CIT should be considered as a continuous risk factor or whether threshold values can be considered to identify subgroups with a relevant excess in the risk of mid-term graft and patient outcomes.12.Opelz G. Döhler B. Multicenter analysis of kidney preservation.Transplantation. 2007; 83: 247-253Crossref PubMed Scopus (166) Google Scholar,16.Kayler L.K. Magliocca J. Zendejas I. et al.Impact of cold ischemia time on graft survival among ECD transplant recipients: a paired kidney analysis.Am J Transplant. 2011; 11: 2647-2656Crossref PubMed Scopus (117) Google Scholar,18.Summers D.M. Johnson R.J. Hudson A. et al.Effect of donor age and cold storage time on outcome in recipients of kidneys donated after circulatory death in the UK: a cohort study.Lancet. 2013; 381: 727-734Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar,20.Salahudeen A.K. Haider N. May W. Cold ischemia and the reduced long-term survival of cadaveric renal allografts.Kidney Int. 2004; 65: 713-718Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar For instance, Salahudeen et al.20.Salahudeen A.K. Haider N. May W. Cold ischemia and the reduced long-term survival of cadaveric renal allografts.Kidney Int. 2004; 65: 713-718Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar demonstrated a significantly worse graft survival for patients with a CIT higher than 30h, whereas Opelz et al.12.Opelz G. Döhler B. Multicenter analysis of kidney preservation.Transplantation. 2007; 83: 247-253Crossref PubMed Scopus (166) Google Scholar described that increasing CIT up to 18h was not associated with an increased risk of graft failure. In addition to the heterogeneity of cutoff values used to define high-risk patients, the definition of these values is often arbitrary. Besides, the majority of the previous studies analyzed graft and patient survival and death-censored graft survival; only Johnson et al.17.Johnson R.J. Fuggle S.V. O’Neill J. et al.Factors influencing outcome after deceased heart beating donor kidney transplantation in the United Kingdom: an evidence base for a new national kidney allocation policy.Transplantation. 2010; 89: 379-386Crossref PubMed Scopus (51) Google Scholar studied the association of the CIT with patient survival, and they showed a nonsignificant association of continuous CIT on patient death. Therefore, taking the opportunity of a large prospective, multicentric, and validated cohort, the aim of this study was to revisit the potential relationship between the CIT and either the graft failure (death-censored) or the patient death using an etiological approach. The mean CIT was 20.6h (range from 6 to 58.6h; s.d.=7.8). The CIT duration was between 6 and 16h for 1274 (33.2%) patients, 16 and 24h for 1531 (39.9%), 24 and 36h for 853 (22.2%), and longer than 36h for 181 patients (4.7%). Figure 1 displays boxplots of CIT for each year of transplantation. Over the past decade, we observed a global decrease in CIT duration (median from 23.0h in 2000 to 16.3h in 2011). However, this progress was more important for prolonged CIT (third quartile from 32.7h in 2000 to 21.8h in 2011) than for short CIT (first quartile from 17.6h in 2000 to 13.1h in 2011). Demographic and baseline characteristics at the time of transplantation according to the CIT are described in Table 1. Patients with a history of cardiovascular diseases (P=0.022), hypertension (P=0.001), those undergoing hemodialysis (P=0.047), those with anticlass II panel reactive antibody (PRA) (P=0.006), those who received kidneys from old donor (P<0.001), or from cerebrovascular deceased donor (P=0.025), and a depleting induction therapy (P=0.003) displayed the longest CIT. As expected, we observed an increased risk of DGF with CIT (P<0.0001): from 22% for CIT between 6 and 16h, to 40% for CIT above 24h. Finally, ECD was distributed differently in the four CIT-based groups (P<0.001). Only 48 patients (1.3%) received kidneys placed under hypothermic machine perfusion, leading to an unbalanced parameter, which was not taken into account in the multivariate modeling.Table 1Description of recipient, donor, and transplantation characteristics of the global study population and according to CIT-based groups (6–16, 16–24, 24–36, and>36h)Missing dataGlobal (N=3839)CIT from 6 to 16h (N=1274)CIT from 16 to 24h (N=1531)CIT from 24 to 36h (N=853)CIT above 36h (N=181)P-valueQuantitative characteristics: mean±s.d. Recipient age (years)051.6±13.251.0±13.452.3±13.351.3±12.751.4±13.40.051 Recipient BMI (kg/m2)3824.4±4.324.3±4.324.4±4.224.5±4.524.3±4.40.894 Donor age (years)050.5±16.248.9±16.452.1±16.150.0±15.550.3±16.6<0.001 Donor serum creatinine (mg/ml)093.9±56.190.7±47.996.0±65.495.6±51.391.6±43.70.061 HLA incompatibilities ABDR873.3±1.33.4±1.23.2±1.33.3±1.43.3±1.3<0.001Categorical characteristics: N (%) Recipient men02365 (61.6)794 (62.3)958 (62.6)504 (59.1)109 (60.2)0.345 Dialysis technique0.047 Pre-emptive transplantation0304 (7.9)120 (9.4)123 (8.0)54 (6.3)7 (3.9) Hemodialysis03251 (84.7)1053 (82.7)1300 (84.9)735 (86.2)163 (90.0) Peritoneal dialysis0284 (7.4)101 (7.9)108 (7.1)64 (7.5)11 (6.1) Detectable anticlass I PRA0750 (19.5)228 (17.9)313 (20.4)165 (19.3)44 (24.3)0.131 Detectable anticlass II PRA0608 (15.8)172 (13.5)243 (15.9)155 (18.2)38 (21.0)0.006 History of cardiovascular diseases01394 (36.3)427 (33.5)557 (36.4)335 (39.3)75 (41.4)0.022 History of hypertension03036 (79.1)981 (77.0)1197 (78.2)700 (82.1)158 (87.3)0.001 History of dyslipidemia01138 (29.6)380 (29.8)450 (29.4)250 (29.3)58 (32.0)0.894 History of diabetes0466 (12.1)144 (11.3)202 (13.2)97 (11.4)23 (12.7)0.398 Donor men02277 (59.3)772 (60.6)894 (58.4)502 (58.9)109 (60.2)0.672 Expanded criteria donor01295 (33.7)379 (29.8)589 (38.5)264 (31.0)63 (34.8)<0.001 Cerebrovascular donor death02169 (56.5)686 (53.9)904 (59.1)470 (55.1)109 (60.2)0.025 Depleting induction171521 (39.6)500 (39.3)567 (37.0)366 (42.9)88 (48.6)0.003 Delayed graft function1101204 (31.4)280 (22.0)474 (31.0)342 (40.1)108 (59.7)<0.001 Machine perfusion048 (1.3)24 (1.9)14 (0.9)10 (1.2)0 (0)0.0579Abbreviations: BMI, body mass index; CIT, cold ischemia time; HLA, human leukocyte antigen; PRA, panel reactive antibody.Quantitative characteristics expressed as the mean and s.d.; categorical characteristics expressed as number (%). Open table in a new tab Abbreviations: BMI, body mass index; CIT, cold ischemia time; HLA, human leukocyte antigen; PRA, panel reactive antibody. Quantitative characteristics expressed as the mean and s.d.; categorical characteristics expressed as number (%). Among the 3839 patients, 449 lost their graft and 238 died with a functioning graft. The cumulative follow-up covered 15,978 patient–years. Graft failure and patient survival curves and their corresponding 95% confidence intervals (CIs) are presented in Figure 2. The graft survivals at 1, 5, and 10 years post transplantation were, respectively, 95%, 88%, and 77%. The patient survivals at 1, 5, and 10 years post transplantation were, respectively, 98%, 93%, and 87%. In addition, the graft failure and death risks appeared higher in the first year post transplantation than afterwards. Indeed, the incidence rates were equal to 54.1 and 20.8 in the first year, respectively, for graft failures and death per 1000 patient–year, whereas the mean incidence rates were 21.3 and 13.9 after the first year post transplantation. For patients with less than 16-h CIT, the corresponding absolute risks were 4, 10, and 20% for graft failure and 1, 6, and 11% for death. Patients with CIT between 16 and 24h had absolute risk close to those with CIT between 24 and 36h (5, 13, and 24% for graft failure and 3, 7, and 13% for death), whereas the risk appears slightly higher for patients with more than 36h CIT (8, 13, and 36% for graft failure and 5, 15, and 22% for death). Whereas we considered a large number of nonlinear associations, we finally retained a proportional relationship between CIT and the risk of graft failure. Results of the unadjusted analyses and the final multivariate model are presented in Table 2 and illustrated in Figure 3a. For each additional hour of ischemia time, the risk of graft failure was multiplied by 1.013 (P=0.035). For instance, patients with 12-h CIT had a risk of graft failure 8% (1.01312-6) higher than patients with 6-h CIT. This relationship is constant regardless of the baseline CIT level, that is, this excess of risk being similar between patients with 30h and those with 24h. This association was independent from other possible confounding factors.Table 2Cox unadjusted analyses and Cox multivariate analysis of graft failure risk (with death-censored)Unadjusted Cox modelsMultivariate Cox frailty modelaRandom effect variance=0.025.uHR95% CIP-valueaHR95% CIP-valueCold ischemia time (h)1.0181.007–1.0300.0011.0131.001–1.0250.035Donor age<0.001<0.001 51–60 Years vs. =50 years1.3621.072–1.7300.0111.2280.943–1.5980.128 61 Years and more vs. =50 years2.0801.675–2.581<0.0012.0361.535–2.699<0.001Donor gender (men vs. women)0.8510.706–1.0260.0910.8430.692–1.0280.091Donor serum creatinine (=15 vs. <15mg/ml)1.2480.974–1.5990.0801.3911.069–1.8090.014Donor cause of death (vascular vs. others)1.2711.052–1.5370.0131.0480.852–1.2890.659Recipient age (=55 vs. <55 years)1.3421.114–1.6150.0020.8780.697–1.1190.280Recipient BMI (=30 vs. <30kg/m2)1.9841.544–2.549<0.0011.8221.408–2.358<0.001Dialysis technique0.2010.052 Peritoneal dialysis vs. pre-emptive transplantation1.2320.714–2.1260.4531.0280.585–1.8060.924 Hemodialysis vs. pre-emptive transplantation1.4050.930–2.1210.1061.3160.865–2.0020.199History of cardiovascular diseases (yes vs. no)1.4361.192–1.730<0.0011.2551.030–1.5290.024PRA anticlass I (detectable vs. undetectable)1.3981.113–1.7550.0041.4291.094–1.8670.009PRA anticlass II (detectable vs. undetectable)1.1150.865–1.4380.3990.9390.697–1.2640.677HLA incompatibility ABDR (>4 vs. =4)1.2760.995–1.6360.0551.3001.003–1.6840.047Induction therapy (depleting vs. non-depleting)1.1590.959–1.4000.1261.0770.875–1.3260.482Recipient gender (men vs. women)0.8650.716–1.0440.129———History of hypertension (yes vs. no)1.0510.829–1.3310.682———History of dyslipidemia (yes vs. no)1.2781.051–1.5530.014———History of diabetes (yes vs. no)1.4191.088–1.8500.010———Expanded criteria donor (yes vs. no)1.8601.542–2.244<0.001———Abbreviations: aHR, adjusted hazard ratio; BMI, body mass index; CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; uHR, unadjusted hazard ratio; PRA, panel reactive antibody.n=3701, 138 observations deleted due to missingness concerning covariates. Each additional hour of cold ischemia time increases significantly the risk of graft failure (HR=1.013, P=0.024), independently of possible confounding factors.a Random effect variance=0.025. Open table in a new tab Abbreviations: aHR, adjusted hazard ratio; BMI, body mass index; CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; uHR, unadjusted hazard ratio; PRA, panel reactive antibody. n=3701, 138 observations deleted due to missingness concerning covariates. Each additional hour of cold ischemia time increases significantly the risk of graft failure (HR=1.013, P=0.024), independently of possible confounding factors. In this model, we also retained a proportional association between CIT and risk of death. Table 3 presents the unadjusted analyses and the final multivariate model. Actually, the risk of death was multiplied by 1.018 (P=0.026) for each supplementary hour of CIT. Figure 3b illustrates this relationship. As an example, patients with 30-h CIT will have 53% (1.01830-6) more risk of death than patients with 6-h CIT. This relationship was constant regardless of the baseline CIT level. This association was independent from the other possible confounding factors.Table 3Cox unadjusted analyses and Cox multivariate analysis of patient death risk (with a functioning graft)Unadjusted Cox modelsMultivariate Cox frailty modelaRandom effect variance=0.002.uHR95% CIP-valueaHR95% CIP-valueCold ischemia time (h)1.0161.000–1.0310.0441.0181.002–1.0350.026Donor age<0.0010.006 51–60 Years vs. =50 years2.3451.664–3.304<0.0011.6631.145–2.4150.008 61 Years and more vs. =50 years3.7352.718–5.133<0.0012.0431.381–3.021<0.001Donor gender (men vs. women)0.8470.655–1.0940.2030.9340.713–1.2230.619Donor serum creatinine (=15 vs. <15mg/ml)0.7580.504–1.1390.1830.8540.555–1.3130.471Donor cause of death (vascular vs. other)1.4961.147–1.9520.0031.0090.758–1.3420.953Recipient age (=55 years vs. <55 years)3.2022.421–4.235<0.0011.7541.253–2.4570.001Dialysis technique0.0430.118 Peritoneal dialysis vs. pre-emptive transplantation1.1550.490–2.7190.7421.0240.434–2.4170.956 Hemodialysis vs. pre-emptive transplantation1.8250.968–3.4410.0631.5860.838–3.0030.157PRA anticlass I (detectable vs. undetectable)1.2980.939–1.7930.1141.5011.035–2.1790.032PRA anticlass II (detectable vs. undetectable)0.9760.675–1.4100.8960.7680.499–1.1820.230History of cardiovascular diseases (yes vs. no)2.7952.154–3.627<0.0012.0281.538–2.672<0.001History of diabetes (yes vs. no)2.8922.153–3.886<0.0011.9421.423–2.651<0.001HLA incompatibility ABDR (>4 vs. =4)1.1420.800–1.6300.4641.0810.751–1.5600.670Induction therapy (depleting vs. non-depleting)0.7690.584–1.0120.0610.7740.582–1.0290.077Recipient BMI (=30 vs. <30kg/m2)1.9901.410–2.809<0.001———Recipient gender (men vs. women)1.4001.025–1.7780.032———History of hypertension (yes vs. no)1.1620.830–1.6270.382———History of dyslipidemia (yes vs. no)1.4861.143–1.9320.003———Expanded criteria donor (yes vs. no)2.6442.047–3.415<0.001———Abbreviations: aHR, adjusted hazard ratio; BMI, body mass index; CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; uHR, unadjusted hazard ratio; PRA, panel reactive antibody.n=3738, 101 observations deleted due to missingness concerning covariates. Each additional hour of cold ischemia time increases significantly the risk of death (HR=1.019, P=0.023), independently of possible confounding factors.a Random effect variance=0.002. Open table in a new tab Abbreviations: aHR, adjusted hazard ratio; BMI, body mass index; CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; uHR, unadjusted hazard ratio; PRA, panel reactive antibody. n=3738, 101 observations deleted due to missingness concerning covariates. Each additional hour of cold ischemia time increases significantly the risk of death (HR=1.019, P=0.023), independently of possible confounding factors. For both survival analyses, no clinically relevant interaction with CIT appeared statistically significant. Specifically, we assumed that prolonged CIT could be more deleterious among patients receiving kidneys from ECD than from non-ECD. However, by testing the interaction between CIT and ECD, we could not demonstrate a significant difference in the relationships between CIT and both graft and patient survivals in patients who received kidney from Extended or Standard Criteria Donor (P=0.319 and P=0.408, respectively, for graft and patient survivals). From our cohort-based analysis, we describe the negative influence of CIT on mid-term outcomes. The originality of these findings is demonstrated in the proportional relationship between CIT and the risk of graft failure. Moreover, we also demonstrated, for the first time to our knowledge, the proportional relationship between CIT and the risk of patient death. More precisely, and for both outcomes, the results suggest that every additional hour of CIT matters. According to the description of CIT trend within the last 10 years (Figure 1), we observed an important decrease in prolonged CIT, whereas this decrease was not so important for short CIT. However, our study highlights that this effort to minimized CIT has to be considered either for short and long CIT. Nevertheless, such management strategies do not exclude also minimizing and/or treating lesions already established in the graft by using therapeutic strategies,25.Neumayer H.H. Kunzendorf U. Schreiber M. Protective effects of calcium antagonists in human renal transplantation.Kidney Int Suppl. 1992; 36: S87-S93PubMed Google Scholar, 26.Dawidson I.J. Sandor Z.F. Coorpender L. et al.Intraoperative albumin administration affects the outcome of cadaver renal transplantation.Transplantation. 1992; 53: 774-782Crossref PubMed Scopus (73) Google Scholar, 27.Powell J.T. Tsapepas D.S. Martin S.T. et al.Managing renal transplant ischemia reperfusion injury: novel therapies in the pipeline.Clin Transplant. 2013; 27: 484-491Crossref PubMed Scopus (31) Google Scholar or by spreading the use of hypothermic machine perfusion.28.Moers C. Kornmann N.S. Leuvenink H.G. et al.The influence of deceased donor age and old-for-old allocation on kidney transplant outcome.Transplantation. 2009; 88: 542-552Crossref PubMed Scopus (49) Google Scholar Even if this proportionality was assumed in other studies,14.Mikhalski D. Wissing K.M. Ghisdal L. et al.Cold ischemia is a major determinant of acute rejection and renal graft survival in the modern era of immunosuppression.Transplantation. 2008; 85: S3-S9Crossref PubMed Scopus (143) Google Scholar,21.Quiroga I. McShane P. Koo D.D.H. et al.Major effects of delayed graft function and cold ischaemia time on renal allograft survival.Nephrol Dial Transplant. 2006; 21: 1689-1696Crossref PubMed Scopus (265) Google Scholar,23.Van der Vliet J.A. Warlé M.C. Cheung C.L.S. et al.Influence of prolonged cold ischemia in renal transplantation.Clin Transplant. 2011; 25: E612-E616Crossref PubMed Scopus (52) Google Scholar it has not been clearly demonstrated that this assumption corresponds to a real relationship. In other words, our analyses highlighted the absence of CIT thresholds, in contrast to numerous manuscripts in which patients are stratified according to CIT intervals.12.Opelz G. Döhler B. Multicenter analysis of kidney preservation.Transplantation. 2007; 83: 247-253Crossref PubMed Scopus (166) Google Scholar,16.Kayler L.K. Magliocca J. Zendejas I. et al.Impact of cold ischemia time on graft survival among ECD transplant recipients: a paired kidney analysis.Am J Transplant. 2011; 11: 2647-2656Crossref PubMed Scopus (117) Google Scholar, 17.Johnson R.J. Fuggle S.V. O’Neill J. et al.Factors influencing outcome after deceased heart beating donor kidney transplantation in the United Kingdom: an evidence base for a new national kidney allocation policy.Transplantation. 2010; 89: 379-386Crossref PubMed Scopus (51) Google Scholar, 18.Summers D.M. Johnson R.J. Hudson A. et al.Effect of donor age and cold storage time on outcome in recipients of kidneys donated after circulatory death in the UK: a cohort study.Lancet. 2013; 381: 727-734Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar The proportional increase in the risk of graft failure and patient death related to CIT avoids loss of information and loss of power that may explain the nonsignificant association between the CIT and the mid-term graft outcomes in previous studies.15.Kayler L.K. Srinivas T.R. Schold J