Effectiveness of serum beta-2 microglobulin as a tool for evaluating donor kidney status for transplantation

Kidney transplantations using expanded criteria donors (ECD) are being increasingly adopted, but no consensus tools are available to evaluate donor kidney status. Beta-2 microglobulin (B2MG) is a marker of kidney function, and herein, we evaluate the usefulness of assessing B2MG to evaluate donor kidney status. Fifty-seven kidney transplantations were performed from March 2017 to April 2019. Medical records were retrospectively reviewed, and relationships between clinical and laboratory variables and transplant outcomes were investigated. Thirty-eight patients received a standard criteria donor kidney and 19 patients an ECD kidney. Ten patients experienced delayed graft function (DGF), but no patient experienced primary nonfunction. Of the parameters studied, only donor renal replacement therapy (RRT) [odds ratio (OR) 24.162; p = 0.018] and donor serum B2MG (OR 22.685; p = 0.022) significantly predicted DGF. The presence of either of these two risk factors can better reflect the condition of the donor than previous classification. However, on their last follow-up creatinine and estimated glomerular filtration rate values in those with or without these risk factors were not significantly different. For an ECD with a B2MG level of <7.18 and no history of RRT, kidney transplantation can be undertaken without considering the possibility of kidney discard.


Risk factors of delayed graft function.
Receiver operating characteristic (ROC) curve analysis showed donor final eGFR, highest Cr, serum cystatin C, serum B2MG level, and cold ischemic time significantly predicted DGF. Areas under curves (AUCs) of parameters are displayed in Table 3. Univariate logistic regression analysis revealed hypoxic-origin brain death, donor RRT, an AKI grade > 2, and donor serum cystatin C and serum B2MG levels significantly predicted DGF. However, multivariate logistic regression analysis revealed that donor RRT [odds ratio (OR) = 12.837; p = 0.025; 95% confidence interval (CI) = 1.374 -119.954] and donor B2MG level of ≥ 7.18 mg/L (OR = 12.719; p = 0.011; 95% CI = 1.785-90.63) were significant independent risk factors of DGF. Logistic regression analysis results are provided in Table 4.
Comparison of the baseline characteristics of recipients reclassified based on the risk factors. Contrary to the traditional SCD/ECD classification, DGF developed more frequently in those recipients with a risk factor (2, 5.4% vs.7, 50%; p = 0.001). All living donor kidney recipients belonged to the group without a risk factor. Laboratory parameters including initial eGFR, highest Cr, final Cr, and final eGFR were much better in the group without risk factors than in the group with risk factors, and the prevalence of DGF and biopsy-proven rejection were lower in recipients without a risk factor. However, recipient last follow-up Cr and eGFR values were similar in the two groups. A summary of the results is provided in Table 5.

Discussion
In this retrospective study, we analyzed the incidence and risk factors of DGF during the immediate postoperative period after KT. DGF is the most common problem after deceased donor KT but is also encountered after living donor KT. The main contributors to the development of DGF are ischemia-reperfusion injury and hemodynamic instability 9 . The likelihood of DGF is the most important consideration when a surgeon decides to use or discard a kidney with AKI from an ECD. Several authors have reported that KT using an ECD with AKI results in comparable short-and long-term outcomes [10][11][12][13][14] , although other studies have shown contrary results 15 . Although many authors have reported it is acceptable to use an ECD kidney with AKI 16 , no practical tools are available to evaluate ECD kidneys with AKI.    www.nature.com/scientificreports www.nature.com/scientificreports/ without AKI or SCD, subgroups that needed RRT or showed Cr level elevation showed poor outcomes 10,17,18 . A significant risk factor of primary nonfunction or poor graft outcome in KT was severe AKI that needed RRT 16 . However, our results showed that RRT was the only risk factor of DGF and did not affect short-term outcomes. Unlike our study, previous studies have included data from multiple centers and donors from multiple centers, and those previous studies did not precisely reveal the final Cr level of donors and estimate AKI status. In our study, although we examined a small number of cases, all donors were procured in our hospital, and all laboratory results were reported by using a sample collected on procurement day. These factors may explain why there are differences in the primary nonfunction incidences and effects on graft outcome between our study and previous studies. B2MG has a central role in cellular immunology, is related to residual kidney function, and its plasma concentration is correlated with GFR. 19 However, careful interpretation of a serum B2MG increase is needed as a B2MG increase can occur in solid organ malignancy, lymphoproliferative and autoimmune diseases [20][21][22][23] , and inflammatory bowel disease 24 . B2MG is associated with the major histocompatibility complex I (MHC-I) in all nucleated cells, and under the above conditions, a large number of cells may be generated and shed B2MG 25 . Furthermore, brain death induces a profound inflammation reaction, which influences multiple transplantable organs and causes kidney tubular damage by generating cascades of activated cytokines 26,27 . Ischemic-reperfusion injury can induce similar events. B2MG is an indicator of systemic inflammation, but not of kidney status, and the outcomes and functions of transplantable organs are significantly influenced by inflammation reactions. Thus, B2MG status, which is strongly correlated with GFR and inflammation, may be a meaningful marker of transplantable organ function, including kidney function. Most B2MG is reabsorbed in proximal tubules, and thus, a high level of urinary B2MG indicates tubule damage 25 . However, Cooper et al. 28 reported that acute tubular injury did not affect transplant outcome. In the present study, urinary B2MG was not predictive of DGF, and the urinary B2MG level was not different between the SCD and ECD groups. In a previous study, the effectiveness of serum B2MG level as a biomarker of adverse outcomes in CKD was investigated. In patients with ESRD or chronic allograft damage after renal transplantation, a previous study reported that compared to urinary B2MG level, serum B2MG level predicted kidney status more accurately 25 . However, there are no reports evaluating donor kidney status or predicting the incidence of DGF by assessing serum B2MG levels. Even though our study was conducted at a single institution and on a small number of patients, serum B2MG showed good predictive power for DGF with RRT, and it reflected the condition of donors more accurately than those derived by using the SCD and ECD classification approaches.
DGF is not an uncommon complication during the period immediately following transplantation and is associated with prolonged hospitalization, increased cost, and the possibility of graft failure 29 . If identification of DGF risk factors before transplantation were possible, recipient outcomes would be improved, and donors would have an additional incentive to prevent kidney discard. The therapeutic interventions used to prevent DGF have been widely investigated and reviewed. A recent Cochrane review showed that machine perfusion provides a better means of preserving donor kidneys than cold storage in terms of preventing DGF and improving outcome 30 . Modifications of immunosuppressant agents have been investigated, and the selection of induction agents is a well-debated issue. Rabbit anti-thymocyte globulin (rATG) produced better results than interleukin-2 receptor blockers in some studies 31,32 , but other studies produced contradictory results 33 . In the present study, rATG was used to a greater extent in the recipients with risk factors group, and its use was not associated with reductions in DGF and rejection rates. However, final creatinine and eGFR levels in the recipients with risk factors group were similar to those in the recipients without risk factor group despite the presence of DGF and rejection. Even though machine perfusion can produce better results, many institutions cannot use machine perfusion due to resource limitations. Thus, modulation of immunosuppressant agents may be a more realistic option in many situations.
In conclusion, we increased the use of ECDs and DCDs to address donor shortage, which means we now attach greater importance to detecting patients at risk to prevent DGF and improve long-term outcomes. In this regard, it is important to find kidney function indicators that can be quickly and repeatedly assessed. Although our study was conducted on a small cohort at a single institution, the results show that if serum B2MG is <7.18 mg/L and therapeutic, and there is no history of donor RRT, KT from an ECD with AKI can be performed without the need to consider kidney discard.

Methods
Patients. From March 2012 to July 2019, 135 patients underwent KT at Wonju Severance Christian Hospital, Wonju, South Korea, with all transplants performed by a single surgeon. B2MG samples were prospectively obtained beginning in March 2017, and the present retrospective study was performed using the data of 57 patients that registered from March 2017 to April 2019. This study was approved beforehand by the Institutional Review Board of Yonsei University Wonju Severance Hospital (CR318083). The requirement for informed consent was waived because of the retrospective nature of the study. All studies were conducted according to the principles of the Declaration of Helsinki. All living and deceased donors were evaluated for infections of hepatitis B or C virus (HBV, HCV) or human immunodeficiency virus (HIV) before transplantation. Transplantation was conducted in cases when donors did not show evidence of HBV, HCV, and HIV infections. All procurements and transplantations were performed by surgeons of our departments who were registered into Korean Network for Organ Sharing (KONOS), under the permission of KONOS. There is no procurement from prisoners.

Variables.
Samples of blood and urine to assess serum cystatin C, creatinine, B2MG, and urine cystatin C levels were collected and evaluated one day before procurement of a deceased donor or living donor kidney. B2MG level was measured by enzyme-linked fluorescent assay (ELFA) using a Vidas analyzer (bioMérieux, France). The eGFR was calculated using the Modification of Diet in Renal Disease four4-variable equation 34 . SCD, ECD, DGF, and acute AKI were defined as we described in a previous article 13 . (2020) 10:8109 | https://doi.org/10.1038/s41598-020-65134-6 www.nature.com/scientificreports www.nature.com/scientificreports/ Recipients at immunologic risk were defined as those that had undergone a second or ABO-incompatible KT, had a panel reactive antibody level > 20%, or had received a kidney from a deceased donor with an AKI grade > 3.
Immunosuppressive agents. When neither donor nor recipient had any risk factor, we chose basiliximab as the induction agent. Briefly, 20 mg of basiliximab were injected on the day of the operation and again at four days after surgery. However, when a recipient was deemed high immunologic risk, rATG was the induction agent and was administered at 4.5 mg/kg/day in the first three days. We used a triple regimen of tacrolimus, corticosteroid, and mycophenolate mofetil as the maintenance immunosuppressant agents. Trough levels of tacrolimus and mycophenolate mofetil dosage were controlled as we previously described 13 with the exception that corticosteroid administration was started as 500 mg of methylprednisolone followed by dose tapering as previously described 13 . We checked cytomegalovirus (CMV) and BK virus infection status repeatedly. CMV infection status was checked postoperatively at 1 day, 1 week, 1 month, and 3 months. The BK virus status was checked postoperatively every three months for 1 year and then every 6 months. If quantitative CMV and BK virus test results were positive, we performed qualitative tests. As needed, oral valganciclovir was used to treat CMV infection, and a reduced level of immunosuppressive agents was applied for BK virus treatment.
Outcomes. The baseline characteristics of SCDs and ECDs were compared and their effects on transplantation outcomes were investigated. Because there was no case of primary nonfunction, we performed analysis to identify risk factors for DGF. We also compared recipient baseline characteristics were compared and transplantation outcomes analyzed according to the absence or presence of risk factors.
Statistics. Statistical analyses were performed using SPSS Ver. 25.0 (SPSS Inc., Chicago, IL, USA). Continuous variables are presented as medians and ranges, and categorical variables as numbers and percentages. Categorical and continuous variables were compared using Fisher's exact t test and the t-test, respectively. Cut off values of factors predictive of DGF were determined by performing receiver operating characteristic (ROC) analysis. Variables significantly associated with DGF were identified by univariate logistic regression analysis. Multivariate analyses were performed using the logistic regression analysis posterior Wald test.