Soluble urokinase plasminogen activator receptor and acute kidney injury in hematopoietic cell transplantation Open Access

Patients who underwent allogeneic hematopoietic cell transplantation (alloHCT) are at risk for developing acute kidney injury (AKI) and chronic kidney disease.^1,2 Some of the risk factors associated with AKI are the conditioning regimens for HCT; engraftment syndrome; graft-versus-host disease (GVHD) involving the kidney; exposure to drugs, such as calcineurin inhibitors for the prevention and/or treatment of GVHD; and antibiotics for prophylaxis and treatment of various infections. AKI in this patient population portends a worse prognosis after HCT and poorer overall outcome.^3-5 The risk of developing AKI is higher after alloHCT than after autologous HCT.⁴ The current risk stratification methods for AKI after HCT lack precision in identifying individual patients at risk of AKI. Biomarkers for AKI may facilitate a more precise risk estimation, thereby enabling the prediction of risk of AKI before the changes become irreversible.⁶ The clinical utility of biomarkers of AKI is not well established in the post-HCT population.

Serum soluble urokinase-type plasminogen activator receptor (suPAR) is a circulating biomarker produced by immature myeloid cells and podocytes and is elevated in malignant diseases, infectious processes, and kidney disease.⁷ Higher serum suPAR levels have been reported in proteinuric kidney disease and are associated with a higher risk of progression to chronic kidney disease and end stage renal disease.^8,9 Elevated serum suPAR levels have also been associated with worse clinical outcomes and higher mortality after alloHCT.¹⁰ Recently, it has been shown that serum suPAR can predict AKI in various clinical contexts, supporting its use as a novel biomarker for AKI.^11-14 

We hypothesized that serum suPAR has potential as a novel biomarker for and independent predictor of AKI requiring dialysis (AKI-D) after HCT.

This was a nested case-control study that used the Blood and Marrow Transplant Clinical Trials Network 1202 study, an observational study of 1709 alloHCT recipients (www.ClinicalTrials.gov identifier: NCT01879072).¹⁵ 

All adults aged ≥18 years who developed AKI-D after HCT were included in the study. Controls were randomly selected from the patients who did not develop AKI during the 2 years of follow-up. This study focused on adults given the higher burden of comorbidities that also affect their HCT-specific comorbidity index score.

Patients <18 years at HCT and those with a baseline serum creatinine of >1.0 mg/dL or baseline estimated glomerular filtration rate of <60 mL/min were excluded from the study.

Frozen serum samples collected before the transplant (baseline) and at day +7, +14, +21, and +28 were used to measure the primary time points of interest. Baseline samples were obtained a week before conditioning regimen initiation. Additional samples at day +42, +56, and +90, when available and collected before the occurrence of AKI, were also analyzed in an additional exploratory analysis. Serum samples were assayed for suPAR, neutrophil gelatinase-associated lipocalin (NGAL), and serum creatinine (sCr). Serum suPAR and NGAL were measured by quantitative sandwich enzyme immunoassay with validated reagents and quality controls (Quantikine enzyme-linked immunosorbent assays DUP00 and DLCN20, respectively) purchased from Research and Development Systems (Minneapolis, MN). Any result above the high standard was reassayed after the appropriate dilution. The sensitivity of the suPAR assay was 33 pg/mL. The sensitivity of the serum NGAL assay was 12 pg/mL. sCr was measured using the Jaffe method (catalog no. KB02-H1; Arbor Assays, Ann Arbor, MI). The assay was validated against our clinical laboratory sCr method.¹⁶ 

The HCT recipients were divided into case and control groups. AKI patients included recipients who were classified on the Center for International Blood and Marrow Transplant Research forms as having stage 3 AKI-D based on the Kidney Disease: Improving Global Outcomes guidelines,¹⁷ and controls consisted of recipients who had no documented AKI during the 2-year follow-up. The controls were also matched 1:1 by age, sex, and conditioning regimen (myeloablative vs nonmyeloablative conditioning). In addition, we evaluated donor type, disease type, and disease status at HCT as covariates and chronic GVHD of the kidneys, the use of nephrotoxic medications including prophylactic antibiotics and calcineurin inhibitors, engraftment, and veno-occlusive disease as time-dependent covariates for suPAR samples being collected at day +21 and later but not before day +21.

This study was reviewed and approved by the NMDP Institutional Review Board.

The demographic and disease characteristics were summarized using descriptive statistics and compared between the case and control cohorts using the standard t test and Wilcoxon signed-rank test for continuous variables and the χ² test for categorical outcomes. Univariate comparison of biomarkers between matched AKI-D patients and controls were made for each biomarker at different time points separately using the paired t test and Wilcoxon signed-rank test. In the multivariate analysis, the association of suPAR levels at each time point with the development of AKI-D was tested separately using conditional logistic regression with stratification by pairing status.

The association of suPAR levels at each time point with overall survival (OS), disease-free survival (DFS), transplant-related mortality (TRM), relapse, acute GVHD grade 2 to 4, acute GVHD grade 3 to 4, chronic GVHD, and renal failure was also tested separately using Cox regression with stratification by pairing status. For these time-to-event outcomes, the time clock started at the time when the biospecimen was acquired and measured.

All biomarkers were examined as either continuous or dichotomized variables. When they were treated as continuous variables, linearity of the biomarkers was tested using the spline approach. To control for other potential risk factors, a stepwise, forward variable selection was performed. The examined variables in the multivariable regression included patient age, time from diagnosis to transplant, disease type, conditioning regimen, GVHD prophylaxis, graft type, donor source, Karnofsky score, HCT-specific comorbidity index, patient cytomegalovirus (CMV) status, donor CMV status, patient sex, donor sex, patient race, acute myeloid leukemia status, and myelodysplastic syndrome status. Only the patient CMV status was selected for relapse outcome at the nominal 0.05 significance level. To adjust for multiple testing, the threshold of P value ≤.01 was used for significance of the biomarkers. The predictive performance of the significant suPAR levels was assessed using the receiver operating characteristic curve and area under the curve measurement.

In addition, the association of other traditional biomarkers (NGAL and sCr) with the development of AKI-D and the time of OS, DFS, etc, was tested. The impact of acute GVHD grade 2 to 4 and grade 3 to 4 and chronic GVHD on the time of AKI-D, engraftment, and veno-occlusive disease was also assessed as time-dependent covariates.

The study included 124 patients with 62 patients in the AKI-D group (patients) and 62 patients in the control group (supplemental Figure 1). There were no differences between the patients with AKI-D and the controls in terms of Karnofsky score and time from diagnosis to transplant (Table 1). The mean age in the AKI-D group was 53 years vs 52.5 years in the control group. The cohort comprised predominantly White patients in both patients and controls. The underlying disease was noted to be acute leukemia/myelodysplastic syndrome in most patients in both the AKI-D and control groups (n = 46 in each group). The primary hematologic malignancy was lymphoma (9 AKI-D patients and 8 controls), chronic leukemia/myeloproliferative neoplasms (4 AKI-D patients and 5 controls), and plasma cell disorder (3 patients and controls). The conditioning regimen was reduced-intensity conditioning for most patients (58% in AKI-D patients and 53% in controls). The other conditioning regimens used were myeloablative conditioning with chemotherapy (27% AKI-D patients and 31% controls) and myeloablative conditioning with total body irradiation (15% patients and 16% controls). A higher proportion of the AKI group (81%) was on a tacrolimus-based regimen for GVHD prophylaxis when compared with the control group (73%).

The median time of follow-up for survivors among controls was 58.8 months (interquartile range [IQR], 12.9-72.0). The median time from transplant to AKI-D in the affected group was 2.6 months (IQR, 0.7-7.5). The major cause of death in the AKI-D group was primary disease (17%), organ failure (14%), and infection (14%). In the control group, the major cause of death was the primary disease (60%).

The median suPAR value among AKI patients at day +7 was higher than among controls (2.7 ng/mL [IQR, 2.1-3.5] vs 2.1 ng/mL [IQR, 1.53-2.59]; P = .002). There were no significant differences in the NGAL and sCr values between AKI patients and controls for any of the samples tested from baseline to day +90 after HCT (Figures 1 and 2).

In the multivariate analysis, the suPAR level at day +7 after HCT was significantly associated with the development of AKI-D (odds ratio, 1.82; 95% confidence interval [CI], 1.16-2.87; P = .009; Table 2). There was no association between the change in suPAR level and AKI-D and OS (supplemental Tables 2 and 3). To assess the overall performance of suPAR as a biomarker for AKI, receiver operating characteristic curves were generated using suPAR as a continuous variable (Figure 3), and this yielded an area under the curve of 0.749. There was no association of NGAL or sCr with AKI at any time point in the multivariate analysis.

In the multivariate analysis, suPAR at day +7 after HCT was also significantly associated with OS (hazard ratio [HR], 1.91; 95% CI, 1.29-2.86; P = .0017). For simple interpretation, an optimal cutoff of 2.48 was determined for suPAR at day +7 based on the maximum likelihood for the development of AKI-D. A higher day +7 suPAR level (level, >2.48) was associated with poorer OS (HR, 4.78; 95% CI, 1.97-11.61; P = .0005) when compared with a suPAR level of ≤2.48 (Table 3).

The OS probabilities were also estimated for both AKI-D patients and controls based on a suPAR level cutoff of 2.48 ng/mL (Figure 4). The survival probability in AKI-D patients with suPAR >2.48 was lower (HR, 1.63; 95% CI, 0.91-2.92; P = .10) than ≤2.48.

After accounting for the pairing, a day +7 suPAR level of >2.48 was also associated with poor DFS (HR, 4.20; 95% CI, 1.70-10.37; P = .0018; supplemental Table 1). Meanwhile, a suPAR level >2.48 tended to be associated with higher relapse (HR, 8.25; 95% CI, 0.91-74.50; P = .0602) and TRM (HR, 2.77; 95% CI, 0.97-7.88; P = .0565). There was no association of NGAL and sCr with OS.

We hypothesized that the suPAR level could be a novel biomarker for AKI-D in the post-HCT population. This study showed that suPAR may be an early marker of AKI-D after HCT with day + 7 suPAR levels predicting AKI-D. In addition, elevated suPAR levels (>2.48 ng/mL) may be associated with lower OS after HCT. There was no association between the change in suPAR level and AKI-D and OS.

The cause and characteristics of AKI after HCT is multifactorial.¹⁸ sCr is the most frequently used biomarker in clinical practice with varying degrees of sCr elevation corresponding to different stages and severity of AKI. A major pitfall of using sCr is the lack of sensitivity and specificity in the diagnosis of AKI. Significant kidney injury may have already occurred before AKI is detected when relying solely on sCr values, because the increase in sCr can lag.^19,20 Thus, sCr is an insensitive marker of renal injury.²¹ 

Other biomarkers, like NGAL, have been studied in AKI in varied clinical settings. Plasma NGAL levels may be elevated when there is a reduction in the glomerular filtration rate, and urinary NGAL may be high because of decreased reabsorption in the proximal tubule or increased NGAL production in the distal tubule and loop of Henle. NGAL is considered to be an early marker of AKI and can be used to differentiate between kidney dysfunction and kidney injury in hemodynamic AKI and acute tubular injury.^20,22 NGAL has shown promise as a noninvasive, early biomarker of AKI with good correlation between serum and urine NGAL levels.^23,24 A significant correlation between urine NGAL level and AKI has been reported.²⁵ Evaluation of NGAL in the prediction and prognosis of AKI after HCT has not provided a clear consensus about the role of NGAL. In addition, urinary NGAL may not be useful as a biomarker for early AKI after HCT.²⁶ These data highlight the challenges with using the currently accepted biomarkers of AKI after HCT.

suPAR is a novel, early biomarker of AKI in patients who underwent a coronary angiography or cardiac surgery, in critically sick patients in the intensive care units, and for AKI risk stratification among older patients who present to the emergency department. In addition, it can also serve as an early biomarker of COVID-19–related AKI.^12,14,27,28 suPAR seems to be a useful marker of morbidity and mortality because of its role in systemic inflammation.^29,30 Elevated suPAR levels indicate a prevalent state of widespread inflammation, which may lead to end organ damage, notably kidney damage.³⁰ In animal models, high suPAR levels have been associated with profound structural damage to the kidney and impairment in kidney function.³⁰ 

Elevated suPAR levels have been associated with lower survival (independent of renal function) in patients after kidney transplantation likely because of a higher burden of cardiovascular disease.³¹ In patients undergoing hemodialysis, the suPAR level predicted both cardiovascular and noncardiovascular mortality.³² 

Elevated suPAR levels in our study were associated with overall lower survival among AKI patients. Control patients who had no AKI were also at higher risk for mortality in the presence of elevated suPAR levels. It is unclear whether this association is related to a higher burden of cardiovascular mortality or other risk factors after HCT.

suPAR holds promise as an early marker of AKI-D after HCT, perhaps as early as day +7, as demonstrated in this study. This finding may lead to therapeutic interventions that target reducing or eliminating suPAR from the circulation to reduce the burden of AKI after HCT.¹⁴ 

This study has several strengths. The study was conducted on a large cohort of HCT recipients with AKI and matched, unaffected controls. Different biomarkers were rigorously tested. To our knowledge, this is the first study that evaluated the role of suPAR as a biomarker for AKI and mortality in patients who underwent HCT. The positive association of the day +7 suPAR level with development of stage 3 AKI provides proof of concept for the utility of the suPAR level as a biomarker. We also found a positive association between suPAR levels >2.48 ng/mL and lower OS after HCT.

The study also has some limitations. It was a retrospective observational study, and we were only able to study adults with AKI-D (stage 3 AKI requiring dialysis), because the data captured on forms maintained by the Center for International Blood and Marrow Transplant Research does not include AKI stage 1 and 2. In addition, among both patients and controls, White people made up the largest proportion of participants and other races were underrepresented. The results of this study need validation in a prospective, independent study across different stages of AKI, as well as across other racial groups. We were not able to measure suPAR in urine samples, because these samples were not available. In addition, this study was done in the era before widespread use of posttransplant cyclophosphamide for GVHD prophylaxis. Posttransplant cyclophosphamide was used at a higher frequency in the AKI-D group and is being adopted widely for GVHD prophylaxis as the regimen of choice, as shown in the Blood and Marrow Transplant Clinical Trials Network 1703 cohort,³³ and its impact on endothelial injury and AKI-D needs further evaluation. This paradigm shift in conditioning regimen and its impact on the prediction of AKI by use of suPAR will need to be evaluated in future studies.

Future studies should also evaluate suPAR as a marker of AKI after HCT in pediatric patients and evaluate suPAR prospectively in serum and urine samples at varying stages of AKI to understand the association of suPAR with the severity of AKI. This may increase the understanding of factors that cause AKI and also target interventions to mitigate these risk factors. The incorporation of this biomarker prospectively may enable the opportunity to better understand the etiology, causative factors, the molecular and pathologic impact of elevated suPAR levels after HCT, and the association with AKI and OS. In addition, data on OS, DFS, TRM, and relapse are hypothesis-generating and need to be validated in future studies.

The authors thank participating investigators who collected data and cared for study patients. The authors also thank Jennifer Motl, RDN, medical writer, of the Medical College of Wisconsin (MCW) for editorial support, which was funded by MCW in accordance with Good Publication Practice guidelines (http://www.ismpp.org/gpp3).

Support for this study was also provided by grants U10HL069294 and U24HL138660 to the Blood and Marrow Transplant Clinical Trials Network from the National Heart, Lung, and Blood Institute and the National Cancer Institute. The study was funded by the Pilot Grant (project number: FP22531) from the Medical College of Wisconsin Cancer Center, Milwaukee, Wisconsin, and the Small Grants Award from the Mayo Clinic Arizona, Scottsdale, Arizona. The authors acknowledge the support received for this work from the Biostatistics Shared Resource at the Medical College of Wisconsin Cancer Center.

Contribution: B.B.-C., T.W., and W.S. were responsible for the conceptualization of this study and drafting, writing, and editing of the manuscript; T.W. was responsible for data curation, statistical design, and analysis; H.R. performed the laboratory analyses, critically reviewed and edited the manuscript, and compiled supplemental Figure 1; and J.E.L., M.-A.P., S.S., and A.B. critically reviewed and edited the manuscript.

Conflict-of-interest disclosure: J.E.L. reports receiving consultant fees from bluebird bio, Editas, Equillium, Inhibrx, Kamada, Mesoblast, Sanofi, and X4 Pharmaceuticals. S.S. reports receiving honoraria and consulting fees from University of Arkansas for Medical Sciences (UAMS), St. Jude Children's Research Hospital (SJCRH), bluebird bio, Banner Hospital, Blackwood Continuing Medical Education (CME), and Beam Therapeutics. M.-A.P. reports receiving honoraria from Adicet, Allogene, AlloVir, Caribou Biosciences, Celgene, Bristol Myers Squibb, Equillium, ExeVir, ImmPACT Bio, Incyte, Karyopharm, Kite/Gilead, Merck, Miltenyi Biotec, MorphoSys, Nektar Therapeutics, Novartis, Omeros, Orca Bio, Sanofi, Syncopation, VectivBio AG, and Vor Biopharma; serving on data and safety monitoring boards for Cidara Therapeutics and Sellas Life Sciences; serving on the scientific advisory board of NexImmune; having ownership interests in NexImmune, Omeros, and Orca Bio; and receiving institutional research support for clinical trials from Allogene, Incyte, Kite/Gilead, Miltenyi Biotec, Nektar Therapeutics, and Novartis. H.R. reports serving as a consultant for Cerium Pharmaceuticals, Corcept Therapeutics, and Novo Nordisk. The remaining authors declare no competing financial interests.

Correspondence: Bhavna Bhasin-Chhabra, Division of Nephrology and Hypertension, Mayo Clinic Arizona, 13400 E Shea Blvd, Scottsdale, AZ 85259; email: bhasin-chhabra.bhavna@mayo.edu.