Scholarly article on topic 'Risk Factors Associated with Kidney Injury and the Impact of Kidney Injury on Overall Survival in Pediatric Recipients Following Allogeneic Stem Cell Transplant'

Risk Factors Associated with Kidney Injury and the Impact of Kidney Injury on Overall Survival in Pediatric Recipients Following Allogeneic Stem Cell Transplant Academic research paper on "Clinical medicine"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{Pediatrics / "Kidney injury" / "Reduced toxicity and myeloablative conditioning" / "Allogeneic stem cell transplant"}

Abstract of research paper on Clinical medicine, author of scientific article — Prakash Satwani, Sejal Bavishi, Zhezhen Jin, Judith S. Jacobson, Courtney Baker, et al.

Pediatric allogeneic stem cell transplant (AlloSCT) patients are at substantial risk of developing kidney injury (KI), and KI contributes to transplant-related morbidity and mortality. We compared the estimated creatinine clearance (eCrCl) at 1, 3, 6, 9, and 12 months post-AlloSCT in 170 patients following reduced toxicity conditioning (RTC) versus myeloablative conditioning (MAC) to baseline. eCrCl was calculated using the Schwartz equation. Patients with ≥50% drop in eCrCl from the baseline were considered to have KI. Patients received tacrolimus and mycophenolate mofetil (MMF) for graft-versus-host disease (GVHD) prophylaxis. The logistic regression model was used for assessing risk factors for KI. Seventy-six patients (median age = 10.6 years) received RTC AlloSCT; 94 patients (median age = 8.5 years) received MAC AlloSCT. The incidence of KI at 1 month post-AlloSCT was significantly higher in MAC versus RTC AlloSCT (43/94 [45.7%] versus 13/76 [17.1%] P < .0001). There was no statistical difference in KI at 3, 6, 9, and 12 months post-AlloSCT between the 2 conditioning groups. On multivariate analysis, only MAC was a significant risk factor for KI (odds radio [OR] 3.44, 95% confidence interval [CI] 1.59-7.42, P = .002). In multivariate analysis for risk factors affecting overall survival (OS), the following were statistically significant: MAC versus RTC (hazard ratio [HR] 2.66, P = .0008), average versus poor-risk disease status (HR 2.09, P = .004), matched sibling donor (MSD) and matched unrelated donor (MUD) versus umbilical cord blood (UCB) (HR 2.31, P = .013), no KI versus KI (HR 2.00, P = .005). In children, MAC is associated with significant risk of KI in the first month after transplant, and KI in the first month post-AlloSCT is associated with a significantly decreased OS.

Similar topics of scientific paper in Clinical medicine , author of scholarly article — Prakash Satwani, Sejal Bavishi, Zhezhen Jin, Judith S. Jacobson, Courtney Baker, et al.

Academic research paper on topic "Risk Factors Associated with Kidney Injury and the Impact of Kidney Injury on Overall Survival in Pediatric Recipients Following Allogeneic Stem Cell Transplant"

ASBMI

American Society for Blood and Marrow Transplantation

Risk Factors Associated with Kidney Injury and the Impact of Kidney Injury on Overall Survival in Pediatric Recipients Following Allogeneic Stem Cell Transplant

Prakash Satwani,1 Sejal Bavishi,1 Zhezhen Jin,2 Judith S. Jacobson,3 Courtney Baker,1 Deirdre Duffy,1 Leora Lowe,1 Erin Morris,1 Mitchell S. Cairo1'4'5

Pediatric allogeneic stem cell transplant (AlloSCT) patients are at substantial risk of developing kidney injury (KI), and KI contributes to transplant-related morbidity and mortality. We compared the estimated creatinine clearance (eCrCl) at 1,3, 6, 9, and 12 months post-AlloSCT in 170 patients following reduced toxicity conditioning (RTC) versus myeloablative conditioning (MAC) to baseline. eCrCl was calculated using the Schwartz equation. Patients with $50% drop in eCrCl from the baseline were considered to have KI. Patients received tacrolimus and mycophenolate mofetil (MMF) for graft-versus-host disease (GVHD) prophylaxis. The logistic regression model was used for assessing risk factors for KI. Seventy-six patients (median age = 10.6 years) received RTC AlloSCT; 94 patients (median age = 8.5 years) received MAC AlloSCT. The incidence of KI at 1 month post-AlloSCT was significantly higher in MAC versus RTC AlloSCT (43/94 [45.7%] versus 13/76 [17.1%] P <.0001). There was no statistical difference in KI at 3, 6, 9, and 12 months post-AlloSCT between the 2 conditioning groups. On multivariate analysis, only MAC was a significant risk factor for KI (odds radio [OR] 3.44, 95% confidence interval [CI] 1.59-7.42, P = .002). In multivariate analysis for risk factors affecting overall survival (OS), the following were statistically significant: MAC versus RTC (hazard ratio [HR] 2.66, P = .0008), average versus poor-risk disease status (HR 2.09, P = .004), matched sibling donor (MSD) and matched unrelated donor (MUD) versus umbilical cord blood (UCB) (HR 2.31, P = .013), no KI versus KI (HR2.00, P = .005). In children, MAC is associated with significant risk of KI in the first month after transplant, and KI in the first month post-AlloSCT is associated with a significantly decreased OS. Biol Blood Marrow Transplant 17: 1472-1480 (2011) © 2011 American Society for Blood and Marrow Transplantation

KEY WORDS: Pediatrics, Kidney injury, Reduced toxicity and myeloablative conditioning, Allogeneic stem cell transplant

INTRODUCTION

Over the last 4 decades, there has been a significant improvement in the field of allogeneic hematopoietic stem cell transplant (AlloSCT). AlloSCT from related

From the 'Departments of Pediatrics; 2Biostatistics; ^Epidemiology, Mailman School of Public Health; 4Medicine; and 5Pathology and Cell Biology, New York-Presbyterian Morgan Stanley Children's Hospital, Columbia University, New York, NY. Financial disclosure: See Acknowledgments on page 1479. Correspondence and reprint requests: Prakash Satwani, MD, Division of Pediatric Blood and Marrow Transplantation, New York-Presbyterian Morgan Stanley Children's Hospital, Columbia University, 3959 Broadway, CHN 10-02, New York, NY 10032 (e-mail: ps2087@columbia.edu). Received September 28, 2010; accepted February 13, 2011 © 2011 American Society for Blood and Marrow Transplantation 1083-8791/$36.00 doi: 10.1016/j.bbmt.2011.02.006

or unrelated histocompatible donors has been well established as a potentially curative therapy for children and adults with various malignant and nonmalignant disorders [1]. The concept of AlloSCT as a treatment option for hematologic malignancies has long been based on the assumption that myeloablative doses of cytotoxic therapy were required for both disease eradication and host immunosuppression. However, over the last decade, there has been a paradigm shift toward curative potential of graft-versus-leukemia or tumor (GVL/GVT) effect [2]. The concept behind reduced toxicity conditioning AlloSCT (RTC AlloSCT) is that instead of eradicating tumors through intensive/ toxic chemoradiation, the SCT donor's immune cells might be used for tumor eradication, relying on alloge-neic GVT effects. Hence, elimination of high-dose cytotoxic therapy would allow medically infirm patients to be treated with AlloSCT [2]. However, the success of AlloSCT is limited by transplant-related mortality

(TRM). One potential factor that could lead or add to TRM is kidney injury (KI) following AlloSCT.

The incidence of acute renal insufficiency in children receiving myeloablative conditioning (MAC) following SCT ranges from 21% to 50% [3-5], and the incidence of chronic renal insufficiency ranges from 11% to 62% [6]. Overall mortality rate is significantly higher in patients with acute renal insufficiency after AlloSCT [7]. Schrier and Parikh [7] compared the incidence of acute renal injury in adult patients receiving autologous myeloablative SCT versus nonmyeloabla-tive AlloSCT and MAC AlloSCT and found the incidence of acute KI to be 21% versus 40% versus 69%, respectively. Not surprisingly, an increase in acute KI correlated with parallel increases in mortality from 34% to 58% at 6 to 12 months, as well as progressive multiorgan involvement [7].

Patients with acute KI are at risk of chronic KI [8], and patients who develop chronic KI are at increased risk of mortality [9]. In a recently published study from Fred Hutchinson Cancer Research Center, Seattle, WA [10], the 30-year cumulative incidence of hypertension in long-term survivors of pediatric hematopoietic cell transplantation was 36%, and acute KI was highly associated with risk of hypertension in long-term survivors. Thus, prevention of acute KI following AlloSCT is needed to decrease early TRM as well as to decrease long-term consequences.

Factors associated with KI in children following AlloSCT include nephrotoxic medications, sepsis, dehydration, sinusoidal obstruction syndrome, hypertension, transplant-associated thrombotic microangiop-athy, viral and fungal infections, and/or its treatment, and total-body irradiation conditioning regimen [3,5]. Understanding of risk factors associated with KI might help to reduce the incidence of KI and may lead to decreases in short- and long-term morbidity and mortality.

There have been a few prospective studies in children and adults as well as retrospective studies in adults regarding the incidence of acute KI. Although there have been a few retrospective studies in adults comparing KI followingMAC and RTC AlloSCT, there is a paucity of parallel data on incidence of KI in children following RTC versus MAC AlloSCT. In the current study, we compared the incidence of KI and risk factors and mortality in children following MAC and RTC AlloSCT.

MATERIALS AND METHODS (DESCRIBED IN PART IN OUR PREVIOUS PUBLICATIONS) [11,12]

Patients included in this retrospective analysis included all consecutive children and adolescents who received a RTC or MAC AlloSCT at the New York-Presbyterian Morgan Stanley Children's

Hospital between January 2001 and December 2007. Indications for transplant included a variety of malignant and nonmalignant conditions. Allogeneic sources of stem cells included bone marrow (BM), peripheral blood stem cells (PBSC), and umbilical cord blood (UCB). All patients were on clinical protocols for AlloSCT approved by the institutional review board at Columbia University Medical Center, and all research protocols were in compliance with the Declaration of Helsinki.

HLATyping and Stem Cell Source

HLA-A and HLA-B antigen typing was performed by intermediate resolution molecular testing, and HLA-C, HLA-DRB1, and HLA-DQB1 allele typing was determined by hybridization of polymerase chain reaction (PCR)-amplified DNA with sequence-specific oligonucleotide probes, as we previously described [11]. Confirmatory typing was performed at Columbia University Medical Center. The criteria for graft matching included at least 4-5/6 loci for UCBT or 8/10 or greater for unrelated donor PB/BM and 56/6 HLA matched sibling. Transplants were classified as HLA-mismatched with 1 or 2 differences if disparities were detected in HLA-A, HLA-B, antigens or HLA-C, HLA-DRB1, and HLA-DQB1 alleles.

Conditioning Regimens

Specific conditioning regimens were protocol driven and disease specific. The RTC regimen included patients who received busulfan (6.4-8 mg/kg), fludara-bine (150-180 mg/m2) 6 ATG (n = 40); or busulfan (12.8-16 mg/kg), fludarabine (150-180 mg/m2) 6 alem-tuzumab (54 mg/m2) (n = 20); and cyclophosphamide 60 mg/kg, fludarabine (150-180 mg/m2) (n = 16). The MAC regimen included busulfan (12.8-16 mg/ kg), cyclophosphamide (120-200 mg/kg), melphalan (135 mg/m2) 6 antithymocyte globulin (ATG) (n = 50); or total body irradiation (TBI), cyclophosphamide (120-200 mk/kg), melphalan (135 mg/m2) 6 ATG (n = 44).

Graft-versus-Host Disease

Graft-versus-host disease prophylaxis and grading

Acute graft-versus-host disease (aGVHD) prophylaxis in the majority of patients consisted of tacrolimus starting at 0.03 mg/kg/day as continuous intravenous (i.v.) infusion or 0.12 mg/kg orally (p.o.) twice a day with dosage adjustment to maintain blood levels between 5 and 20 ng/mL and mycophenolate mofetil (MMF) at 15-30 mg/kg every 6 to 12 hours either p.o. or i.v., as we have previously described [13,14]. Tacrolimus was started either on the first day of conditioning regimen or day — 1 and MMF began on

day +1. Tacrolimus and/or MMF were tapered if patients had equal to or less than grade II aGVHD on day +30 for malignant diseases and day +180 for nonmalignant diseases [14]. Acute GVHD and chronic GHVD (cGVHD) were graded according to Seattle criteria [15]. Patients who received unrelated AlloSCT also received methotrexate (MTX) 15 mg/m2 administered i.v. on day +1 followed byMTX 10 mg/m2 slow i.v. push on days +3, +6, and +11.

Infection Prophylaxis and Supportive Care

All patients were hospitalized in protective isolation defined as single hospital rooms with high efficiency particulate air filtration system and reverse isolation requiring strict hand degerming and mask use by staff for unrelated hematopoietic SCT (HSCT) patients. All patients received sargramostim (250 mg/m2/day) i.v. daily from day 0 until the white blood cells (WBCs) reached $300/mm3 x 2 days and then were switched to filgrastim (10 mg/kg/day) either i.v. or subcutaneously until an absolute neutrophil count of $2500/mm3 was achieved for 3 days as we previously described [16]. Herpes simplex virus (HSV) prophylaxis consisted of acyclovir (250 mg/ m2) i.v. every 8 hours from day —5 until engraftment and equal to or less than grade II mucositis. Pneumocystis carinii prophylaxis consisted of trimethoprim/sulfa-methoxazole until day — 2 and then resumed 3 times weekly after myeloid engraftment. Patients unable to tolerate trimethoprim/sulfamethoxazole received i.v. pentamidine prophylaxis every 2 weeks. Fungal prophylaxis consisted of liposomal amphotericin B (3 mg/kg/day) i.v. starting on day 0 through day +100, as we previously described [17]. Cytomegalovirus (CMV) prophylaxis was administered as we have recently described [18]. Briefly, patients with risk of acquiring CMV infection (CMV+ donor and/or recipient) after achieving an absolute neutrophil count >750/mm3 after AlloSCT received prophylaxis with foscarnet (90 mg/kg/dose) every other day, alternating with ganciclovir (5 mg/kg/dose) every other day until day +100.

Kidney Function Monitoring

In all patients, serum creatinine was measured daily in the inpatient unit for the first 30 days, and those patients who were discharged before day +30 had serum creatinine measured 2 or 3 times a week.

TRM was defined as death because of any cause other than disease relapse.

Poor-risk malignant patients were defined as: che-moresistant malignant disease, third complete remission or greater, induction failure, progressive disease, and/or receiving second allograft. All other patients with malignant diseases and nonmalignant diseases were defined as average risk.

Definitions

Estimated creatinine clearance (eCrCl) was calculated using the Schwartz equation [19]. We compared eCrCl at 1, 3, 6, 9, and 12 months post-AlloSCT to baseline (#2 weeks pre-AlloSCT).

Kidney injury was defined as $50% decrease in eCrCl from the baseline pretransplant value [20].

Statistical Methods

The continuous variables were summarized by mean 6 standard deviation, and categoric variables were summarized by percentage. The comparisons between 2 treatment groups were done by t-test for continuous variables and by chi-square test for categoric variables. For KI at 1, 3, 6, and 12 months following AlloSCT the eCrCl was compared to the baseline eCrCl measured prior to the start of conditioning regimen. The logistic regression model was used to investigate the effects of treatment on the first month KI. The multivariate logistic regression model was built with those factors significant at level 0.2 in the univariate analysis. Survival was examined with the KaplanMeier method, and comparison of survival between groups was carried out by the log-rank test. In addition, the Cox proportional hazards model was used for the analysis of risk factors on survival, and multivar-iate analysis was carried out based on those factors significant at level 0.2 in the univariate analysis. Patients who were alive were censored at subsequent transplant, or end of follow-up. Risk factors analyzed for KI were age, gender, baseline eCrCl, CMV risk status, fungal infection, disease status disease type (malignant versus nonmalignant), veno-occlusive disease (VOD), UCB, MAC, and stem cell source, liver dysfunction (alanine aminotranferase > 5 times of upper limit of normal and bilirubin >2 mg/dL) in the first 30 days post-AlloSCT and use of foscarnet in the first 30 days post-AlloSCT; and risk factors analyzed for overall survival (OS) were age, gender, CMV risk status, disease type and disease status, VOD, fungal infections, stem cell source, bacterial and viral infections, and aGVHD and cGVHD. The SAS 9.2 was used for data analysis.

RESULTS

Patients and Demographics

A total of 170 patients received AlloSCT within the study time period (106 male, 64 female; age: 9.5 6 6.55 years). The RTC group contained 76 patients (age 10.66 6 6.66 years) of whom 44 (57.8%) received transplants for malignant conditions (acute myeloid leukemia/myelodysplastic syndrome n = 14; neuroblastoma n = 10; Hodgkin lymphoma n = 9; chronic myeloid leukemia n = 5; non-Hodgkin

lymphoma n = 6), and 32 (42.2%) for nonmalignant conditions (hemoglobinopathies n = 11; severe aplastic anemia n = 12; immunodeficiency disorders n = 3; histiocytic disorders/hemophagocytic lymphohistio-cytosis n = 3; metabolic disorder n = 3). The MAC group contained 94 patients (age 8.5 6 6.33 years), 64 (67%) received transplants for malignant conditions (acute myeloid leukemia/myelodysplastic syndrome n = 28; acute lymphoblastic leukemia n = 30; non-Hodgkin lymphoma n = 4; neuroblastoma n = 2), and 30 (33%) received transplants for nonmalig-nant conditions (hemoglobinopathies n = 9; histio-cytic disorders/hemophagocytic lymphohistiocytosis n = 8; severe aplastic anemia n = 10; and immunodeficiency disorders n = 3). There was no statistical difference when we compared gender (P = .07), CMV risk status (P = .96), and patients with malignant diseases (P = .17), between MAC versus RTC AlloSCT. However, children in the RTC AlloSCT group were older (P = .03), and there were more children who received UCB (P < .001) and were in the poor-risk group (P = .009) in the MAC AlloSCT group (Table 1).

Donor source

The donor source in the RTC regimen consisted of UCB (n = 36), HLA matched siblings (n = 31) and unrelated donors (n = 9). Among patients who received a MAC AlloSCT the donor source consisted of UCB (n = 74), HLA-matched siblings (n = 15), and unrelated donors (n = 5).

Engraftment

The time to neutrophil and platelet engraftment in the RTC AlloSCT group was 19.7 6 10.4 days and 43.3 6 37.3 days, respectively. The time to neutrophil and platelet engraftment in the MAC AlloSCT group was 18.4 6 11.3 days and 30.5 6 26.9 days, respectively.

The probability of developing aGVHD grade II-IV was 30.1% (CI95 19.6-40.7) in RTC versus MAC AlloSCT 52.8% (CI95 42.3-63.4) P = .003. The probability of developing cGVHD in RTC versus MAC AlloSCT was 21.6% (CI95 1 0.8-32.4%) versus 15.9% (CI95 6.6%-25.2%) P = .431 (ns).

Kidney injury

Patients were considered to have KI if they had $50% drop in eCrCl when compared to baseline eCrCl performed 2 weeks prior to starting the conditioning regimen. eCrCl 2 weeks prior to starting conditioning regimen between MAC and RTC AlloSCT group was comparable (P = .31). The median eCrCl was 139.8 mL/min/1.72 m2 and mean eCrCl was 145.1 6 47.0 mL/min/1.72 m2, in the MAC AlloSCT group; and median eCrCl was 131.5 mL/min/1.72 m2, and mean eCrCl was 132.5 6 36.2 mL/min/1.72 m2 in the RTC AlloSCT group. Estimated CrCl (mL/min per 1.72 m2) of 71 to 100 was noted in 17% versus

Table 1. Demographics of Pediatric AlloSCT Recipients Who Received RTC versus MAC

Characteristic All MAC RTC P Value

n 5 170 n 5 94 n 5 76

Age (years) 9.5 ± 6.55 8.49 ± 6.33 10.66 ± 6.66 .031

Gender .074

Male 106 (62.3%) 53 (56.4%) 53 (69.7%)

Female 64 (37.7%) 41 (43.6%) 23 (30.3%)

Disease type .230

Malignant 108 (63.2%) 64 (68.1%) 45 (59.2%)

Nonmalignant 62 (36.8%) 30 (31.9%) 31 (40.8%)

Disease status .11

Average-risk group 135 (79.3%) 68 (72%) 67 (88.2)

Poor-risk group 35 (20.7%) 26 (28.0%) 9 (11.8%)

CMV risk status

Donor or recipient seropositive 116 (68.2%) 64 (68.1%) 52 (68.4%) .963

Donor and recipient seronegative 54 (31.8%) 30 (31.9%) 24(31.6%)

Donors

MSD 46 (27%) 15 (16.0%) 31 (40.7%) <.001

MUD 14 (8.2%) 5 (5.3%) 9 (11.8%)

UCB 110 (64.7%) 74 (78.7%) 36 (48.6%)

Pre-AlloSCT estimated CrCl*

#70 1 (0.6%) 1 (1%) 0 .307

71-100 35 (20.6%) 16 (17%) 19 (25%)

>100 134 (78.8%) 78(81.9%) 57 (75%)

Median 133.9 139.8 131.5

Mean 139.5 ± 42.8 145.1 ± 47 132.5 ± 36.2

CMV indicates cytomegalovirus; MSD, matched sibling donor; MUD, matched unrelated donor; UCB, umbilical cord blood; MAC, myeloablative conditioning; RTC, reduced toxicity conditioning. *CrCl = mL/min per 1.72 m2.

25% of patients in the MAC versus RTC AlloSCT group, respectively, and eCrCl >100 was noted in 81.9% versus 75% of patients in the MAC versus RTC AlloSCT group, respectively (Table 1).

Thirty days following AlloSCT, 33% patients developed KI. The incidence of KI at 1 month was 45.7% in MAC AlloSCT versus 17.1% in RTC AlloSCT, P < .0001. However, the incidence of KI at 3 months was 55% (59% in MAC AlloSCT versus 50.7% in RTC AlloSCT, P = .3) (Table 2). The incidence of KI at 6 months was 58% (66.7% in MAC AlloSCT versus 50% in RTC AlloSCT, P = .05), and the incidence of KI at 12 months was 36% (46% in MAC AlloSCT versus 28.9% in RTC AlloSCT, P = .10). The incidence of KI was consistently higher in children who received MAC AlloSCT when compared to RTC AlloSCT during the first year post-AlloSCT. KI was statistically significantly higher at 1 month following AlloSCT. We subsequently performed an analysis comparing the incidence of >75% drop in eCrCL at 1 month between MAC versus RTC AlloSCT; 15/94 (16.0%) MAC AlloSCT and 2/76 (2.6%) RTC AlloSCT patients developed >75% drop in eCrCL from the baseline (P = .004).

During the first year post-AlloSCT, 25/170 (14.7%) required continuous renal replacement therapy (CRRT) with continuous veno-venous hemofiltra-tion. The incidence for CRRT was 17/94 (18%) for MAC AlloSCT and 8/76 (11%) for RTC AlloSCT (P = ns); 24/25 of the patients requiring CRRT ultimately died. Among 17 patients receiving CRRT in the MAC AlloSCT group, 15/17 (88%) deaths were related to TRM and 2/17 (12%) were related to disease relapse. Among 8 patients receiving CRRT in the RTC AlloSCT group, 1 patient survived, and 4/7 (57%) deaths were related to TRM and 3/7 (43%) deaths were disease related.

Table 2. Incidence of KI Following RTC versus MAC in Pediatric AlloSCT Recipients

Time Drop in eCrCL MAC-AlloSCT RTC-AlloSCT P Value

1 month n 5 94 n 5 76 <.0001

3 months <50% $50% 51 (54.3%) 43 (45.7%) n 5 78 63 (82.9%) 13 (17.1%) n 5 73 .306

6 months <50% $50% 32 (41.0%) 46 (59.0%) n 5 63 36 (49.3%) 37 (50.7%) n 5 70 .052

9 months <50% $50% 21 (33.3%) 42 (66.7%) n 5 48 35 (50.0%) 35 (50%) n 5 58 .088

12 months <50% $50% 21 (43.7%) 27 (56.3%) n 5 37 35 (60.3%) 23 (39.7%) n 5 52 .098

<50% $50% 20 (54.0%) 17 (46.0%) 37 (71.1%) 15 (28.9%)

eCrCl indicates estimated creatinine clearance; MAC, myeloablative conditioning; RTC, reduced toxicity conditioning.

Survival analysis

Kidney injury and various other transplant-related factors could lead to mortality in patients following Al-loSCT. We performed univariate analysis for risk factors associated with OS following RTC versus MAC AlloSCT; age, gender, CMV risk status, disease type, and VOD, fungal infections, bacterial and viral infections, and cGVHD were not statistically significant. However, disease status, UCB as a stem cell source, MAC regimen, aGVHD, and KI at 1 month were statistically significant on univariate analysis (Table 3). On multivariate analysis, poor-risk disease status (hazard ratio [HR] = 2.008, 95% confidence interval [CI] 1.261-3.458, P = .0042), UCB as a stem cell source (HR = 2.312, 95% CI 1.194-4.478, P = .0129), MAC regimen (HR = 2.666, 95% CI 1.499-4.743, P = .0008), and KI at 1 month (HR = 2.000, 95%CI 1.230-3.252, P = .0052) were independently associated with poor OS (Table 4).

Table 3. Univariate Analysis for Risk Factors Associated with OS Following RTC versus MAC in Pediatric AlloSCT Recipients

Variable Ratio Confidence Limits P Value

Age 1.009 0.973 1.047 .613

Gender

Female 1

Male 0.788 0.486 1.280 .336

CMV risk status

D or R seropositive 1

D and R seronegative 0.933 0.571 1.525 .782

Malignant disease

Yes 1.150 0.703 1.880 .578

Baseline eCrCI 1.003 0.997 1.008 .394

Disease risk status

Average 1

Poor 2.735 1.673 4.471 <.0001

Yes 1.921 0.696 5.303 .208

Fungal infection

Yes 1.276 0.633 2.574 .496

Stem cell source

MSD/MUD 1

UCB 3.161 1.656 6.034 .0005

Acute GVHD*

Yes 1.652 1.005 2.714 .048

Chronic GVHD*

Yes 0.628 0.225 1.757 .376

Conditioning

MAC 3.839 2.214 6.650 <.0001

1 month

Drop in eCrCl

<50% 1

$50% 2.425 1.507 3.900 .0003

CMV indicates cytomegalovirus; D/R, donor/recipient; VOD, veno-occlusive disease; MSD/MUD, matched sibling donor/matched unrelated donor; UCB, umbilical cord blood; GVHD, graft-versus-host disease. *Time-dependent analysis.

Table 4. Multivariate Analysis for Significant Risk Factors Affecting OS Following RTC versus MAC in Pediatric AlloSCT Recipients

Hazard 95% Hazard Ratio

Variable Ratio Confidence Limits P Value

MAC 2.666 1.499 4.743 .0008

Average 1

Poor 2.088 1.261 3.458 .0042

Donor source

MSD and MUD 1

UCB 2.312 1.194 4.478 .0129

Acute GVHD*

Yes 1.282 1.230 3.252 .3356

1 month

Drop in eCrCl

<50% 1

$50% 2.000 1.230 3.252 .0052

MAC indicates myeloablative conditioning; RTC, reduced toxicity conditioning; MSD, matched sibling donor; MUD, matched unrelated donor; UCB, umbilical cord blood; eCrCl, estimated creatinine clearance. *Time-dependent analysis.

Figure 1 shows significant impact of KI at 1 month on OS. We further analyzed our data to determine that KI was associated with an increase in TRM. In univariate analysis, KI at 1 month was associated with increased risk of TRM, odds ratio 2.065 (CI95 0.9764.368, P = .0578).

Univariate and multivariate analysis for KI at 1 month

We performed a univariate analysis with various factors (age, gender, baseline eCrCl, CMV risk status, fungal infection, disease status, disease type

[malignant versus nonmalignant], VOD, liver dysfunction, foscarnet, UCB, MAC, and stem cell source) that could result in KI at 1 month (Table 5). On univariate analysis, poor disease risk status (odds ratio [OR] = 2.325, 95% CI 1.087-4.972, P = .030), VOD (OR = 8.689, 95% CI 0.948-79.65, P = .056), UCB (OR = 1.782, 95% CI 0.885-3.591, P = .106), and MAC (OR = 4.086, 95% CI 1.985-8.410, P = .0001), were significant risk factors of KI at 1 month post-AlloSCT. On multivariate analysis with the significant risk factors, only MAC (OR = 3.438, 95% CI 1.5947.418, P = .002) remained as significant (Table 6).

DISCUSSION

AlloSCT has curative potential for children with malignant and nonmalignant diseases. However, Al-loSCT is associated with short-term and long-term morbidity and mortality. Acute and chronic KI is an important cause of morbidity and can lead to mortality following AlloSCT. Understanding causes for KI is of paramount importance to prevent KI. Reduction in KI might lead to improvement in AlloSCT outcomes.

There are no uniform definitions and time points for assessment of KI following AlloSCT [21], which makes comparisons of various studies difficult and hampers our ability to perform meaningful comparative analysis. Therefore, there is a need for development of uniform guidelines to assess kidney function following AlloSCT. Schwartz equation is 1 of the most common methods of estimating changes in acute renal function in critically ill pediatric patients [20,22]. Similarly, there are also various definitions used for chronic KI [23-26]. Because we assessed KI longitudinally for 1 year, we used the consistent

Figure 1. Probability of OS

following KI at 1 month versus no KI at 1 month in pediatric allogeneic stem cell transplantation recipients.

Table 5. Univariate Analysis for Risk Factors Associated with KI at 1 Month in Pediatric AlloSCT Recipients

Odds Risk 95% Confidence

Variable Ratio Limits P Value

Age i.0i7 0.969-i.068 .494

Gender

Female i.0

Male i.270 0.65i-2.478 .484

Disease risk status

Average i.0

Poor 2.325 i.087-4.972 .030

Malignancy

Yes i.0

No i.05 0.539-2.044 .886

CMV risk status

D or R seropositive i.0

D and R seronegative 0.679 0.345-i.334 .26i

Stem cell source

MSD/MUD i

UCB i.782 0.885-3.59i .i06

Conditioning

MAC 4.086 i.985-8.4i0 .000i

Baseline eCrCI i.004 0.997-i.0ii .293

Fungal infection

No i.0

Yes i.4i7 0.543-3.693 .476

Yes 8.689 0.948-79.65 .056

Foscarnet

Yes i.i95 0.630-2.268 .586

ALT* $5 x ULN

Yes 0.943 0.36i-2.464 .905

Bilirubin* >2 mg/dL

Yes i.438 0.7i9-2.874 .304

CMV indicates cytomegalovirus; D/R, donor/recipient; VOD, veno-occlusive disease; MSD/MUD, matched sibling donor/matched unrelated donor; UCB, umbilical cord blood; RTC, reduced toxicity conditioning; MAC, myeloablative conditioning; foscarnet therapy in first 30 days post-AlloSCT, ALT, alanine aninotransferase; ULN, upper limit of normal.

*ALTand biliubin values in first 30 days post-AlloSCT.

definition for KI for early and late KI, so that the data could be compared at various time points.

In the current retrospective study, we compared the incidence of KI in children receiving RTC versus MAC AlloSCT over a period of 1 year. One month following AlloSCT, a significantly higher number of children had KI after MAC AlloSCT when compared to RTC AlloSCT. One month following AlloSCT, 33% of patients developed KI. In a study published by Van Why et al. [5], the incidence of acute renal insufficiency in 64 children who underwent MAC AlloSCT was 50%, which is comparable to the incidence of KI in our children following MAC AlloSCT. The incidence of KI in the MAC AlloSCT group in our study at 1, 3, and 6 months was 45.7%, 59%, and 66%, respectively. However, a retrospective study published by Kist-van Holthe et al. showed a 34% incidence of acute renal insufficiency in children (n = 142) following MAC AlloSCT [4]. In a follow-up prospective study from

Table 6. Multivariate Analysis for Risk Factors Associated with KI at 1 Month in Pediatric AlloSCT Recipients

Odds Risk 95% Confidence

Variable Ratio Limits P Value

Disease risk status

Average i

Poor i.760 0.777-3.988 .i75

Yes 4.892 0.5i8-46.i8 .i66

Stem cell source

MSD/MUD i

UCB i.062 0.484-2.33i .88i

Conditioning

MAC 3.438 i.594-7.4i8 .002

VOD indicates veno-occlusive disease; MSD/MUD, matched sibling donor/matched unrelated donor; UCB, umbilical cord blood; RTC, reduced toxicity conditioning; MAC, myeloablative conditioning.

the same group, the incidence of acute renal insufficiency in 66 children following MAC AlloSCT was 21% [3], with the decrease in KI attributed to aggressive monitoring of renal function. Incidence of KI in our study is relatively higher when compared to the Kist-van Holthe et al. study, and 1 probable reason for that could be our aggressive antifungal and antiviral prophylaxis with liposomal amphotericin B and foscarnet, respectively. Based on our analysis, we have modified our antifungal prophylaxis by decreasing the dose (1.5 mg/kg from 3 mg/kg) and duration of liposomal amphotericin-B: after 45 days following AlloSCT liposomal, amphotericin-B is replaced by micafungin [27]. Maribavir or CMX 001 (Chimerix) could be a potential alternative to foscarnet and ganciclovir for CMV prophylaxis in high-risk patients at risk for CMV reactivation and disease.

In our study, the incidence of CRRT was 18% for MAC AlloSCT and 11% for RTC AlloSCT. The reported incidence of CRRT after MAC in pediatric patients is variable between 5% and 24%. The incidence of CRRT in pediatric patients after RTC AlloSCT has never been reported. However, in adult RTC AlloSCT the incidence of CRRT is 3% to 4% [28]. In our study, OS of children requiring CRRT was dismal (4%). Various single-center studies in children and adults have reported OS rates between 0% and 45% in patients requiring CRRT after MAC AlloSCT [29-31].

Recently, a report from the prospective pediatric CRRT registry group demonstrated that 45% of children post-AlloSCT survived to get discharged from the intensive care unit [29]. These results are encouraging, and in this study, 88% of the patients received CRRT because of fluid overload. It seems that investigators in this observational multicenter study were initiating CRRT early in order to prevent severe fluid overload. In a recent observational study by Sutherland et al. [32], children who developed $20% fluid overload at CRRT initiation had significantly higher

mortality (65.6%) than those who had 10% to 20% fluid overload (43.1%) and those with <10% fluid overload (29.4%). This study also demonstrated a 3% increase in mortality for each 1% increase in severity of fluid overload and patients with $20% fluid overload had an adjusted mortality odds ratio of 8.5 (95% CI, 2.8-25.7). Aggressive and appropriate management of fluid balance is critical for children following AlloSCT.

Kidney injury is not only associated with TRM, but also associated with inferior OS. Kidney injury after RTC and MAC AlloSCT is an independent predictor of relapse and TRM. In our study, at 3 years, OS was significantly higher in patients without KI at 1 month versus KI at 1 month (75% versus 45%, P < .0001). In a study by Lopes et al. [33], adult patients with KI following RTC AlloSCT experienced poorer long-term survival. In this study 5-year OS of KI patients was 41.6%, compared with 67.1% for those who did not develop KI (P 5 .028). In another study, Kersting et al. [34] reported similar findings that adult patients following RTC AlloSCT with KI had significantly higher mortality compared to no acute KI (P = .006). In a meta-analysis by Parikh et al. [35], in adult patients after MAC AlloSCT there was doubling in relative risk (RR = 2.31; CI [1.473.65], P < .01) of death in patients who had at least a doubling of serum creatinine from baseline. Data comparing impact of KI versus no KI in children following RTC or MAC AlloSCT is very limited, and our study highlights the critical nature of impact of KI on TRM and OS.

When assessing the risk of KI in children undergoing AlloSCT, one needs to consider various other factors beyond the conditioning regimen. Compared with heavily pretreated patients, patients who undergo AlloSCT for nonmalignant disease potentially have a lower risk of KI post-AlloSCT because of an intact immune system, absence of chemotherapy-related KI, and lack of significant end-organ damage. Donor source also may have an impact on the risk of KI. Compared to children undergoing AlloSCT with UCB (delayed immune reconstitution) or an unrelated donor (increased risk of GVHD), children undergoing HLA-matched sibling AlloSCT have a lower risk of GVHD and relatively faster immune reconstitution, which may lead to a decreased incidence of fungal and viral infections. The infections and/or its treatment may lead to KI. In univariate analysis for risk factors for KI, poor risk disease status, VOD, UCB transplant, and MAC were statistically significant. In our multivariate analysis, we controlled for individual risk factors, and were unable to demonstrate the individual impact of these factors on the risk of KI at 1 month post-AlloSCT other than MAC, that was a significant independent risk factor for KI at 1 month. We also performed a multivariate analysis for significant

risk factors affecting OS following AlloSCT. Myeloa-blative conditioning, poor-risk disease status, UCB transplant, and KI at 1 month were independent predictors of poor survival. However, the primary goal of this study was to compare KI between RTC versus MAC AlloSCT and not the OS.

The major limitation of this retrospective study is the small sample size in both groups, lack of prospective randomization, similar donor sources, and disease risk status. Because subjects were heterogeneous in terms of underlying disease, conditioning regimen, and hematopoietic stem cell donor sources, these preliminary analyses should be replicated in additional studies of children with more homogenous diagnoses and cell sources. Despite these limitations, we believe that the central conclusions are valid. The fact that this study was conducted at a single institution ensures that consistent policies and protocols were followed in terms of anti-infection prophylaxis, isolation precautions, and diagnosis of GVHD, VOD, and infections.

In summary, a significant number of children sustain KI following AlloSCT. MAC AlloSCT is a significant risk factor for KI at 1 month. Kidney injury may lead to an increase in TRM and impact OS. Patients undergoing AlloSCT, especially MAC AlloSCT, should have close monitoring of kidney function, judicious use of nephrotoxic medication and drug level monitoring, and early intervention of fluid overload. There is need for prospective studies for prevention of KI, validation of KI biomarkers, and role of early initiation of CRRT.

ACKNOWLEDGMENTS

This work was presented in part at the American Society of Blood and Marrow Transplantation (ASBMT) Meeting, Feb 2010, Orlando, Florida. This work was supported in part by grants from the Pediatric Cancer Research Foundation, Andrew J. Gargiso Foundation, Marisa Fund, Sonia Scaramella Fund, Britanny Baron Fund, Paul J. Luisi Foundation, and Dream Discovery and Cure Fund.

Financial disclosure: All authors declare no conflicts of interest.

REFERENCES

1. Thomas ED. Karnofsky Memorial Lecture. Marrow transplantation for malignant diseases. J Clin Oncol. 1983;1:517-531.

2. Satwani P, Harrison L, Morris E, Del Toro G, Cairo MS. Reduced-intensity allogeneic stem cell transplantation in adults and children with malignant and nonmalignant diseases: end of the beginning and future challenges. Biol Blood Marrow Transplant. 2005;11:403-422.

3. Kist-van Holthe JE, Goedvolk CA, Brand R, et al. Prospective study of renal insufficiency after bone marrow transplantation. Pediatr Nephrol. 2002;17:1032-1037.

4. Kist-van Holthe JE, van Zwet JM, Brand R, van Weel MH, Vossen JM, van der Heijden AJ. Bone marrow transplantation in children: consequences for renal function shortly after and 1 year post-BMT. Bone Marrow Transplant. 1998;22:559-564.

5. Van Why SK, Friedman AL, Wei LJ, Hong R. Renal insufficiency after bone marrow transplantation in children. Bone Marrow Transplant. 1991;7:383-388.

6. Tarbell NJ, Guinan EC, Chin L, Mauch P, Weinstein HJ. Renal insufficiency after total body irradiation for pediatric bone marrow transplantation. Radiother Oncol. 1990;18(Suppl 1):139-142.

7. SchrierRW, ParikhCR. Comparison of renal injury in myeloabla-tive autologous, myeloablative allogeneic and non-myeloablative allogeneic haematopoietic cell transplantation. Nephrol Dial Transplant. 2005;20:678-683.

8. Hingorani S, Guthrie KA, Schoch G, Weiss NS, McDonald GB. Chronic kidney disease in long-term survivors of hematopoietic cell transplant. Bone Marrow Transplant. 2007;39:223-229.

9. Cohen EP, Piering WF, Kabler-Babbitt C, Moulder JE. Endstage renal disease (ESRD) after bone marrow transplantation: poor survival compared to other causes of ESRD. Nephron. 1998;79:408-412.

10. Hoffmeister PA, Hingorani SR, Storer BE, Baker KS, Sanders JE. Hypertension in long-term survivors of pediatric hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2010;16:515-524.

11. Bradley MB, Satwani P, Baldinger L, et al. Reduced intensity al-logeneic umbilical cord blood transplantation in children and adolescent recipients with malignant and non-malignant diseases. Bone Marrow Transplant. 2007;40:621-631.

12. Satwani P, Baldinger L, Freedman J, et al. Incidence of viral and fungal infections following busulfan-based reduced-intensity versus myeloablative conditioning in pediatric allogeneic stem cell transplantation recipients. Biol Blood Marrow Transplant. 2009;15:1587-1595.

13. Bhatia M, Militano O, Jin Z, et al. An age-dependent pharmaco-kinetic study of intravenous and oral mycophenolate mofetil in combination with tacrolimus for GVHD prophylaxis in pediat-ric allogeneic stem cell transplantation recipients. Biol Blood Marrow Transplant. 2010;16:333-343.

14. Osunkwo I, Bessmertny O, Harrison L, et al. A pilot study of tacrolimus and mycophenolate mofetil graft-versus-host disease prophylaxis in childhood and adolescent allogeneic stem cell transplant recipients. Biol Blood Marrow Transplant. 2004;10: 246-258.

15. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation. 1974;18: 295-304.

16. Waxman IM, Militano O, Baldinger L, et al. Sequential administration of sargramostim and filgrastim in pediatric allogeneic stem cell transplantation recipients undergoing myeloablative conditioning. Pediatr Transplant. 2009;13:464-474.

17. Roman E, Osunkwo I, Militano O, Cooney E, van de Ven C, Cairo MS. Liposomal amphotericin B prophylaxis of invasive mold infections in children post allogeneic stem cell transplantation. Pediatr Blood Cancer. 2008;50:325-330.

18. Shereck EB, Cooney E, van de Ven C, Della-Lotta P, Cairo MS. A pilot phase II study of alternate day ganciclovir and foscarnet in preventing cytomegalovirus (CMV) infections in at-risk pedi-atric and adolescent allogeneic stem cell transplant recipients. Pediatr Blood Cancer. 2007;49:306-312.

19. Schwartz GJ, Brion LP, Spitzer A. The use of plasma creatinine concentration for estimating glomerular filtration rate in infants,

children, and adolescents. Pediatr Clin North Am. 1987;34: 571-590.

20. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care. 2004;8: R204-R212.

21. Hahn T, Yao S, Dunford LM, et al. A comparison of measured creatinine clearance versus calculated glomerular filtration rate for assessment of renal function before autologous and allogeneic BMT. Biol Blood Marrow Transplant. 2009;15: 574-579.

22. Goldstein SL. Kidney function assessment in the critically ill child: is it time to leave creatinine behind? Crit Care. 2007;11: 141.

23. Cohen EP, Lawton CA, Moulder JE, Becker CG, Ash RC. Clinical course of late-onset bone marrow transplant nephropathy. Nephron. 1993;64:626-635.

24. Kist-van Holthe JE, Bresters D, Ahmed-Ousenkova YM, et al. Long-term renal function after hemopoietic stem cell transplantation in children. Bone Marrow Transplant. 2005;36:605-610.

25. Leblond V, Sutton L, Jacquiaud C, et al. Evaluation of renal function in 60 long-term survivors of bone marrow transplantation. J Am Soc Nephrol. 1995;6:1661-1665.

26. Patzer L, Ringelmann F, Kentouche K, et al. Renal function in long-term survivors of stem cell transplantation in childhood. A prospective trial. Bone Marrow Transplant. 2001;27:319-327.

27. El-Mallawany N, George D, Jin Z, et al. Adenovirus (ADV) infection in pediatric allogeneic stem cell transplantation (AlloSCT) recipients is a major independent factor for significantly increasing the risk of treatment related mortality. Biol BloodMarrow Transplant. 2010;16:S258.

28. Parikh CR, Yarlagadda SG, Storer B, Sorror M, Storb R, Sandmaier B. Impact of acute kidney injury on long-term mortality after nonmyeloablative hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2008;14:309-315.

29. Flores FX, Brophy PD, Symons JM, et al. Continuous renal replacement therapy (CRRT) after stem cell transplantation. A report from the prospective pediatric CRRT Registry Group. Pediatr Nephrol. 2008;23:625-630.

30. Gruss E, Bernis C, Tomas JF, et al. Acute renal failure in patients following bone marrow transplantation: prevalence, risk factors and outcome. Am J Nephrol. 1995;15:473-479.

31. Lane PH, Mauer SM, Blazar BR, Ramsay NK, Kashtan CE. Outcome of dialysis for acute renal failure in pediatric bone marrow transplant patients. Bone Marrow Transplant. 1994;13: 613-617.

32. Sutherland SM, Zappitelli M, Alexander SR, et al. Fluid overload and mortality in children receiving continuous renal replacement therapy: the prospective pediatric continuous renal replacement therapy registry. Am J Kidney Dis. 2010;55: 316-325.

33. Lopes JA, Jorge S, Goncalves S, et al. Contemporary analysis of the influence of acute kidney injury (AKI) after myeloablative hematopoietic cell transplantation on long-term patient's survival. Bone Marrow Transplant. 2008;42:139-141.

34. Kersting S, Dorp SV, Theobald M, Verdonck LF. Acute renal failure after nonmyeloablative stem cell transplantation in adults. Biol Blood Marrow Transplant. 2008;14:125-131.

35. Parikh CR, McSweeney P, Schrier RW. Acute renal failure independently predicts mortality after myeloablative allogeneic he-matopoietic cell transplant. Kidney Int. 2005;67:1999-2005.