Systematic Review of Risk factors and Incidence of Acute Kidney Injury Among Patients Treated with CAR-T Cell Therapies

Swetha R. Kanduri, Wisit Cheungpasitporn, Charat Thongprayoon, Tananchai Petnak, Yi Lin, Karthik Kovvuru, Sandhya Manohar, Kianoush Kashani and Sandra M. Herrmann Department of Medicine, Ochsner Medical Center, New Orleans, Louisiana, USA; Division of Nephrology and Hypertension, Department of Medicine, Mayo Clinic, Rochester, Minnesota, USA; Division of Pulmonary and Pulmonary Critical Care Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand; Division of Pulmonary and Critical Care Medicine, Department of Medicine, Mayo Clinic, Rochester, Minnesota, USA; and Division of Hematology, Department of Medicine, Mayo Clinic, Rochester, Minnesota, USA

I mmunotherapy has grown significantly in the management of hematological and solid organ tumors. Indepth knowledge of T-cell pathways has led to the development of adoptive cell transfer techniques and subsequent evolution of chimeric antigen receptor Tcell (CAR-T) therapies. 1 CAR-T cells are biologically engineered cells with a CAR receptor that recognizes a tumor antigen. T-cell receptors are equipped with extracellular domain and variable costimulatory domains yielding divergent T-cell response. 2 Along with their utility among refractory B cell lymphomas, adult diffuse large B-cell lymphoma (DLBCL), and pediatric acute lymphoblastic leukemia (ALL), CAR-T therapies are now being explored in the management of solid organ tumors and multiple myeloma. 3 Cytokine release syndrome (CRS), macrophage activation syndrome (MAS)/ hemophagocytic lymphohistiocytosis (HLH), and immune effector cellÀassociated neurotoxicity syndrome (ICANS) are among the most alarming complications of CAR-T cells. S1 The incidence of acute kidney injury (AKI) reported in the literature among patients with CAR-T therapies varies from 5% to 33%. S1ÀS22 The mechanism of AKI in patients with CAR-T cell therapy is not completely understood. However, it is proposed that AKI could be secondary to inflammation associated with cytokine release, potentially leading to acute tubular injury. S19 Even though high-grade CRS is associated with AKI, S18 the correlation between severe AKI requiring renal replacement therapies (AKI-RRT) and high grades of CRS is not widely known. We conducted a current systematic review of literature and meta-analysis of articles reporting the incidence of AKI among patients receiving CAR-T therapies and the correlation between severity of CRS and AKI-RRT incidence.

MATERIALS AND METHODS
We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) S23 statement in conducting systematic review. EMBASE, Cochrane, and Ovid MEDLINE databases were systematically searched from database inception through May 2020. Full details are provided in the Supplementary Material.

RESULTS
A total of 481 potentially relevant articles were identified and screened. In all, 32 articles were assessed in detail, of which 22 cohort studies S1ÀS22 with 3376 patients (Table 1) were included in our systematic review ( Figure 1).

Incidence of AKI Among Adults
Subgroup analysis among adults treated with CAR-T therapies resulted in the pooled estimated incidence of AKI at 17.0% (95% CI ¼ 12.8%À22.2%, I2 ¼ 73%) (Supplementary Figure S2), and the incidence of AKI-RRT was 2.9% (95% CI ¼ 0.9%À9.4%, I2 ¼ 41%) (Supplementary Figure S3). Upon analyzing the results using Kidney Disease: Improving Global Outcomes 32 Potentially relevant articles included for full-length article review 10 Articles were excluded because they did not provide data on outcomes of interest

Study name Statistics for each study Event rate and 95% CI
(KDIGO) criteria, we report that the pooled estimated AKI incidence among adults was 24.1% (95% CI ¼ 14.9%À36.5%, I2 ¼ 49%) (Supplementary Figure S4). The incidence of AKI among adults on CAR-T therapies was slightly higher by KDIGO criteria as compared to overall AKI incidence.

Publication Bias Funnel plots (Supplementary Figures S10ÀS13) and
Egger regression asymmetry tests were performed to evaluate for publication bias. We found no significant publication bias in the meta-analysis evaluating AKI incidence (P $ 0.05 for all analysis).

DISCUSSION
In the current era, the utility of CAR-T therapies has extended widely among patients with various cancers. In our study, the pooled incidence of any AKI among all patients with CAR-T therapies was 19%, and the incidence of AKI-RRT was 4%. Upon analyzing studies that defined AKI by KDIGO criteria, the pooled incidence of AKI among adults was slightly higher at 24%. This observation could be secondary to using a definition with a higher sensitivity in identifying patients with AKI. Interestingly, most patients experienced only mild AKI, with a rise in serum creatinine $0.3 mg/dl to1.5 times baseline creatinine, and a smaller proportion of patients reached AKI stage 3 or AKI-RRT. Our findings are in concordance with the results of the largest published studies on AKI incidence among patients with CAR-T therapies. 4,S14, S18, S19 We found that the pooled incidence of CRS among patients with CAR-T therapies was 75%, which is in agreement with major published studies. S1,S3,S6ÀS9,S13,S17ÀS20 Our analysis reports a strong correlation between AKI-RRT incidence and severe CRS. In addition, our study indicated a higher incidence of AKI and AKI-RRT in the pediatric population than in older adults (22.5% vs. 17% and 6% vs. 2.9%, respectively). The current observation could be secondary to increased CRS incidence among the pediatric population as reported by Lee et al. and Maulde et al. S3,S8 The underlying mechanisms for a higher incidence of CRS among the pediatric population are unclear. However, it is hypothesized to be related to a higher dose, excessive tumor burden, and association of ALL with higher blast count leading to increased CRS risk, 5 and, finally, immaturity of the immune system among pediatric patients and young adults.
Our study has several strengths. It is the first systematic review looking at the pooled incidence of AKI among adults and pediatric subgroups receiving CAR-T therapies and reporting a correlation between incidence of AKI-RRT and severe CRS. However, our systematic review also has a few limitations. We included a few retrospective cohort studies with smaller sample sizes that reported associations rather than a causal relationship between AKI incidence and CAR-T cell therapies. There is modest heterogeneity among various studies in defining AKI. In addition, we do not have data on baseline kidney function except for a few studies S18,S19 that included patients with median and mean baseline GFR >60 ml/min per m 2 respectively.
In conclusion, we report that AKI incidence among patients with CAR-T therapies varies among pediatric and adult cohorts. We additionally report a strong correlation between the severity of CRS and AKI-RRT. As AKI-RRT is associated with higher mortality and morbidity, it would be helpful to undertake additional preventive strategies along with the addition of tocilizumab in patients at high risk for severe CRS to mitigate the risk of AKI. Future research on identifying models or biomarkers that could help to predict AKI among patients on CAR-T therapies and the impact of preventive measures on the incidence of AKI, AKI-RRT, and severe CRS is warranted.

DISCLOSURE
All the authors declared no competing interests. and by a Mary Kathryn and Michael B. Panitch Career Development Award.

AUTHOR CONTRIBUTIONS
SK, TP, and SM performed an independent literature search using search terms. WC, CT, and KK collected the data. SK, SH, and KK interpreted the data. SK, WC, CT, TP, YL, KK, SM, KK, and SH prepared the manuscript and approved the current version of the manuscript.

SUPPLEMENTARY MATERIAL
Supplementary File (PDF) Supplementary Methods Figure S1. Incidence of severe AKI or AKI-RRT among patients on CAR-T therapies. Figure S2. Incidence of overall AKI among adults on CAR-T therapies. Figure S3. Incidence of AKI-RRT among adults on CAR-T therapies. Figure S4. Incidence of overall AKI (only KDIGO definition) among adults on CAR-T therapies. Figure S5. Incidence of overall AKI among pediatric patients and young adults on CAR-T therapies. Figure S6. Incidence of severe AKI-RRT among pediatric patients and young adults. Figure S7. Incidence of overall AKI (only KDIGO definition) among pediatric patients and young adults. Figure S8. Overall incidence of CRS among patients on CAR-T therapies. Figure S9. Relation between CRS severity of incidence of AKI-RRT among patients on CAR-T therapies. Figure S10. Funnel plot of adults with AKI on CAR-T therapies. Figure S11. Funnel plot of adults with severe AKI-RRT on CAR-T therapies. Figure S12. Funnel plot of pediatric patients and young adults with AKI on CAR-T therapies. Figure S13. Funnel plot of pediatric and young adult patients with severe AKI-RRT on CAR-T therapies. Supplementary References