Third-Party BK Virus-Specific Cytotoxic T Lymphocyte Therapy for Hemorrhagic Cystitis Following Allotransplantation

Abstract

Purpose: BK virus-associated hemorrhagic cystitis (BKV-HC) is a common complication of allogenic hematopoietic stem cell transplantation (AHSCT), particularly in recipients of alternative donor transplants, which are being performed in increasing numbers. BKV-HC typically results in painful hematuria, urinary obstruction, and renal dysfunction, without a definitive therapeutic option.

Methods: We performed a clinical trial (ClinicalTrials.gov identifier: NCT02479698) to assess the feasibility, safety, and efficacy of administering most closely HLA-matched third-party BKV-specific cytotoxic T lymphocytes (CTLs), generated from 26 healthy donors and banked for off-the-shelf use. The cells were infused into 59 patients who developed BKV-HC following AHSCT. Comprehensive clinical assessments and correlative studies were performed.

Results: Response to BKV-CTL infusion was rapid; the day 14 overall response rate was 67.7% (40 of 59 evaluable patients), which increased to 81.6% among evaluable patients at day 45 (40 of 49 evaluable patients). No patient lost a previously achieved response. There were no cases of de novo grade 3 or 4 graft-versus-host disease, graft failure, or infusion-related toxicities. BKV-CTLs were identified in patient blood samples up to 3 months postinfusion and their in vivo expansion predicted for clinical response. A matched-pair analysis revealed that, compared with standard of care, after accounting for prognostic covariate effects, treatment with BKV-CTLs resulted in higher probabilities of response at all follow-up timepoints as well as significantly lower transfusion requirement.

Conclusion: Off-the-shelf BKV-CTLs are a safe and effective therapy for the management of patients with BKV-HC after AHSCT.

FIG 1.

Characteristics of BKV-CTLs following ex vivo expansion. (A and B) Cytokine production (IFN-γ, TNF-α, and IL-2) and degranulation (CD107a) of BKV-CTLs following 2 weeks of ex vivo expansion with single BKV pepmix (BKV alone) or with a multivirus pepmix composed of CMV, adenovirus-hexon, and BKV. (A) Representative FACS plots from a single donor in the CD4 (upper panels) and CD8 (lower panels) compartments. (B) Bar graphs summarizing data from three different donors in the CD4 (left panel) and CD8 (right panel) compartments; expansion with BKV pepmix alone represented by the blue bars, and expansion with multivirus pepmix represented by the red bars. P values are indicated at the top of each graph. (C) Bar graph showing the distribution of naive, CM (CD62L+ CD45RA−), EM (CD62L-CD45RA−), and TDEM (CD45RA+ CD62L−) in the CD4 (blue bars) and CD8 (red bars) from donor VSTs (n = 9) generated using the single virus (BKV pepmix) approach. (D) Pie chart showing the distribution of single (blue), dual (red), or triple (green) cytokine producing (IFN-γ, TNF-α, and IL-2) expanded BKV-CTLs (n = 8) generated using the single virus approach following ex vivo stimulation with BKV pepmix. BKV, BK virus; CM, central memory; CMV, cytomegalovirus; CTL, cytotoxic T lymphocyte; EM, effector memory; FACS, fluorescence-activated cell sorting; IFN, interferon; IL-2, interleukin-2; ns, not significant; TDEM, terminally differentiated effector memory; TNF, tumor necrosis factor; VST, viral-specific T cells.

FIG 2.

Patient response at different time points after BKV-CTL infusion. Bar graph showing the percentage of response, CR in green, PR in red, and NR in blue, at the different time points following infusion. BKV, BK virus; CR, complete response; CTL, cytotoxic T lymphocyte; NR, nonresponder; PR, partial response.

FIG 3.

Viral response in patients with BKV-HC following BKV-CTL infusion. Plots showing the (A) viral load and (B) the log reduction in BKV viral load in urine (measured by qPCR) after infusion of CTLs in patients who responded by day 45 (CR plus PR) compared with nonresponders. Statistical significance is indicated as **P ≤ .01 and ***P ≤ .001. BKV, BK virus; BKV-HC, BKV-hemorrhagic cystitis; CR, complete response; CTL, cytotoxic T lymphocyte; PR, partial response; qPCR, quantitative polymerase chain reaction.

FIG 4.

Expansion of BKV-reactive T cells following adoptive infusion. (A) Bar graph showing the mean absolute number of CD4 (blue bars) and CD8 (red bars) BKV-reactive T cells in the responders versus nonresponders over time following infusion (n = 32). BKV-specific T-cell responses were detectable for up to 3 months postinfusion. (B) Bar graphs showing the distribution of T-cell subsets among BKV-CTLs in the CD4 compartment (left panel) and CD8 compartment (right panel) over time following infusion. Bars show mean absolute counts and lines indicate the standard deviation. The BKV reactive CD4 and CD8 T cells composed mostly of effector memory (CD62L-CD45RA) with a minority of central memory (CD62L+ CD45RA−) cells (n = 32). (C) Bar graphs showing HLA flow chimerism performed on the peripheral blood from three patients by gating on IFN-γ and IL-2+ BKV-CTLs and using HLA antibodies that could distinguish the third-party BKV-CTL donor from recipient (n = 3). Each bar represents the percentages of BKV-reactive T cells (y-axis) from the third-party donor (blue) and recipient (red) over time after BKV-CTL infusion. BKV, BK virus; CM, central memory; CTL, cytotoxic T lymphocyte; EM, effector memory; IFN, interferon; IL-2, interleukin-2; TDEM, terminally differentiated effector memory.