Infusion of donor-derived CD19-redirected virus-specific T cells for B-cell malignancies relapsed after allogeneic stem cell transplant: a phase 1 study

Abstract

Autologous T cells expressing a CD19-specific chimeric antigen receptor (CD19.CAR) are active against B-cell malignancies, but it is unknown whether allogeneic CD19.CAR T cells are safe or effective. After allogeneic hematopoietic stem cell transplantation (HSCT), infused donor-derived virus-specific T cells (VSTs) expand in vivo, persist long term, and display antiviral activity without inducing graft-vs-host disease; therefore, we determined whether donor VSTs, engineered to express CD19.CAR, retained the characteristics of nonmanipulated allogeneic VSTs while gaining antitumor activity. We treated 8 patients with allogeneic (donor-derived) CD19.CAR-VSTs 3 months to 13 years after HSCT. There were no infusion-related toxicities. VSTs persisted for a median of 8 weeks in blood and up to 9 weeks at disease sites. Objective antitumor activity was evident in 2 of 6 patients with relapsed disease during the period of CD19.CAR-VST persistence, whereas 2 patients who received cells while in remission remain disease free. In 2 of 3 patients with viral reactivation, donor CD19.CAR-VSTs expanded concomitantly with VSTs. Hence CD19.CAR-VSTs display antitumor activity and, because their number may be increased in the presence of viral stimuli, earlier treatment post-HSCT (when lymphodepletion is greater and the incidence of viral infection is higher) or planned vaccination with viral antigens may enhance disease control.

Trial registration: ClinicalTrials.gov NCT00840853.

Figures

Figure 1

Characteristics of CD19.CAR-VSTs. (A) Transduction of VSTs with the CD19.CAR is shown as both percentages of T cells on flow cytometry and number of copies on Q-PCR. Each symbol represents a patient, with the marker style corresponding to the same patient across the characteristics measured. (B) Phenotypic characteristics of CD19.CAR-VSTs as measured by flow cytometry. (C) Cytotoxic specificity of CD19.CAR-VSTs as measured by 51Cr release assays against different targets: control PHA blasts, PHA blasts loaded with virus-derived peptides or CD19+ cells. (D) IFNγ ELISpot of CD19.CAR-VSTs against different viral-antigen–loaded and CD19+ target cells.

Figure 2

Outcome after CD19.CAR-VST infusions in patients infused with disease. Panels A-F represent single patients. Panels A-C illustrate patients with sustained objective clinical response, and panels D-F illustrate patients with transient or no response. In all panels, black arrows denote the time of CD19.CAR-VST infusions, black circles indicate the detection of CD19.CAR-VSTs in the PB by Q-PCR, and red triangles represent the values of normal B-cell counts in the PB. (A) Data from patient 1. Detection of bcr/abl transcripts are indicated as light gray squares. The gray arrow indicates the increase in dose of dasatinib (TKI). (B-C) Data from patients 2 and 3, respectively. In both subjects, circulating B-CLL cells (CD19+CD5+) were measured by flow cytometry and are shown as light gray squares. (D-F) Data from patients 4, 5, and 6, respectively. Leukemic cells, measured in the PB by flow cytometry, are represented as light gray squares. The star indicates the time when patients received additional/alternative treatments or were considered out of study.

Figure 3

Outcome after CD19.CAR-VST infusions without evidence of disease. Each panel represents a single patient. In all panels, black arrows denote the time of CD19.CAR-VST infusions, black circles indicate the detection of CD19.CAR-VSTs in the PB by Q-PCR, and red triangles represent the values of normal B-cell counts in the PB. (A-B) Data from patients 7 and 8 who received CD19.CAR-VSTs 8 months and 3 months after allogeneic HSCT, respectively.

Figure 4

In vivo trafficking of CD19.CAR-VSTs. (A) CD19.CAR-VSTs accumulate at the sites of disease. Immunohistochemistry stain of B-CLL cells (top row) and CD3+ T cells (bottom row) from patient 3. (B) Summary table documenting the presence of detectable CD19.CAR-VSTs at the sites of disease as assessed by Q-PCR compared with concomitant PB samples measured as copies per 1000 ng of DNA.

Figure 5

Viral reactivation. (A) AdV reactivation and monitoring of the specific immune response in patient 1. Detection of AdV-specific T-cell responses by IFNγ ELISpot (striped bars) and CD19.CAR transgene by Q-PCR (black circles) are illustrated. The light gray diamond denotes the time of adenovirus antigen detection in stool. (B-C) EBV reactivation and immune responses in patients 7 and 8, respectively. Detection of EBV-DNA viral load by Q-PCR (light gray squares), CD19.CAR transgene by Q-PCR (black circles), and EBV-specific T-cell responses (striped bars) by IFNγ ELISpot in the PB are illustrated. Arrows denote time of CD19.CAR-VSTs infusions.