Long-term control of recurrent or refractory viral infections after allogeneic HSCT with third-party virus-specific T cells

Abstract

Donor-derived adoptive T-cell therapy is a safe and effective treatment of viral infection posttransplant, but it is limited by donor serostatus and availability and by its personalized nature. Off-the-shelf, third-party virus-specific T cells (VSTs) appear promising, but the long-term safety and durability of responses have yet to be established. We conducted a prospective study of 30 allogeneic hemopoietic stem cell transplant (HSCT) patients with persistent or recurrent cytomegalovirus (CMV) (n = 28), Epstein-Barr virus (n = 1), or adenovirus (n = 1) after standard therapy. Patients were treated with infusions of partially HLA-matched, third-party, ex vivo-expanded VSTs (total = 50 infusions) at a median of 75 days post-HSCT (range, 37 to 349 days). Safety, viral dynamics, and immune recovery were monitored for 12 months. Infusions were safe and well tolerated. Acute graft versus host disease occurred in 2 patients, despite a median HLA match between VSTs and the recipient of 2 of 6 antigens. At 12 months, the cumulative incidence of overall response was 93%. Virological control was durable in the majority of patients; the reintroduction of antiviral therapy after the final infusion occurred in 5 patients. CMV-specific T-cell immunity rose significantly and coincided with a rise in CD8+ terminal effector cells. PD-1 expression was elevated on CD8+ lymphocytes before the administration of third-party T cells and remained elevated at the time of viral control. Third-party VSTs show prolonged benefit, with virological control achieved in association with the recovery of CD8+ effector T cells possibly facilitated by VST infusion. This trial was registered at www.clinicaltrials.gov as #NCT02779439 and www.anzctr.org.au as #ACTRN12613000603718.

Conflict of interest statement

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Figures

Graphical abstract

Figure 1.

Cumulative incidence of response at 12 months post–VST infusion. (A) The cumulative time to best overall response (OR) (either CR or PR) for all patients. (B) The cumulative time to CR for all patients.

Figure 2.

Outcome of CMV enterocolitis in patient 5 post–CMV-specific VST infusion. (A) Viral load, lymphocyte count, and colonic biopsy in patient 5, who was treated with a single VST infusion. The patient had persistent symptomatic biopsy–proven CMV enterocolitis after a month of antiviral therapy. Post–VST infusion, the symptoms resolved, the patient developed an absolute lymphocytosis, and viral load fluctuated between negative and below the level of quantitation (<150 cp/mL). (B) CMV pp65-specific T-cell response as measured by IFN-γ ELIspot. (C-D) Photomicrograph of colonic biopsy with CMV immunostain pre–VST (C) and post–VST (D) infusion (original magnification ×200, Olympus microscope, model BX43). LLQ, lower limit of quantitation.

Figure 3.

Antiviral therapy pre– and post–final VST infusion. The numbers to the left of the red bars refer to the patient number. The red bars on the left show the cumulative days of antiviral therapy administered before the final infusion (including therapy before the first VST infusion). The blue bars to the right show the duration and number of blocks of antiviral therapy administered over the course of follow-up after the final VST infusion.

Figure 4.

CMV response and reactivations post–final VST infusion. CMV viral load (cp/mL) after final VST infusion in the 28 patients treated for CMV reactivation.

Figure 5.

T-cell subset responses post–VST infusion. (A) Median cell number for CD3+, CD4+, and CD8+ T cells for time points post–first VST infusion (× 109/L). (B) Median cell number for CD8+ T-cell memory subsets defined as T terminal effector (CD45RA+62L–), T naive (CD45RA+62L+), T central memory (CD45RA–62L+), and T effector memory (CD45RA–62L–). (C) Median cell number for the CD4+ T-cell memory subsets. (D) Percentage of CD8+ T cells expressing PD-1. Bars and lower and upper whiskers represent the median, 25th, and 75th percentiles, respectively. The results for healthy individuals at steady state are indicated by the dashed line (median) and gray shading (range).

Figure 6.

Correlation between the time to CR with timing of first VST infusion post-HSCT. To assess whether virological responses may be explained by natural immune recovery post-HSCT, linear regression and Pearson’s correlation were performed to determine any significant relationship between the timing of the response and the timing of first VST infusion post-HSCT in the 23 patients who achieved CR. No correlation was observed.

Figure 7.

Peak CMV pp65-specific T-cell responses pre– and post–VST infusion. Peak SFC count as measured by IFN-γ ELIspot after the first VST infusion in 23 patients. Bars and lower and upper whiskers represent the median, 25th, and 75th percentiles, respectively. Significance was determined by the Wilcoxon matched-pairs signed rank test.