Safety and Clinical Efficacy of Rapidly-Generated Virus-Specific T Cells with Activity Against Adv, EBV, CMV, HHV6 and BK Virus Administered after Allogeneic Hematopoietic Stem Cell Transplant

We have previously shown that small numbers of ex vivo-expanded trivirus-specific T-cells (tVSTs) targeting EBV, CMV, and Adv are safe, effective and protective in vivo. However, broader implementation is limited by a manufacturing process that is prolonged (8-12wks), complex and requires infectious virus (EBV) and vector (Adv). Moreover, antigenic competition limits extension to additional viruses. We have now made T-cell lines with activity against 5 common post-transplant viruses (EBV, CMV, Adv, BK, HHV6), using a simplified 10-day procedure that excludes viral components, and show that a single line can be clinically effective against multiple viruses. With NHLBI-PACT support 45 clinical-grade pentavalent (p)VSTs have been generated. By exposing 30x10⁶ PBMCs to overlapping peptide libraries spanning Adv (Hexon, Penton), CMV (pp65, IE1), EBV (LMP2, EBNA1, BZLF1), BK (Large T, VP1) and HHV6 (U11, U14, U90) antigens we expanded a median of 345x10⁶ cells (range 99-825x10⁶) over 9-11 days. The lines were polyclonal, comprising both CD4+ (58±5%) and CD8+ (34±5%) cells, that expressed activation and memory markers. pVST specificity was dependent on the donor’s prior viral exposure; 42/45 lines had Adv activity (Hexon: 453±159; Penton: 367±199 SFC/2x10⁵), 24/45 against CMV (IE1: 315±135; pp65: 969±442), 34/45 against EBV (LMP2: 145±78; EBNA1: 115±46; BZLF1: 106±78), 24/45 against BK (Large T: 139±65; VP1: 223±93) and 26/45 against HHV6 (U90: 114±82; U11: 38±17; U14: 90±27). None of the lines reacted against recipient cells. To date 11 allogeneic HSCT recipients have received 0.5-2x10⁷ pVSTs/m² without adverse events. Three patients received the cells prophylactically and remained infection free up to 3 months post-pVSTs. The other 8 patients were treated for one or more active infections. Based on viral load measurements a single infusion successfully controlled active infections associated with all our targeted viruses: CMV (2 CR, 1 PR); EBV (4 CR, including a frank PTLD case); Adv (1 CR); HHV6 (2 CR) and BK (4 CR, 1 PR, 1 NR). Of note, all 3 patients with BK hemorrhagic cystitis had marked improvement/disappearance of hematuria post-pVSTs. One subject had transient but severe bladder pain with inflammation, seen on cytoscopy, and a 6-log fall in urine BK viral load with detection of BK-specific T-cells in his bladder. Our only partially non-responding patient had 3 viral infections (EBV, HHV6, BK) and cleared EBV and HHV6 but not BK following the infusion of a line that lacked specificity for this virus, likely reflecting the seronegative status of the donor. Thus, infusion of pVSTs has been safe and clinically effective against up to four simultaneous/sequential infections in a single HSCT recipient. We are currently extending this platform to include other clinically relevant viruses and are planning to assess the activity of “off the shelf” 3^rd party pVSTs for broader implementation.