Background:  Some aviremic HIV patients on HAART continue to have low CD4 counts. Low-level cryptic viral replication in mucosal tissues may deplete progenitor cells, contributing to poor immunologic outcomes. Previous CD4 adoptive transfer therapies showed limited persistence of the infused cells. We hypothesized gene modification of circulating CD4 cells may make them resistant to HIV. Preclinical studies showed zinc finger nuclease (ZFN) mediated (M) CCR5 (R5) disrupted (D) (ZFN-M-R5-D) CD4 cells provide protection against R5 tropic infection. This study assesses the safety, engraftment, persistence, CD4 count, and homing to gut mucosa of SB-728-T in aviremic HIV subjects.

Methods:  In an open-label, single-arm study, 6 aviremic HIV subjects on HAART with CD4 counts of 200 to 500 cells/mm³ were enrolled in 2 cohorts: 1x10¹⁰ and 2x10¹⁰ total cells. A third cohort (n = 3) will be dosed at 3x10¹⁰ cells. Autologous R5-disrupted T cells were expanded ex vivo with mean R5 modification rates of 26% (14 to 36%). Subjects were followed weekly for 1 month and then monthly for 11 months post-infusion.

Results:  Infusions were safe and well tolerated with only mild adverse events (AE) (median f/u Cohort 1 [n = 3] and 2 [n = 2] is 24 [16 to 28] and 4 weeks, respectively). The percentage of R5 disruption in the peripheral blood (PB) of the 5 subjects was 6, 3, 1, 2, and 2% at day 14 and persisted for the duration of follow up. CD4 counts increased in all subjects at day 14 (35 to 1038 cells/mm³) and were sustained at all time points (mean increase of 208, 86, 233, 911, 210 cells/mm³). R5-disrupted cells were detected in the rectal mucosa of all subjects at all assessed time points with levels of R5 disruption approximating that of PB when normalized for CD4 cells within each compartment.

Conclusions:  ZFN-disrupted CD4 T cells can be processed at clinical scale achieving R5 modification rates of up to 36%. Infusion of both cell doses was safe and well tolerated. Successful engraftment of SB-728-T was observed in all subjects, with R5-disrupted CD4 cells accounting for up to 6% of PB in 1 subject. The levels of persistent engraftment on day 90 are 6- to 40-fold greater than previously reported. Homing of these cells to the gut mucosa was observed in all subjects tested, with R5 disruption levels similar to PB suggesting these cells traffic normally. Importantly, increases in total CD4 counts were seen in all 5 subjects at all time points. These preliminary data suggest that ZFN-M-R5-D CD4 T cells can bolster CD4 cell counts even in HIV subjects with undetectable viral load.