ORLANDO, Fla. – New phase I/II data from Autolus Therapeutics plc announced at the American Society of Hematology’s (ASH) annual conference show that AUTO-3, the first bicistronic CAR T targeting CD19 and CD22 followed by an anti-PD1, was well-tolerated in a phase I/II study.

The clinical trial, Alexander, is in patients with relapsed/refractory diffuse large B-cell lymphoma (DLBCL) and pediatric acute lymphoblastic leukemia (ALL). The new, preliminary data show 90% of patients had a complete response (CR) and 100% had an overall survival at 12 months in CAR T naïve pediatric ALL patients.

“CD19 CARs are highly active in relapsed and refractory DLBCL, however, a significant number of lymphomas relapse by losing the CD19 antigen, which makes them invisible to the current CD19 CARs,” Kirit Ardeshna, the study’s lead author, a hematologist at University College London Hospital, told BioWorld. “By targeting two antigens on the lymphoma cells we can make it difficult for the lymphoma to relapse as it will be more difficult to lose two antigens.”

In addition, in the current study we also address another tumor escape mechanism, which is driven by PD-L1 upregulation, which can cause early T-cell exhaustion, he said.

“By adding a few doses of anti-PD1 therapy we can prevent this mechanism of tumor escape as well. The AUTO-3 approach therefore addresses two escape mechanisms of lymphoma,” Ardeshna added.

AUTO-3 is designed to reduce antigen-driven relapse by targeting CD19 and CD22. The Alexander trial began in September 2017 and is expected to wrap up in March 2021. The phase I’s primary outcome, safety in the incidence of grade 3-5 toxicities, along with confirmation of the phase II’s dose and schedule, is in line with the preliminary data. The phase II primary outcome measure is the overall response rate using the Lugano criteria, with a time frame of up to two years.

In the dose escalation phase, of 16 treated patients, four were dosed at 50 x 10⁶ cells without pembrolizumab, 11 patients received escalating doses of AUTO-3 with pembrolizumab administered at day 14: three at 50 x 10⁶ cells, four at 150 x 10⁶ cells, four at 450 x 10⁶ of AUTO-3, and one patient was dosed with 450 x 10⁶ cells with pembrolizumab administered a day before AUTO-3 infusion. In one month, 14 of the patients were evaluable.

AUTO-3 was well-tolerated, with no patients experiencing ≥ grade 3 cytokine release syndrome (CRS) with primary infusion and one of 14 experienced grade 3 neurotoxicity, quickly resolved with steroids. No pembrolizumab immune-related toxicities were found. The majority of grade 3 or higher adverse events were hematological. Low levels of serum cytokines were consistent with the observed low levels of CRS and neurotoxicity.

Across all the tested doses, five patients achieved a CR. Four of the five CRs were ongoing, with the longest at 18 months. All CRs did not need steroid or tocilizumab-based management or ICU level care.

While in initial clinical testing of AUTO-3 only one relapse was attributed to CD19 loss, there was less long-term persistence compared to AUTO-1, which is the company’s lead product. In September, Autolus released data indicating that using a lower affinity CD19-specific binder results in fast binding to the antigen target, while the residence time is comparable to that of normal T cells, avoiding over-activation of the immune system. Modulating the on-rate and off-rate in this way reduced the severity of CRS and the incidence of neurotoxicity in a 14-patient pediatric trial of AUTO-1 in treating ALL. AUTO-1 T cells also showed enhanced expansion and prolonged persistence compared to the approved CAR T cell therapy for ALL, Novartis' Kymriah (tisagenlecleucel), which uses a higher affinity binder, FMC63.