AUTO3 CAR T-Cell Product Plus Pembrolizumab Shows Clinical Activity, Tolerability in Relapsed/Refractory DLBCL

The investigational CAR T-cell product AUTO3 in combination with pembrolizumab was found to have a tolerable safety profile and elicit durable complete responses in patients with relapsed/refractory diffuse large B-cell lymphoma.

The investigational CAR T-cell product AUTO3 in combination with pembrolizumab (Keytruda) was found to have a tolerable safety profile and elicit durable complete responses (CRs) in patients with relapsed/refractory diffuse large B-cell lymphoma (DLBCL), according to results of the single-arm, phase 1/2 ALEXANDER study that were presented during the virtual 2020 ESMO Congress.1

Findings showed that across 4 cohorts with varying doses of AUTO3 alone or in combination with pembrolizumab, the objective response rate (ORR) was 68% (n = 19/28) and the CR rate was 54%. Four patients had a partial response (PR), and 9 patients had progressive disease (PD).

When stratified by a cohort that comprised the recommended phase 2 dose, which was 150-450 x 106 AUTO3 plus pembrolizumab on day minus 1 (n = 20), the ORR was 71%, including a 64% CR rate. Here, 1 patient achieved a PR and 4 patients had PD.

“AUTO3 is safe and well tolerated, with its best-in-class safety profile. The complete remission rates of the recommended phase 2 cohort appear promising, with overall durable remissions noted for the product,” lead study author Eleni Tholouli, MD, consultant hematologist at NHS Foundation Trust, said in a virtual presentation during the meeting.

AUTO3 is a first-in-human bicistronic CD19- and CD22-directed CAR T-cell therapy. The agent was designed in an effort to reduce the risk of relapse that is due to single antigen loss in patients with B-cell malignancies.2

The single-arm open-label, multicenter, phase 1/2 study comprised a dose-escalation cohort and an efficacy cohort. In the phase I dose-escalation cohort (n = 30), preconditioning fludarabine and cyclophosphamide were given, and patients were treated with AUTO3 at varying doses, along with pembrolizumab on either day 14 for 3 doses, or on day minus 1 for 1 dose. AUTO3 was infused at a starting dose of 50 x 106 CD19/CD22 CAR T cells.

In the phase 2 efficacy stage, there were 2 cohorts. Cohort 1 was comprised of patients with DLBCL not otherwise specified (NOS), high-grade B-cell lymphoma, and transformed DLBCL (tDLBCL) from follicular lymphoma; all received more than 2 prior lines of therapy. Cohort 2 included patients with primary mediastinal large B-cell lymphoma and tDLBCL from other indolent non-Hodgkin lymphoma; these patients also received over 2 lines of previous treatment.

All except the first 3 patients in cohort 1 received consolidation preconditioning therapy with pembrolizumab, Tholouli added.

In total, investigators enrolled 35 patients with DLBCL (n = 25), comprising both germinal center B-cell ([GCB], n = 17) and non-GCB (n = 8); tDLBCL from follicular lymphoma (n = 7) and from marginal zone lymphoma (n = 1); and high-grade B-cell lymphoma (n = 2). Most patients had stage IV disease (n = 22), followed by stage III (n = 9), stage II (n = 3), and unknown (n = 1).

The median age was 59 years (range, 28-83), and 23 patients were male. Nineteen patients had both relapsed and refractory disease, 7 patients had refractory, and 9 patients had relapsed disease. Additionally, patients had low-risk disease as assessed by International Prognostic Index (n = 4), low-intermediate risk (n = 6), high-intermediate risk (n = 10), high-risk (n = 4), and unknown (n = 11). The median number of prior therapies was 3 (range, 1-10), and 8 patients underwent prior autologous stem cell transplant. The median sum of the product of diameters (SPD) was 19.31 cm (2.1-260.84).

“The median SPD shows that the tumor burden is, in this trial, similar to that of other CD19[-directed] CAR T trials,” said Tholouli.

The varying dose levels were as follows: 50 x 106 AUTO3 with no pembrolizumab (n = 4), 50 x 106 AUTO3 with pembrolizumab on day 14 (n = 3), 150-450 x 106 AUTO3 on day 14 with pembrolizumab (n = 8), and 150-450 x 106 AUTO3 with pembrolizumab on day minus 1 (n = 20).

No severe cytokine release syndrome (CRS) was reported in any patients on study. Grade 1 CRS was reported in 8 patients (22.9%) and occurred in 3 of the 4 dose levels: 50 x 106 AUTO3 (n = 1), 150-450 x 106 AUTO3 with pembrolizumab on day 14 (n = 2), and 150-450 x 106 AUTO3 with pembrolizumab on day minus 1 (n = 5). Grade 2 CRS occurred in 4 patients (11.4%) and was reported in 2 patients each of both 150 to 450 x 106 AUTO3 cohorts.

Tholouli noted that 1 patient who had no CRS with primary infusion later developed grade 3 CRS with severe hypoxia with re-treatment 1 year later, which occurred in a setting of no CAR T-cell expansion and significant disease burden in the lung, which had been treated with radiation.

No prophylactic measures were given, but 5 patients (14%) required tocilizumab (Actemra) for CRS. The median time to CRS was 6 days (range, 1-36), and the median duration of CRS was 3 days (range, 1-19).

“Based on our experience, this profile appears clearly superior to that of currently approved CD19-directed CAR T-cell products,” Tholouli said.

Three cases (n = 8.6%) of neurotoxicity were reported across the population, 2 were grade 3 or higher and occurred at the 50 x 106 AUTO3 level (n = 1) and the 150 to 450 x 106 AUTO3 with pembrolizumab on day minus 1 dose (n = 1). Similar to the CRS data, no prophylactic measures were needed.

Tholouli added that no any-grade neurotoxicity occurred in patients who achieved a CR, which were also patients who had robust expansion. All cases of neurotoxicity were atypical in context of tumor progression, with no to minimal CAR-T expansion in the peripheral blood.

In the patient with grade 2 neurotoxicity, their mental status was altered and associated with sepsis and narcotics and resolved. In the patients with grade 3 neurotoxicity, Tholouli said that they experienced facial and muscle weakness that was onset on day 53 but resolved. Finally, in the patient with grade 4 neurotoxicity, encephalopathy was experienced and was associated with sepsis, hyponatremia overcorrection to hypernatremia, metabolic acidosis, and multiorgan failure. The patient died from disease progression and multiorgan failure.

Twenty-eight of the 35 patients were evaluable for response, and responses were reported at all dose levels. In the 50 x 106 AUTO3 cohort with no pembrolizumab (n = 4), 2 patients achieved a response, with 1 CR and 1 PR; 2 patients experienced PD. In the cohort of those who received 50 x 106 AUTO3 with pembrolizumab on day 14 (n = 3), 2 responses were observed—1 CR and 1 PR, but no patients experienced PD. In patients who were treated with 150 to 450 x 106 AUTO3 with pembrolizumab on day 14 (n = 8), 5 patients had a response, which comprised of 4 CRs and 1 PR; 3 patients experienced PD.

At a median follow-up of 6 months (range, 1-24), 93% of patients who achieved CRs did not develop PD.

“Most interestingly, CRs appear to be durable,” Tholouli said. “Only 1 patient developed PD after achieving a CR, and this was patient treated at the lowest dose of [50 x 106 AUTO3] and without pembrolizumab.”

Based on these findings, Tholouli concluded that an ongoing expansion cohort with the recommended phase 2 dosing in the outpatient setting is currently enrolling in the United States and the United Kingdom.

References

  1. Tholouli E, Ardeshna K, Ramakrishnan A, et al. Phase I Alexander study of AUTO3, the first CD19/22 dual targeting CAR T cell therapy, with pembrolizumab in patients with r/r DLBCL. Presented at: 2020 ESMO Congress; September 19-21, 2020; virtual. Abstract 890M0.
  2. Autolus Therapeutics presents additional data on AUTO3 in DLBCL during the ASCO20 Virtual Scientific Program. News release. Autolus Therapeutics. May 29, 2020. Accessed September 23, 2020. https://bit.ly/2FX2GgK.