Allogeneic double‐negative T cells as a novel adoptive cellular immunotherapy for relapsed AML post allo‐HSCT : a first‐in‐human phase I study

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is widely used as a potentially curative treatment option for high-risk and refractory or relapsed acute myeloid leukemia (AML) patients.1 However, about 40% of allo-HSCT patients experiences disease relapse.1 Traditionally, allo-HSCT patients with recurrent disease may be treated with salvage therapies such as chemotherapy, donor lymphocyte infusion (DLI), or secondary allo-HSCT.1 However, disease relapse remains one of the main causes of death due to low response rate to the commonly used salvage therapies.1 Hence, safe and effective treatment options are urgently needed for allo-HSCT patients. Double-negative T cells (DNTs) are a rare subset of mature T cells that comprises 1%–5% of peripheral leukocytes and defined by the expression of CD3 but not CD4 and CD8.2 Preclinically, ex vivo expanded donor-derived DNTs show potent activity against an array of AML targets without off-tumor toxicities toward healthy cells and tissues. Further, non-genetically modified allogeneic DNTs (allo-DNTs) fulfill the requirements of an off-the-shelf cellular immunotherapy, including not causing graft-versus-host disease (GvHD), resistance to host-versus-graft (HvG) rejection, scalability, and storability.3 Hence, we undertook a first-in-human phase I trial (registered in the Chinese Clinical Trial Registry; ChiCTR-IPR-1900022795) to assess the feasibility, safety, and potential efficacy of healthy unrelated donor-derived allo-DNTs as a treatment for AML patients who relapsed after allo-HSCT. In this trial, 12 patients were consecutively enrolled in accordance with the protocol (Data S1). Of these patients, one patient withdrew before the initiation of treatment owing to the absence of detectable disease, and one patient (patient #4) withdrew after the second dose of DNT infusion for personal reasons. Thus, 10 patients received three planned doses of allo-DNTs given 1 week apart (Figure 1A). The patient clinical characteristics are shown in Table S1. Among the 10 patients, 2 patients had received stem cells from human leukocyte antigen (HLA)-matched siblings, and 8 patients received umbilical cord blood transplant. The patient median age was 24.1 (range 4–44 years). The median time from allo-HSCT to relapse was 9.0 months (6.0–15.4 months), and the median time from relapse to DNT therapy was 5.1 months (1.7–11.7 months). After disease relapse, all patients discontinued immunosuppressants and received 1–5 lines of salvage therapies prior to DNT treatment as described in Table S2. Salvage therapies were stopped at least 1 week prior to preconditioning chemotherapy. DNTs were derived from third-party donors and successfully manufactured as described in Supplementary methods for all planned treatments at targeted dosage with high purity and anti-leukemic activity against AML cell lines (Figure S1). Each patient received with a total of 3 doses of third-party donor-derived DNTs at escalating doses of 5 × 107, 1 × 108 and 2–4 × 108 cells per kilogram of body weight after lymphodepleting preconditioning. The patients were monitored for adverse events (AEs) from the time of DNT infusion until at least 30 days after the last DNT infusion or until AEs were relieved (Table S3). All DNT-treated patients showed symptoms of grade 1 or 2 cytokine release syndrome (CRS) such as fever and hypotension. The CRS symptoms were controlled within 24 h in all patients with nonsteroidal anti-inflammatory drugs such as ibuprofen to control fever. The levels of the CRS-related cytokines interleukin (IL)-6, IL-10, macrophage inflammatory protein (MIP)-1α, MIP-1, and monocyte chemokine (MCP)-1 were elevated in 8 evaluable patients after each DNT infusion. The cytokine levels of all patients returned to baseline within 24 h after each DNT infusion (Figure S2). No signs of GvHD or severe neurotoxicity were observed in any DNT-treated patients. All patients experienced grade 3–4 neutropenia and lymphopenia 1 week after the 1st DNT infusion, and approximately half of the patients also had grade 3–4 thrombocytopenia and/or anemia (Table S3). Two etiology workup-negative patients with febrile neutropenia were considered to have infection. The other allo-DNT treatment-related grade 1 or 2 toxicities included headache (3/11), vomiting (2/11), and muscle pain (1/11). Although the doses of infused DNTs were almost one or more orders of magnitude higher than those used for autologous CAR-T cell therapy, none of the patients showed >grade 2 AE related to allo-DNT treatment at a maximum dose of 4 × 108 DNT cells/kg; hence, the maximum tolerated dose was not reached. The patient status was first assessed 28-days after DNT treatment. Two patients were in minimal residual disease (MRD)− complete remission (CR), two patients were in CR with incomplete hematologic recovery (CRi), and one was in CR with evidence of residual disease (Figure 1B and Table S4). The other five patients were experiencing progressive disease. After the first patient assessment, all except one patient received post-DNT maintenance or salvage therapy starting 28–70 days after the last DNT infusion as described in Table S5. Specifically, the five CR patients received IFNα and decitabine as maintenance therapy. All surviving patients were followed up until July 31, 2021, and the median follow-up time was 20.1 (range 15.0–24.3) months. Five of the 10 DNT-treated patients remained alive, and among those, 4 remained in CR for 16.5–24.3 months post-DNT therapy. We did not observe any significant correlation between the interval from allo-HSCT to relapse, the interval from relapse to DNT-treatment, patient age, and the bone marrow blast level pre-DNT treatment between patients that are alive at the last follow-up versus the deceased patients (Figure S3). However, 4 out of 5 live patients were in MRD+ CR prior to DNT-treatment (Table S4), suggesting that patients with low disease burden may be more likely to benefit from DNT-treatment. Treatments given after DNT-infusion, particularly hypomethylating agent (HMA)-based maintenance therapy, can attribute to the superior survival of DNT-treated patients.4 To assess the efficacy of DNT-treatment, allo-HSCT patients who survived for at least 5.1 months after the disease relapse (the median time from relapse to DNT infusion for the trial patients) and treated with conventional or HMA-based salvage therapy from the same medical center were chosen as the historical controls. The clinical characteristics of DNT-treated patients and the historical control groups were comparable (Table S6). For the DNT-treated patients, 1-year progression free survival was 50.0% (95% CI, 18.4%–75.3%; Figure S4), and 1-year OS was 60.0% (95% CI, 25.3%–82.7%). For the patients treated with conventional salvage therapy and HMA-based therapy, 1-year OS was 35.2% (95% CI, 13.3%–58.2%) and 35.2% (95% CI, 11.2%–60.7%), respectively (Figure 1C). However, the differences were not statistically significant. The expansion and persistence of infused cell products may contribute to the success and durable responses of adoptive T cell therapy.5 To assess the persistence of DNTs, the frequency and number of total and donor DNTs in patient peripheral blood (PB) were monitored before lymphodepleting preconditioning and at various time points after DNT infusions. Shortly after DNT infusion, the frequencies of DNTs among T cells and donor DNTs among total DNTs significantly increased (Figure S5A). Further, the total DNT number peaked after each infusion (Figure 1D), with the average DNT counts higher than the preinfusion level by 1.6-, 3.0-, and 4.9-fold 5 days after the first, second, and third DNT infusions, respectively. On day 35, the average DNT number in PB was 4.71-fold higher than the baseline level. The total DNT number was not related to the dose level (Figure S5B). DNT levels continued to be higher than the average preinfusion level (23.3 × 106 ± 8.0 × 106 cells/L) at the last follow-up in the five patients who achieved CR after DNT treatment (36.58 × 106 cells/L to 593.4 × 106 cells/L). The number of donor DNTs appeared to be highest in the 2–4 × 108 cells/kg groups, but was not significantly different between the three doses due to the small case number (Figure S5C). Donor DNTs were detectable as early as 6 h after infusion, decreased slightly at 24 h and expanded at 48 h after the first infusion in most patients and slightly increased after the second and third infusions, and the increased levels lasted for 2 weeks after the third infusion (Figure S5C). The peak peripheral DNT levels did not correlate with the treatment response or patient status at the last follow up (Figure S6). DNTs are potent producers of inflammatory cytokines, such as IFNγ and TNFα, upon encountering leukemic blasts.2 To assess antileukemic activity in patients, serum samples were collected before and at various time points after DNT treatment and serum IFNγ and TNFα levels were compared to the preinfusion levels. Notable increases in serum IFNγ (2.03 ± 0.28-fold) and TNFα (1.93 ± 0.33-fold) levels were observed 6 h after each infusion and persisted for up to 24 h, supporting the induction of antileukemic activity by DNTs (Figure 1E). Overall, this study demonstrates that allo-DNT therapy is a safe and practical treatment option for AML patients who relapse after allo-HSCT. Similar to the historical control groups from the study, the 2-years OS of relapsed AML patients after allo-HSCT was 19% in a retrospective study conducted on 1265 patients by Yanada et al.6 Hence, third-party donor-derived DNT therapy potentially showed signal for activity for this patient population with poor prognosis and limited treatment options. However, the small number of DNT-treated patients and the heterogeneity of patient characteristics limit the assessment of the antileukemic response mediated by DNTs. Therefore, the results from this trial support the need for a next-phased clinical trial to assess the efficacy of third-party DNTs as treatment of relapsed AML after allo-HSCT in a larger patient cohort. Patient consent: Written informed consent was obtained from all the patients or guardians in accordance with the Declaration of Helsinki. Clinical trial registration: The trial was registered in the Chinese Clinical Trial Registry; ChiCTR-IPR-1900022795. We thank all the donors and patients who participated in this study. We thank Dr Mark D. Minden for his excellent comments and critically reading and editing the manuscript. We thank Dr Xiaojun Xu from Zhejiang Children's Hospital for patient referral. Li Zhang is a co-inventor on several DNT cell technology related patents and intellectual properties (IPs), received a grant and consulting fees from WYZE Biotech Co. Ltd. WYZE Biotech Co. Ltd licensed the Chinese right for one of the patents; and Li Zhang owns shares in WYZE Biotech Co. Ltd. Liming Yang, JongBok Lee, and Aaron D. Schimmer are co-inventor on DNT cell technology related patents/IPs. Liming Yang, Zhiqiang Xiang, Meijuan Tu and Pingping Teng own shares in WYZE Biotech Co., Ltd and are employees of WYZE Biotech Co., Ltd. The others declare no competing interests. Xiaoyu Zhu, Li Zhang, Zimin Sun, JongBok Lee, Baolin Tang and Liming Yang conceived and designed the study. Xiaoyu Zhu, Baolin Tang, Siqi Cheng, Tianzhong Pan, Wen Yao, Dongyao Wang, Meijuan Tu, Xiandeng Chu, Liangquan Geng, Ping Qiang, Guangyu Sun, Huilan Liu, Jian Wang, and Zimin Sun enrolled patients. Zhiqiang Xiang and Pingping Teng recruited and screened donors and manufactured DNT products. Baolin Tang, Siqi Cheng, Tianzhong Pan, and Pingping Teng performed experiments and collected data. JongBok Lee, Xiaoyu Zhu, and Li Zhang analyzed and interpreted data, and wrote the manuscript. Aaron D. Schimmer provided feedback and edited the manuscript. Xiaoyu Zhu, Zimin Sun, Liming Yang, and Li Zhang provided funding support. The data that support the findings of this study are available from the corresponding author upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1. Supporting Information. Data S2. Supporting Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.