Clinical trials of CAR-T cells in China

Currently, there are 121 trials reported and/or registered at ClinicalTrials.​gov from China (Table 1). The trials are mainly carried out in leading hospitals from Beijing, Shanghai, Guangzhou, and Chongqing. CAR-T trials are started in hospitals throughout China. In this study, to avoid duplication of trials that can lead to miscalculation, those trials in Chinese registries were not included. It is possible that the number of institutions carrying out CAR-T trials will increase at a slower pace once regulatory policies are in place. We believe these CAR-T cells should be regulated as drugs [39].

T cell receptors (TCRs) are engineered by incorporating a specific antigen-targeting element and CD3 element to form a completely novel TCR structure, the chimeric antigen receptor (CAR) [35, 40]. In addition, several co-stimulating sequences have been used to facilitate the expansion of the CAR-T cells [41]. CAR-engineered T lymphocytes have been in active clinical development to treat patients with advanced leukemia, lymphoma, and solid tumors [42–45].

One of the major hurdles in CAR-targeted cellular therapy has been the limited cell dose due to the lack of adequate in vivo cell expansion. Co-stimulatory signals can enhance immune responses of effector T cells [46]. Inducible co-stimulatory signal (ICOS), 4-1BB (CD137), CD28, OX40 (CD134), CD27, and DAP10, along with CD3ζ, have been investigated [31, 47–50]. Among these, 4-1BB (CD137), CD28, and CD3ζ are the most commonly used COS elements in the CARs (Tables 2, 3, and 4) [51, 52].

Most CARs in the CAR-T trials in China are second-generation CAR constructs, which have one co-stimulatory signal [41]. A trial of CAR-T cells containing a third-generation CAR construct with both CD28 and CD137 co-stimulatory signals is still recruiting patients with relapsed/refractory ALL (NCT02186860). Fourth-generation CARs have incorporated additional elements in the CAR constructs, such as an inducible caspase-9 gene element that can lead to self-destruction by apoptosis of the CAR-T cells [53]. A total of 10 trials of CAR-T cells contain a fourth-generation CAR (Table 5). Among these, five trials are evaluating CARs with an inducible caspase-9 suicide switch.

The recombinant CAR cassette is typically packaged into a pseudo-lentivirus vector which can efficiently incorporate into the genome of T cells. To date, the lentiviral vector is the most commonly used vector in CAR-T cells. The other vector commonly used is the retroviral vector (Tables 2, 3, and 4).

By altering a specific antigen-targeting element, the specificity of the CAR-T cells can be easily re-directed to a specific type of malignancy. This makes the CAR-T cell therapy highly versatile. A number of antigens have been targeted in this way. More and more antigens are being engineered into CAR-T cells, leading to a large repertoire of CAR-T cells that are being explored for the therapy of both solid and hematological malignancies (Tables 3 and 4).

CD19 is the most commonly targeted antigen to date (Table 2). Out of the 121 trials, 57 trials have CD19 as a target. Currently, there are 19 clinical trials in China targeting non-CD19 antigens, including CD20, CD22, CD30, CD33, CD38, CD123, CD138, BCMA, and Lewis Y antigen for hematological malignancies (Table 3). Dual- and multi-specificity CAR-T cells have also been in clinical trials in China.

The most common type of diseases in CAR-T trials are B cell malignancies, including leukemia, lymphoma, and myeloma.

The CD19-targeted autologous CAR-T product, tisagenlecleucel, was recently approved by FDA for therapy of refractory/relapsed (r/r) B cell ALL. In 30 patients including children and adults who received this product, 90% of them achieved complete remission (CR) [54]. Severe cytokine-release syndrome (CRS) was reported in 27% of the patients. This product has been in clinical trials for CD19+ B cell malignancies, including CLL, ALL, and lymphoma [21–24, 54, 55]. In a Chinese study (NCT 02813837), 30 patients (5 children and 25 adults) with r/r ALL were treated with autologous CD-19 CAR-T cells [56]. In this 2017 report of preliminary results of a seven-center clinical trial, CR was 86% and severe CRS was seen in 26% of the patients [56]. Successful outcome has been reported with other CAR-T cells against CD19 antigen in r/r ALL [29, 32, 57–59].

The CD19-specific CAR-T cells, axicabtagene ciloleucel (axi-cel, KTE-C19), have been reported to be safe for treatment of aggressive lymphomas including r/r diffuse large cell lymphoma (DLBCL) [25]. In the phase II part of the ZUMA-1 trial, overall response rate (ORR) was 76% (47% CR and 29% PR) at the time of report in the cohort 1 of 51 patients [60]. This product is currently under evaluation by FDA.

CD33 and CD123 are targets on myeloid leukemias. Currently, there are three trials on CAR-T cells targeting CD33 and two trials targeting CD123 antigen in China (Table 3). In the USA, three CAR-T trials targeting CD123 were either terminated (NCT02623582) or suspended (UCART123, NCT02159495, and NCT03190278) at this time.

B cell maturation antigen (BCMA) is an antigen target on myeloma cells. Currently, three trials on BCMA-targeted CAR-T cells are being done in r/r myeloma in China (Table 3). In one of the trials of CAR-T cells targeting BCMA in China, 19 patients with r/r multiple myeloma were evaluable and 7 of the patients were followed for more than 6 months at the time of the report [61]. CRS was observed in 14 (74%) patients. The ORRs were close to 100% in the evaluable r/r myeloma patients. The outcome from the preliminary report was highly encouraging. Complete response was also reported in a case of r/r myeloma patient who received autologous CTL019 cells, even though 99.95% of the myeloma cells were negative for CD19 [38, 62]. It appears therefore that multiple myeloma is highly sensitive to immunotherapy.

There are also a few registered clinical trials that are testing two or more CARs either simultaneously or sequentially. In the trial NCT02846584, patients receive intravenously infused autologous anti-CD19 or anti-CD20 CAR-T cells to treat B cell malignancies. Another trial, NCT02737085, is to explore the sequential therapeutic effect of anti-CD19 and anti-CD20 CAR-T cells in the treatment of DLBCL.

The trial NCT02903810 was planned with a treatment scheme of infusion of equal numbers of anti-CD19 and anti-CD22 CAR-T cells in the treatment of refractory hematologic malignancies. Two trials (NCT03097770 and NCT03098355) target two antigens simultaneously with one CAR construct (Table 2). These trials are ongoing at this time.

Multiple solid tumors are being studied in CAR-T clinical trials. At the time of this report, 20 different antigens are being targeted in solid tumor trials (Table 4). GPC3, mesothelin, epidermal growth factor receptor (EGFR), and EpCAM were the most targeted antigens (Table 4). This is consistent with reports from international trials [63–68]. Liver cancer remains the most commonly studied solid tumor in China [69]. In a preliminary report of a trial of CAR-T cells against CD133+ epithelial tumors (NCT02541370), 24 patients were enrolled, including 14 patients with sorafenib-refractory hepatocellular carcinoma (HCC), 7 with pancreatic carcinomas, 2 with colorectal carcinomas, and 1 with cholangiocarcinoma [69]. The number of CAR-T cells was found to be inversely related to the CD133+ epithelial cells in peripheral blood. There was a separate report treating refractory cholangiocarcinoma with sequential infusion of two different types of CAR-T cells targeting EGFR and CD133 [70].

Two trials in China are evaluating GD2 antigen-targeted CAR-T cells in neuroblastoma (Table 4). Another two trials are evaluating CAR-T cells against EGFRvIII+ glioblastoma. There was one case report in the literature on rapidly progressing refractory glioblastoma that showed dramatic CR to IL13Rα2-targeted CAR-T cells after repeated infusion [71]. In a separate report, nine patients with refractory EGFRvIII+ glioblastoma received autologous CART-EGFRvIII cells in a pilot study [66]. Interestingly, there was no CRS observed. CAR-T cell infiltration was shown in the resected tumor specimen. This study suggested that the CAR-T cells are safe and immunologically active with tracking capability to the cancer cells in the brain.

Multiple antigens are being explored as targets in solid tumors for CAR-T cells (Table 4). Preliminary reports have been presented and published throughout the world [64, 65, 67, 72]. Outcomes from larger sample size and longer follow-up are clearly needed from these trials.

There is currently one clinical trial of autologous CAR-T19 cells for patients with systemic lupus erythematosus (NCT03030976, Table 2). This trial is designed to infuse 1 × 10⁶ cells/kg. More trials are expected to come for non-malignant diseases.