Safety and Efficacy of NY-ESO-1 Antigen-Specific T-Cell Receptor Gene-Transduced T Lymphocytes in Patients with Synovial Sarcoma: A Phase I/II Clinical Trial

This article is featured in Selected Articles from This Issue, p. 4991

Synovial sarcoma (SS) is a malignant, rare type of cancer that accounts for approximately 5% to 10% of all soft-tissue sarcomas (STS) and primarily presents in the extremities (1, 2). The incidence of SS, which has been estimated to be around 70 cases per year in Japan, has remained almost constant (3); it peaks in the third decade of life with around 30% of cases occurring in patients under 20 years old (2, 4, 5). SS is associated with a poor prognosis because of late local recurrence and distant metastases (1, 5), which occur in about 50% of cases and mainly involve the lung (6–8).

The standard therapeutic approach for SS, as for other STS, is wide surgical excision combined with radiotherapy and/or chemotherapy (8). Although localized sarcomas can be cured through surgical treatment, controversies exist over the optimal management of patients with metastatic SS, for which curative treatment is rarely achievable (9). Anthracycline-based chemotherapy, either alone or combined with ifosfamide, is generally the recommended first-line therapy for advanced disease (8–10). In addition, recently approved drugs, such as pazopanib (11), trabectedin (12), and eribulin (13), have been administered as second-line or later treatment options for patients with advanced STS, including SS (10, 14). Understanding that each chemotherapy-based treatment has a different efficacy and safety profile is necessary to maximize the benefit to patients and improve treatment options.

Genetically engineered autologous T cells expressing cloned T-cell receptors (TCR) directed against tumor-associated antigens have attracted attention as a potential next generation of adoptive immune therapy (15). The New York esophageal squamous cell carcinoma 1 (NY-ESO-1) antigen is a hydrophobic cancer-testis antigen encoded on the CTAG1B gene on the X chromosome and its expression has been reported in a wide range of tumor types, including neuroblastoma, myeloma, metastatic melanoma, as well as bladder, esophageal, hepatocellular, head and neck, prostate and breast cancer (16). Further, expression of NY-ESO-1 has been reported in a high proportion of SS samples (9, 17–19). Preliminary but promising results of adoptive cellular immunotherapy using autologous T cells expressing anti–NY-ESO-1-specific TCR (anti–NY-ESO-1 TCR) for SS and melanoma have been shown in previous studies (20, 21). However, there are potential drawbacks of using modified lymphocytes, such as mispairing of the introduced TCR α/β chains with the endogenous TCR subunits, which can impair cell surface expression of the transduced TCR, resulting in insufficient function and potential generation of autoreactive T cells (22). To help overcome this drawback, new techniques involving retroviral vectors (siTCR vector, Takara Bio Inc., Shiga, Japan) encoding siRNA constructs that specifically downregulate endogenous TCR have been developed (23). Such new technologies further raise expectations for adoptive cellular immunotherapy to treat patients with recurrent or advanced SS expressing NY-ESO-1.

Here, we report the results from a phase I/II trial of autologous T lymphocytes expressing NY-ESO-1 antigen-specific TCR gene (product code: TBI-1301, nonproprietary name: mipetresgene autoleucel) infused to patients with advanced or recurrent SS who were not suitable for surgical resection and resistant to anthracyline. Phase I primarily assessed the safety of a single-dose, split infusion of TBI-1301, whereas phase II assessed the efficacy of the TBI-1301 infusion dose confirmed to be safe in phase I.

This open-label, multicenter study (NCT03250325) consisted of phase I followed by phase II and was conducted between October 2017 (first patient consent) and January 2020 (last patient observation). The study comprised a screening and treatment induction period (3 days), a treatment period (2 days, day 0 and day 1), and an observation period (consisting of an early observation period from day 2 to day 28 and a late observation period from day 29 to week 52). Late toxicity information such as replication-competent retrovirus (RCR), clonality assessed by linear-amplification mediated-PCR (LAM-PCR), survival information, and secondary carcinogenesis in patients were collected during a follow-up period up to 15 years (Supplementary Fig. S1). At the beginning of the screening period, patients who gave written informed consent were registered (primary registration) and TBI-1301 production was started. Patients were registered again (secondary registration) with written informed consent obtained again before the treatment induction period.

The safety of the infusion of TBI-1301 was evaluated for up to 28 days after infusion during phase I of the study and was conducted by the Efficacy and Safety Evaluation Committee of this clinical trial. Safety was determined by measuring dose-limiting toxicity (DLT) assessed by CTCAE version 4.0 (Japanese version; ref. 24). In phase I of the study, progression to phase II was allowed if DLT did not occur in the first three patients. If a patient developed DLT, three additional patients were included and, if they did not develop DLT, progression to phase II was allowed. If at least one of the three additional patients developed DLT or if at least two of the first three patients developed DLT, the study was discontinued. DLT was not determined in phase II but was treated as an adverse event.

The trial was approved by the institutional ethical review board at participating institutions and conducted in accordance with the Declaration of Helsinki as well as relevant Japanese Ministry of Health, Labour and Welfare ordinances such as Good Clinical Practice.

Patients were 18 years old or older at the time of consent, had histologically diagnosed advanced or recurrent SS that was unable to be surgically resected and had received between one and four systemic chemotherapy regimens, including anthracycline therapy. Also, at the primary registration, it was confirmed that the HLA type of each patient was either HLA-A*02:01 or *02:06 or both, and that tumor tissue was positive for NY-ESO-1 expression by using immunohistochemistry. NY ESO-1 expression was judged positive when any staining was observed. The main exclusion criteria were serious complications (e.g., unstable angina, myocardial infarction, uncontrollable diabetes mellitus, active infection), active autoimmune disorders that require systemic corticosteroids or immunosuppressants, active metastatic disease in the CNS, or active “double cancer.” Full inclusion and exclusion criteria (at both the primary and secondary registration) are detailed in Supplementary Table S1 and Supplementary Table S2, respectively.

Patients who underwent secondary registration received intravenous cyclophosphamide 750 mg/m² once daily on days -3 and -2 of the induction period. During the subsequent treatment period, 5×10⁹ (±30%) TBI-1301 cell suspension was divided and delivered by infusion of 2.5×10⁹ (±30%) cells on day 0 and day 1. During the follow-up period, patients were observed annually for survival, information about late toxicities, RCR, LAM-PCR, cell kinetics of TBI-1301, and quantification of provirus copies by real-time PCR. Tocilizumab (8 mg/kg over 1 hour by IV administration) was made available in the event of cytokine release syndrome (CRS; see Supplementary Table S3 for grading of CRS severity; refs. 25, 26).

Preparation of TBI-1301 cells for infusion was conducted before the induction period as follows. Peripheral blood mononuclear cells (PBMC) were obtained from peripheral blood (up to 200 mL) of each patient by specific gravity centrifugation. PBMC were then cultured with IL2, anti-CD3 antibody (RRID: AB_2904535), and RetroNectin (Takara Bio Inc., Shiga, Japan). Proliferating lymphocytes were transduced with the retroviral vector MS3II-NYESO1-siTCR. MS3II-NYESO1-siTCR was constructed from DNA encoding HLA-A*02:01/*02:06-restricted NY-ESO-1_157–165-specific TCR-α and -β chains. This TCR is an affinity-enhanced TCR with amino acid substitutions (26, 27). The retrovirus vector also contained interfering RNA constructs that specifically downregulate endogenous TCR (23). The transducing TCR-α and -β are composed of codon-optimized sequences, which are resistant to siRNA for endogenous TCR. The method for TCR T-cell preparation is described elsewhere (26, 28). The cell counts expressing transduced TCR were calculated from the IFNγ productivity upon stimulation with NY-ESO-1–pulsed T2 cells (RRID: CVCL_2211) in the quality tests. Each lot that passed the quality tests was used for infusion.

The primary endpoints for phase I were safety-related and consisted of adverse events graded according to NCI-CTCAE (version 4.0), laboratory values, presence or absence of RCR confirmed by PCR, and clonality assessed by LAM-PCR. Adverse events were also assessed for the degree of causal relationship to TBI-1301 (i.e., related, probably related, possible connection, no association, unknown) and outcome (i.e., recovery, remission, not recovered, recovered with sequelae, death, unknown).

Cell kinetics of TBI-1301 in peripheral blood were measured by real-time quantitative PCR according to Cycleave PCR Core Kit, Product Code CY501, Takara Bio Inc. DNA samples were extracted from isolated PBMC. The PCR primers for proviral DNA were complementary to the retroviral packaging signal region found in TCR-transduced cells. The PCR primers for internal control DNA were complementary to the human IFNγ gene. Both PCR primer sets were from Takara Bio Inc. (Provirus Copy Number Detection Primer Set, Human, Product Code 6167). The copy number of NY-ESO-1 siTCR DNA in the PBMC was represented by the ratio of proviral DNA to IFNγ DNA values (26).

The primary endpoint of phase II was objective response rate (ORR), which was the percentage of patients in whom tumor response was assessed according to RECIST version 1.1 (29). Immune-related RECIST (irRECIST) was also used to evaluate the percentage of patients with tumor response assessed as immune-related complete response (irCR) or immune-related partial response (irPR; refs. 19, 30). Tumor response was measured by the investigator and by central assessment, with the evaluation based on at least one imaging study after TBI-1301 infusion.

The secondary efficacy endpoints were the ORR (phase I only), progression-free rate (PFR), progression-free survival (PFS), and overall survival (OS). PFR was the percentage of patients with a tumor response of CR, PR, or stable disease (SD) at 12 weeks according to RECIST version 1.1. PFS was the time from secondary entry to objective progressive disease (PD), based on RECIST, or to all-cause death, whichever occurred first. PFR assessed as irCR, irPR, or irSD at 12 weeks as well as PFS were also determined according to irRECIST. The secondary safety endpoints in phase II were the same as those of the primary endpoints in phase I.

Efficacy-evaluable patients were those who received at least 2.5×10⁹ (±30%) TBI-1301 cells on day 0 and had their results judged centrally by diagnostic imaging specialists and clinical oncology specialists independently of facility judgments. Tumor response assessment based on imaging diagnosis was performed at least once after infusion of TBI-1301 by central judgment.

For the efficacy evaluation of the total phase I/II data, the sample size was determined on the basis of the effectiveness evaluation using the ORR from phase I and II combined. The response rate for pazopanib (5.7%) in the PALETTE study of patients with STS was used as an external control and compared via a Bayesian approach to evaluate the target disease (11). A Bayesian posterior probability exceeding the response rate to pazopanib of 80% was considered a success, and the number of patients to be analyzed for establishing the efficacy of TBI-1301 was set at eight patients for the full analysis set (FAS) obtained by pooling the results from phase I and phase II. At least three patients were required in phase I for evaluation of safety as already described.

For the evaluation of tumor response (i.e., ORR and PFR), a two-sided 95% confidence interval (CI) was calculated using the Clopper–Pearson method (31) based on a binomial distribution. Further, a Bayesian posterior probability of ORR for TBI-1301 exceeding that of pazopanib for STS in the PALETTE study was calculated (11). A beta distribution with parameters (0.5, 0.5) was used as noninformative prior distribution for both treatments. Kaplan–Meier survival curves were created for OS and PFS. For OS, a Bayesian calculation was performed assuming an exponential distribution. A gamma distribution was used as a noninformative prior distribution for the parameter of the exponential distribution. The posterior probability that the hazard for TBI-1301 was lower than that of pazopanib was calculated. The reported median OS for pazopanib (12.5 months) in the PALETTE study was used as a historical control (11). SAS (SAS version 9.4 or later) was used for statistical analysis.

All clinical trial data presented in the article are not publicly available, as it could compromise patient privacy or consent, but are available upon reasonable request by contacting the corresponding author.

Baseline demographics for all eight patients comprising the FAS and included in phase I (n = 3) and phase II (n = 5) are summarized in Table 1. Among the patient cohort, there were seven males and one female. The median (min, max) age was 53.0 (21, 61) years. Representativeness of this study patients is described in Supplementary Table S4. All cases of SS were of stage IV at entry and, of the patients previously treated for SS, seven patients had a history of surgery, three patients had a history of radiation therapy, and eight patients had a history of anticancer drug treatment as outlined in Supplementary Table S5. All patients had a history of treatment with anthracycline, and none violated the registration conditions.

In phase I, we screened five patients who provided informed consent. Three patients were eligible for primary registration and underwent blood sampling for TBI-1301 production. All three patients were eligible for secondary registration and were infused with TBI-1301 after cyclophosphamide administration during the lymphodepletion period. Of three patients receiving TBI-1301 infusion, one patient completed the 52-week observation period, and two patients discontinued before 52 weeks due to worsening of their condition, according to the discontinuation criteria of this study.

In phase II, we screened 12 patients who provided informed consent. Five patients were eligible for primary registration and underwent blood sampling for TBI-1301 production. All five patients were eligible for secondary registration and received TBI-1301 infusion. Of these five patients, one patient completed the study and four patients discontinued before 52 weeks due to worsening of their condition, according to the discontinuation criteria of this study. As judged by the investigator, one patient in phase II received a half dose of TBI-1301 (i.e., 2.5×10⁹ cells in total) due to the patient's systemic condition related to CRS including fever of 40°C.

Regarding all cases from phase I and phase II, the ORR according to RECIST version 1.1 by central assessment for the FAS (n = 8) was 50.0% (n = 4/8; 95% CI, 15.7–84.3) and consisted of best overall responses as follows: CR, n = 0; PR, n = 4; SD, n = 1; PD, n = 3. The Bayesian posterior probability that the ORR of TBI-1301 exceeded that of pazopanib for STS in the PALETTE study (i.e., 5.7%) was 100.0%. The ORR according to irRECIST was also 50.0% (n = 4/8; 95% CI, 15.7–84.3) and consisted of best overall responses as follows: irCR, n = 0; irPR, n = 4; irSD, n = 2; irPD, n = 2. The patient (TBI1301–03–08) who received a half dose of TBI-1301 was evaluated as having a PR.

PFR for the FAS based on both RECIST version 1.1 and irRECIST was 62.5% (n = 5/8 patients; 95% CI, 24.5–91.5). PFS and OS for the FAS is shown as a Kaplan–Meier curve in Fig. 1A and B, respectively. The median PFS according to both RECIST version 1.1 and irRECIST was 227.0 days. The median OS was 650.0 days. The Bayesian posterior probability that the hazard of TBI-1301 was lower than that of STS in the PALETTE study was 97.7%.

The spider plot of change in tumor size for individual patients in the FAS is shown in Fig. 2A (based on the facility judgment). In seven of eight patients, tumor regression was observed, which continued for more than 36 weeks in four patients. The swimmer plot shown in Fig. 2B (based on the central judgment) reveals the duration and type of response as well as final outcome for individual patients. One patient (TBI1301–03–08) completed the 1-year observation period without developing PD. Representative CT scan images of lung metastases that occurred in one patient (a 21 year-old male, TBI1301–03–02) are shown in Fig. 3 and highlight the extent of tumor shrinkage throughout 36 weeks of the observation period.

As listed in Supplementary Table S6, adverse events occurred in eight patients (100.0%). Adverse events for which a causal relationship to TBI-1301 could not be excluded (i.e., adverse drug reactions) occurred in seven of eight patients (87.5%), and adverse events associated with cyclophosphamide occurred in all eight patients (100.0%). There were no deaths and no cases of study discontinuation were attributable to adverse events. Specific adverse events classified according to severity (all grades/grade ≥3) are detailed in Table 2. Most grade 3 or higher adverse events were due to the pretreatment drug cyclophosphamide, and only one case was related to TBI-1301 (TBI-1301–03–04: decrease of platelet count, decrease of neutrophil count, hypophosphatemia, increase in pancreatic enzymes, and acute cholangitis). Serious adverse events, greater than grade 4 severity, were observed in four of eighr patients (50.0%). The most common serious adverse events were decreased lymphocyte count, decreased neutrophil count, decreased white blood cell count, and increased pancreatic enzymes. No adverse event was considered as DLT by the investigator, and the transition to phase II was considered appropriate.

CRS occurred in four of eight patients (50.0%) and consisted of one patient with grade 1 CRS and three patients with grade 2 CRS. The median (min, max) period from the infusion of TBI-1301 to the onset of CRS was 2 (1, 2) days. The median (min, max) period to recovery from CRS was 9 (5, 12) days. All patients who developed CRS recovered with prespecified treatment. Of four patients, one patient with grade 1 CRS was only treated with symptomatic treatment. Two of three patients with CRS greater than grade 2 were administered tocilizumab, and one of two patients was also administered steroids. One patient (TBI1301–03–01) was administered 8 mg/kg tocilizumab immediately after CRS development (grade 2) at 2 days after the second infusion of TBI-1301. This patient recovered 6 days after the onset of CRS. The other patient (TBI1301–03–04) developed CRS 1 day after the first infusion, progressed to grade 2 on day 3, and then was given 8 mg/kg tocilizumab on day 4, and approximately 0.1 mg/kg dexamethasone on day 5. This patient with CRS recovered 10 days after onset. Further details of CRS, including treatment and disease course, for individual patients are outlined in Supplementary Table S7.

Laboratory parameter values that decreased from the secondary registration at certain periods were: white blood cell count (day 0–14), red blood cell count (day 0–9), hemoglobin (day 0–9), hematocrit (day 0–9), and platelets (day 0–3). Blood glucose, on the other hand, tended to increase (day 3–7). In the leukocyte fraction, the percentage and absolute lymphocyte count tended to decrease until day 3. In lymphocyte subsets, both the percentage and absolute count of CD19-positive cells tended to decrease until day 28. Changes in lymphocytes at different time points are detailed in Supplementary Table S8. Overall, changes in laboratory parameters for the total population in both phase I and phase II were temporary and eventually recovered to levels similar to those at the secondary registration by day 28. Analysis of vital signs found that body temperature increased after TBI-1301 infusion (mainly on day 0–3), and blood pressure decreased; also, a decrease in percutaneous oxygen saturation occurred, especially in subjects with CRS. No patients had immune effector cell–associated neurotoxicity syndrome (ICANS). Neither RCR nor clonal dominance were detected in any subjects throughout the study period and the follow-up period.

Figure 4 shows the cell kinetics of TBI-1301 in peripheral blood in individual patients. Following a rise by 7 to 10 days, TCR-gene transduced cell counts fell down and became below the limit of detection in most patients at day 28. Overall, the pharmacokinetic parameters of TBI-1301 were as follows: T_max median (min, max) was 7.00 (6.8–9.9) days, T_last median (min, max) was 17.90 (8.9–58.9) days, C_max mean (SD) was 6,350 (4,090) copies/10⁵ cells, and AUC_t mean (SD) was 61,900 (49,200) copies/10⁵ cells × day.

This open-label, multicenter phase I/II trial showed that, following lymphodepletion treatment with cyclophosphamide, patients with SS who received an infusion of TBI-1301 demonstrated an ORR of 50.0%. A Bayesian approach was adopted to evaluate the efficacy of TBI-1301 because this study targets a rare cancer such as SS. The results of this trial strongly outperformed the efficacy of pazopanib (ORR, 5.7%) used as a historical control (11). It also showed that the OS of patients treated with TBI-1301 was longer than those treated with pazopanib. Although the incidence of adverse events (n = 8/8, 100.0%), serious adverse events (n = 4/8, 50.0%), and adverse drug reactions (n = 7/8, 87.5%) were high, no deaths due to adverse events were observed. Most adverse events occurred early after infusion of TBI-1301 and a low frequency of grade 1 to 2 adverse events occurred in the late phase after the treatment. All eight patients experienced adverse events that were not ruled out to be related to cyclophosphamide but all recovered. Finally, CRS was observed in four of eight patients and they all recovered with prespecified treatment.

Overall, the results of this study generally align with those of previous studies using adoptive transfer of NY-ESO-1-specific TCR engineered T cells for advanced SS (20, 21, 32). In an open-label phase I/II study of 42 patients with SS, an affinity-enhanced TCR against an NY-ESO-1–specific HLA-A*02-restricted peptide led to good response (32). Detailed results from the initial cohort of 12 patients using this affinity-enhanced TCR confirmed an ORR of 50% comprising one patient with CR and five patients with PR, which was also associated with tumor shrinkage over several months (20).

Despite apparent similarities, cellular immunotherapy associated with TBI-1301 employed methods distinct from those of previous studies. Adoptive cellular immunotherapies using NY-ESO-1–specific TCR has room to improve (22), such as by addressing mispairing of the introduced TCR α/β chains or manufacturing TCR-engineered T cells. In this phase I/II study, new technologies, including the use of siTCR vectors (23) and RetroNectin (28) were implemented in the manufacturing process of TBI-1301. Using the previously developed siTCR technology, it was possible to suppress the expression of endogenous TCR and avoid mispairing (23). Using a unique culture method incorporating RetroNectin, TBI-1301 was produced from a small amount (200 mL) of collected blood without an apheresis process, which typically involves the processing of 12 to 15 L of blood (33). The manufacturing method using whole blood is also preferable to apheresis in terms of burden of treatment and cost. The targeted number of cells was obtained in all cases, and it was thereby possible to infuse TBI-1301 intravenously to all eight patients. As a result, the success rate of the TBI-1301 preparation used in this study was 100%.

In cases of unresectable and/or metastatic STS, anticancer drugs used as first-line treatment include doxorubicin, ifosfamide, and dacarbazine, which are administered alone or in combination to prolong OS (2, 10, 34). However, significant differences in the OS obtained from these combination treatments and doxorubicin monotherapy have not been consistently observed (35), which may partly explain the lack of change in patient survival over recent decades (3, 36). Moreover, the use of combination therapy may be limited by adverse effects and reduced feasibility. In the 2010s, new treatment options like pazopanib, trabectedin, and eribulin have been used second-line or later for patients with advanced STS (9, 14), although trabectedin and eribulin have not been approved for the treatment of SS in the U.S. (37, 38). They have not been tested for efficacy in SS but have been shown in subgroup analyses to be effective (10). In this study, the response rate of treatment could be recognized only for SS, and good PFS was also shown despite the study design with eight cases. TBI-1301 could be also used second-line or later for patients with advanced STS.

In clinical studies of CAR T cells, CRS occurred in over 50% of patients, and grade 3 or higher CRS occurred in more than 20% of patients (39, 40). In studies of TCR T cells, one study reported CRS (20), whereas another reported no CRS where all patients had at least one adverse event considered to be a potential symptom of CRS, and grade 3 severity occurred in eight of 25 patients (41). In this trial, CRS was reported in 50.0% of patients, although all patients with CRS were grade 1 or 2 in severity and adequately managed. Pretreatment with cyclophosphamide in this study was milder than in other trials of CAR T and TCR-T engineered T cells (32, 39–41). Pretreatment for lymphocyte depletion is one risk factor associated with the development of CRS (42), and a milder regimen might affect the severity. Regarding the mechanism for CRS in CAR T therapy, cytokines may be produced by monocytes and/or macrophages in the host (43), although cytokine analysis of TCR-T has not yet been reported in detail. The pathogenesis of CRS may also depend on tumor burden (43).

TCR T-cell therapy has potential risk to develop an autoimmune reaction derived from mispairing of the introduced TCR with endogenous TCR (44). TBI-1301 was manufactured using siTCR vectors to suppress the expression of endogenous TCR and mispairing, which may reduce the potential risk of autoimmune reaction. Suppression levels of endogenous TCR were confirmed by a quantitative PCR method in the whole cell products of three lots, and the results were 34.3% to 54.8% and 26.1% to 44.3% in TCR-α and TCR-β, respectively. Throughout the study, no symptoms suggestive of autoimmune reaction were observed.

Neurologic symptoms can result from a variety of pathologic processes (e.g., hepatotoxicity, electrolyte imbalances) following adoptive cellular immunotherapy. Adverse events caused by the development of ICANS include encephalopathy, seizure, dysphasia, tremor, headache, confusion, depressed level of consciousness, and cerebral edema (25, 45, 46). ICANS has been reported in other clinical studies using engineered T cells (39, 40), and in some studies using anti–NY-ESO-1 TCR T-cell therapy (41). Considering the mechanism for the development of ICANS (46), it may develop if CRS becomes severe (45, 46). Also, the development of Guillain-Barre syndrome has been recently reported in a clinical study (47), although no symptom was reported in this trial.

Regarding cell kinetics of TBI-1301, the median T_last was 17.90 days. By contrast, another study reported that the infused anti–NY-ESO-1 TCR T cells persisted in the blood for at least 6 months (20). A milder lymphodepletion regimen using cyclophosphamide alone may be one cause associated with the shorter persistence, so stronger regimens, such as adding fludarabine have a potential to increase the persistence and enhance the efficacy. Despite the brief circulation period in blood, the effect of TBI-1301 appeared to continue even after levels in the blood were below the limit of detection. That could be due to efficient infiltration of TBI-1301 into the tumor microenvironment (48) and enhancement of the immune effect by antigen spreading (49). Also, the development of self-renewing T-cell populations, such as memory cells, may have a positive effect on efficacy (50).

In conclusion, TBI-1301 consists of autologous lymphocytes genetically engineered with a NY-ESO-1-reactive TCR designed to selectively target NY-ESO-1–positive tumor. In this study, TBI-1301 was infused to patients with advanced or recurrent SS that could not be surgically resected and was resistant to anthracycline-based regimens. The efficacy of the treatment, in terms of ORR and OS, was strongly superior to that of pazopanib and in line with results of other adoptive immunotherapy regimens directed against the NY-ESO-1 antigen that have been previously reported. Regarding the safety profile, including CRS, TBI-1301 was considered acceptable because most adverse events were manageable with procedures prepared in advance.

Additional confirmation, ideally with larger numbers of patients, is needed, but this study and the phase I study of TBI-1301 that demonstrated favorable efficacy against SS (26) support that adoptive immunotherapy based on TBI-1301 infusion is expected to be a promising new strategy for the treatment of SS.

A. Kawai reports personal fees from Takara Bio, Boehringer Ingelheim, and Otsuka Pharmaceutical during the conduct of the study and personal fees from Daiichi Sankyo outside the submitted work. M. Ishihara reports personal fees from Eisai, Chugai Pharma, MSD, Ono Pharmaceutical, Eli Lilly, and Daiichi Sankyo outside the submitted work. S. Kitano reports grants and personal fees from AstraZeneca, Pfizer, Boehringer Ingelheim, MSD, Eisai, GlaxoSmithKline, Daiichi Sankyo, Chugai, Takeda, and Eli Lilly Japan K.K.; personal fees from Taiho, Novartis, Sumitomo Pharma, Bristol-Myers Squibb, Rakuten Medical, ImmuniT Research Inc., Merck KGaA, United Immunity, Novartis, and Janssen; and grants from Astellas, Takara Bio Inc., Incyte, LOXO Oncology, and AbbVie outside the submitted work. K. Takada reports grants from Takara Bio Inc. during the conduct of the study; personal fees from Chugai Pharmaceutical Co. Ltd., Eli Lilly Japan K.K., Eisai Co. Ltd., Janssen Pharmaceutical K.K., Ono Pharmaceutical Co. Ltd., Takeda Pharmaceutical Co. Ltd., and Daiichi Sankyo Co. Ltd. outside the submitted work. K. Kato reports receiving consulting fees from AbbVie, AstraZeneca, Celgene, Chugai Pharmaceutical Co. Ltd., Daiichi Sankyo Co. Ltd., Eisai Co. Ltd., Janssen, Novartis, and Ono Pharmaceutical Co. Ltd.; honoraria (e.g., lecture fees) from Takeda Pharmaceutical Co. Ltd., MSD Pharmaceuticals, Kyowa Kirin Co. Ltd., Janssen, Celgene, Ono Pharmaceutical Co. Ltd., Mundipharma Pty Ltd., Sumitomo Dainippon Pharma Co. Ltd., and Bristol Myers Squibb; and research funding from AbbVie, Celgene, Chugai Pharmaceutical Co. Ltd., Daiichi Sankyo Co. Ltd., Eisai Co. Ltd., Janssen, Kyowa Kirin Co. Ltd., Novartis, Ono Pharmaceutical Co. Ltd., and Takeda Pharmaceutical Co. Ltd. outside the submitted work. M. Endo reports personal fees from Takara Bio Inc. during the conduct of the study and personal fees from Bayer AG, Daiichi Sankyo Co. Ltd., Kyowa Kirin Co. Ltd., Eisai Co. Ltd., Taiho Pharmaceutical Co. Ltd., Boehringer Ingelheim International GmbH, and Novartis AG outside the submitted work. Y. Miyahara reports other support from Takara Bio Inc. during the conduct of the study. K. Morino reports personal fees from Takara Bio Inc. during the conduct of the study. S. Takahashi reports personal fees from Takara Bio Inc. during the conduct of the study; personal fees from Takara Bio Inc. outside the submitted work. F. Matsuo reports personal fees from Statcom Co. Ltd. during the conduct of the study. No disclosures were reported by the other authors.

A. Kawai: Investigation. M. Ishihara: Investigation. T. Nakamura: Investigation. S. Kitano: Investigation. S. Iwata: Investigation. K. Takada: Investigation. M. Emori: Investigation. K. Kato: Investigation. M. Endo: Investigation. Y. Matsumoto: Investigation. S. Kakunaga: Investigation. E. Sato: Investigation. Y. Miyahara: Investigation. K. Morino: Project administration. S. Tanaka: Project administration. S. Takahashi: Project administration. F. Matsuo: Formal analysis. A. Matsumine: Conceptualization. S. Kageyama: Supervision. T. Ueda: Investigation.

The authors thank all clinicians for their involvement and contribution to the clinical trial. This article is dedicated to the memory of Drs. Hiroshi Shiku and Yasuo Ohashi. Dr. Hiroshi Shiku made a great contribution for his valuable and constructive suggestions during the development of TBI-1301. He passed away in 2022. Dr. Yasuo Ohashi made a great contribution to the design of this study, especially to biostatistical field using Bayesian methods. He passed away in 2021.

The authors also thank Masanobu Kimura and Maki Tanaka for helping with manuscript preparation.

This study was funded by Takara Bio Inc.

This work was supported by Takara Bio Inc., Shiga, Japan.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.