Efficacy and safety of autologous CD5-KO anti-CD5 CAR-T cells in relapsed/refractory CD5⁺ hematological malignancies

Summary

Chimeric antigen receptor (CAR)-T cell therapy targeting antigens shared with normal T cells requires genetic modifications to prevent fratricide. This phase 1 trial evaluates autologous CD5-targeting CAR-T cells with CD5 gene deletion (CT125A) in seven patients with relapsed/refractory CD5⁺ hematologic malignancies. The overall response rate is 85.7%, including four complete responses. All patients experience cytokine release syndrome (six grade 1–2, one grade 3), and two patients develop immune effector cell-associated neurotoxicity syndrome. The most common grade ≥3 adverse events are cytopenia and infection, with unique observations of rash and autoimmune-related events. Post-infusion immunophenotyping shows persistent depletion of CD5⁺ T cells and CD19⁺ B cells, with reduced CD4/CD8 ratios. The human CD5 knockin murine model reveals skin lesions without significant vital organ involvement. These findings demonstrate CT125A’s therapeutic potential in CD5⁺ malignancies while highlighting the need for safety optimization. The trial has been registered at ClinicalTrials.gov (NCT04767308).

Graphical abstract

Highlights

• •CT125A achieves 85.7% response rate in relapsed/refractory CD5⁺ malignancies
• •CD5 gene deletion prevents fratricide and enhances CAR-T cell persistence
• •Prolonged CD5⁺ T cell aplasia associates with infections and autoimmune events
• •Mouse model reveals on-target, off-tumor effects primarily affecting skin tissue

Cheng et al. report a phase 1 trial of autologous CD5-targeting CAR-T cells with CD5 gene deletion (CT125A) in seven patients with relapsed/refractory CD5⁺ malignancies. CT125A achieves an 85.7% response rate but causes prolonged immunosuppression, infections, and autoimmune events, highlighting the need for safety optimization strategies.

Introduction

CD5, a cysteine-rich scavenger receptor, plays a multifaceted role in modulating immune responses. Its expression is mainly restricted to thymocytes, T cells, and a small subpopulation of B cells (B1 cells).^1,2 Functionally, CD5 modulates immune responses by tuning T cell receptor signaling,³ priming the effect of dendritic cells (DCs) on T cells,⁴ and shaping the regulatory functions of B1 cells.⁵ CD5 is broadly expressed across hematologic malignancies, most notably in 85% of T cell malignancies,⁶ nearly all mantle cell lymphomas (MCLs),⁷ and a subset of diffuse large B cell lymphomas (DLBCLs).⁸ Its presence in B cell malignancies is associated with more aggressive clinical phenotypes and inferior outcomes, underscoring its potential as both a prognostic marker and a therapeutic target.

Given this, CD5 has become an attractive target for chimeric antigen receptor (CAR) T cell therapy, particularly in the context of T cell malignancies. A phase 1 clinical trial (NCT03081910) assessed the safety profile and preliminary efficacy of autologous T cells engineered to express a murine-derived single-chain variable fragment (scFv)-based CAR targeting CD5 in patients with refractory or relapsed disease. Among the nine evaluable patients, four patients demonstrated an objective response, including three patients who achieved complete remission (CR) and one patient who achieved partial remission (PR). However, these responses were often not durable in the absence of bridging hematopoietic stem cell transplantation (HSCT).⁶ Notably, several patients relapsed with CD5-positive disease despite detectable CAR transgene expression for up to 9 months, highlighting the need for strategies that enhance in vivo persistence and long-term efficacy of CD5 CAR-T cells.

Preclinical studies have demonstrated that CD5 deletion or incorporation of tyrosine kinase inhibitors (TKIs) into the CAR-T cell culture process may enhance the in vivo persistence and anti-tumor efficacy of CAR-T cell therapies.^9,10,11 Translating these findings into the clinic, the follow-up clinical trial NCT05032599 explored the use of allogeneic CD5 gene-deleted anti-CD5 CAR-T cells in patients with relapsed/refractory (r/r) T cell acute lymphoblastic leukemia (T-ALL).¹² Notably, this study reported a remarkable 100% CR rate across all 19 enrolled patients. Furthermore, relapses were infrequent, occurring in only three patients during a median follow-up duration of 14.3 months, and were characterized by the loss of CD5 antigen expression, thereby indicating the potent anti-tumor activity of the CD5 CAR-T cell therapy. However, despite the high CR rates and reduced relapse, most patients underwent HSCT consolidation. Additionally, these regimens were complicated by treatment-related toxicities, primarily life-threatening infections, which warrant careful consideration in clinical applications.

In this study, we present findings from the phase 1 clinical trial (NCT04767308) evaluating the safety and efficacy of CT125A in patients with r/r CD5⁺ hematologic malignancies, without the requirement for subsequent HSCT. Notably, this approach allows for the assessment of CD5 CAR-T cell therapy as a standalone intervention, potentially expanding treatment access. Beyond safety and efficacy outcomes, we delve into the uncommon adverse events (AEs) observed and characterize post-infusion immune reconstitution using patient data and human CD5 (hCD5) knockin murine models. These comprehensive investigations aim to optimize the therapeutic profile and enhance the safety of CD5-targeted CAR-T cell therapies.

Results

Patient enrollment and CT125A manufacturing

As outlined in the consort diagram (Figure 1A), from February 2021 to October 2021, a total of seven patients were screened, resulting in the enrollment of seven patients with r/r CD5-positive T cell or B cell malignancies. All enrolled patients had CT125A successfully manufactured and received a single-dose infusion as per the dosing levels of 1.0 × 10⁵ cells/kg, 1.0 × 10⁶ cells/kg, and 2.0 × 10⁶ cells/kg. Detailed characteristics of the CAR-T cell products, such as cell viability, percentage of CAR-positive T cells, average CAR copy numbers, and IFN-γ secretion, are provided in Table S1. The median time from the start of manufacturing to product release was 11 days (range: 10–12 days). The median percentage of CAR-positive cells was 28.33% (range: 21.10%–49.60%). Before infusion, all products were confirmed to be free of infections by sterility tests.

Figure 1.

The efficacy of CT125A

(A) Flow chart showing the processes from patient enrollment to treatment. Three dosing levels were initially designed. However, as the trial progressed, 1e5/kg was tested.

(B) The follow-up and efficacy of each patient. NA, not assessed; SD, stable disease; PD, progression disease; PR, partial remission; CR, complete remission; HSCT, hematopoietic stem cell transplantation.

(C) Flow cytometry results of bone marrow (BM) from patient 2 at the indicated time points. P1 represents abnormal B cells. Kappa and lambda show the clonality of gated cells.

(D) Flow cytometry results of lymph node (the upper row) and abdominal effusion (the lower row) of patient 3. P1 represents abnormal B cells. The cells from P1 were lambda positive, indicative of monoclonal cells.

(E) PET/CT images of patient 6 before and 1 month after CAR-T cell treatment.

See also Figure S1.

Baseline characteristics of patients enrolled

Seven patients enrolled, presenting a range of disease types: one with anaplastic large cell lymphoma (ALCL), two with angioimmunoblastic T cell lymphoma (AITL), one with subcutaneous panniculitis T cell lymphoma (SPTCL), two with MCL, and one with DLBCL. Key demographic and clinical characteristics are summarized in Table 1. Patients 4, 5, and 6 correspond to patients 2, 1, and 3, respectively, in the previously published article.¹³ The cohort consisted of four males and three females, with a median age of 47 years (range: 17–67 years old). All patients had an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 at the enrollment. The median line of prior therapies was 3 (range: 2–6). Notably, patient 2 had undergone prior anti-CD19 and anti-CD22 CAR-T cell therapy, and patient 3 had received autologous HSCT. At the time of enrollment, six patients had progression disease (PD), while one patient (patient 5) had stable disease (SD). The expression of CD5 on tumor cells was validated as positive by flow cytometric analysis or immunohistochemical staining across all cases (Figures 1 and S1). Six patients underwent genetic testing at diagnosis or relapse, and all harbored high-risk genetic abnormalities. These included mutations in TP53, KMT2D, NOTCH1, TET2, RHOA, IDH2, DNMT3A, CARD11, and SETD2 genes. Additionally, breakage or rearrangement was observed in CCND1 and IgH genes, and amplifications were detected in BCL2 and IgH genes. All patients were asymptomatic for infectious diseases at the time of CT125A administration, with the exception of patient 2, who presented with recurrent fever 10 days prior to CAR-T cell infusion and was subsequently diagnosed with John Cunningham (JC) virus polyomavirus infection through next-generation sequencing (NGS) analysis of a peripheral blood (PB) sample.

Table 1.

Baseline characteristics of patients

Characteristics	Pt 1	Pt 2	Pt 3	Pt 4	Pt 5	Pt 6	Pt 7
Age (year)	17	50	36	47	49	31	67
Gender	M	M	M	F	M	F	F
Diagnosis	ALCL	MCL	MCL	AITL	AITL	SPTCL	DLBCL
ECOG score	2	2	2	2	2	2	2
Prior lines of therapy	3	6	4	2	3	6	3
Prior HSCT?	no	no	yes	no	no	no	no
Prior CAR-T therapy?	no	yes	no	no	no	no	no
Disease status at enrollment	PD	PD	PD	PD	SD	PD	PD
BM blasts (FCM)	0%	24.50%	0%	0%	0%	0%	0%
EMD	LN	LN	GI, LN	LN	LN, Subq tissues	Subq tissues	LN
High-risk genetic abnormalities	N/A	IgH/CCND1 rearrangement	TP53 and KMT2D mutations	BCL2 and IGH expression	TET2, RHOA, IDH2, DNMT3A, and CARD11 mutations	TET2 and SETD2 mutations	CCND1 rearrangement
Microbial detection	Neg	JC virus	Neg	Neg	Neg	Neg	Neg
Infused dosage (cells/kg)	1.16E+06	2.15E+06	2.20E+06	1.18E+06	8.10E+05	1.02E+06	1.00E+05
Product status	Cryo	Cryo	Cryo	Cryo	Cryo	fresh	fresh

M, male; F, female; ALCL, anaplastic large cell lymphoma; MCL, mantle cell lymphoma; AITL, angioimmunoblastic T cell lymphoma; SPTCL, subcutaneous panniculitis T cell lymphoma; DLBCL, diffuse large B cell lymphoma; PD, progression disease; FCM, flow cytometry; EMD, extramedullary disease; LN, lymph node; GI, gastrointestinal tract; Subq, subcutaneous; Neg, negative; Cryo, cryopreserved.See also Figure S1; Table S1.

The clinical efficacy of CAR5^5ko-T cells in patients

The overall response rate was 85.71%, with CR observed in four patients and PR in two patients (Figure 1B). Patient 1 remains alive and disease-free for more than 902 days at the end of follow-up. Patients 2 and 6, who initially achieved CR and PR, respectively, eventually experienced disease relapse at day 170 and day 182 following CAR-T cell infusion. Figure 1C illustrates the clearance of tumor cells in the bone marrow of patient 2, as confirmed by flow cytometry (FCM) 2 months post infusion. Figure 1E presents the positron emission tomography/computed tomography (PET/CT) imaging results of patient 6, pre-treatment and 1-month post treatment, which revealed a substantial reduction in tumor burden and metabolic activity after therapy. At relapses, the PB CAR copy number was 2,925 and 1,603 copies/μg DNA for patients 2 and 6, respectively, accompanied by persistent CD5⁺ T cell aplasia. Moreover, CD5 was negative in the recurrent lesion for patient 6 (Figure S1) and not evaluated for patient 2. For patient 3, disease was initially controlled at day 14 post CAR-T cell therapy but progressed rapidly at day 25, with 90.4% CD5-negative monoclonal tumor cells detected in the ascites (Figure 1D). Meanwhile, the CAR copy number was 5,402 copies/μg DNA in blood and 341 copies/μg DNA in ascites, and CD5⁺ T cells were absent in blood. Patient 4, who had achieved CR, died of suspected infection at day 124 post therapy. Specifically, during the follow-up, patient 4 had vomiting (6–8 times per day, with gastric contents) and diarrhea (8–10 times per day, with watery yellow stool) for a week and developed fever (highest temperature of 41°C) for 4 days before admission to our hospital. The subsequent laboratory examinations identified markedly elevated inflammatory markers with procalcitonin at 14.7 ng/mL and C-reactive protein exceeding 320 mg/L, but no bacteria grew in the blood culture. One day after supportive and antibiotic therapy, the patient died with multi-organ failure. Patient 5 proceeded to allogeneic HSCT after achieving PR (assessed by ultrasound evaluation), at day 109 post CT125A infusion, and was subsequently withdrawn from the clinical trial. Patient 7 achieved CR; however, this patient subsequently developed severe pneumonia that progressed to multiple organ failure, resulting in mortality at day 106 post treatment. Overall, the median progression-free survival (PFS) and overall survival (OS) were 170 days and 205 days, respectively, with follow-up duration ranging from 43 to 902 days.

Safety profile of patients treated with CT125A

The administration of CT125A was well tolerated, with no immediate toxicities observed within the first 4 h post infusion. All patients experienced cytokine release syndrome (CRS), with six cases of grade 1–2 and one case of grade 3. Immune effector cell-associated neurotoxicity syndrome (ICANS) occurred in two patients (28.6%) at grade 1 (Table 2) and resolved within 15 days. Cytokine levels, including ferritin, IL-6, and TNF-α, fluctuated moderately within 30 days post infusion (Figure S2A), coinciding with the occurrence of CRS.

Table 2.

Safety profile of all patients

Adverse event	Any grade, n (%)	Grades 3–5, n (%)
CRS	7 (100%)	1 (14.3%)
ICANS	2 (28.6%)	0 (0%)
Hematologic event
Anemia	7 (100%)	5 (71.4%)
Leukopenia	7 (100%)	7 (100%)
Neutropenia	7 (100%)	7 (100%)
Lymphopenia	7 (100%)	7 (100%)
Thrombocytopenia	7 (100%)	5 (71.4%)
Coagulopathy	7 (100%)	2 (28.6%)
Gastrointestinal event
Loss of appetite	4 (57.1%)	0 (0%)
Nausea/vomiting/diarrhea	3 (42.9%)	0 (0%)
Abdominal bloating	0 (0%)	0 (0%)
Abdominal pain	3 (42.9%)	0 (0%)
Cardiovascular event
Chest pain	0 (0%)	0 (0%)
Tachycardia/arrhythmia	2 (28.6%)	0 (0%)
Heart failure	0 (0%)	0 (0%)
Hypotension	0 (0%)	0 (0%)
General condition
Fever	7 (100%)	0 (0%)
Fatigue	4 (57.1%)	0 (0%)
Headache	2 (28.6%)	0 (0%)
Hypoxia	1 (14.3%)	0 (0%)
Laboratory values
ALT increase	4 (57.1%)	1 (14.3%)
AST increase	4 (57.1%)	1 (14.3%)
GGT increase	6 (85.7%)	5 (71.4%)
Bilirubin increase	1 (14.3%)	0 (0%)
Creatinine increase	3 (42.9%)	0 (0%)
LDH increase	6 (85.7%)	0 (0%)
HypoNa/HypoCl/HypoCa	7 (100%)	2 (28.6%)
Infection
Bacterial infectiona	6 (85.7%)	6 (85.7%)
CMV activation	4 (57.1%)	0 (0%)
EBV activation	6 (85.7%)	0 (0%)
BK infection	4 (57.1%)	0 (0%)
JC infection	1 (14.3%)	0 (0%)
Others
Rash	5 (71.4%)	1 (14.3%)
Skin tingling	5 (71.4%)	0 (0%)
Dry eyes	1 (14.3%)	0 (0%)
Xerostomia	2 (28.6%)	0 (0%)
Hypothyroidism	1 (14.3%)	0 (0%)
Adrenal insufficiency	1 (14.3%)	0 (0%)
Blurry vision	1 (14.3%)	0 (0%)

See also Figure S2.

^aTwo patients were clinically diagnosed with bacterial infection, without bacteremia detected.

Other general AEs, such as loss of appetite, nausea, vomiting, diarrhea, abdominal pain, tachycardia, fever, fatigue, headache, and hypoxia, were generally mild and transient. Some patients exhibited abnormal laboratory values, including elevated alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), bilirubin, creatinine, lactate dehydrogenase (LDH), and hyponatremia/hypochloremia/hypocalcemia, with grade 3 or higher severity observed in individual patients (Table 2). All these AEs were resolved with supportive care.

Cytopenia was the most common grade 3 or higher AE, including leukopenia (100%), neutropenia (100%), lymphopenia (100%), thrombocytopenia (71.4%), and anemia (71.4%). Specifically, five patients had first episode of leukopenia, and four experienced recurrent grade 4 leukopenia 1 month post infusion, gradually recovering during follow-up (Figure 2A). Two episodes of grade 4 leukopenia were observed within 2 months, with the first nadir occurring approximately 10 days post infusion and the second episode after 1 month (Figure 2B). Neutrophil counts exhibited a similar trend, while lymphopenia was more severe and prolonged (Figures 2C–2F). After preconditioning, all patients experienced grade 4 lymphopenia, lasting for a median of 10 days. Recovery was inconsistent, with most patients continuing to exhibit grade 3–4 lymphopenia for up to 2 months post infusion. Grade 4 thrombocytopenia was identified in two patients within 1 month post therapy and 3 patients after 1 month (Figure 2G). Generally, thrombocytopenia was not clinically significant within the first month but tended to manifest between day 30 and day 60 (Figure 2H).

Figure 2.

The changes of complete blood count within 2 months post CT125A infusion

(A, C, E, and G) The degree of leukopenia (A), neutropenia (C), lymphopenia (E), and thrombocytopenia (G) of individual patient at different days from pre-fludarabine and cyclophosphamide (FC) conditioning to 60 days post CAR-T infusion. Each column represents a day. Each row represents a patient. Red block, grade 4 aplasia; blue block, grade 3 aplasia; pink block, grade 2 aplasia; green block, grade 1 aplasia; light purple block, normal range; light gray, missing data.

(B, D, F, and H) The median level (red line) and range (the pink shadow) of the amounts of white blood cells (B), neutrophil (D), lymphocyte (F), and platelet (H) of the seven treated patients from pre-FC to 60 days post CAR-T infusion. The upper dashed line represents a grade 3 reduction in cells, while the lower dashed line represents a grade 4 reduction in cells.

Infection emerged as a significant complication associated with prolonged cytopenia. All seven patients experienced either cytomegalovirus (CMV) or Epstein-Barr virus (EBV) reactivation within 2 months post infusion, with six cases of EBV and four cases of CMV reactivation (Table 2). Additionally, four patients had BK virus (BKV) infection, and one patient had concurrent JC virus infection. Bacterial infections were confirmed in four patients and suspected in two patients (patient 4 and 7), resulting in two deaths. Timely identification of pathogens facilitated effective management, as the four confirmed cases survived from the infection after antibiotic therapy.

Several unexpected AEs were observed, including rash (71.4%), skin tingling (71.4%), dry eyes (14.3%), xerostomia (28.6%), hypothyroidism (14.3%), and adrenal insufficiency (14.3%) (Table 2). Rash and skin tingling were managed with cetuximab to switch-off CAR-T cell expansion and ruxolitinib for immunosuppression.¹³ Autoimmune-related symptoms primarily affected two patients. One developed xerostomia, with salivary gland scintigraphy showing deficient secretory function (Figure S2B), suspected of having Sjögren’s syndrome. The other presented with dry eyes, dry mouth, decreased adrenocorticotropic hormone (ACTH) and thyroid hormones, and refractory hyponatremia, suspected of having hypophysitis. Laboratory analysis revealed increased anti-nuclear antibodies (ANAs) and anti-Sjögren’s syndrome antigen A (SSA) antibodies, supporting the diagnosis of Sjögren’s syndrome. Before CAR-T cell infusion, neither patient had reported these symptoms nor undergone related examinations.

Pharmacokinetic profile of CT125A in treated patients

The in vivo dynamics of CT125A were monitored simultaneously using droplet digital PCR (ddPCR) and FCM. Both methodologies yielded consistent results. The results of ddPCR exhibited that following infusion, CAR-T cells expanded rapidly, reaching a median peak copy number of 87,178 per μg DNA (range: 31,618–118,000 copies/μg DNA) at a median of 14 days (range: 11–21 days) (Figure 3A). FCM data revealed a median peak concentration of CAR-T cells of 176.165 cells/μL PB (range: 37.534–1,072.674 cells/μL), with a median time to peak concentration of 14 days (range: 11–21 days) (Figure 3B). During a median follow-up period of 124 days, CAR-T cells were persistently detectable in all patients. This persistence was observed even in instances where cetuximab was administered to eliminate CAR-T cells.¹³ Comparative analysis of patients who achieved sustained remission versus those who experienced relapse, as well as those who succumbed to infection versus those who did not, revealed no statistically significant differences in either the peak CAR-T cell expansion or the area under the curve (AUC) of CAR transgene levels over 28 days (Figure S3). Notably, CAR-T cells persisted for more than 809 days post infusion in patient 1, underscoring the superior persistence of CT125A.¹⁴

Figure 3.

Dynamic changes of CT125A and lymphocyte subsets in patients

(A and B) The copy number of the CAR transgene per microgram of genomic DNA (A) determined by ddPCR and the counts of CAR-T cells per microliter of peripheral blood (B) determined by flow cytometric analysis for each patient across various time points post CAR-T cell infusion.

(C–G) The levels of the indicated immune cell populations (CD3⁺ T cells, CD19⁺ B cells, NK cells, and NKT cells) at different time points post CAR-T infusion, as well as the percentage of CD5⁺ T cells and the ratio of CD4⁺ to CD8⁺ T cells.

See also Figure S3.

Dynamics of lymphocyte subset populations in the PB of patients

To dissect the immune reconstitution processes following CAR-T cell infusion, we monitored the dynamics of key immune cell populations in PB, including CD3⁺ T cells, CD4⁺ T cells, CD5⁺ T cells, CD8⁺ T cells, natural killer (NK) cells (CD3⁻CD56⁺), NKT cells (CD3⁺CD56⁺), and B cells (CD19⁺). Following the transient nadir induced by preconditioning, CD3⁺ T cells recovered to pre-lymphodepletion levels within 1 month and remained stable subsequently. This indicates that CT125A therapy did not reduce CD3⁺ T cell numbers in PB (Figure 3C). Notably, the reconstituted T cells expressed no membrane CD5 protein persistently, in both T cell and B cell tumor patients (Figure 3D). Comparative analyses between patients achieving sustained remission versus those experiencing relapse, as well as between infection-related fatalities versus survivors, demonstrated no statistically significant differences in (1) time to CD5⁺ T cell depletion, (2) duration of CD5⁺ T cell aplasia, or (3) CAR-T cell persistence characteristics (Table S2). Additionally, the CD4/CD8 ratio decreased promptly following CT125A infusion and remained below 1 throughout the entire follow-up, suggesting predominant proliferation of CD8⁺ T cells (Figure 3E). Remarkably, CD19⁺ B cells were rapidly depleted post-CAR-T cell infusion and failed to recover during long-term follow-up in all patients (Figure 3F). NK cells and NKT cells rebounded to pre-lymphodepletion levels after a transient decline attributable to preconditioning (Figure 3G).

Effects of CD5 CAR-T cells in hCD5 knockin murine models

To elucidate the potential mechanisms underlying atypical AEs such as rash and autoimmune reactions, we conducted comprehensive validation studies in hCD5 knockin mice, wherein the extracellular domain of the murine CD5 protein was replaced with the corresponding hCD5 structure. This model enables us to mitigate the impact of CD5 knockout (KO) in clinical settings while preventing fratricide among the manufactured CAR-T cells.

Mouse T cells isolated from GFP knockin mice were transduced with a retrovirus harboring the hCD5-targeting CAR (identical sequence used in clinical trials), resulting in a 35.22% transduction efficiency (Figure S4A). These engineered cells demonstrated robust proliferation and potent anti-tumor efficacy against human CD5⁺ cell lines in vitro (Figure S4B). After in vitro expansion, GFP-positive T cells and hCD5-targeting CAR-T cells were infused into hCD5 knockin syngeneic mice 2 h after receiving 5 Gy total body irradiation (Figure S4C). Radiotherapy was adopted as pre-conditioning in the murine model, which offers a stable, rapid, and convenient approach to lymphocyte clearance, thereby facilitating CAR-T engraftment and expansion.^15,16 PB samples were collected at multiple time points and subjected to spectral FCM to evaluate the temporal dynamics of key immune cell subpopulations. Additionally, hematoxylin and eosin (H&E) and immunohistochemistry (IHC) staining were performed on multi-organ tissue sections at the endpoint (Figure 4A).

Figure 4.

Effects of CD5 CAR-T cells in hCD5 knockin murine models

(A) Schematic illustration of the generation of CD5 CAR-T cells and post-infusion monitoring in murine models.

(B) Hematoxylin and eosin (H&E) staining results of various organs, including the heart, liver, spleen, lung, kidney, skin, and adrenal gland, from mice at the experimental endpoint. Black arrows indicate minor inflammatory cell infiltration; red arrows indicate hair follicles. Scale bars: 200 μm in heart and liver, 400 μm in spleen, lung, and kidney, and 100 μm in skin and adrenal gland.

(C and D) Dynamic changes in the levels of human CD5⁺ T cells, murine CD3⁺ T cells, murine CD19⁺ B cells, and the ratio of CD4⁺ to CD8⁺ T cells in mice measured by spectral flow cytometry. Data are represented as mean ± SEM (n = 5 for CAR-T groups, and n = 2 for the mock T group).

(E) Flow cytometric characterization of T cell subpopulations in vivo at different time points. Data are represented as mean ± SEM (n = 5 for CAR-T groups, and n = 2 for the mock T group).

See also Figures S4–S6.

The infused mCD3⁺GFP⁺ cells were consistently higher in the CAR-T groups (including CAR-T cells and mock T cells) compared with mock T groups (only mock T cells), suggesting that infused CAR-T cells expanded and persisted in the PB of mice (Figure S5A). Mice receiving hCD5 CAR-T cell infusion developed black patches on their skin after hair removal (Figure S5B), although no obvious changes in appearance were observed. Histopathological examination revealed no discernible pathological changes or tissue damage in the heart, liver, adrenal gland, spleen, lung, or kidney. However, skin biopsy exhibited an increased number of hair follicles in the dermal layer and minimal inflammatory cell infiltration (Figure 4B). Further IHC staining revealed that hCD5 was expressed in the dermis layer of the skin but absent in the heart, liver, kidney, lung, and adrenal gland. In the spleen, partial hCD5 was positivity observed, consistent with presence of the resident T cells (Figure S5C). Moreover, CAR-T cells were specifically detectable in the skin section of the CAR-T groups by immunofluorescence analysis (Figure S6). Thus, we postulate that the skin-related AEs in patients resulted from on-target, off-tumor effects of anti-CD5 CAR-T cells.

Following infusion, the endogenous hCD5⁺ T cells were almost completely eliminated by day 4 and gradually recovered during the follow-up period. By day 36, when the percentage of endogenous mouse CD3⁺ T (GFP⁻ mCD5⁻ mCD3⁺) cells in the CAR-T cell group became similar to that in the T cell group, 34.44% of T cells were CD5 positive (Figure 4C). At days 16 and 36, both the CD4/CD8 ratio and the percentage of CD19⁺ cells were decreased in the CAR-T group compared to the T cell group (Figure 4D), consistent with clinical observations.

Given the substantial hCD5-negative T cell populations in mice receiving hCD5 CAR-T cell infusion, we further analyzed the constitution of mCD3-positive T cells. Generally, the proportion of terminal differentiated effector memory re-expressing CD45RA (EMRA) T cells was sustained at elevated levels in hCD5 CAR-T-treated mice over time, compared to the control T cell group (Figure 4E). Specifically, at day 16, the proportions of EMRA cells in the hCD5 CAR-T cell group were significantly higher compared to the T cell control group, with CD4 EMRA cells at 16.50% versus 6.66% and CD8 EMRA cells at 21.72% versus 8.61%. By day 36, this trend persisted, with CD4 EMRA cells in the CAR-T group reaching 19.23% compared to 5.66% in the control group and CD8 EMRA cells at 19.57% versus 13.88% in the T cell group. CD4⁺ and CD8⁺ EMRA T cells were associated with chronic inflammation, immune senescence, or autoimmune diseases.^17,18,19 Concurrently, the proportions of naive and central memory CD4⁺ and CD8⁺ T cells in both groups gradually decreased (Figure 4E). These findings suggest a shift in the reconstituted T cell compartment from stemness toward effector function, which may account for the dysregulated immune functions of the patients.

Discussion

In this study, we demonstrated the feasibility of generating second-generation CD5-KO anti-CD5 CAR-T cells from peripheral blood mononuclear cells (PBMCs) of heavily pretreated patients with hematological malignancies, even in those relapsed after CAR-T cell or HSCT therapy. Furthermore, we demonstrated the therapeutic efficacy and safety of CT125A in patients with CD5-positive r/r malignancies. Infections emerged as a critical issue in this clinical context. Notably, unusual AEs, primarily autoimmune-related symptoms, were reported in CAR-T cell clinical trials and were hypothesized to be associated with the failure of CD5-positive T cell recovery in patients. Additionally, our in-depth study using an hCD5 knockin murine model revealed that CAR5-T cells exhibit a certain degree of on-target, off-tumor effects, particularly in the skin.

Enhanced response rate, accompanied by superior expansion and persistence of CAR-T cells, was identified in this study. When engineering CAR-T cells to target self-expressed antigens such as CD5, CD7, and CD70, deletion of the target antigen prevents CAR-T cells from fratricide, thereby enhancing their ex vivo expansion and in vivo anti-tumor efficacy by reducing activation-induced exhaustion. Furthermore, CD5 blockade has been demonstrated to augment the antitumor efficacy of CD5-specific CAR-T cells, as well as other target-specific CAR-T cell therapies.^10,20 Other factors, such as the CAR structure, immunogenicity of CAR molecules, and the manufacturing procedure, also influence the clinical responses, which warrants further investigation.

Risks for severe infection were highlighted in this report. It can be occurred within 30 days post CAR-T cell infusion, as well as 60 days later, reminding us to keep this issue in mind during the long-term follow-up. Severe cytopenia may account for the incidence of infection, as suggested in Pan’s study. Pre-FC chemotherapy probably was the main contributor of cytopenia within 30 days post therapy, which was identified in several CAR-T clinical trials.²¹ The second-round decline in neutrophils and lymphocytes, which occurred 30 days post infusion, was suspected to be attributable to hematopoietic toxicity associated with the CD5-targeted CAR-T cell therapy. Furthermore, infection-induced inflammatory processes may exacerbate hematological toxicity, which subsequently compromises the patient’s anti-microbial immune functions, resulting in uncontrolled infection. Therefore, preventing the recurrent decline of blood profile is crucial for reducing the incidence of severe infection in the later stage of this therapy. Despite reducing the therapeutic dose to 1.0 × 10⁵ cells/kg during the trial (patient 7), severe infections persisted, accompanied by robust CAR-T expansion and recurrent grade 4 lymphopenia. Consequently, mitigating the hematological toxicity associated with the CT125A product remains an imperative priority for optimizing clinical outcomes.

Immune functions of these patients were also an essential aspect requiring further inspection, as notified that all patients experienced viral infection or reactivation. The total counts of CD3⁺ T cells in PB recovered to baseline levels 30 days post infusion but with CD5⁺ T cells and B cells persistently absent. A versatile inspection of the anti-infectious functions of these compensatorily expanded CD5-negative lymphocytes contributes to identifying patients at risk for infection. Studying the association between the infection and the characteristics of patients, such as the intrinsic feature of immune cells, the baseline hemogram, the age, etc., may also help to predict high-risk patients. Quick precision diagnosis of infected microbials leads to early effective treatment, which perhaps decreases the associated fatality.

Reasons behind the unusual AEs, including rash and autoimmune-associated symptoms (dry eyes, xerostomia, hypothyroidism, adrenal deficiency, and syndrome of inappropriate secretion of antidiuretic hormone), need to be clarified. In Pan’s study, rash was also observed and was suggested to be a manifestation of graft-versus-host disease (GVHD). However, our study utilized autologous CAR-T cells, and similar symptoms were observed. During our clinical trial, the patients’ rash symptoms were alleviated by controlling the expansion of CT125A with cetuximab.¹³ Notably, CAR-T cells were not fully eliminated by administering cetuximab, which may be due to reduced NK cells in PB and, therefore, diminished killing of CT125A through cetuximab-dependent NK cell-mediated cytotoxicity effects. Additionally, CD5 was positive in the skin tissues in our hCD5 mice model. These collectively indicated that the on-target, off-tumor effects of CAR-T cells may account for these AEs. For the AEs involving hypophysitis and Sjögren’s syndrome, we speculate that they correlate with the immune dysregulation caused by unrecovered normal CD5⁺ T cells and B cells due to long-term persistence of CT125A in vivo. CD5 null mice are known to develop autoimmune diseases,²² and upregulation of CD5 expression in T cells and B cells protects against autoimmunity.²³ This hypothesis was further validated in the hCD5 knockin mouse model, where CAR-T cells did not undergo CD5 molecule KO, and the mice’ own CD5⁺ T cells were able to recover. In this model, no autoimmune-related AEs were observed. Therefore, timely and precise deactivation of CAR-T cells in patients, allowing the reconstruction of their own CD5-positive lymphocytes, is crucial for improving this situation.

In summary, our findings highlight the therapeutic potential and favorable safety profile of CT125A in the treatment of hematologic malignancies. While CAR-T cell therapy has demonstrated robust in vivo expansion and significant anti-tumor activity, the emergence of post-remission complications, including viral reactivation and severe infections, poses substantial challenges to the long-term viability of this immunotherapeutic approach. These complications may critically impair the durability of therapeutic responses and patient safety. Consequently, our data underscore the importance of modulating the persistence duration of CD5 CAR-T cells in patients and advocate for HSCT as a consolidative therapy for those achieving remission. These strategies serve as crucial safeguards to mitigate the risks of prolonged immunodeficiency and enhance the overall therapeutic framework, thereby optimizing the clinical outcomes and ensuring patient safety.

Limitations of the study

This study is subject to several limitations. First, the restricted cohort size and pronounced interpatient variability at baseline diminish the reliability of the findings, hindering a thorough evaluation of clinical determinants influencing therapeutic outcomes, hematologic toxicities, and infectious complications. Additionally, the association between CAR-T cell kinetics and CD5⁺ T cell aplasia with the clinical outcomes cannot be well analyzed. Subsequent trials should be initiated following refinement of the treatment protocol. Second, relapse mechanisms were not comprehensively examined. Tumor samples were unavailable for patient 2 who relapsed, and the precise mechanisms behind the antigen negative relapses in patients 3 and 6 were not further investigated. It may represent either clonal selection under CAR-T pressure or acquisition of CD5 gene mutation during treatment. Third, the functional role of CD5-negative T cells in the mature peripheral T cell population remains inadequately characterized, requiring more extensive mechanistic studies. Finally, the etiologies underlying fatal infections, persistent cytopenia, and rare AEs have yet to be fully elucidated. Addressing these gaps could yield valuable insights for optimizing the efficacy and safety of CAR-T cell therapy.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Xiaojian Zhu (zhuxiaojian@hust.edu.cn).

Materials availability

Plasmids generated in this study will be available on request through completion of a material transfer agreement.

Data and code availability

• •Relevant data supporting the findings of this study are available within the article and supplemental materials.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

X. Zhu was supported by a grant from the National Natural Science Foundation of China (grant no. 82270183). W.M. was supported by a grant from the National Natural Science Foundation of China (grant no. 82100247) and two grants from Tongji Hospital (grant nos. 2024A30 and 24-2KYC13060-23). S.D. was supported by a grant from the National Natural Science Foundation of China (grant no. 82100241).

This paper is dedicated to the late Professor Jianfeng Zhou, who passed away on March 27, 2022. His profound mentorship, pioneering work, and unwavering dedication to cellular immunotherapy continue to inspire all who worked with him. We honor his legacy and significant contributions to the field.

The authors extend their special thanks to the patients who participated in the study, as well as the physicians, nurses, researchers, and other team members at the hospital for their invaluable contributions and support, which made this research possible.

Author contributions

X. Zhu designed and supervised the clinical trial; X.Z., Y.X., and J.W. recruited and provided patient care; W.M. and J.C. oversaw the CAR-T cell production and quality control; J.C., L.Z., S.D., and X.M. collected and/or analyzed clinical data and performed analyses; J. Wu, W.M., J.C., and L.Z. performed the experiments and illustrated data in murine models. W.M. and J.C. wrote the manuscript; and W.M., J.C., Y.Z., and X. Zhu revised the manuscript. All authors approved the final version of the manuscript and are accountable for all aspects of the work.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
V500-c anti-human CD45 (clone 2D1)	BD Bioscience	Cat# 662912
PercP-Cy5.5 anti-human CD3 (clone SK7)	Agilent	Cat# 8931015
FITC anti-human CD4 (clone SK3)	BD Bioscience	Cat# 340133; RRID: AB_400007
APC-Cy7 anti-human CD8 (clone SK1)	BD Bioscience	Cat# 663521
PE anti-human CD56 (clone NCAM16.2)	BD Bioscience	Cat# 652825
BV605 anti-human CD5 (clone L17F12)	Biolegend	Cat# 364020; RRID: AB_2565941
APC anti-human CD19 (clone HIB19)	Agilent	Cat# 8930007
PE-Cy7 anti-human CD19 (clone HIB19)	Agilent	Cat# 8930015
APC-Cy7 anti-human CD20 (clone L27)	BD Bioscience	Cat# 335829; RRID: AB_2868690
APC anti-human CD22 (clone S-HCL-1)	BD Bioscience	Cat# 340933; RRID: AB_400545
PercP-Cy5.5 anti-human CD38 (clone HIT2)	Biolegend	Cat# 303522; RRID: AB_893314
FITC and RPE anti-human Kappa/Lambda (Polyclonal)	Dako	Cat# FR481
Alexa Fluor 488 anti-human EGFR (clone AY13)	Biolegend	Cat# 352908; RRID: AB_11126165
PercP anti-human CD45 (clone 2D1)	BD Bioscience	Cat# 652803
APC anti-human CD3 (clone SK7)	BD Bioscience	Cat# 652815
Spark Blue™ 550 anti-mouse CD3 (clone 17A2)	Biolegend	Cat# 100259; RRID: AB_2819767
APC anti-mouse CD3 (clone 17A2)	Biolegend	Cat# 100236; RRID: AB_2561456
PE anti-mouse CD4 (clone GK1.5)	Biolegend	Cat# 100407; RRID: AB_312692
BV510 anti-mouse CD5 (clone 53–7.3)	Biolegend	Cat# 100627; RRID: AB_2563930
Percp-Cy5.5 anti-mouse CD8a (clone 53–6.7)	BD Bioscience	Cat# 551162; RRID: AB_394081
BV750 anti-mouse CD19 (clone 6D5)	Biolegend	Cat# 115561; RRID: AB_2813978
Alexa Fluor 700 anti-mouse CD44 (clone IM7)	Biolegend	Cat# 103025; RRID: AB_493712
BV570 anti-mouse CD62L (clone MEL-14)	Biolegend	Cat# 104433; RRID: AB_10900262
Anti-CD5 antibody	Abcam	Cat# ab75877; RRID: AB_1310059
Goat Anti-Rabbit IgG H&L (HRP)	Abcam	Cat# ab205718; RRID: AB_2819160
Anti-CD3 antibody	Abcam	Cat# ab237721; RRID: AB_3662950
Anti-mCherry antibody	Abcam	Cat# ab213511; RRID:AB_2814891
Chemicals, peptides, and recombinant proteins
ACK Lysis Buffer	Thermo Fisher	Cat# A1049201
β-mercaptoethanol	Gibco	Cat# 21985023
MEM non-essential amino acids (NEAA)	Gibco	Cat# 11140050
Sodium Pyruvate (100mM)	Gibco	Cat# 11360070
HEPES	Gibco	Cat# 15630080
Recombinant Human IL-2	PeproTech	Cat# 200-02-100ug
GlutaMAX™	Gibco	Cat# 35050061
Retronectin	TAKARA	Cat# RetroNectin®-C T100AC
Lipofectamine™ 2000 Transfection Reagent	Invitrogen	Cat# 11668019
iFluor® 647 tyramide	aatbio	Cat# 11066
iFluor® 488 tyramide	aatbio	Cat# 11060
Critical commercial assays
EasySep™ Mouse T cell Isolation Kit	Stemcell Technologies	Cat# 19851
Dynabeads™ Mouse T-Activator CD3/CD28	Gibco	Cat# 2683905
D-luciferin sodium	Solarbio	Cat# D9390
DAB Horseradish Peroxidase ColorDevelopment Kit	beyotime	Cat# P0203)
Experimental models: Cell lines
Molt-4 cell line	ATCC	Cat# CRL-1582
SUP-T1 cell line	ATCC	Cat# CRL-1942
Plat-E cell line	ATCC	Cat# CRL-1573
Molt-4-luciferase cell line	Constructed by our lab	N/A
SUP-T1-luciferase cell line	Constructed by our lab	N/A
Experimental models: Organisms/strains
Mouse: C57BL/6-Tg (CAG-EGFP)1Osb/J	Shulaibao Biotechnology	N/A
Mouse: C57BL/6JGpt-Cd5em1Cin(hCD5)/Gpt	GemPharmatech	N/A
Recombinant DNA
Plasmid: mCAR	Constructed by our lab	N/A
Software and algorithms
ImageJ v1.53	National Institutes of Health (NIH)	https://imagej.net/ij/
Graphpad prism 9	GraphPad Software Inc	https://www.graphpad.com
FlowJo V10	FlowJo	https://www.flowjo.com/
R	The R Project for Statistical Computing	https://www.r-project.org/
Other
Clinical Trial Registry	ClinicalTrials.gov	NCT04767308
RPMI 1640	Gibco	Cat# 11875093
DPBS	Gibco	Cat# 14190144
PBS	Biosharp	Cat# BL302A
Opti-MEM	Gibco	Cat# 31985070
Fetal Bovine Serum	Gibco	Cat# A5669701
DMEM	Gibco	Cat# 11965092

Experimental model and study participant details

Human subjects

The study reported the outcomes of a phase I clinical trial (NCT04767308), which included 7 subjects. Female and male adults with refractory or relapsed CD5-positive hematological malignancies were included and received a single dose of CT125A infusion. The detailed demographic information including age and gender for each participant was provided in Table 1. The patients 4, 5 and 6 in this manuscript were described as Patients 2, 1 and 3 in our previously published study.¹³ The study was approved by the Institutional Review Board of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, and conducted according to the principles of the Declaration of Helsinki. All participants gave informed consent prior to enrollment.

Cell lines

CD5-positive T lymphoblast cell lines, Molt-4 and SUP-T1, were cultured in RPMI1640 (Gibco) supplemented with 10% fetal bovine serum (FBS) (Gibco). Platinum-E (Plat-E) cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco) supplemented with 10% FBS (Gibco). Firefly luciferase (ffLuc) expressed cell lines, Molt-4-ffLuc and SUP-T1-ffLuc, were created by transducing parental cells with lentivirus encoding ffLuc and puromycin-resistant genes, and cultured in the media with puromycin 1 μg/mL.

Primary mouse T cell culture

Murine T cells were isolated from spleen using the EasySep Mouse T cell Isolation Kit (Stemcell Technologies) per the manufacturer’s instructions. Post-isolation, T cells were activated with Dynabeads Mouse T-Activator CD3/CD28 (Gibco) at a concentration of 10 μL per 1×10⁶ cells and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) (Gibco), 50 μM β-mercaptoethanol, 1% non-essential amino acids (NEAA), 1% sodium pyruvate, 1% HEPES, 1% GlutaMax, and 200 IU/mL interleukin-2 (IL-2).

Animal experiments

GFP knock-in (C57BL/6-Tg (CAG-EGFP)1Osb/J) and hCD5 knock-in (C57BL/6JGpt-Cd5em1Cin(hCD5)/Gpt) mice were purchased from Shulaibao Biotechnology Co., Ltd. and GemPharmatech Co., Ltd, respectively. Mice were housed in SPF facility. Gender was not considered a factor influencing the results. Six-to eight-week-old female mice were used for experiments. Five and two hCD5 mice received 4.0 × 10⁶ anti-human CD5 murine CAR-T cells and control GFP-positive T cells intravenously 2 h after 5 Gy total body irradiation. Peripheral blood was collected at defined time points for flow cytometric analysis. Meanwhile, parameters including body weight, coat condition, skin integrity, and behavior were assessed. At the experimental endpoint (day 42), mice were euthanized for tissue collection and histopathological examination.

The animal study was approved by the Animal Ethics Committee of Tongji Hospital of Huazhong University of Science and Technology, China (Approval ID: T1-2024-08-074), and compliant with the Institutional Animal Care and Use Committee guidelines.

Method details

Clinical trial design

This dose-escalation study evaluated the safety and efficacy of CAR5^5ko T cell therapy in CD5-positive hematological malignancies. Participants received a single CAR5^5ko T cell infusion without subsequent HSCT to assess the investigational product’s safety profile and therapeutic potential. Detail information is provided in the Data S1.

Patients received CAR5^5ko-T cell infusion after a three-day lymphodepletion regimen of fludarabine (30 mg/m²/day) and cyclophosphamide (500 mg/m²/day), followed by a one-day rest. Cryopreserved or freshly prepared CAR5^5ko-T cells were then administered at pre-defined dose levels: 1.0 × 10⁶ cells/kg, 2.0 × 10⁶ cells/kg, and 1.0 × 10⁵ cells/kg. Adverse events (AEs) were documented post CAR-T cell administration. AE severity was classified using the Common Terminology Criteria for Adverse Events version 5.0 (CTCAE v5.0). Cytokine release syndrome (CRS) and CAR-T cell-related encephalopathy syndrome (CRES) were graded according to the 2019 American Society for Transplantation and Cellular Therapy (ASTCT) consensus guidelines. Therapeutic response was assessed per the Lugano 2014 criteria. Evaluations were conducted at 4- and 12-week post CAR-T cell infusion, with subsequent assessments determined by the treating physician. Progression-free survival (PFS) was calculated from CAR-T infusion to the first documented relapse or death from any cause. Overall survival (OS) was defined as the time from CAR-T infusion to death from any cause.

Plasma levels of cytokines, including IL-6, TNF-α, and ferritin, were measured at indicated time points using an electrochemiluminescence assay during CT125A treatment. CAR-T cell expansion and persistence were assessed by droplet digital PCR and flow cytometry, respectively. The frequencies of immune cell subsets were assessed by flow cytometry.

Manufacturing of CT125A

Enrolled patients underwent lymphapheresis for peripheral blood mononuclear cell (PBMC) collection and subsequent cryopreservation. The manufacture of CT125A was performed by the Good Manufacture Production Center at Nanjing IASO biotherapeutics, according to Good Clinical Practice and Guideline for Good Clinical Practice of International Conference on Harmonisation. The manufacturing process of CT125A is generally outlined as follows: (1) Isolation and activation of CD3⁺ T cells; (2) CRISPR-Cas9-mediated CD5 gene knockout using the Celetrix electroporator; (3) Lentiviral transduction of the CAR5 construct; (4) Ex vivo expansion of CAR5^5ko-T cells; (5) Formulation, cryopreservation, or fresh preparation of CT125A for patient administration.

Generation of murine CAR5-T cells

Genes encoding anti-human CD5 CAR were synthesized and cloned into transfer plasmids. Retrovirus production was accomplished in Platinum-E cells using target plasmids and Lipofectamine 2000 Transfection Reagent (Lipo2000) (Invitrogen). Specifically, replace fresh culture medium for Plat-E cells when they are at 85% confluent. Add 14.1 μg target plasmid and 21.6μL Lipo2000 into two tubes of 1.5 mL Opti-MEM (Gibco), separately. Mix the two tubes together, incubate for 10 min at room temperature, and add the mixture into Plat-E plates. After 16 h culture, replace the Plat-E cells with fresh culture medium (8mL for 10 cm plates). 16 h later, the supernatant containing packaged virus was collected. After centrifugation for 5 min at 1200rpm, 4°C, the supernatant was ready for transduction.

After two days of culture, freshly isolated murine T cells were seeded at 1 × 10⁶ cells/mL onto retronectin (Takara, T100AC)-coated non-tissue culture-treated plates (Corning, 351146) with retroviral supernatant. Following centrifugation at 1200 rpm for 10 min, the plates were incubated (37°C, 5% CO₂), maintaining a cell density of approximately 1 × 10⁶ cells/mL. Please refer to the STAR protocol published by Ekaterina Eremenko et al. for more detailed information.²⁴

Cytolysis assay

Ten thousand firefly luciferase-expressed target cells were seeded in 96-well flat-bottom clear plate (Corning). According to the pre-designed effector-to-target ratios, certain number of murine CAR-T cells or mock T cells were added, and cultured in RPMI 1640 medium supplemented with 10% FBS. The plates were then placed in a cell incubator at 37°C for 24 h. The luciferase activity in each well was then measured after adding D-luciferin sodium (Solarbio) by Synergy H1 microplate reader (BioTek), to reflect the quantity of remaining target cells. The cytotoxicity was calculated using the formula: lysis efficacy = 1 − luminescence _{effector T cells + targets}/luminescence _{targets only}.

Flow cytometry assay

For cultured cell samples, approximately 1×10⁶ cells were used for analysis. After washing with 2% FBS phosphate buffer saline (PBS), antibodies were added in 100uL PBS with 2% FBS and incubated for 15 min at room temperature. After two rounds of washing using PBS with 2% FBS, samples were prepared for acquisition. For mouse blood samples, a volume of 100uL was desired for analysis. Antibodies were added directly after 10 min co-incubation with 1:500 diluted Rat-originated anti-mouse CD16/CD32 (BD Biosciences, 553142). After 15min antibody incubation, 900uL red blood cell lysis reagent (Thermo Fisher) was added, and incubated for 10 min. After two rounds of washing using PBS with 2% FBS, samples were prepared for acquisition.

Hematoxylin and eosin (H&E) staining

Samples were fixed in 10% neutral-buffered formalin for 24 h, processed, embedded in paraffin, and sectioned at 4–6 μm. Hematoxylin and eosin-stained sections were analyzed by a blinded board-certified hematopathologist at Baiqiandu Biotechnology using light microscopy (×20.0).

Immunohistochemical assay

Tissue samples were fixed in 10% neutral-buffered formalin, processed, and embedded in paraffin. Sections (4–6 μm) were deparaffinized and rehydrated. Heat-induced antigen retrieval was performed in citrate buffer (pH 6.0). Endogenous peroxidase activity was blocked with 3% H₂O₂, followed by blocking with 3% BSA. Sections were then incubated overnight at 4°C with 1:500 diluted rabbit anti-human CD5 primary antibody (Abcam, ab75877). After washing, a 1:2000 diluted goat anti-rabbit IgG HRP-conjugated secondary antibody (Abcam, ab205718) was applied. Signal was visualized using a DAB substrate kit (Beyotime, P0203), with reaction time monitored microscopically. Finally, sections were counterstained with hematoxylin, dehydrated, cleared, and mounted. Staining was assessed using a microscope. Negative controls, omitting the primary antibody, were included to ensure specificity.

Immunofluorescence assay

The paraffin embedded 5μm skin sections were first deparaffinized in xylene and rehydrated through a graded ethanol series. Antigen retrieval was performed by incubating the slides in EDTA buffer (pH 9.0) under high-temperature and high-pressure conditions. Subsequently, the sections were sequentially incubated with 3% hydrogen peroxide for 30 min at room temperature, 10% goat serum for 30 min at 37°C. Then, sections were prepared for antibody staining. Firstly, 1:8000 diluted rabbit anti-mouse CD3 primary antibody (Abcam, ab237721) were incubated overnight at 4°C. After washing, 1:4000 diluted goat anti-rabbit IgG HRP-conjugated secondary antibody (Abcam, ab205718) and 1:600 diluted iFluor 488-conjugated tyramide (aatbio, 11060) were applied for 15 min and 10 min at room temperature, separately. After 5 min antigen retrieval and 30 min 10% goat serum incubation, the sections were ready for second round of antibody staining. 1:400 diluted rabbit anti-mCherry antibody (abcam, ab213511) were incubated overnight at 4°C, followed by secondary antibody staining and 1:600 diluted iFluor 647-tyramide (aatbio, 11066) detection. Cell nuclei were counterstained with DAPI. After mounting, the slides were scanned using a 3D HISTECH Pannoramic SCAN II slide scanner to acquire high-resolution whole-slide images. Appropriate controls, including omission of primary antibodies, were included to confirm staining specificity.

Quantification and statistical analysis

Descriptive statistics were used to summarize the patient and disease characteristics (related with Tables 1 and 2). Median and range were usually reported. PFS and OS were calculated using GraphPad Prism (version 9). The AUC_0-28 of the CAR copy numbers was calculated with MATLAB. Figures were produced using GraphPad Prism. For statistical analysis in Figure S3, a two-sided unpaired non-adjusted t test was employed with significance level alpha set at 0.05 (5%) by using GraphPad Prism.

Additional resources

The study was registered at https://clinicaltrials.gov/study/NCT04767308, with the number of NCT04767308.

Published: February 2, 2026