This study retrospectively analyzed aBCL patients who had received CD19 CART therapy (clinical Trials#: ChiCTR2100055062) and then progressed (relapsed/refractory to CD19 CART) at Beijing Gobroad Boren Hospital from December 2019 to May 2022. This study was approved by the Institutional Review Board of Boren Hospital (approval number: 20191225-PJ-003) and was conducted according to the principles of the Declaration of Helsinki. Written informed consent was obtained from each patient.

The inclusion criteria of this study were: 1) ≥18 years old; 2) patients with diffuse large B-cell lymphoma, not otherwise specified (DLBCL, NOS), Burkitt lymphoma (BL), transformed follicular lymphoma (tFL), high-grade B-cell lymphomas (HGBL) (15) after two or more lines of systemic therapy received murinized second-generation anti-CD19 CAR-T cells with signals provided by costimulatory molecules 4–1BB and CD3-zeta, and 3) patients who did not achieve complete remission (CR) or partial remission (PR) after CD19 CART treatment (refractory to CD19 CART), or experienced disease progression after initially achieving CR/PR (relapsed to CD19 CART). The exclusion criteria were 1) patients showing sustained remission following CD19CART treatment, 2) patients who received CD19 CART and other drugs simultaneously, and 3) patients who experienced progression after CD19 CART without subsequently receiving anti-cancer therapy.

The selection of therapy is determined by the positive expression status of tumor tissue antigens (CD20 positive, ≥90%), physical condition (ECOG <4 or ECOG=4 but demanding further anti-tumor treatment), and wiliness of included patients. Patients were divided into using CAR T-cells targeting a second different antigen (CD20) (CD20-SD-CART) group and non-CART group, depending on the subsequent anti-cancer treatment.

CD20CART cells were manufactured using cryopreserved autologous peripheral blood mononuclear cells stored when patients enrolled/consented to CD19 CART therapy. The details of the CART cell manufacturing process have been described in previous studies (16–19). Bridging treatment (platinum and/or anthracycline combinations with BTKi, ICPi, or BCL2i) was permitted prior to CD20CART infusion in cases of bulky disease ( Supplementary Figure S1 ). Lymphodepleting chemotherapy was fludarabine (30 mg/m2, d-5 to d-3) monotherapy or in combination with cyclophosphamide (CTX, 300 mg/m2, d-5 to d-3). And if patients who have previously used fludarabine would receive lymphodepletion chemotherapy composed of bendamustine (90 mg/m2 d-5, -4). The infusion was performed on day 0. The infusion dose was administered at target number 2 ×10⁶cells/kg, minimum accepted dosage of CAR T cells (1 × 10⁵/kg) and maximum accepted dosage of CAR T cells (10×10⁶/kg).

Transduction efficiency and cell viability were examined at the time of CD20-SD-CART cell infusion. Transduction efficiency was defined as the ratio of CAR-T to CD3+ T cells, determined by flow cytometry (FCM). When the harvest of CD20 CAR T-cells was less than 1×10⁵/kg, we defined it as CAR manufacturing failure. A multicolor flow cytometer (FACSCalibur, BD, USA) was employed to detect CART cells on day 3 (d3), day 7, day 14, day 21, and day 28 post-transfusion. After that, monitoring continued monthly until the assay reached a lower quantitation limit. Enhanced CT/MRI for patient evaluation was performed once per month for the first six months. The first PET/CT was performed in the third month and then every 3 months until disease progression.

Non-CART therapy consisted of polatuzumab (pola-based) and non-polatuzumab (non-pola-based) therapies. Polatuzumab was given a dose of 1.6–1.8mg/kg intravenously every 3–4 weeks. The non-polatuzumab therapy in our study included lenalidomide, immune checkpoint inhibitors (ICPi), B cell lymphoma/leukemia-2 inhibitors (BLC2i), Bruton tyrosine kinase inhibitors (BTKi), and chemotherapy (anthracycline and/or platinum-based). The treatment regimen involved one cycle every 28 days, with enhanced CT or MRI assessment conducted before the next treatment cycle. Additionally, PET/CT monitoring was performed every two cycles.

The primary outcome was overall survival (OS). Secondary outcomes included the best overall response rate (best ORR), progression-free survival (PFS), and the incidence of treatment-related adverse reactions. Response assessment was conducted according to the Lugano 2014 criteria, which categorized responses as complete remission (CR), partial remission (PR), stable disease (SD), or progression disease (PD) (20). The best ORR was defined as the percentage of patients achieving a response, including the complete remission rate (CRR) and partial remission rate (PRR). PFS in the CD20-SD-CART group was defined as the time from infusion of CD20CART cells to disease progression, last follow-up, or death. PFS in the non-CART group was defined as the time from initiating a new line of treatment after CD19 CART therapy to disease progression, last follow-up, or death. OS was defined as the time from the date of CD20CART cell infusion for the cellular group or the initiation of a new line of treatment for non-CART after CD19 CART therapy to the date of death or last follow-up. Treatment-related adverse reactions, such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxic syndrome (ICANS), were graded and managed according to the American Society of Transplant and Cellular Therapy grading criteria (21).

Medical records at the time of progression following CD19 CART therapy were screened. The following data for each patient were extracted: age, sex, medical histology, number of extranodal site lesions, bulky disease (bulky disease was defined as maximal diameter ≥7.5cm measured in either the transverse or coronal plane on computed tomography), Eastern Cooperative Oncology Group performance status (ECOG PS) (22), disease stage(Ann Arbor stage), International Prognostic Index (IPI) score (23), lactate dehydrogenase level (LDH), TP53 genes mutation by targeted next-generation sequencing, therapies pre-CD19 CART, response/nonresponse to CD19 CART, anti-cancer treatment therapy following CD19 CART progression, and clinical response. We used a combination of telephone and outpatient visits for follow-up observations and followed up to August 31, 2022.

Analyses were performed by using SAS 9.4 (SAS Institute Inc, Cary, NC). The continuous data conforming to the normal distribution were expressed as means ± standard deviation and compared by independent sample t-test. When analyzing continuous variables that did not conform to a normal distribution, we described them using medians and IQR, and assessed the difference using the rank sum test. Categorical variables were reported in numbers and percentages and analyzed using the chi-square (χ²) or Fisher’s exact test. Logistic regression was used for multivariable analyses of the response. Estimates of survival were calculated according to the Kaplan-Meier method. Factors associated with PFS and OS were analyzed using Cox proportional hazards regression models. Factors entered the multivariable model according to clinical significance and the p-value of the univariate analysis (p<0.1). A two-sided P ≤ 0.05 was considered statistically significant.

A total of 93 patients who had previously undergone CD19 CART treatment were included in this study after careful screening from a pool of 179 patients ( Figure 1 ). The median age of the included patients was 51 years (interquartile range [IQR], 42–64), and 52 (55.9%) were male. Among the 93 patients, 40 experienced relapsing after CD19 CART treatment (CR, n=13; PR, n=27), while 53 were refractory. After infusion, the median time to CD19 CART treatment failure was 2.3 months (IQR, 1.35–3.06). At progression or relapse following CD19CAR T-cell therapy, rebiopsies and CD19 evaluation by IHC/FCM were available in 49 patients (52.7%). 14 (28.6%) were found to be CD19-negative, while 35 (71.4%) remained CD19-positive. The median time of CD19CART treatment failure occurrence was 2.27 months (IQR, 1.25–3.31) in the CD19-negative group and 2.17 months (IQR, 1.48–2.87) in the CD19-positive group (p=0.56). Of the 93 patients, 54(relapsed n=23, refractory n=31) received CD20CART immunotherapy, while 39 (relapsed n=17, refractory n=22) received non-CART therapy including pola-based(n=15), targeted drugs (n=14: BCL2 inhibitors n=5, lenalidomide n=3, ICPi n=3, and BTKi n=3),and chemotherapy alone(n=10). The baseline characteristics of the patients are presented in Table 1 . The two groups had no significant differences in baseline characteristics (all p>0.05).

Among the 93 patients enrolled in the study, one patient declined to undergo an imaging examination and was not rated. The best ORR(PRR+CRR) of the primary cohort was 32/92 (34.8%) and CRR was 18/92 (20%). Of patients who had no response to CD20CART(n=32), 19 (59.3%) ones were refractory to CD19CART. A median follow-up period of 18.54 months (95% confidence interval [CI], 14.2–25.8) was conducted for the patients included in this study. The overall median PFS and median OS were 3.02 months (95%CI, 2.3–3.88) and 4.77 months (95%CI, 3.68–7.27), respectively ( Figures 2A, D ). At the cutoff date, 17 (53.1%) patients had an ongoing response for the entire cohort, and 26 (81.3%) patients were still alive.

Comparisons were made between the CD20-SD-CART and non-CART groups in ORR, PFS, and OS. There was no significant difference observed in the best ORR between the CD20-SD-CART group (22/54, 40.7%) and the non-CART group (10/38, 26.3%) (p=0.15). However, the CRR in the CD20-SD-CART group (15/54, 27.8%) was significantly higher than that in the non-CART group (3/38, 7.9%) (p=0.03). Among patients treated with CD20-SD-CART, the median PFS was 4.04 months (95% CI, 3.06–6.41), and the median OS was 8.15 months (95% CI, 5.62-NE). In contrast, for patients receiving non-CART treatment, the median PFS was 2.27 months (95% CI, 1.87–3.02), and the median OS was 3.02 months (95% CI, 1.91–4.21) ( Figures 2B, E ). Notably, patients in the CD20-SD-CART group demonstrated significantly longer PFS (HR, 0.52; 95% CI, 0.32–0.86; p=0.003) and OS (HR, 0.38; 95% CI, 0.22–0.66; p<0.0001) compared to patients in the non-CART group. The Cox proportional hazard model was used to reduce confounding factors. Multivariate regression analysis revealed that the CD20-SD-CART has independently associated with better PFS (HR 0.46, 95% CI 0.27–0.8, p=0.005) and OS(HR 0.28, 95% CI 0.16–0.51, p<0.0001) (factors enter the cox model are shown in Table 2 ).

Among the 39 patients who received non-CART treatment, 15 patients (38.5%) received pola-based therapy. In the pola-based group, the best ORR was 7/15 (46.7%) with a CRR of 2/15(13.3%) ( Figure 3 ). When comparing the pola-based group to the CD20-SD-CART group, no statistical differences were observed in ORR (46.7% vs. 40.7%, P=0.77), CRR (13.3% vs. 27.8%, P=0.33), PFS(3.62months vs. 4.04 months p=0.69), and OS(5.59 months vs. 8.15 months p=0.28). However, a trend for CD20-SD-CART to be more effective was observed ( Supplementary Figure S2 ). Additionally, among patients with the ongoing response (n=17) at the cutoff day, 12(70.6%) were in the CD20-SD-CART group, whereas 3(17.6%) were in pola-based group. Among the survivors (n=26), 18(69.2%) were in the CD20-SDCART group, while 5 were in pola-based group.

We also found that in the CD20-SD-CART group, the PFS for the 22 responders was 27.9 months (95% CI, 10.42-NE), while that for the 32 non-responders was only 2.2 months (95% CI, 1.97–3.62) (HR, 0.14, 95% CI 0.06–0.3, p<0.0001). The median OS for the responders to CD20-SD-CART was not reached, while that for the non-responders was only 3.88 months (95% CI, 2.96–6.35) (HR 0.08, 95% CI 0.03–0.24, p<0.0001). No statistically significant difference was found in the median PFS (HR 1.05, 95% CI 0.64–1.71, p=0.86) and OS (HR 0.74, 95% CI 0.45–1.23, p=0.24) between patients who did not respond to CD20-SD-CART therapy and those in the non-CART group. However, for patients who responded to CD20-SD-CART, the median PFS (HR 0.14, 95% CI 0.06–0.31, p<0.0001) and OS (HR 0.06, 95% CI 0.02–0.18, p<0.0001) were significantly longer compared to the non-CART group ( Figures 2C, F ).

Among the 54 patients treated with CD20-SD-CART therapy, 37 (68.5%) received bridging treatment with a CRR of 1/37 before starting lymphodepletion chemotherapy. Moreover, 49 patients (90.7%) received lymphodepletion chemotherapy (Flu-based n=32, Ben-based n=17). The median infusion dose was 1.21 ×10⁶cells/kg (IQR, 0.4–2.1). The infusion dose for 37 patients (68.5%) fell below the target level. After CD20-SD-CART infusion, CART cells were detectable in the peripheral blood by FCM from most patients (45/54), indicating a high in vivo expansion rate. The expansion peaked from days 12 to 15 after infusion, and the median expansion peak of CD20-SD-CART cells reached 7.76×10⁶cells/L (IQR 2.56–29.65). Despite receiving a comparable CD20-SD-CART infusion dose to CD19 CART (1.21 ×10⁶cells/kg vs.1.28×10⁶cells/kg, p=0.65), we observed lower CART cell peak expansion after CD20-SD-CART than that after CD19 CART (7.76×10⁶cells/L vs. 52.6×10⁶cells/L, p<0.0001) ( Supplementary Figures S1A, B ).

Concerning the safety of CD20-SD-CART, we found that 5 patients (9.2%) experienced severe CRS (≥grade 3), and 4 patients (7.4%) had severe ICANS (≥grade 3) ( Supplementary Figures S1C, D ). Dexamethasone was administered to treat CRS and/or ICANS in 13 patients (24%) after CD20-SD-CART. The patients who developed severe ICANS and CRS completely recovered, and no treatment-related deaths occurred. We noted comparable proportions of patients with sCRS (9.2% vs. 10.9%, p>0.999) and proportions of patients who developed sICANS (7.4% vs. 8.7%, p>0.999) between the CD20-SD-CART and CD19 CART.

Serum cytokine markers of systemic inflammation such as interleukin 6(IL-6), tumor necrosis factor-α (TNFα), and ferritin, exhibited elevation during CAR T-cell treatment. The changes in cytokines (IL6, TNFα, IL10, and ferritin) are shown in Supplementary Figure S4 . Toxicities assessed by CTCAE 5.0 grading were depicted in Supplementary Figure S4 . Overall, CD20-SD-CART regimens did not escalate toxicity, and no Grade 5 events were reported.

Using a univariable logistic regression model to assess factors related to the response of CD20-SD-CART, the findings showed that a higher infusion dose (OR 1.74, 95% CI, 1.1 - 2.86, p=0.03) was associated with a higher ORR, while IPI score ≥3 (OR 0.3, 95% CI, 0.10 - 0.93, p=0.04) and bridging treatment before CD20-SD-CART (OR 0.23, 95% CI, 0.07 - 0.78, p=0.02) were associated with a lower ORR. However, no correlation with response was detected when these factors were incorporated in a multifactor logistic regression ( Supplementary Table S1 ). Further multivariable analyses of the PFS and OS in CD20-SD-CART group illustrated that shorter PFS was associated with ECOG PS ≥3(HR, 3.12, 95% CI, 1.23–7.94, p=0.02), while longer PFS was associated with increased infusion dose (HR, 0.68, 95% CI, 0.5–0.92, p=0.01). Additionally, the OS was associated with ECOG ≥3 (HR, 4.13, 95% CI, 1.64–10.5, p=0.003) and refractory to CD19 CART (HR, 2.19, 95% CI, 1.01–4.76, p=0.0047) ( Table 3 ).