Enver Aydilek

Abstract CD19‐directed chimeric antigen receptor (CAR)‐T cell therapy has become a standard treatment for relapsed/refractory diffuse large B‐cell lymphoma (r/r DLBCL). While the benefits of CAR‐T cell treatment are clear in the general patient population, there remains a relative scarcity of real‐world evidence regarding its efficacy and toxicity in patients (pts) aged ≥70 years with DLBCL. We conducted a multicenter retrospective analysis including 172 r/r DLBCL pts with CAR‐T cell treatment, axicabtagene ciloleucel or tisagenlecleucel, between 2019 and 2023 at three tertiary centers. Pts were grouped by age at CAR‐T infusion (<70 vs. ≥70 years). Subsequently, descriptive and survival analyses, including propensity score matching, were performed to compare outcomes between both age groups. We identified 109 pts aged <70 and 63 pts aged ≥70 years. Overall response rates for both age groups were comparable (77.7% vs. 78.3%; p = 0.63). With a median follow‐up of 8.3 months, median progression‐free survival was 10.2 months (95% confidence interval [CI]: 6.5–21.8) and 11.1 months (95% CI: 4.9–NR) ( p = 0.93) for both cohorts. Median overall survival reached 21.8 months (95% CI: 11.8–NR) and 34.4 months (95% CI: 10.1–NR) ( p = 0.97), respectively. No significant differences in the incidence of cytokine release syndrome ( p = 0.53) or grade ≥3 neurotoxicity ( p = 0.56) were observed. Relapse and nonrelapse mortality were not significantly different between both groups. Our findings provide additional support that CAR‐T cell therapy is feasible and effective in patients with r/r DLBCL aged 70 years or older, demonstrating outcomes comparable to those observed in younger patients. CAR‐T cell therapy should be not withheld for elderly patients with r/r DLBCL.

ABSTRACT: Patients with large B-cell lymphoma (LBCL) who experience relapsed disease after CD19-directed chimeric antigen receptor (CAR) T-cell (CAR-T) therapy have a poor prognosis. Bispecific antibodies (BsAbs) induce complete remissions in ∼35% of these cases. Hypothesizing overlapping LBCL-intrinsic resistance mechanisms as well as common poor prognosis predictors to CAR-T and BsAb therapy, we conducted a multicenter retrospective analysis including 92 patients with relapsed/refractory (R/R) LBCL treated with BsAbs after CAR-T failure. Overall response rate (ORR) was 43%, with a progression-free survival (PFS) of 2.8 months. Patients receiving BsAbs during early relapse (≤3 months) achieved a significantly worse outcome (ORR, 29%; PFS, 2.2 months) compared with patients with an intermediate (4-6 months; ORR, 54%; PFS, 3.7 months) or a late relapse (>6 months; ORR, 60%; PFS, 10.5 months). The benefit of later relapse was particularly notable in patients receiving BsAbs as first salvage therapy compared with those receiving a BsAb in subsequent lines (PFS not reached vs 2.7 months; overall survival not reached vs 9.1 months, respectively). In addition to early R/R state before BsAbs, elevated lactate dehydrogenase and higher International Prognostic Index score were significant predictors of poor outcomes to BsAb in multivariate Cox regression analyses. The finding that patients with early relapse after CAR-T respond particularly poorly to BsAb highlights the necessity for alternative treatment options in this high-risk patient cohort.

First-line immune-chemotherapy induces long-lasting remissions for most patients with mantle cell lymphoma (MCL). In case of progression, response rates and overall survival decline.1,2 In 2020, chimeric antigen receptor (CAR)-T cell therapy (Brexucabtagene-autoleucel, brexu-cel; Tecartus) was approved by the FDA and EMA for the treatment of patients suffering from refractory/relapsed (r/r) MCL, who had been previously treated with 2 lines of therapy, including Bruton-tyrosine kinase inhibition. Data on treatment strategies for r/r MCL after CAR-T cell therapy are scarce. Here, we report the clinical courses and molecular observations of 4 patients with r/r MCL, who received a second infusion of the initial CAR-T cell product (second bag infusion) between 4 and 8 months after the first CAR-T treatment. All patients were treated with brexu-cel at, the University Medicine Göttingen and the Knappschaftskrankenhaus Bochum (Ruhr University Bochum) between 2021 and 2023. Patient characteristics are summarized in Table 1. Before each CAR-T infusion, CD19+ MCL was verified by flow cytometry and/or immunohistochemistry. Lymphodepleting chemotherapy before brexu-cel infusions followed standard protocols applying fludarabine (90 mg/m2) and cyclophosphamide (1.500 mg/m2). Response was assessed by PET/CT according to the Lugano classification.3 CAR-T cell expansions were quantified by specific digital droplet-PCR (ddPCR) from cell-free DNA (cfDNA) as previously described.4,5 In 1 patient, diagnostic follow-up was performed by ddPCR analysis of peripheral blood monocytes (PBMC).6 Molecular disease monitoring was performed by ddPCR assays for IGH-CCND1 fusions (t11;14), according to the previous reports.7 Table 1 - Baseline Characteristics Patient 1 Patient 2 Patient 3 Patient 4 Age at initial diagnosis, y 42 57 62 66 Sex (Male/female) Female Female Male Male Histology Lymphocytic Blastoid Blastoid n.a. Initial MIPI-Score n.a. 5.4(Low) 6(Intermediate) 5.98(Intermediate) Initial Ki-67 10% 30% 15% n.a. Bone marrow involvement >90% 50% 50% 1% Ann-Arbor stage at CAR-T - First IVa IVb IVb IVa - Second IV IVb IV IV Quotient to ≥ UNV of LDH at CAR-T - First 0.48 0.93 1.02 0.89 - Second 0.62 1.32 1.11 1.10 Spleen size diameter (cm) after CAR-T - First <11 <11 16 19.4 - Second <11 <11 n.a. 13.9 Number of prior therapies - First 3 3 2 3 - Second 4 5 3 4 Prior HDT + ASCT Yes Yes Yes Yes Prior allogeneic SCT No No No No Prior radiation therapy Yes No No Yes Bridging before CAR-T - First R-DHAP/RTx/Dex R-Bendamustin None RTx - Second R-Bendamustin R-Bendamustin Vincristin/Prednisolone R-Bendamustin Response status at CAR-T infusion - First PD PD SD PD - Second SD PD PR PD Time to CAR-T from initial diagnosis (mo) - First 54 45 56 58 - Second 62 51 60 63 Time from first to second CAR-T (days) 247 196 111 147 Toxicity post-CAR-T: CRS - First Grade II Grade II Grade II Grade II - Second None Grade III Grade III–IV Grade III Toxicity post-CAR-T: ICANS - First Grade IV None None None - Second None Grade I None None Best response to CAR-T (mo) - First PR (1) CR (1) CR (3) PR (2) - Second SD (1) CR (1) n.a. PR (2) Cause of death Progressive disease - CRS/infection Infection ASCT = autologous stem-cell transplant; CAR = chimeric antigen receptor; cm = centimeter; CR = complete response; CRS = cytokine release syndrome; HDT = high-dose chemotherapy; ICANS = immune effector cell-associated neurotoxicity syndrome; LDH = lactate dehydrogenase; MIPI = Mantle Cell Lymphoma International Prognosis Index; n.a. = not available; PD = progressive disease; PR = partial response; SD = stable disease; UNV = upper normal value. The first patient presented with lymphoma refractory to second-line ibrutinib. Salvage-chemotherapy (R-DHAP) and radiotherapy resulted in partial response (PR) before leukapheresis, with rapid progression thereafter. Following the first CAR-T cell infusion, the patient suffered from grade IV immune effector cell-associated neurotoxicity syndrome (ICANS) and grade II cytokine release syndrome (CRS). In a PET/CT scan at day +30, a PR was documented, which was ongoing in the 3- and 6-month follow-up PET/CT scans. In ddPCR analysis of cfDNA, CAR-T cells’ expansion was low (peak 0.1 copies/µL blood) and the amount of cfDNA (reference cfDNA) did not increase after therapy, indicating insufficient cytotoxicity of the CAR-T cells (Figure 1A).4 Clinical disease progression occurred 8 months after the initial CAR-T infusion. Subsequent to a dexamethasone bridging therapy and standard lymphodepletion, the second bag infusion was administered without any signs of CRS or ICANS. Response was a stable disease for only 1-month and rapid MCL progression thereafter. In line with this, no increase in the amount of reference cfDNA was measured, after the second CAR-T infusion (Figure 1A). Next, allogeneic stem-cell transplant (alloSCT) was performed, which induced only a short-lasting remission (<100 days). Relapse affecting the central nervous system occurred at day +90 and the patient was referred to palliative care.Figure 1.: Analysis of CAR-T cells, reference DNA, and disease-specific t(11;14) in vivo by digital droplet-PCR. CAR-T cell infusions are depicted as dotted lines. Values are shown as copies/µL blood.4 , 8 (A) Analysis of cell-free DNA of patient 1 displays a moderate increase of the CAR-T signal within the first 14 days, which is not accompanied by a robust increase of total cell-free DNA (reference, bottom). Around day +100, there was an increase of reference DNA for unknown reasons. The second CAR-T therapy followed around day +250 (dotted line). Again, modest CAR-T increment was not accompanied by an increase of the reference DNA, suggesting insufficient cytotoxicity of CAR-T cells in this patient. (B) Analysis of cfDNA samples of patient 2. After the first CAR-T infusion, robust CAR-T increase is accompanied by an increase of the t(11;14) signal and reference DNA, suggesting killing activity of the CAR-T cells and lysis of lymphoma cells with release of cfDNA into plasma. After day +100, the t(11;14) signal and CAR-T cells again increase. However, disease progression was evident. After bridging therapy and the second CAR-T infusion, there is again robust CAR-T expansion and increase of the reference DNA. Disease specific t(11;14) was not measurable in the following. (C) In cfDNA samples of patient 3, increase of CAR-T cell signal is accompanied by an increase of reference DNA and t(11;14) signal, suggesting killing activity of the CAR-T cells. The peak expansion of CAR-T cells per ml blood is high, however no clearance of t(11;14) is achieved, despite the CR in PET/CT scan. CAR-T signal is lost beyond day 30 and subsequently, the t(11;14) signal increase. The patient died shortly after the second CAR-T cell infusion. (D) Analysis of PMBC samples of patient 4. After CAR-T treatment, the CAR-T cell signal increase but no clearance oft(11;14) could be achieved in the following, reflecting the PR in PET/CT scans. After bridging therapy around day 100, t(11;14) is no more detectable. Rapidly after the second CAR-T treatment, the t(11;14) signal increase. The patient died from neutropenic sepsis due to prolonged neutropenia and persistent disease. CAR = chimeric antigen receptor; cfDNA = cell-free DNA; CR = complete response; PBMC = peripheral blood monocyte; PR = partial response.The second patient had disease progression 6 months after initiation of second-line ibrutinib. After leukapheresis, bridging therapy with rituximab/bendamustine was administered. However, the disease progressed, with large malignant pleural effusions and ascites. Following CAR-T cell infusion, grade II CRS without ICANS occurred. Treatment resulted in a CR at day +30 as confirmed by PET/CT (Figure 2A). In the ddPCR analysis, coincident peaks of the CAR-T expansion, the t(11;14) signal, and the reference cfDNA were evident at day +10, indicating lymphoma lysis due to CAR-T cells’cytotoxicity.4 At first, clearance of the t(11;14) translocation was obvious from day +30 on, but it increased again 3 months after therapy (Figure 1B). In a PET/CT scan on day +117, suspect uptake was apparent in the colon with marginal ascites (Figure 2B). The patient had no clinical signs of disease relapse, and colonoscopy and subsequent biopsies were negative for MCL. At day +180, CD19+ relapse was proven by biopsies after a follow-up PET/CT (Figure 2C). Interestingly, a CAR-T cell signal could be measured by ddPCR again at this time, after clearance at day +50. Despite the presence of CAR-T cells, the lymphoma progressed. Because of CD19+ relapse and the missing peak of reference cfDNA, indicating CAR-T cells’ cytotoxic activity, progression was probably accountable to T-cell exhaustion. After salvage-chemotherapy with rituximab/bendamustine and lymphodepleting chemotherapy, the second bag infusion was administered. Grade III CRS and grade I ICANS occurred, subsequently. On day +30 after the second CAR-T cell infusion, no lymphoma was detected in a PET/CT (Figure 2D). Accordingly, CAR-T cell expansion was substantial and liquid biopsy revealed again clearance of the specific t(11;14) signal (Figure 1B). The patient remains in CR 3 months after therapy.Figure 2.: Longitudinal PET/CT scans from patient 2. (A) PET/CT scan at day +30 after the first CAR-T cell therapy showing complete remission of pleural effusion and ascites. (B) Hypermetabolic lesions in the colon and ascites (red arrows) in a PET/CT scan at day +117, without clinical symptoms. Colonoscopy and biopsy showed no signs of lymphoma, and paracentesis was not possible, due to low amounts of ascites. At this time, an increase of the t(11;14) signal in cfDNA was evident (Figure 1B). (C) PET/CT scan at day +180. The patient now suffered from night-sweats and abdominal pain. Progression of the ascites, as well as paracardial effusions and hypermetabolic extranodal manifestations were evident (green arrows). Paracentesis confirmed disease relapse. (D) PET/CT scan at day+30 after the second CAR-T cell infusion showing again complete remission. The patient suffered from unspecific thyreoiditis, which resolved without intervention (blue arrow). CAR = chimeric antigen receptor; cfDNA = cell-free DNA.The third patient also relapsed after second-line ibrutinib. Subsequent treatment with brexu-cel resulted in a CR at day +30. Grade II CRS without signs of ICANS occurred after the first infusion. Despite the negative PET/CT scan and substantial CAR-T cell expansion, a disease-specific signal for t(11;14) was still detectable in ddPCR analysis at day +30 (Figure 1C). Relapse became evident in a PET/CT scan 3 months after therapy. A single dose of vincristin/prednisolone was administered, shortly followed by standard lymphodepletion before the second bag infusion. At the same day of the second brexu-cel infusion, the patient developed systemic inflammatory response syndrome (SIRS). CRS treatment with tocilizumab, dexamethasone, and broad-spectrum antibiotics were given. After initial response, rapid progression of SIRS with multiorgan failure occurred 5 days after the infusion. In a respiratory swab, the endemic coronavirus subtype NL63 was detected. The patient succumbed to SIRS due to multiorgan failure 6 days after second CAR-T infusion. The fourth patient relapsed after second-line ibrutinib and suffered from disease progression at the time of CAR-T therapy, with leukemic spread of lymphoma cells. After CAR-T cell therapy, grade II CRS without ICANS occurred. As shown in ddPCR of PBMCs, CAR-T cells initially expanded but were no longer detectable at day +30 (Figure 1D). Best response was a PR 2 months after infusion. The follow-up CT scan 3 months after therapy showed progressive disease. Following bridging therapy with rituximab/bendamustine, a second CAR-T therapy was administered after which the patient developed grade III CRS. CT scans displayed a PR 2 months after the second CAR-T cell infusion. However, the patient suffered from persistent neutropenia and died 2 months later due to neutropenic sepsis. To our knowledge, this is the first report on repeated infusionsof brexu-cel given at short intervals for the treatment of r/r MCL. Also, this is the first report about analyzing the disease-specific IGH-CCND1 fusion transcript (t11;14) by ddPCR as marker for minimal residual disease (MRD) in the context of MCL CAR-T cell therapy. Currently, there are no guidelines on treatment options for r/r MCL after CAR-T cell therapy. Case series describe clinical responses to allogeneic SCT and bispecific antibodies following CAR-T failure.9–11 Administering the second bag of CAR-T cells was intended under the premise that r/r CD19+ MCL is still susceptible to CD19-CAR-T cells, despite disease progression after the first infusion. As we report, molecular follow-up may help to guide clinical decision-making. While the CAR-T cells engraft and expanded after the first infusion in all patients (Figure 1), patients 1 and 4 had persistent disease in PET/CT scans (Table 1) and IGH-CCND1 fusion signals could consistently be measured. Patient 3 had a persistent t(11;14) signal in cfDNA, despite a negative PET/CT. Only patient 2 showed clearance of the t(11;14) in cfDNA analysis and a CR in PET/CT scans, but eventually relapsed. Primary and secondary resistance toward CAR-T cell therapy can be CAR-T cell associated (expansion and exhaustion) or tumor associated (eg, resistance to apoptosis).12,13 In ddPCR analysis of patient 2, the coincident peaks of CAR-T cells, the t(11;14) signal and reference cfDNA indicate susceptible MCL disease and the CAR-T cells’ cytotoxicity (Figure 2B). As described above, we believe disease progression in patient 2 was due to CAR-T cell exhaustion, which could be overcome by the second bag infusion. In contrast, the CAR-T cells’ cytotoxicity was inadequate in patient 1 after the first infusion. This can be explained by an insufficient product or tumor inherent resistance. From this point of view, failure of the second bag infusion was likely, despite CD19+ relapse. Furthermore, adverse events were of higher severity following the second brexu-cel infusion. In 3 of 4 patients, grade III–IV CRS occurred after the second infusion. ICANS occurred in 2 patients after the second brexu-cel infusion, and in only 1 patient after the first. Severe infections occurred in 2 of 4 patients. Based on our findings, there is no adaption toward the inflammatory effects of CAR-T cells. A previous report with a different CAR-T cell product also found increased toxicity after a second CAR-T infusion, while another report did not.14,15 More research is needed, to clarify this issue. In conclusion, a second infusion of brexu-cel can facilitate disease control in the event CAR-T cell failure and may be considered as treatment option including bridging to further therapy, such as allogeneic SCT. Analyzing cfDNA and MRD may help to guide therapy decisions toward a second CAR-T cell infusion or the need for a more aggressive treatment, for example, allogeneic SCT in patients with insufficient CAR-T cell cytotoxicity. Our data emphasizes the need for detailed disease monitoring by MRD in the context of CAR-T cell therapy, to gain further insight into resistance mechanisms in future research. ACKNOWLEDGMENTS We thank Christina Eilert-Micus and Sabine Bachmann for excellent technical support in ddPCR analysis. AUTHOR CONTRIBUTIONS EA, SK-S, JT, RS, GW, and TM wrote the article. EA, SK-S, JT, V N-E, and DV obtained and processed the patient samples. SK-S and TM analyzed and interpreted the ddPCR data. EA, RS, GW, and TM designed the study. DISCLOSURES The authors have no conflicts of interest to disclose. SOURCES OF FUNDING The authors declare no sources of funding.

Introduction Patients (pts) with relapsed or refractory (r/r) large B-cell lymphoma (LBCL) after CD19 chimeric antigen receptor (CAR)-T cell therapy are challenging to treat. CD20xCD3 T-cell-engaging bispecific antibodies (BsAbs) have emerged as a promising therapeutic option for these pts. In the pivotal trials, glofitamab and epcoritamab demonstrated notable efficacy, with an overall response rate (ORR) of at least 50% and a complete response rate (CR) of at least 34% after CAR-T cell therapy. However, data on the efficacy and safety of BsAbs post-CAR-Ts in the real-world setting are scarce. Methods To assess the efficacy and feasibility of BsAbs in the post CAR-T population, we conducted a retrospective, multicenter, multinational analysis, enrolling 85 pts with r/r LBCL treated with BsAbs across 20 centers in Germany, Austria, and Switzerland. Pts were eligible for the study if they had experienced CAR-T cell failure. We defined early ((within first 3 months (mo)), intermediate (4-6 mo) and late (&amp;gt;6 mo) CAR-T cell failure and assessed efficacy (ORR, CR, PR, mPFS, mOS) and feasibility of BsAbs in the real-world setting depending on these time points. Furthermore, information on treatment between CAR-T and following BsAb was collected. Results The study cohort included de novo diffuse large B-cell lymphoma (DLBLC) (55/85), transformed indolent lymphoma (18/85), high grade B-cell lymphoma (HGBCL) ((10/85; MYC &amp; BCL2/BCL6 rearrangements (5/10); not otherwise specified (NOS) (4/10); or 11q deletion (1/10)), primary mediastinal B-cell lymphoma (PMBCL) (1/85) and T-cell/histiocyte-rich large B-cell lymphoma (TCRLBCL) (1/85). Before receiving BsAbs, 43 (51%) pts presented with early (mo 1-3), 24 (28%) with intermediate (mo 4-6), and 18 (21%) pts with late (&amp;gt;6 mo) relapse post-CAR-Ts. Moreover, 51 out of 85 pts (60%) received BsAbs directly after CAR-T failure (immediate BsAb group), while 34 (40%) had a median of one therapy (range 1-4) prior to BsAbs (non-immediate BsAb group). Glofitamab (75/85) was the most commonly administered BsAb followed by epcoritamab (9/85) and mosunetuzumab (1/85). Overall, 42% of pts responded to BsAbs, with 23% achieving CR and 19% partial response (PR). The ORR varied based on the timing of BsAbs application (47% immediate vs. 35% non-immediate, p=0.282) and CAR-T cell failure (30% vs. 50% vs. 61% in early, intermediate and late relapse post-CAR-Ts, p=0.019). The highest CR rate was observed in pts with late CAR-T failure (44%), followed by intermediate (25%) and early (14%) groups (p=0.037). The median progression-free survival (mPFS) was 3.27 mo, while the median overall survival (mOS) was 6.83 mo. Estimated mPFS was significantly better in the immediate than in the non-immediate BsAb group (4.37 mo vs. 2.48 mo, p=0.026) while mOS did not differ significantly (7.73 mo vs 5.10 mo, p=0.100). Additionally, mPFS and mOS differed significantly in early (2.13 and 4.10 mo), intermediate (3.73 and 7.73 mo) and late (10.46 mo and not reached) CAR-T failure groups, accordingly (p≤0.004). In the multivariate analysis, sensitivity to the last treatment prior BsAbs (PFS - [HR=2.31, p=0.004]; OS - [HR=2.81, p=0.006]), the timepoint of CAR-T failure (PFS - [HR=1.48, p=0.05]; OS - [HR=1.94, p=0.011]), additional therapies between CAR-T and BsAb (PFS - [HR=1.86, p=0.039]) and LDH levels before application of BsAb considered as continuous variable (p=0.049 for PFS) were the strongest predictors of survival after initiation of BsAbs. The frequency of cytokine release syndrome (CRS) was 31%, with grades 3-4 documented in 3%. Severe immune effector cell-associated neurotoxicity syndrome (ICANS) occurred only in one patient (1%) (grade 3). Infections were documented in 36% of pts, with grades 3 in 11% and three pts with grade 5 (all associated with COVID-19). Conclusions In summary, BsAbs show efficacy and manageable safety profiles in r/r LBCL pts following CAR-T failure in real-world settings. However, efficacy of BsAb is significantly impaired in early CAR-T refractory pts underscoring the necessity of novel therapeutic approaches in these pts.