Abstract
Follicular lymphoma with progression of disease within 24 months (POD24) is associated with poor prognosis and represents clinical challenges. Therefore, we performed a systematic review and pooled analysis of patients with POD24. Twenty-one trials involving 1242 participants were included, assessing the overall response rate (ORR), complete response (CR), duration of response, and progression-free survival. In some trials, we compared pooled response rates between POD24 and non-POD24 populations with the same treatment regimen. Four trials evaluated chimeric antigen receptor (CAR) T-cell therapy in patients with POD24. Pooled analysis showed an ORR of 91.2% (95% confidence interval [CI], 83.7-98.7) with significant heterogeneity (P = .0414; I2 = 68.61%) and a CR of 75.7% (95% CI, 55.1-96.4) with significant heterogeneity (P < .0001; I2 = 93.99%). The specific response rates for different bispecific antibodies in POD24 were pooled analysis, the ORR was 81.6% (95% CI, 75.9-87.3) with no heterogeneity (P = .6958; I2 = 0%), and the CR was 65.7% (95% CI, 57.1-74.3) with moderate heterogeneity (P = .2148; I2 = 34.99%). For anti-CD19 antibody-drug conjugates (ADCs)/monoclonal antibodies (mAbs), the ORR and CR rate for loncastuximab plus rituximab and tafasitamab plus R2 (lenalidomide + rituximab) were 100% and 79.3%, and 87.5% and 43.2%, respectively. Phosphatidylinositol 3-kinase inhibitors and anti-CD20 mAb-containing regimens were also analyzed in pooled analyses. Our results demonstrated that anti-CD19 CAR T-cell therapy achieved the highest CR rate. Additionally, bispecific antibodies, anti-CD19 ADCs/mAbs, and the combination of lenalidomide with obinutuzumab or rituximab also exhibited excellent efficacy. Notably, lenalidomide plus obinutuzumab showed superior efficacy compared with R2.
Introduction
Follicular lymphoma (FL) is the most common indolent B-cell lymphoma, characterized by a generally favorable prognosis with median survival often extending over decades. However, ∼15% to 20% of patients treated with rituximab-based chemotherapeutic or nonchemotherapeutic regimens experience progression of disease within 24 months (POD24) of initiating first-line therapy, which is associated with a 5-year overall survival (OS) rate of ∼50%, compared with over 90% in patients without POD24.1,2 This was demonstrated in multiple independent cohorts, including those treated with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone), R-FM (rituximab, fludarabine, and mitoxantrone), or others.1 A pooled analysis of 5225 patients from 13 randomized clinical trials further confirmed that POD24 remains the most significant independent prognostic factor for reduced survival.3
However, POD24 serves as a posttreatment surrogate end point and is not suitable for guiding initial treatment decisions.4 Its prognostic significance lies in aiding the selection of subsequent treatment strategies for patients with relapsed or refractory (R/R) FL. POD24 predicts poor outcomes after autologous stem cell transplantation (ASCT), indicating persistent aggressive biology, although early ASCT may benefit some high-risk subgroups. Optimal therapeutic strategies for patients experiencing early progression after rituximab-containing chemotherapy, particularly the potential role of high-dose therapy with ASCT, have not been adequately investigated.1 Emerging therapeutic approaches, such as chimeric antigen receptor (CAR) T-cell therapy, bispecific antibodies (BsAbs), and anti-CD19 antibody-drug conjugates (ADCs), have shown highly encouraging results in this patient population. These treatments have achieved overall response rates (ORR) ranging from 85% to 100% and complete response (CR) rates between 57% and 87.5%. Notably, approximately half of these patients maintained remission at the 23-month follow-up.5, 6, 7 Alternatively, targeted therapies such as phosphatidylinositol 3-kinase (PI3K) inhibitors, tazemetostat, zanubrutinib-obinutuzumab, and R2 (lenalidomide-rituximab), have demonstrated ORR of 60% to 85%. However, these therapies are associated with modest progression-free survival (PFS) outcomes.8, 9, 10, 11, 12 Promising new data from the inMIND trial, which investigated the addition of tafasitamab combined with R2 in patients with FL and POD24, revealed a PFS of 19.2 months.13
Day et al investigated treatment patterns and clinical outcomes in 220 patients with FL and POD24 from the Lymphoma Epidemiology of Outcomes Consortium, most of whom received anti-CD20 antibody–based therapy or traditional chemotherapy. This real-world evidence demonstrated that certain treatment categories yielded more favorable outcomes. Among these, initial treatment with bendamustine-rituximab or rituximab alone was common, whereas R-CHOP was the predominant regimen used in clinical practice in patients with POD24.14
In the current landscape of CAR T-cell therapy, BsAbs, ADCs, novel monoclonal antibodies (mAbs), and small-molecule inhibitors, patients with FL and POD24 have more treatment options.15 However, there is currently no consensus on the optimal therapy for this patient subset. Therefore, we conducted a systematic review and pooled analysis of patients with POD24 enrolled in 21 clinical trials.
Methods
Data sources and search strategy
This study has been registered with PROSPERO (CRD420251074708). A comprehensive search strategy was applied using specific MeSH (Medical Subject Headings) terms and keywords relevant to the study objectives. Our search focused on novel, biologically targeted therapies in recent years, including anti-CD19 CAR T-cell therapy, BsAbs, anti-CD19 ADCs, and lenalidomide plus anti-CD20 antibodies (rituximab or obinutuzumab). We systematically searched for research articles in the following databases: PubMed, Web of Science, Scopus, Embase, and Cochrane Library from 1 January 2019 to 10 July 2025. Additionally, we reviewed conference abstracts published by the American Society of Clinical Oncology, European Hematology Association, and American Society of Hematology.
The search terms were based on MeSH and included the following: lymphoma, indolent; follicular; CAR-T; axicabtagene ciloleucel; tisagenlecleucel; lisocabtagene maraleucel; antibodies, bispecific; mosunetuzumab; glofitamab; epcoritamab; odronextamab; phosphatidylinositol 3-kinase inhibitor; idelalisib; copanlisib; duvelisib; umbralisib; linperlisib; parsaclisib; TQ-B3525; HPML-689; loncastuximab; tazemetostat; tafasitamab; atezolizumab; zanubrutinib; magrolimab; MIL62; obinutuzumab; rituximab; and lenalidomide. For PubMed, the Cochrane highly sensitive search strategy for identifying clinical trial reports was used. Additionally, we manually screened the reference lists of all included studies to identify any potentially relevant trials that may have been missed in the initial search.
Selection criteria
Participants meeting the following inclusion and exclusion criteria were selected for this study. We included full articles and conference abstracts that reported the efficacy of therapeutic regimens in FL with POD24. Our inclusion criteria were as follows: (1) studies conducted in adult populations; (2) studies focusing on FL with POD24; (3) studies evaluating anti-CD19 CAR T-cell therapies, including tisagenlecleucel (tisa-cel), lisocabtagene maraleucel (liso-cel), or axicabtagene ciloleucel (axi-cel); BsAbs, including mosunetuzumab, odronextamab, or epcoritamab; loncastuximab; PI3K inhibitors, including TQ-B3525 or linperlisib; tazemetostat; atezolizumab; magrolimab; zanubrutinib; obinutuzumab; rituximab; or lenalidomide; (4) studies with at least 10 patients in the study arm; (5) studies providing sufficient data, including ORR, CR, or PFS; and (6) prospective clinical trials.
Data extraction and quality assessment
We performed a systematic review and extracted the following study characteristics: publication year, first author’s name, sample size, ClinicalTrials.gov identifier, trial phase, number of prior treatment lines, FL grade, and study design. Furthermore, we collected treatment-related outcomes, including ORR, CR, duration of response (DOR), PFS, and follow-up duration.
Two reviewers (Z.Z. and H.M.) independently screened all relevant studies, and any discrepancies were resolved through consensus. The quality of the trials was evaluated using the Joanna Briggs Institute critical-appraisal checklist for systematic reviews and research syntheses.
Outcome measures
The primary end point was ORR as assessed by the independent review committee. Secondary end points included independent review committee–assessed CR, DOR, and PFS. DOR was defined as the time from the initial achievement of ORR until disease progression or death in patients who achieved an ORR. PFS was defined as the time from the initiation of treatment to either disease progression or death, based on the Lugano criteria for all patients. The definition of PFS was consistent across all trials.
Statistical analysis
All statistical analyses were performed using RevMan version 5.4 (Cochrane), Stata version 16 (StataCorp), and R software (R Foundation for Statistical Computing), with the metan and metaprop packages incorporated. A random-effects model was used to calculate the pooled estimates of CR and ORR, along with their corresponding 95% confidence intervals (CIs) across intervention groups. Between-study heterogeneity was assessed using the Higgins I2 statistic, with the following thresholds for interpretation: ≤25% (low heterogeneity), 26% to 50% (moderate heterogeneity), and >50% (substantial heterogeneity). Publication bias was assessed using the Egger test and funnel plot.
This analysis is based exclusively on previously published data and does not involve the collection of new data from human participants or animal models. Ethical approval was not required for this study because all data included in this analysis were derived from publicly accessible sources or studies that complied with relevant ethical guidelines and obtained appropriate consent from participants where applicable.
Results
Description of trials and patient characteristics
This systematic review and meta-analysis were conducted according to the reporting guidelines in Preferred Reporting Items for Systematic Reviews and Meta-Analyses and Meta-Analyses of Observational Studies in Epidemiology. After eligibility screening, 21 studies met the inclusion criteria (Figure 1). The systematic review ultimately included 15 peer-reviewed articles5,6,8, 9, 10, 11,16, 17, 18, 19, 20, 21, 22, 23, 24 and 6 conference abstracts,7,12,13,20,21,25 collectively comprising 21 interventional trials that satisfied the predefined eligibility criteria. The included trials were conducted between 2014 and 2023, with publication timelines extending beyond the study period: results from 2 trials were reported in 2025, 10 in 2024, 4 in 2023, 2 in 2022, 1 in 2021, 1 in 2020, and 1 in 2019. Three trials addressed anti-CD19 CAR T-cell therapy,5,16,17 4 trials CD20×CD3 BsAbs,6,18,19,25 2 trials anti-CD19 ADCs/mAbs,7,13 1 trial anti-CD47–containing regimens,26 1 trial zanubrutinib-containing regimens,11 3 trials PI3K inhibitor–containing regimens,8,9,20 2 trials enhancer of zeste homolog 2 (EZH2) inhibitor–containing therapy,10,26 3 trials R2,12,13,22 and 4 trials evaluated lenalidomide plus obinutuzumab (LG).20,23,24,27
Figure 1.
Flowchart depicting the identification, screening, eligibility assessment, and inclusion of studies in the systematic review and meta-analysis. Records were identified through electronic database searches and additional sources.
A total of 1242 patients with FL and POD24 were included in our analysis. The number of prior treatment lines ranged from 1 to 18, and the median follow-up duration was 9.7 to 49.4 months. Table 1 summarizes the characteristics of the included patients.
Table 1.
Characteristics of patients included in trials
| Category | Reference | Clinical identification | Phase | LOPT | Treatment | POD24 | Patient number | ORR, n (%) | CR, n (%) | mDOR | mPFS | Follow-up (mo) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Anti-CD19 CAR T-cell therapy | 16 | NCT03568461 | 2 | 4 (2-13) | Tisa-cel | Yes | 61 | 50 (82) | 36 (59) | — | — | 29 |
| No | 33 | 31 (93.9) | 28 (84.8) | — | — | |||||||
| 17 | NCT04245839 | 2 | 2 (1-10) | Liso-cel | Yes | 71 | 68 (95.8) | 67 (94.4) | 12-mo, 78% | 12-mo, 78.4% | 18.9 | |
| No | 52 | 51 (98.1) | 49 (94.2) | 12-mo, 92.1% | 12-mo, 90.3% | |||||||
| 5 | NCT03105336 | 2 | 3 (2-4) | Axi-cel | Yes | 57 | 53 (93) | 41 (72) | 18-mo, 59.6% | 18-mo, 55.3% | 17.5 | |
| No | 36 | 33 (92) | 30 (83) | 18-mo, 77.6% | 18-mo, 84.1% | |||||||
| CD20×CD3 BsAbs | 18 | NCT03888105 | 2 | 3 (2-13) | Odronextamab | Yes | 63 | 51 (81) | 46 (73) | — | — | 20.1 |
| No | 65 | 52 (80) | — | — | — | |||||||
| 6 | NCT02500407 | 2 | 3 (2-7) | Mosunetuzumab | Yes | 47 | 38 (80.9) | 27 (59.6) | 45-mo, 50.9% | 48-mo, 38.9% | 49.4 | |
| No | 43 | 32 (74) | 27 (63) | 45-mo, 49% | 48-mo, 37.8% | |||||||
| 19 | NCT03625037 | 2 | 3 (2-4) | Epcoritamab | Yes | 67 | 53 (79) | 43 (64) | — | — | 17.4 | |
| No | 61 | 52 (85) | 37 (61) | — | — | |||||||
| 25 | NCT04663347 | 2 | 1 (1-7) | Epcoritamab + R2 | Yes | 42 | — | 33 (79) | — | 27.6 mo (23.7 to NR) | 25.3 | |
| No | 69 | — | 64 (93) | — | — | |||||||
| Anti-CD19 ADC/mAb | 7 | NCT04998669 | 2 | 1 (1-6) | Loncastuximab + R | Yes | 16 | 16 (100) | 14 (87.5) | — | — | 9.7 |
| No | 11 | 10 (90.9) | 9 (81.8) | — | — | |||||||
| 13 | NCT04680052 | 3 | 1 (1-10) | Tafasitamab + R2 | Yes | 121 | 96 (79.3) | 52 (43.2) | — | 19.2 mo (14.8 to NE) | 14.1 | |
| No | 142 | 124 (87.3) | 78 (55) | — | 24.0 mo (22.3 to NE) | |||||||
| Anti-CD47–containing regimens | 26 | NCT02953509 | 1b/2 | 3 (1-9) | Magrolimab + R | Yes | 26 | 15 (57.7) | 9 (34.6) | — | — | 36.7 |
| No | 20 | 9 (45) | 5 (25) | — | — | |||||||
| Zanubrutinib-containing regimens | 11 | NCT03332017 | 2 | 3 (2-11) | Zanubrutinib + G | Yes | 50 | 30 (60) | — | — | — | 20.2 |
| No | 74 | 55 (74) | — | — | — | |||||||
| PI3K inhibitor–containing regimens | 8 | NCT04324879 | 2 | 3 (2-7) | TQ-B3525 | Yes | 60 | 51 (85) | — | NR (9.2 to NE) | NR (9.2 to NE) | 24.9 |
| No | 22 | 20 (90.9) | — | 16.8 mo (5.6 to NE) | 18.5 mo (7.4 to NE) | |||||||
| 9 | NCT04370405 | 2 | 4 (2-18) | Linperlisib | Yes | 61 | 47 (77.1) | — | 13 mo (9.3 to NR) | 13.7 mo (11.1 to 19.4) | 16.7 | |
| No | 23 | 20 (87) | — | 12.3 mo (6.6 to NR) | 11.5 mo (8.2 to NR) | |||||||
| 20 | NCT03269669 | 2 | ≥2 | Umbralisib + G | Yes | 23 | — | 12 (52) | 19 mo | 48-mo, 28% | 49 | |
| EZH2 inhibitor–containing regimens | 10 | NCT01897571 | 2 | ≥2 | Tazemetostat | Yes | 51 (EZH2mut) | 12 (63) | 2 (11) | 6.6 mo (2.1 to NE) | 13.8 mo (7.3 to NE) | 22 |
| Yes | 32 (EZH2WT) | 8 (25) | 1 (3) | 13.0 mo (0.5 to NE) | 5.8 mo (2.0-16.7) | 35.9 | ||||||
| 21 | NCT04224493 | 1b | 1 (1-4) | Tazemetostat + R2 | Yes | 12 | 11 (91.7) | — | NR | NR | 11.2 | |
| Anti-CD20–containing regimens | 12 | NCT01996865 | 3b | ≥2 | R2 | Yes | 133 | 86 (65) | 43 (32) | 37.0 mo (24.9 to NR) | 37.4 mo (20.3-41.6) | 40.6 |
| No | 261 | 193 (74) | 121 (46) | NR (NR to NR) | NR (41.4 to NR) | |||||||
| 13 | NCT04680052 | 3 | 1 (1-10) | R2 | Yes | 128 | 83 (64.7) | 40 (31.4) | — | 11.5 mo (8.5-13.9) | 14.1 | |
| No | 140 | 113 (80.7) | 69 (49.2) | — | 18.2 mo (14.4 to NE) | |||||||
| 22 | NCT04110301 | 1b/2 | 1 (1-2) | Lenalidomide + MIL62 | Yes | 31 | 26 (84) | 9 (29) | 9-mo, 68% | 12-mo, 68% | 12.3 | |
| 23 | NCT01995669 | 1b/2 | 1 (1-7) | LG | Yes | 27 | 27 (100) | 18 (67) | — | 44.2 mo (22.5 to NR) | 42.7 | |
| 24 | NCT01582776 | 2 | 2 (1-7) | LG | Yes | 24 | 18 (75) | — | — | NR (1.4 to NR) | 31.2 | |
| 20 | NCT03269669 | 2 | ≥2 | LG | Yes | 27 | — | 14 (52) | NR | 48-mo, 49% | 49 | |
| 27 | NCT02631577 | 1b/2 | ≥1 | LG + atezolizumab | Yes | 12 | 7 (58.3) | 6 (50) | — | — | 35.9 | |
| No | 20 | 18 (90) | 17 (85) | — | — |
LOPT, lines of prior treatment; mDOR, median DOR; G, obinutuzumab; mut, mutant; NE, not estimated; R, rituximab; WT, wild-type.
Risk of bias of included trials
All trials were open-label; 3 trials were controlled and 18 trials were single-arm. One trial was phase 1b, 4 were phase 1b/2, 14 were phase 2, and 2 were phase 3; the primary end point was assessed by an independent review committee. Consequently, the risk of attrition bias and selective outcome reporting bias across all trials was minimal. The Egger test indicated no publication bias for ORR (P = .549 and P = .087) and CR (P = .625 and P = .148). The funnel plot also confirmed the absence of publication bias for both ORR and CR (supplemental Figures 1, 3, 5, and 7). Sensitivity analysis of ORR and CR demonstrated that the data were stable (supplemental Figures 2, 4, 6, and 8).
Comparison of response rates between populations with and without POD24 with the same treatment regimen
Fourteen trials provided results comparing response rates between POD24 and non-POD24 populations with 15 treatment groups.5, 6, 7, 8, 9,11, 12, 13,16, 17, 18, 19,26,27 The specific response rates for different anti-CD19 CAR T-cell therapy in patients with and without POD24 were as follows: tisa-cel: ORR (82% vs 93.9%) and CR (59% vs 84.8%); liso-cel: ORR (95.8% vs 98.1%) and CR (94.4% vs 94.2%); and axi-cel : ORR (93% vs 92%) and CR (72% vs 83%). In pooled analyses of ORR and CR (POD24 vs non-POD24), odds ratios (ORs) were 0.565 for ORR (P = .4456; I2 = 0%) and 0.447 for CR (P = .3350; I2 = 8.55%). The specific response rates for different BsAbs among patients with and without POD24 were as follows: odronextamab: ORR (81% vs 80%) and CR (73% vs not available); mosunetuzumab: ORR (80.9% vs 74%) and CR (59.6% vs 63%); and epcoritamab: ORR (79% vs 85%) and CR (64% vs 61%). In pooled analyses of ORR and CR (POD24 vs non-POD24), ORs were 0.987 for ORR (P = .5064; I2 = 0%) and 0.994 for CR (P = .5093; I2 = 0%). The specific response rates for different anti-CD19 ADCs/mAbs among patients with and without POD24 were as follows: loncastuximab plus rituximab: ORR (100% vs 90.9%) and CR (87.5% vs 81.8%); and tafasitamab plus R2: ORR (79.3% vs 87.3%) and CR (43.2% vs 55%). In pooled analyses of ORR and CR (POD24 vs non-POD24), ORs were 0.860 for ORR (P = .2128; I2 = 35.58%) and 0.648 for CR (P = .4083; I2 = 0%). The specific response rates for PI3K inhibitors among patients with and without POD24 were as follows: TQ-B3525: ORR (85% vs 90.9%); and linperlisib: ORR (77.1% vs 87%). In pooled analysis of ORR (POD24 vs non-POD24), the OR was 0.529 (P = .9126; I2 = 0%). The specific response rates for 2 R2 treatment groups among patients with and without POD24 were as follows: group 1: ORR (65% vs 74%) and CR (32% vs 46%); group 2: ORR (64.7% vs 80.7%) and CR (31.4% vs 49.2%). In pooled analyses of ORR and CR (POD24 vs non-POD24), ORs were 0.553 for ORR (P = .2968; I2 = 8.12%) and 0.514 for CR (P = .6217; I2 = 0%). The specific response rates for anti-CD20 mAb-containing combinations among patients with and without POD24 were as follows: zanubrutinib plus obinutuzumab: ORR (60% vs 74%); magrolimab plus rituximab: ORR (57.7% vs 45%) and CR (34.6% vs 25%); and LG plus atezolizumab: ORR (58.3% vs 90%) and CR (50% vs 85%). In pooled analyses of ORR and CR (POD24 with non-POD24), ORs were 0.593 for ORR (P = .0798; I2 = 60.44%) and 0.565 for CR (P = .0415; I2 = 75.93%) (Figures 2 and 3).
Figure 2.
Bar graph showing ORR in patients with vs without POD24 receiving the same treatment regimen. ORs (POD24 vs non-POD24) for patients receiving anti-CD19 CAR T-cell therapy; CD20 and CD3 BsAbs; anti-CD19 ADCs/mAbs; PI3K inhibitors; R2; and anti-CD20 mAb-containing regimens combined with zanubrutinib, magrolimab, or lenalidomide plus atezolizumab were 0.565, 0.987, 0.860, 0.529, 0.553, and 0.593, respectively. Error bars represent 95% CIs. Data were analyzed using the χ2 or the Fisher exact test, as appropriate. df, degrees of freedom; G, obinutuzumab; IV, inverse variance; R, rituximab.
Figure 3.
Bar graph showing CR rates in patients with vs without POD24 receiving the same treatment regimen. ORs (POD24 vs non-POD24) for patients with anti-CD19 CAR T-cell therapy; CD20 and CD3 BsAbs; anti-CD19 ADCs/mAbs; R2; and anti-CD20 mAb-containing regimens combined with zanubrutinib, magrolimab, or lenalidomide plus atezolizumab were 0.447, 0.994, 0.648, 0.514, and 0.565, respectively. Error bars represent 95% CIs. Data were analyzed using the χ2 or the Fisher exact test, as appropriate. df, degrees of freedom; G, obinutuzumab; IV, inverse variance; R, rituximab.
Comparisons of efficacy between different types of treatment regimens in patients with POD24
The 17 treatment regimens involved in the 21 trials were divided into 6 types of treatment regimens, and the efficacy of different types of treatment regimens was compared. For anti-CD19 CAR T-cell therapies,5,16,17 the ORR and CR of tisa-cel, liso-cel, and axi-cel were 82%, 95.8%, and 93%, and 59%, 94.4%, and 72%, respectively. Upon pooled analysis, the overall ORR was 91.2% (95% CI, 83.7-98.7) with significant heterogeneity (P = .0414; I2 = 68.61%), and the CR was 75.7% (95% CI, 55.1-96.4) with significant heterogeneity (P < .0001; I2 = 93.99%). For CD20×CD3 BsAbs,6,18,19 the ORR and CR of odronextamab, mosunetuzumab, and epcoritamab were 81%, 80.9%, and 79%, and 73%, 59.6%, and 64%, respectively. Upon pooled analysis, the overall ORR was 80.3% (95% CI, 74.4-86.1) with no heterogeneity (P = .9587; I2 = 0%), and the CR was 65.7% (95% CI, 57.1-74.3) with moderate heterogeneity (P = .2148; I2 = 34.99%). For anti-CD19 ADCs/mAbs,7,13 the ORR and CR of loncastuximab plus rituximab and tafasitamab plus R2 were 100% and 79.3%, and 87.5% and 43.2%, respectively. Upon pooled analysis, the overall ORR was 89.6% (95% CI, 69.3-100) with significant heterogeneity (P = .0002; I2 = 92.89%), and the CR was 64.7% (95% CI, 21.1-100) with significant heterogeneity (P < .0001; I2 = 95.53%). For PI3K inhibitors,8,9 the ORR of TQ-B3525 and linperlisib were 85% and 77.1%, respectively. Upon pooled analysis, the overall ORR was 81.5% (95% CI, 73.8-89.2) with low heterogeneity (P = .2620; I2 = 20.53%). For R2 regimens,12,13,22 the ORR and CR of the 3 trials involving the R2 regimen were 65%, 64.7%, and 84%, and 32%, 31.4%, and 29%, respectively. Upon pooled analysis, the overall ORR was 70.1% (95% CI, 58.7-81.5) with significant heterogeneity (P = .0305; I2 = 71.35%), and the CR was 31.5% (95% CI, 26.2-36.8) with no heterogeneity (P = .9337; I2 = 0%). For the LG regimen,20,23,24,27 the ORR and CR of the 4 trials involving LG with or without atezolizumab were 100%, 75%, not reached (NR), and 58.3%, and 67%, NR, 52%, and 50%, respectively. Upon pooled analysis, the overall ORR was 80.6% (95% CI, 56.7-100) with significant heterogeneity (P = .0005; I2 = 86.68%), and the CR was 58% (95% CI, 46.3-69.8) with no heterogeneity (P = .4430; I2 = 0%) (Figures 4 and 5).
Figure 4.
Bar graph depicting ORR of different treatment regimens in patients with POD24. ORR in POD24 population receiving anti-CD19 CAR T-cell therapy, CD20 and CD3 BsAbs, anti-CD19 ADCs/mAbs, PI3K inhibitors, R2, and LG were 91.2%, 80.3%, 89.6%, 81.5%, 70.1%, and 80.6%, respectively. df, degrees of freedom; G, obinutuzumab; IV, inverse variance; R, rituximab.
Figure 5.
Bar graph depicting CR rates of different treatment regimens in patients with POD24. CR in POD24 population receiving anti-CD19 CAR T-cell therapy, CD20 and CD3 BsAbs, anti-CD19 ADCs/mAbs, R2, and LG were 75.7%, 65.7%, 64.7%, 31.5%, and 58.0%, respectively. df, degrees of freedom; G, obinutuzumab; IV, inverse variance; R, rituximab.
Secondary outcomes
The median duration of follow-up ranged from 9.7 to 49.4 months. Data from 8 trials were included in the analysis of median DOR as follows: liso-cel: 12 months, 78%; axi-cel: 18 months, 59.6%; TQ-B3525, NR; linperlisib, 13 months; tazemetostat (EZH2mut), 6.6 months; tazemetostat (EZH2WT), 13.0 months; tazemetostat plus R2, NR; R2, 37.0 months; and lenalidomide plus MIL62: 9 months, 68%; respectively.8, 9, 10,12,16,17,22,26 Eleven trials were available for median PFS (mPFS) analysis: liso-cel: 12 months, 78.4%; axi-cel: 18 months, 55.3%; TQ-B3525, NR; linperlisib, 13.7 months; tazemetostat (EZH2mut), 13.8 months; tazemetostat (EZH2WT), 5.8 months; tazemetostat plus R2, NR; R2, 11.3 and 37.4 months (from 2 separate trials); tafasitamab plus R2, 19.2 months; LG: 44.2 months, NR; and lenalidomide plus MIL62: 12 months, 68%; respectively8, 9, 10,12,13,16,17,20, 21, 22,26 (Table 1).
Discussion
Patients who experience early progression after frontline therapy represent a biologically distinct and high-risk subgroup, often characterized by intrinsic resistance to conventional immunochemotherapy, genomic instability, increased tumor burden, immune evasion mechanisms, and immunosuppressive tumor microenvironment. Despite this aggressive phenotype, emerging evidence suggests that these patients may derive substantial benefit from novel immunotherapies, such as anti-CD19 CAR T cells, BsAbs, and ADCs, because of their ability to overcome key mechanisms of treatment resistance. Although the underlying disease biology remains challenging, these immunotherapies offer mechanistically distinct approaches that can restore immune surveillance and induce deep responses even in high-risk populations; however, long-term remissions remain a challenge and warrant further investigation.
CAR T-cell therapy and BsAbs are approved for patients with R/R FL in the third- or later-line setting. Patients with POD24 achieved higher CR rates in the liso-cel study (94.4%) than in the tisa-cel (59%) and axi-cel (72%) studies. The CR rate was similar between patients with and without POD24 in the liso-cel study (94.4% vs 94.2%), whereas CR rates were numerically lower in patients with vs without POD24 in the tisa-cel (59% vs 84.8%) and axi-cel (72% vs 83%) studies.5,16,17 The liso-cel study included second-line patients (12 cases), all of whom had POD24 after diagnosis and treatment; fewer patients who were heavily pretreated enrolled in the axi-cel than in the tisa-cel study (median lines of prior therapy, 3 vs 4).5,16,17 In the axi-cel study, CR rates among patients receiving 2, 3, and ≥4 prior therapies (with and without POD24) were 72%, 73%, and 76%, respectively,5 indicating that the number of prior therapies may not affect the efficacy of CAR T-cell therapy. Furthermore, a tisa-cel study indicated that CAR T-cell dose was not significantly associated with clinical outcomes, whereas baseline clinical factors, tumor microenvironment, blood soluble factors, and circulating blood cells were correlated with CAR T-cell therapy clinical response. Low levels of tumor-infiltrating LAG3+CD3+ exhausted T cells and higher baseline levels of naive CD8+ T cells were associated with improved outcomes.16
Regarding durability of response, the PFS of patients with POD24 after CAR T-cell therapy was significantly shorter than that of patients without POD24 (liso-cel: 12 months, 78.4% vs 90.3%; axi-cel: 18 months, 55.3% vs 84.1%).5,15,16 Mosunetuzumab-treated patients with POD24 had PFS similar to those without POD24 (mPFS, 26.3 vs 21.7 months), which indicated the persistence of the therapeutic effect of mosunetuzumab. Besides mosunetuzumab, odronextamab, and epcoritamab monotherapy also achieved comparable ORR and CR rates between patients with and without POD24, although there are no long-term follow-up survival results.18,19 In the absence of head-to-head data, Dahiya et al conducted a matching-adjusted indirect comparison to compare the efficacy of liso-cel (NCT04245839) and epcoritamab (NCT03625037) among patients with third-line or later R/R FL. Liso-cel was associated with a significantly higher ORR (OR, 1.13) and CR rate (OR, 1.45) compared with epcoritamab. Liso-cel also demonstrated significantly improved DOR (hazard ratio [HR], 0.41), PFS (HR, 0.33), and OS (HR, 0.33).28 In another matching-adjusted indirect comparison of efficacy between liso-cel (NCT04245839) and mosunetuzumab (NCT03625037), liso-cel was associated with higher ORR (OR, 3.78) and CR (OR, 6.46) rates, and improved PFS (HR, 0.28).29
In previous reports on R/R FL, PI3K inhibitors idelalisib, duvelisib, copanlisib, and umbralisib achieved ORR rates of 42.2% to 59%.30, 31, 32, 33 In the POD24 population, a prospective study showed that a combination of umbralisib with obinutuzumab was inferior to R2 (48-month PFS, 28% vs 49%).20 Both TQ-B3525 and linperlisib achieved high ORR rates in patients with POD24 comparable to those observed in patients without POD24.8,9 In addition, improved DOR (HR, 0.69) and PFS (HR, 0.68) were observed among patients with POD24 than those without POD24,8 and progression incidence rates were only 3.3% for patients with POD24 and 7.2% for those without POD24.9 Taken together, TQ-B3525 and linperlisib are effective regimens for the poor-prognosis POD24 subgroup. EZH2 inhibitors had lower CR rates even in patients with EZH2-mutated POD24, resulting in a PFS of only 13.8 months at a follow-up period that was almost as long as that in the TQ-B3525 study.8,11 Another small-molecule inhibitor, zanubrutinib, combined with obinutuzumab and recommended as third-line therapy for R/R FL, achieved an ORR rate of 60% in patients with POD24; however, data on PFS were not available. The ongoing MAHOGANY study is comparing the efficacy of zanubrutinib plus obinutuzumab and R2 in R/R FL.
The AUGMENT study opened the era of R2 regimen for the treatment of R/R FL without chemotherapy.34 The R2 regimen achieved a CR rate of 31.4% to 32% and demonstrated an mPFS of up to 44.2 months among patients with POD24.12,13 MIL62, a novel glycoengineered type II anti-CD20 mAb, combined with lenalidomide, achieved efficacy similar to that of R2.22 When obinutuzumab was used instead of rituximab in combination with lenalidomide, higher CR rates of 50% to 67% were achieved,20,23,27 including regimens incorporating atezolizumab (programmed cell death-ligand 1 monoclonal antibodies [PD-L1 mAb]), with a 48-month mPFS of up to 49%.20 These findings suggest that obinutuzumab is more effective than rituximab plus lenalidomide in patients with POD24. Nevertheless, the addition of R2 to certain treatments has shown enhanced efficacy. For instance, epcoritamab combined with R2 resulted in a higher CR rate compared with epcoritamab alone (79% vs 64%).19,25 Similarly, tafasitamab with R2 achieved a superior CR rate compared with R2 alone (43.2% vs 31.4%) and extended PFS (19.2 vs 11.5 months).13 Tazemetostat paired with R2 also demonstrated an increased ORR compared with tazemetostat alone (91.7% vs 24%) and a longer PFS.10,21 Therefore, R2 remains a cornerstone regimen for patients with POD24.
Several limitations of our analysis merit consideration. The retrospective and observational design of the study limits the ability to make direct comparisons between therapies because of potential uncontrolled selection bias. The presence of significant heterogeneity in certain pooled analyses of ORR and CR rates may impact the interpretation of the findings. The routine use of BsAbs and CAR T-cell therapies is currently limited because they are only approved for use after ≥2 lines of therapy. Their inclusion across different lines of treatment represents a limitation of our study.
Conclusion
In summary, the management of patients with R/R FL and POD24 remains a critical unmet clinical need in hematologic oncology. Our pooled analysis of 21 independent clinical cohorts demonstrated that anti-CD19 CAR T-cell therapy achieved the highest CR rate, followed by BsAbs, anti-CD19 ADCs/mAbs, LG, and R2, respectively. R2, a cornerstone of treatment for these patients, combined with epcoritamab, tafasitamab, or tazemetostat, also achieved very good efficacy.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Acknowledgments
All claims expressed in this article are solely those of the authors and do not necessarily reflect the views of their affiliated organizations, the publisher, the editors, or the reviewers. Any product evaluated in this article or any claim made by its manufacturers should not be assumed to have been endorsed by the aforementioned parties. The manufacturer neither guarantees nor endorses the content, and the publisher assumes no responsibility.
Authorship
Contribution: J.S. and J.L. contributed to conceptualization, design, and drafting of the manuscript; Jingyi Zhang and Z.Z. were involved in data curation, reviewing, and editing; H.M. and Junpeng Zhang were involved in data curation and formal analysis; F.Z. and H.T. contributed to data curation, writing, and editing; and all authors contributed to the article and approved the submitted version.
Footnotes
Deidentified data generated during this study are available from the corresponding author, JingHua Liu (mtljh7646@163.com), on request. The data will be accessible for 36 months after publication.
The full-text version of this article contains a data supplement.
Supplementary Material
References
- 1.Casulo C, Byrtek M, Dawson KL, et al. Early relapse of follicular lymphoma after rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone defines patients at high risk for death: an analysis from the National LymphoCare Study. J Clin Oncol. 2015;33(23):2516–2522. doi: 10.1200/JCO.2014.59.7534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lansigan F, Barak I, Pitcher B, et al. The prognostic significance of PFS24 in follicular lymphoma following first line immunotherapy: a combined analysis of 3 CALGB trials. Cancer Med. 2019;8(1):165–173. doi: 10.1002/cam4.1918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Casulo C, Dixon JG, Le-Rademacher J, et al. Validation of POD24 as a robust early clinical end point of poor survival in FL from 5225 patients on 13 clinical trials. Blood. 2022;139(11):1684–1693. doi: 10.1182/blood.2020010263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Limbrick-Oldfield EH, Kanters S, Ray MD, et al. The prognostic value of POD24 in relapsed/refractory follicular lymphoma-a SCHOLAR-5 analysis. EJHaem. 2025;6(1) [Google Scholar]
- 5.Jacobson CA, Chavez JC, Sehgal AR, et al. Axicabtagene ciloleucel in relapsed or refractory indolent non-Hodgkin lymphoma (ZUMA-5): a singlearm, multicentre, phase 2 trial. Lancet Oncol. 2022;23(1):91–103. doi: 10.1016/S1470-2045(21)00591-X. [DOI] [PubMed] [Google Scholar]
- 6.Shadman M, Bartlett NL, Matasar M, et al. Mosunetuzumab continues to demonstrate clinically meaningful outcomes in patients with relapsed and/or refractory follicular lymphoma after ≥2 prior therapies including those with a history of POD24: 4-year follow-up of a pivotal phase Ⅱ study. EHA. 2024;144(suppl 1):4407. [Google Scholar]
- 7.Alderuccio JP, Alencar AJ, Schatz JH, et al. Limited duration loncastuximab tesirine with rituximab induces high complete metabolic response rate in high-risk relapsed/refractory follicular lymphoma—a phase 2 study. Blood. 2023;142(suppl 1):984. [Google Scholar]
- 8.Wang HQ, Feng JF, Liu YY, et al. Phase II study of novel orally PI3Kα/δ inhibitor TQ-B3525 in relapsed and/or refractory follicular lymphoma. Signal Transduct Target Ther. 2024;9(1):99–108. doi: 10.1038/s41392-024-01798-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wang TY, Sun XH, Qiu LH, et al. The oral PI3Kδ inhibitor linperlisib for the treatment of relapsed and/or refractory follicular lymphoma: a phase II, single-arm, open-label clinical trial. Clin Cancer Res. 2023;29(8):1440–1449. doi: 10.1158/1078-0432.CCR-22-2939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Morschhauser F, Tilly H, Chaidos A, et al. Tazemetostat for patients with relapsed or refractory follicular lymphoma: an open-label, single-arm, multicentre, phase 2 trial. Lancet Oncol. 2020;21(11):1433–1442. doi: 10.1016/S1470-2045(20)30441-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zinzani PL, Mayer J, Flowers CR, et al. ROSEWOOD: a phase II randomized study of zanubrutinib plus obinutuzumab versus obinutuzumab monotherapy in patients with relapsed or refractory follicular lymphoma. J Clin Oncol. 2023;41(33):5107–5117. doi: 10.1200/JCO.23.00775. [DOI] [PubMed] [Google Scholar]
- 12.Lansigan F, Andorsky DJ, Coleman M, et al. P1156: magnify phase 3b study of lenalidomide + rituximab (. R2) followed by maintenance in relapsed/refractory indolent non-Hodgkin lymphoma: complete induction phase analysis. Hemasphere. 2022;6(suppl):1043–1044. [Google Scholar]
- 13.Sehn LH, Hübel K, Luminari S, et al. Post hoc analysis of outcomes by POD24 status from the inMIND study of tafasitamab plus lenalidomide and rituximab in relapsed or refractory follicular lymphoma. Hematol Oncol. 2025;43 [Google Scholar]
- 14.Day JR, Larson MC, Durani U, et al. Treatment patterns and outcomes in follicular lymphoma with POD24: an analysis from the LEO Consortium. Blood Adv. 2025;9(5):1013–1023. doi: 10.1182/bloodadvances.2024014053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Caridà G, Martino EA, Bruzzese A, et al. Relapsed/refractory follicular lymphoma: current advances and emerging perspectives. Eur J Haematol. 2025;114(5):775–784. doi: 10.1111/ejh.14401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Dreyling M, Fowler NH, Dickinson M, et al. Durable response after tisagenlecleucel in adults with relapsed/refractory follicular lymphoma: ELARA trial update. Blood. 2024;143(17):1713–1725. doi: 10.1182/blood.2023021567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Morschhauser F, Dahiya S, Palomba ML, et al. Lisocabtagene maraleucel in follicular lymphoma: the phase 2 TRANSCEND FL study. Nat Med. 2024;30(8):2199–2207. doi: 10.1038/s41591-024-02986-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kim TM, Taszner M, Novelli S, et al. Safety and efficacy of odronextamab in patients with relapsed or refractory follicular lymphoma. Ann Oncol. 2024;35(11):1039–1047. doi: 10.1016/j.annonc.2024.08.2239. [DOI] [PubMed] [Google Scholar]
- 19.Linton KM, Vitolo U, Jurczak W, et al. Epcoritamab monotherapy in patients with relapsed or refractory follicular lymphoma (EPCORE NHL-1): a phase 2 cohort of a single-arm, multicentre study. Lancet Haematol. 2024;11(8):e593–e605. doi: 10.1016/S2352-3026(24)00166-2. [DOI] [PubMed] [Google Scholar]
- 20.Barr PM, Li H, Schoder H, et al. Randomized phase Ⅱ trial comparing lenalidomide/ obinutuzumab and umbralisib/obinutuzumab with chemoimmunotherapy in early progressing follicular lymphoma (POD24) ASH. 2025 [Google Scholar]
- 21.Salles G, Park SI, Phillips TJ, et al. Tazemetostat in combination with lenalidomide and rituximab in patients with relapsed/refractory follicular lymphoma: updated phased 1b results of symphony-1 with 22.5 months follow-up. Blood. 2023;142(suppl 1) [Google Scholar]
- 22.Shi Y, Zhou KS, Zhou H, et al. Efficacy and safety of MIL62, a novel glycoengineered type Ⅱ anti-CD20 monoclonal antibody, combined with lenalidomide in patients with relapsed/refractory follicular lymphoma or marginal zone lymphoma: a multicentre, single-arm, phase 1b/2 trial. eClinicalMedicine. 2024;73 [Google Scholar]
- 23.Gurumurthi A, Chin CK, Feng L, et al. Safety and activity of lenalidomide in combination with obinutuzumab in patients with relapsed indolent non-Hodgkin lymphoma: a single group, open-label, phase 1/2 trial. eClinicalMedicine. 2024;74 [Google Scholar]
- 24.Morschhauser F, Gouill SL, Feugier P, et al. Obinutuzumab combined with lenalidomide for relapsed or refractory follicular B-cell lymphoma (GALEN): a multicentre, single-arm, phase 2 study. Lancet Haematol. 2019;6(8):e429–e437. doi: 10.1016/S2352-3026(19)30089-4. [DOI] [PubMed] [Google Scholar]
- 25.Falchi L, Sureda A, Leppa S, et al. Fixed-duration epcoritamab + R2 drives deep and durable responses in patients with relapsed or refractory follicular lymphoma: 2-year follow-up from arm 2 of the EPCORE NHL-2 trial. Blood. 2024;144(suppl 1):342. doi: 10.1182/blood.2024025365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mehta A, Popplewell L, Collins GP, et al. Magrolimab plus rituximab in relapsed/refractory indolent non-Hodgkin lymphoma: 3-year follow-up of a phase 1/2 trial. Blood Adv. 2024;8(22):5855–5863. doi: 10.1182/bloodadvances.2024013277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Morschhauser F, Ghosh N, Lossos IS, et al. Obinutuzumab-atezolizumab- lenalidomide for the treatment of patients with relapsed/refractory follicular lymphoma: final analysis of a phase Ib/II trial. Blood Cancer. 2021;11(8):147. [Google Scholar]
- 28.Dahiya S, Kumar J, Wang P, et al. Matching-adjusted indirect comparison of lisocabtagene maraleucel versus epcoritamab in patients with third-line or later relapsed or refractory follicular lymphoma. Exp Hematol Oncol. 2025;14(1):30. doi: 10.1186/s40164-025-00610-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Nastoupil LJ, Bonner A, Wang P, et al. Matching-adjusted indirect comparison (. MAIC) of efficacy and safety of lisocabtagene maraleucel (liso-cel) and mosunetuzumab for the treatment (Tx) of third line or later (3L+) relapsed or refractory (R/R) follicular lymphoma (FL) Blood. 2023;142(suppl 1):2338. [Google Scholar]
- 30.Gopal AK, Kahl BS, de Vos S, et al. PI3Kd inhibition by idelalisib in patients with relapsed indolent lymphoma. N Engl J Med. 2014;370(11):1008–1018. doi: 10.1056/NEJMoa1314583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Dreyling M, Santoro A, Mollica L, et al. Long-term safety and efficacy of the PI3K inhibitor copanlisib in patients with relapsed or refractory indolent lymphoma: 2-year follow-up of the CHRONOS-1 study. Am J Hematol. 2020;95(4):362–371. doi: 10.1002/ajh.25711. [DOI] [PubMed] [Google Scholar]
- 32.Fowler NH, Samaniego F, Jurczak W, et al. Umbralisib, a dual PI3Kd/CK1e inhibitor in patients with relapsed or refractory indolent lymphoma. J Clin Oncol. 2021;39(15):1609–1618. doi: 10.1200/JCO.20.03433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Flinn IW, Miller CB, Ardeshna KM, et al. DYNAMO: a phase II study of duvelisib (. IPI-145) in patients with refractory indolent non-Hodgkin lymphoma. J Clin Oncol. 2019;37(11):912–922. doi: 10.1200/JCO.18.00915. [DOI] [PubMed] [Google Scholar]
- 34.Leonard JP, Trneny M, Izutsu K, et al. AUGMENT: a phase III study of lenalidomide plus rituximab versus placebo plus rituximab in relapsed or refractory indolent lymphoma. J Clin Oncol. 2019;37(14):1188–1199. doi: 10.1200/JCO.19.00010. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.