第一部分

一线危险度分层治疗改善具有临床或生物学高危特征的年轻弥漫大B细胞淋巴瘤患者的疗效：中国单中心队列报道

【目的】 在过去的20余年中，利妥昔单抗联合阿霉素、环磷酰胺、长春新碱和泼尼松的方案（R-CHOP）显著改善了弥漫大B细胞淋巴瘤（Diffuse large B-cell lymphoma, DLBCL）患者的预后。然而即便接受“标准”R-CHOP方案治疗，仍有30-40%的患者面临原发难治或治疗后复发，整体长期生存率约60%；尤其在国内，医疗水平的差异导致整体预后更差。多年来，通过联合新药的“R-CHOP+X”方案多数未取得成功，且新药物的添加仅使特定的患者亚组受益；新兴的治疗方案或药物如CAR-T或双特异性抗体等尚难以普及。早期国内、外一些研究显示强化免疫化疗（IIC）可以改善部分高危患者预后且耐受性良好，基于此，我们中心于2012年起开展了一项针对年轻初治DLBCL患者、基于危险度分层调整免疫化疗强度的前瞻性观察性队列研究，旨在探索适应中国国情的整体治疗方案。本项目为分析该队列整体治疗方案的疗效和临床应用价值。

【方法】 本研究纳入了2012年4月至2021年4月来自中国医学科学院血液病医院淋巴瘤诊疗中心的310例连续的初诊DLBCL患者。入组标准为：年龄在14–65岁之间，经活检证实组织学为DLBCL亚型，并且在入组前未接受过既往治疗。患者根据年龄调整国际预后指数评分（aaIPI）被分配到高危（HR）组（aaIPI≥2分）或低危（LR）组（aaIPI=0–1分）；此外利用免疫组化（IHC）和荧光原位杂交（FISH）判断有无高危生物学（bio-HR）因素。HR组和LR组伴有bio-HR患者接受IIC，LR组且不伴bio-HR患者接受R-CHOP治疗。利用SPSS（版本27.0）和R语言（版本4.2.1）对数据进行统计学处理和作图。

【结果】 
本研究共纳入310名65岁以下初诊患者，其中HR组128例，LR组182例。中位年龄49岁（范围14-65岁）。所有患者接受化疗的中位疗程数为6个（2–8）。总体缓解（OR）和完全缓解（CR）率分别为91.2%和79.9%。中位随访42.8个月（2.8–118.40），预计整体5年无进展生存（PFS）和总体生存（OS）率分别为75.1%（95% CI：69.8–80.4%）和84.4%（95% CI：79.9–88.9%）。对于HR组的患者，86.7%（111/128）接受了IIC；其中DA-EP(D)OCH-R是最常用的方案（n=101)，其次是R-hyperCVAD/MA（n=10）。117（91.4%）例患者完成了至少4个疗程诱导治疗，OR和CR率分别为85.9%（110/128）和65.6%（84/128）。预计5年PFS和OS率分别为63.5%（95% CI：54.5–72.5%）和73.5%（95% CI：65.1–81.9%），优于R-CHOP治疗的历史数据。此外，亚组分析显示IIC克服了部分bio-HR特征，如DE、P53表达和CD5表达等。对于LR组的患者，分别有86例（47.3%）和96例（52.7%）患者接受了IIC和R-CHOP治疗，162例患者（90.0%）获得了CR。预计5年PFS和OS率分别为83.7%（95% CI: 78.0–89.4%）和92.2%（95% CI: 87.9–96.5%）。亚组分析显示，合并至少一个bio-HR特征的患者与不伴有bio-HR特征的患者相比PFS（P = 0.56）和OS（P = 0.91）均无显著差异。多因素分析显示只有R-CHOP治疗和aaIPI=1分与较短的PFS独立相关（P = 0.038和P = 0.003）。

【结论】 整体预后危险度分层的治疗策略对于初诊年轻DLBCL患者来说是安全可控且有效的，可提高整体人群的长期生存结局。IIC，尤其是DA-EP(D)OCH-R方案，可降低临床或生物学高危因素的患者复发风险。

第二部分

循环肿瘤DNA动态监测在复发难治性大B细胞淋巴瘤患者CAR-T治疗中的预后意义

【目的】 近些年来，针对复发/难治（Relapsed/Refractory, R/R）大B细胞淋巴瘤（Large B-cell lymphoma, LBCL）患者的创新疗法—嵌合抗原受体T细胞（Chimeric antigen receptor T-cell, CAR-T）治疗获得了巨大成功。然而长期随访显示超过50%的患者在接受CAR-T细胞治疗后未能获得持久缓解，因此早期识别复发或进展仍然是一个重大挑战。一种新兴非侵入性检测手段-循环肿瘤DNA(Circulating tumor DNA, ctDNA)被证实可在一线或挽救性免疫化疗的LBCL患者队列中预测风险分层，但在接受CAR-T细胞治疗的患者中证据较少。同时考虑到CAR-T耐药的肿瘤内在因素尚未阐明，因此在本项研究中，我们首次在接受靶向CD19 CAR-T细胞治疗的亚洲R/R LBCL患者群体中前瞻性地研究循环肿瘤DNA（ctDNA）动态监测的预后价值，并追踪肿瘤突变图谱和克隆演变。

【方法】 本研究纳入23例在中国医学科学院血液病医院淋巴瘤诊疗中心接受靶向CD19 CAR-T细胞治疗成人R/R LBCL患者，这些患者均入组了注册性前瞻性临床试验（NCT04586478和CTR20211683）。病理组织诊断亚型包括弥漫性大B细胞淋巴瘤（DLBCL）、高级别B细胞淋巴瘤（HGBL）和滤泡淋巴瘤转化的LBCL。所有患者在清除淋巴细胞治疗前和CAR-T细胞输注后的多个时间点（第14天、28天、60天、90天和120天后或进展时）采集序贯血浆样本。对血浆样本提取血浆游离DNA后使用靶向二代测序（覆盖188个淋巴瘤相关基因）的方法进行检测，通过口腔或外周血样本过滤胚系突变。对每个样本的ctDNA水平进行定量（Log hGE/mL）和定性；ctDNA阳性（ctDNA+）定义为治疗后重新检测到任何基线突变，而ctDNA阴性（ctDNA-）定义为治疗后所有基线突变均被清除。

【结果】 
本研究入组了23例R/R LBCL患者，共收集了101份外周血血浆样本。治疗前ctDNA水平与多种临床肿瘤负荷特征呈正相关，且高ctDNA水平的患者无进展生存期（PFS）（P = 0.031）和总生存期（OS）（P = 0.023）均更短。治疗后第14天（D14）ctDNA无法检测到(ctDNA–)的患者的3个月CR率令人印象深刻，达到了77.8%，而可检测到ctDNA (ctDNA+)的患者的3个月CR率仅有22.2%（P = 0.015），第28天（D28）观察到的结果类似。与持续ctDNA+的患者相比，D28时ctDNA–的患者1年PFS率（90.9% vs. 27.3%；P = 0.004）和OS率（90.9% vs. 49.1%；P = 0.003）显著更高。值得注意的是，本研究首次报道了ctDNA片段大小在CAR-T细胞治疗 LBCL患者中的预后价值，我们发现ctDNA短片段（<170bp)与较差的PFS（D14的P = 0.031；D28的P = 0.002）和OS（D14的P = 0.013；D28的P = 0.008）显著相关。此外，多个基线基因突变在治疗后进展的患者中更常见，包括TP53、IGLL5、PIM1、BTG1、CD79B、GNA13和P2RY8，其中伴有IGLL5基因突变的患者的PFS（P = 0.008）和OS（P = 0.014）显著更差。

【结论】 治疗前ctDNA水平有望成为评估肿瘤负荷的替代标志物；CAR-T输注后的ctDNA动力学可有效预测治疗反应和生存结局；纵向ctDNA监测的基因组分析可揭示CAR-T治疗的耐药突变并解析R/R LBCL患者的克隆演变模式。我们的研究证实ctDNA作为一种非侵入性的监测方法，在接受CD19 CAR-T细胞治疗的R/R LBCL患者人群的疗效预测和挽救治疗决策方面具有非常广阔的前景。

Part I

Favorable outcomes of front-line risk-adapted therapy in young patients with diffuse large B-cell lymphoma with clinically or biologically high-risk features: Results from a single-center consecutive cohort

Abstract

Objective: Over the past two decades, the regimen of rituximab combined with cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) has significantly improved the prognosis of patients with diffuse large B-cell lymphoma (DLBCL). However, even with "standard" R-CHOP treatment, 30-40% of patients still face primary refractory disease or relapse after treatment, with an overall long-term survival rate of approximately 60%. Particularly in China, disparities in healthcare result in a poorer overall prognosis. Over the years, combination regimens such as "R-CHOP+X" with novel drugs have mostly failed to achieve success, and the addition of new drugs only benefits specific patient subsets. Emerging therapies or drugs such as CAR-T or bispecific antibodies are still not widely accessible. Earlier domestic and foreign studies have shown that intensified immunochemotherapy (IIC) can improve the prognosis of some high-risk patients and is well-tolerated. Based on this, our center initiated a prospective observational cohort study in 2012 targeting young, newly diagnosed DLBCL patients, adjusting the intensity of immunochemotherapy based on risk stratification, aiming to explore a comprehensive treatment regimen adapted to the Chinese context. This project aims to analyze the efficacy and clinical application value of the overall treatment regimen in this cohort.

Methods: A total of 310 consecutive patients with newly diagnosed DLBCL at the Lymphoma Center of the Institute of Hematology & Blood Disease Hospital from April 2012 to April 2021 were included. Inclusion criteria mandated participants to be aged between 14 and 65 years, histologically confirmed with the DLBCL subtype, and without prior treatment before enrollment. Patients were categorized into high-risk (HR) or low-risk (LR) groups based on the age-adjusted international prognostic index score (aaIPI); additionally, high-risk biological (bio-HR) factors were determined through immunohistochemistry (IHC) and Fluorescence in situ hybridization (FISH). Patients in the HR group and LR group with bio-HR features received IIC, while LR group patients without bio-HR features received R-CHOP treatment. The data underwent statistical processing and graphical representation using SPSS (version 27.0) and R (version 4.2.1).

Results: A total of 310 newly diagnosed patients under the age of 65 were enrolled in this study, comprising 128 patients in the HR cohort and 182 patients in the LR cohort. The median age was 49 years (range 14-65 years). The median number of chemotherapy courses administered to all patients was 6 (range: 2–8). The overall response (OR) and complete response (CR) rates were 91.2% and 79.9%, respectively. With a median follow-up duration of 42.8 months (range: 2.8–118.40), the estimated 5-year progression-free survival (PFS) and overall survival (OS) rates were 75.1% (95% CI: 69.8–80.4%) and 84.4% (95% CI :79.9–88.9%), respectively. Among patients in the HR cohort, 86.7% (111/128) received IIC, with the DA-EP(D)OCH-R regimen being the most frequently utilized (n = 101), followed by R-hyperCVAD/MA (n = 10). Notably, 91.4% (117/128) of patients completed a minimum of 4 induction therapy cycles, yielding OR and CR rates of 85.9% (110/128) and 65.6% (84/128), respectively. The estimated 5-year PFS and OS rates were 63.5% (95% CI: 54.5–72.5%) and 73.5% (95% CI: 65.1–81.9%), respectively, surpassing historical R-CHOP treatment benchmarks. Furthermore, subgroup analysis demonstrated that IIC mitigated certain bio-HR features, including DE, P53 expression, and CD5 expression. Among patients in the LR cohort, 86 (47.3%) and 96 (52.7%) received IIC and R-CHOP treatment, respectively, with 162 patients (90.0%) achieving CR. The estimated 5-year PFS and OS rates were 83.7% (95% CI: 78.0–89.4%) and 92.2% (95% CI: 87.9–96.5%), respectively. Subgroup analysis revealed no significant differences in PFS (P = 0.56) and OS (P = 0.91) between patients with at least one bio-HR feature and those without bio-HR features. Multivariate analysis indicated that only R-CHOP treatment and an aaIPI score of 1 were independently associated with shorter PFS (P = 0.038 and P = 0.003, respectively).

Conclusions: The risk-adapted strategy proves to be a reasonable and cost-effective strategy for treating newly diagnosed young patients with DLBCL, demonstrating potential for enhanced long-term survival outcomes across the broader population. IIC, particularly the DA-EP(D)OCH-R regimen, holds promise in mitigating the risk of recurrence among patients harboring clinical or biological high-risk factors.

Part II

The prognostic significance of dynamic monitoring of circulating tumor DNA in patients with relapsed/refractory large B-cell lymphoma undergoing CAR T-cell therapy

Abstract

Objective: In the past several years, innovative therapy targeting relapsed/refractory (R/R) large B-cell lymphoma (LBCL) patients—Chimeric antigen receptor T-cell (CAR-T) therapy—has achieved tremendous success. However, long-term follow-up shows that over 50% of patients fail to achieve durable remission after CAR T-cell therapy, making early identification of relapse or progression still a major challenge. A novel emerging non-invasive detection method - circulating tumor DNA (ctDNA) - has been shown to predict risk stratification in frontline or salvage immunotherapy-treated LBCL patient cohorts, but there is limited evidence in patients receiving CAR T-cell therapy. Considering that intrinsic factors of CAR T-cell resistance in tumors have not been elucidated, in this study, we prospectively investigate the prognostic value of dynamic monitoring of circulating tumor DNA (ctDNA) in an Asian cohort of R/R LBCL patients receiving targeted CD19 CAR T-cell therapy and track tumor mutation profiles and clonal evolution.

Methods: This study enrolled 23 adult R/R LBC patients who received targeted CD19 CAR T-cell therapy at the Lymphoma and Myeloma Center of the Institute of Hematology & Blood Disease Hospital. These patients were all enrolled in registered prospective clinical trials (NCT04586478 and CTR20211683). Pathological diagnoses included diffuse large B-cell lymphoma (DLBCL), high-grade B-cell lymphoma (HGBL), and follicular transformed LBCL. Sequential plasma samples were collected from all patients at baseline and at multiple time points after CAR T-cell infusion (day 14, day 28, day 60, day 90, and day 120 or at progression). Detection was performed by next-generation sequencing (covering 188 lymphoma-related genes) after plasma cell-free DNA extraction, filtering germ line mutations through oral or peripheral blood samples. Quantitative (Log hGE/mL) and qualitative assessments were conducted for ctDNA levels in each sample; ctDNA positivity (ctDNA+) was defined as the re-detection of any baseline mutation after treatment, while ctDNA negativity (ctDNA-) was defined as the clearance of all baseline mutations after treatment.

Results: This study included 23 R/R LBCL patients, with a total of 101 peripheral blood plasma samples collected. The baseline ctDNA levels were positively correlated with various clinical tumor burden characteristics, and patients with high ctDNA levels had shorter progression-free survival (PFS) (P = 0.031) and overall survival (OS) (P = 0.023). The three-month complete response (CR) rate of patients without detectable ctDNA (ctDNA-) at day 14 (D14) post-treatment was impressive at 77.8%, while the three-month CR rate of patients with detectable ctDNA (ctDNA+) was only 22.2% (P = 0.015), with similar results observed at day 28 (D28). Compared to patients with persistent ctDNA+, patients with ctDNA- at D28 had significantly higher one-year PFS rates (90.9% vs. 27.3%; P = 0.004) and OS rates (90.9% vs. 49.1%; P = 0.003). Notably, this study first reported the prognostic value of ctDNA fragment size in LBCL patients treated with CAR T-cell therapy; we found that short ctDNA fragments (<170bp) were significantly associated with poorer PFS (P = 0.031 at D14; P = 0.002 at D28) and OS (P = 0.013 at D14; P = 0.008 at D28). Furthermore, multiple baseline gene mutations were more common in patients with progression after treatment, including TP53, IGLL5, PIM1, BTG1, CD79B, GNA13, and P2RY8, with patients harboring mutations in the IGLL5 gene having significantly poorer PFS (P = 0.008) and OS (P = 0.014).

Conclusions: Baseline ctDNA levels are promising as an alternative marker for assessing tumor burden; the dynamics of ctDNA after CAR T-cell infusion can effectively predict treatment response and survival outcomes; genomic analysis of longitudinal ctDNA monitoring can reveal resistant mutations to CAR T-cell therapy and elucidate clonal evolution patterns in R/R LBCL patients. Our study confirms ctDNA as a non-invasive monitoring method with broad prospects for efficacy prediction and salvage treatment decisions in R/R LBCL patients undergoing CD19 CAR T-cell therapy.