第一部分 CHN1在B细胞非霍奇金淋巴瘤中的预后价值、表型研究及机制探索

背景

B细胞非霍奇金淋巴瘤（B-cell non-Hodgkin lymphoma, B-NHL）是最常见的非霍奇金淋巴瘤，具有高度异质性。因此，如何判断B-NHL尤其是高侵袭性B-NHL患者预后并探索其可能的发病机制，是目前亟待解决的问题。α-chimaerin（CHN1）是Rho小GTP酶家族成员Rac1的特异性GTP酶激活蛋白。既往研究表明，CHN1与B-NHL的侵袭性相关，并通过生信分析证实CHN1是弥漫大B细胞淋巴瘤（diffuse large B-celly lymphoma, DLBCL）预后较好的生物标志物。然而，该研究结果尚未在临床患者中进行验证，其具体的作用机制也尚不明确。因此，本研究旨在通过回顾性分析DLBCL患者病理标本中CHN1的表达水平，验证CHN1与DLBCL患者预后的相关性；并在体外研究中探索CHN1在B-NHL中可能的作用机制。

方法

1. DLBCL患者的病理标本收集与CHN1的免疫组化染色；

2. DLBCL患者临床特征与预后资料收集；

3. 采用慢病毒感染法构建CHN1稳定过表达的B-NHL细胞系；

4. 采用直接计数法检测细胞的增殖能力；

5. 采用Transwell法检测细胞的迁移和侵袭能力；

6. 采用Annexin V/7-AAD双染法检测细胞的凋亡水平；

7. 采用Western Blotting法检测蛋白表达水平；

8. 采用Pull-Down法检测Rac1蛋白活性；

9. 对CHN1稳定过表达细胞的DLBCL细胞系进行转录组测序，采用GO富集分析、KEGG富集分析和GSEA富集分析等方法，探索CHN1的下游通路；

10. 统计学分析。

结果

1. 通过对DLBCL患者病理标本进行CHN1的免疫组化染色，以CHN1阳性率10%作为截断值，将患者分为CHN1高表达与低表达两组。其中，CHN1高表达组客观缓解（objective response, OR）率和完全缓解（complete response, CR）率分别为86.2%和65.5%，均显著高于CHN1低表达组（OR率：39.1%，p=0.001；CR率：21.7%，p=0.002）。CHN1高表达组的疾病进展（progression disease, PD）率（10.3%）显著低于CHN1低表达组（52.2%，p=0.001）。Kaplan-Meier生存分析结果表明，CHN1表达水平与患者的总生存期（overall survival, OS; p=0.036）和无进展生存期（progression-free survival, PFS; p=0.002）显著相关。

2. 分别构建CHN1稳定过表达的Burkitt淋巴瘤细胞系Raji^OE及DLBCL细胞系OCI-LY3^OE。与对照组细胞相比，CHN1过表达抑制了Raji细胞和OCI-LY3细胞的增殖（p<0.001），促进了Raji细胞（p<0.001）和OCI-LY3细胞（p=0.002）的凋亡。另外，Western Blotting结果表明，CHN1抑制了抗凋亡蛋白BCL-2的表达（OCI-LY3^OE：p=0.029；Raji^OE：p=0.036），促进了促凋亡蛋白Bax的表达（OCI-LY3^OE：p=0.047；Raji^OE：p=0.029）。

3. 通过Transwell法证实，CHN1过表达抑制了OCI-LY3细胞和Raji细胞的迁移（OCI-LY3，p=0.004；Raji，p=0.008）和侵袭（OCI-LY3，p=0.012；Raji，p=0.007）。Rac1是调节细胞迁移和侵袭能力的关键蛋白，CHN1过表达抑制了Rac1活性。

4. 转录组测序结果表明，CHN1过表达的OCI-LY3细胞中，共有867个基因表达上调，1035个基因表达下调。GO富集分析、KEGG富集分析及GSEA富集分析结果表明，CHN1过表达可能会影响细胞黏附、细胞代谢、细胞囊泡形成、细胞连接等多个生物学过程，并可能影响PI3K/Akt信号通路、IL6/JAK/STAT3通路、IL2/STAT5信号通路等多条信号通路。

5. Western Blotting结果证实，CHN1抑制了Akt和mTOR的磷酸化。使用Akt抑制剂Sc79处理CHN1过表达的OCI-LY3细胞和Raji细胞，Akt和mTOR的磷酸化水平部分恢复，并部分恢复细胞的增殖能力、迁移能力和侵袭能力。

结论

1. CHN1与DLBCL患者的预后相关，CHN1高表达患者具有更高的OR率和CR率，并有更长的OS和PFS；

2. CHN1通过抑制B-NHL细胞的增殖、迁移和侵袭，从而抑制疾病进展；

3. Akt/mTOR信号通路可能是CHN1在B-NHL细胞中的下游通路之一。

第二部分 简化老年评估在中国DLBCL患者中的验证和优化

背景

弥漫大B细胞淋巴瘤（diffuse large B-cell lymphoma, DLBCL）是最常见的恶性淋巴瘤，以老年人发病为主，中位发病年龄为66岁。老年患者由于高龄、合并症等因素，对标准免疫化疗方案耐受性差，预后不良，难以通过标准治疗方案获益。老年综合评估结合了年龄、体能状况、合并症等多个指标，可以对老年DLBCL患者进行危险分层，评估患者预后并指导治疗方案选择。近年来，意大利淋巴瘤基金会构建了简化老年评估（simplified Geriatric Assessment, sGA）用于老年DLBCL患者的预后评估，并在此基础上，将sGA与国际预后指数（International Prognostic Index, IPI）评分和血红蛋白水平结合，构建了老年DLBCL患者特异性的预后评估工具老年预后指数（Elderly Prognostic Index, EPI）。EPI与老年DLBCL患者的总生存期（overall survival, OS）、无进展生存期（progression-free survival, PFS）和早期死亡相关，但该结果尚未在中国患者群体中得到验证。因此，本研究旨在验证sGA与EPI在中国老年DLBCL患者中的预后价值，探索它们在治疗分层方面的应用价值，并对它们进行可能的优化。

方法

1. 老年DLBCL患者临床资料收集与预后随访；

2. 依据sGA与EPI对老年DLBCL患者进行分组；

3. 统计学分析。

结果

1. 根据sGA的分类标准，适合组、不适合组和脆弱组患者分别占45.1%、35.4%和19.5%。sGA分类与患者年龄（p<0.001）、东部肿瘤协作组体能状况评分（Eastern Cooperative Oncology Group Performance Status, ECOG PS;  p<0.001）、国际预后指数（International Prognostic Index, IPI; p=0.002）、B症状（p=0.002）、血红蛋白水平（p<0.001）和白蛋白水平（p<0.001）相关。依据EPI的分类标准，低危组、中危组、高危组分别占15.3%、40.0%和44.7%。EPI分类与患者年龄（p<0.001）、Ann Arbor分期（p<0.001）、结外受累部位（p<0.001）、ECOG PS（p<0.001）、LDH水平（p<0.001）、IPI（p<0.001）、B症状（p<0.001）、血红蛋白水平（p<0.001）和白蛋白水平（p<0.001）相关。

2. sGA、EPI与老年DLBCL患者的治疗反应相关。卡方检验结果表明，采用sGA分类时，适合组、不适合组和脆弱组患者的客观缓解率（objective response rate, ORR）分别为85.3%、73.4%和75.0%（p=0.104），完全缓解率（complete response rate, CRR）分别为69.7%、59.5%和45.0%（χ²=7.863，p=0.020）。采用EPI分类时，低危组、中危组和高危组患者的ORR分别为88.9%、84.0%和70.8%（χ²=7.470，p=0.024），CRR分别为83.3%、64.9%和51.0%（χ²=12.174，p=0.002）。

3. sGA、EPI与老年DLBCL患者的OS、PFS和早期死亡相关。Kaplan-Meier生存曲线分析表明，sGA与老年DLBCL患者OS（χ²=7.470，p<0.001）和PFS（χ²=8.358，p=0.015）相关，适合组、不适合组、脆弱组患者5年OS分别为66.5%、54.9%和44.7%，5年PFS分别为62.6%、39.5%和43.5%。EPI与老年DLBCL患者OS（χ²=11.085，p=0.004）和PFS（χ²=8.565，p=0.014）相关。低危组、中危组和高危组5年OS分别为78.4%、61.6%和47.2%，5年PFS分别为69.5%、49.0%和44.1%。另外，卡方检验结果证实，sGA（χ²=12.365，p=0.001）、EPI（χ²=7.626，p=0.016）还与老年DLBCL患者早期死亡相关。

4. 单因素与多因素Cox回归分析表明，白蛋白是中国老年DLBCL患者OS的独立预后因素（p=0.006）。将sGA与白蛋白水平结合构建新的老年DLBCL预后评估模型sGA-A。Kaplan-Meier生存曲线表明，sGA-A与患者的OS（p<0.001）和PFS（p<0.001）相关。卡方检验证实，sGA-A与患者的早期死亡相关（χ²=7.626，p<0.001）。使用ROC曲线比较sGA-A与sGA、EPI预测老年DLBCL患者预后的能力。sGA-A能够预测患者的OS（AUC：0.647；95%CI：0.581-0.714；p<0.001）、PFS （AUC：0.577；95%CI：0.506-0.648；p=0.046）和早期死亡（AUC：0.769；95%CI：0.691-0.848；p<0.001），其预测患者OS能力与sGA（p=0.346）和EPI（p=0.232）相当，预测患者早期死亡能力与sGA相当（p=0.520）而优于EPI（p=0.027）。值得注意的是，sGA和EPI均无法预测患者PFS。

结论

1. sGA与EPI与中国老年DLBCL患者的治疗反应、OS、PFS和早期死亡相关。

2. 将sGA与白蛋白水平结合构建新的老年DLBCL患者的预后评估模型sGA-A。sGA-A能够用于预测中国老年DLBCL患者的OS和早期死亡，并弥补了sGA与EPI无法预测老年DLBCL患者PFS的不足。

Part I Prognostic Value, Phenotypic Study, and Mechanistic Exploration of CHN1 in B-cell Non-Hodgkin Lymphoma

Background

B-cell non-Hodgkin lymphoma (B-NHL), especially aggressive B-NHL, is the most common subtype of non-Hodgkin lymphoma and exhibits high heterogeneity. Therefore, evaluating the prognosis of B-NHL patients, particularly those with aggressive subtypes, and exploring its potential pathogenesis are urgent issues to be addressed. α-chimaerin (CHN1) is a specific GTPase-activating protein for Rac1, a member of the Rho small GTPase family. Previous studies have suggested that CHN1 is associated with the aggressiveness of B-NHL, and bioinformatics analysis has identified CHN1 as a favorable prognostic biomarker in diffuse large B-cell lymphoma (DLBCL). However, these findings have not been validated in clinical patients, and its specific molecular mechanisms remain unclear. This study aimed to validate the correlation between CHN1 expression and prognosis in DLBCL patients through retrospective analysis of pathological specimens and to explore the potential mechanisms of CHN1 in B-NHL using in vitro experiments.

Methods

1. Collection of DLBCL pathological specimens and immunohistochemical staining of CHN1.

2. Collection of clinical characteristics and prognostic data of DLBCL patients.

3. Construction of CHN1-stably overexpressing B-NHL cell lines (Raji and OCI-LY3) using lentiviral infection.

4. Detection of cell proliferation by direct counting.

5. Detection of cell migration and invasion using Transwell assays.

6. Measurement of apoptosis by Annexin V/7-AAD double staining.

7. Analysis of protein expression levels via Western blotting.

8. Detection of Rac1 protein activity using Pull-Down assays.

9. Transcriptome sequencing of CHN1-overexpressing DLBCL cells (OCI-LY3^OE), followed by GO, KEGG, and GSEA enrichment analyses to identify downstream pathways.

10. Statistical analysis.

Results

1. Based on CHN1 immunohistochemical staining in DLBCL specimens (cut-off value: 10% positivity), patients were divided into CHN1-high and CHN1-low groups. The CHN1-high group showed significantly higher objective response (OR: 86.2% vs. 39.1%, p=0.001) and complete response (CR: 65.5% vs. 21.7%, p=0.002) rates compared to the CHN1-low group. The disease progression (PD) rate was significantly lower in the CHN1-high group (10.3% vs. 52.2%, p=0.001). Kaplan-Meier survival analysis revealed that CHN1 expression was significantly associated with overall survival (OS, p=0.036) and progression-free survival (PFS, p=0.002).

2. CHN1-stably overexpressing Burkitt lymphoma (Raji^OE) and DLBCL (OCI-LY3^OE) cell lines were successfully established. CHN1 overexpression significantly inhibited proliferation (p<0.001) and promoted apoptosis in both Raji (p<0.001) and OCI-LY3 cells (p=0.002). Western blotting demonstrated that CHN1 downregulated the anti-apoptotic protein BCL-2 (OCI-LY3^OE: p=0.029; Raji^OE: p=0.036) and upregulated the pro-apoptotic protein Bax (OCI-LY3^OE: p=0.047; Raji^OE: p=0.029).

3. Transwell assays confirmed that CHN1 overexpression suppressed migration (OCI-LY3: p=0.004; Raji: p=0.008) and invasion (OCI-LY3: p=0.012; Raji: p=0.007) in both cell lines. CHN1 overexpression also inhibited Rac1 activity.

4. Transcriptome sequencing of CHN1-overexpressing OCI-LY3 cells identified 867 upregulated and 1,035 downregulated genes. Enrichment analyses (GO, KEGG, GSEA) suggested that CHN1 may regulate biological processes such as cell adhesion, metabolism, vesicle formation, and cell junctions, and modulate pathways including PI3K/Akt, IL6/JAK/STAT3, and IL2/STAT5 signaling.

5. Western blotting revealed that CHN1 suppressed phosphorylation of Akt and mTOR. Treatment with the Akt inhibitor Sc79 partially restored Akt/mTOR phosphorylation and rescued the proliferative, migratory, and invasive capacities of CHN1-overexpressing cells.

Conclusions

1. CHN1 is associated with prognosis in DLBCL patients, with high expression correlating with higher OR and CR rates, as well as prolonged OS and PFS.

2. CHN1 inhibits disease progression by suppressing proliferation, migration, and invasion in B-NHL cells.

3. The Akt/mTOR signaling pathway is a downstream mechanism of CHN1 in B-NHL.

Part II The Validation and Modification of Simplified Geriatric Assessment in Chinese Older Diffuse Large B-cell Lymphoma

Background

Diffuse large B-cell lymphoma (DLBCL) is the most common type of malignant lymphoma, primarily affecting the elderly with a median onset age of 66. Elderly patients, due to advanced age and comorbidities, often have poor tolerance to standard immunochemotherapy regimens, leading to adverse prognoses and limited benefits from standard treatments. Geriatric assessment, which integrates age, performance status, and comorbidities, facilitates risk stratification and guides treatment decisions. The Italian Lymphoma Foundation developed the Simplified Geriatric Assessment (sGA) for prognostic evaluation in elderly DLBCL patients. Building on this, the sGA was combined with the International Prognostic Index (IPI) and hemoglobin levels to create the Elderly Prognostic Index (EPI), which correlates with overall survival (OS), progression-free survival (PFS), and early mortality in elderly DLBCL patients. However, its applicability in Chinese patient populations remains unvalidated. This study aims to evaluate the prognostic value of sGA and EPI in Chinese elderly DLBCL patients, explore their role in treatment stratification, and optimize these tools.

Methods

1. Collection of clinical data and prognostic follow-up for elderly DLBCL patients.

2. Stratification of patients based on sGA and EPI.

3. Statistical analysis.

Results

1. According to sGA criteria, patients were classified into fit (45.1%), unfit (35.4%), and frail (19.5%) groups. sGA categories correlated with age (p<0.001), ECOG PS (p<0.001), IPI (p=0.002), B symptoms (p=0.002), hemoglobin (p<0.001), and albumin levels (p<0.001). Using EPI criteria, patients were divided into low-risk (15.3%), intermediate-risk (40.0%), and high-risk (44.7%) groups. EPI categories correlated with age (p<0.001), Ann Arbor stage (p<0.001), extranodal involvement (p<0.001), ECOG PS (p<0.001), LDH (p<0.001), IPI (p<0.001), B symptoms (p<0.001), hemoglobin (p<0.001), and albumin levels (p<0.001).

2. Both sGA and EPI were associated with treatment responses. Chi-square tests showed that under sGA classification, objective response rates (ORR) were 85.3%, 73.4%, and 75.0% for fit, unfit, and frail groups (p=0.104), while complete response rates (CRR) were 69.7%, 59.5%, and 45.0% (χ²=7.863, p=0.020). Under EPI classification, ORR were 88.9%, 84.0%, and 70.8% for low-, intermediate-, and high-risk groups (χ²=7.470, p=0.024), with CRR of 83.3%, 64.9%, and 51.0% (χ²=12.174, p=0.002).

3. sGA and EPI were linked to OS, PFS, and early mortality. Kaplan-Meier analysis indicated that sGA correlated with OS (χ²=7.470, p<0.001) and PFS (χ²=8.358, p=0.015). 5-year OS rates were 66.5%, 54.9%, and 44.7% for fit, unfit, and frail groups; 5-year PFS rates were 62.6%, 39.5%, and 43.5%. EPI also correlated with OS (χ²=11.085, p=0.004) and PFS (χ²=8.565, p=0.014). 5-year OS rates were 78.4%, 61.6%, and 47.2% for low-, intermediate-, and high-risk groups; 5-year PFS rates were 69.5%, 49.0%, and 44.1%. Chi-square tests confirmed associations of sGA (χ²=12.365, p=0.001) and EPI (χ²=7.626, p=0.016) with early mortality.

4. Univariate and multivariate Cox regression analyses identified albumin as an independent prognostic factor for OS (p=0.006) in Chinese elderly DLBCL patients. A new prognostic model, sGA-A, was developed by combining sGA with albumin levels. Kaplan-Meier curves showed sGA-A correlated with OS (p<0.001) and PFS (p<0.001). Chi-square tests confirmed sGA-A's association with early mortality (χ²=7.626, p<0.001). ROC curve analysis demonstrated sGA-A's ability to predict OS (AUC: 0.647; 95%CI: 0.581-0.714; p<0.001), PFS (AUC: 0.577; 95%CI: 0.506-0.648; p=0.046), and early mortality (AUC: 0.769; 95%CI: 0.691-0.848; p<0.001). sGA-A matched sGA and EPI in predicting OS and outperformed EPI in predicting early mortality (p=0.027). Notably, neither sGA nor EPI predicted PFS.

Conclusions

1. sGA and EPI are associated with treatment responses, OS, PFS, and early mortality in Chinese elderly DLBCL patients.

2. The novel prognostic model sGA-A, combining sGA with albumin levels, predicts OS and early mortality in Chinese elderly DLBCL patients and addresses the limitation of sGA and EPI in predicting PFS.