Objective: To compare the prognostic value of two predictive models based on C-reactive protein (CRP) and albumin (ALB), namely the CRP to ALB ratio (CAR) and the Glasgow prognostic score (GPS), in newly diagnosed patients with diffuse large B-cell lymphoma (DLBCL).
Methods: The data of newly diagnosed DLBCL patients admitted to our center from May 2014 to January 2022 were reviewed. A total of 111 patients who completed at least 4 cycles of R-CHOP or R-CHOP-like chemotherapy with detailed clinical, laboratory data and follow-up information were included. The receiver operating characteristic (ROC) curve was performed to evaluate the predictive value of pre-treatment CAR on disease progression and survival. Furthermore, the association between CAR and baseline clinical, laboratory characteristics of patients was evaluated, and progression-free survival (PFS) and overall survival (OS) were compared between different CAR and GPS subgroups. Finally, the univariate and multivariate COX propor-tional hazard regression models were used to analyze the factors affecting disease outcomes.
Results: ROC curve showed that the area under the curve (AUC) of CAR predicting PFS and OS in DLBCL patients was 0.687 (P =0.002) and 0.695 (P =0.005), respectively, with the optimal cut-off value of 0.11 for both predicting PFS and OS. Compared with the lower CAR (<0.11) group, the higher CAR (≥0.11) group had more clinical risk factors, including age >60 years (P =0.025), ECOG score ≥2 (P =0.004), Lugano stage III-IV (P < 0.001), non-germinal center B-cell-like (non-GCB) subtype (P =0.035), elevated lactate dehydrogenase (LDH) ( P < 0.001), extranodal involved site >1 (P =0.004) and IPI score >2 (P < 0.001). The interim response evaluation of patients showed that the overall response rate (ORR) and complete response rate (CRR) in the lower CAR group were both significantly better than those in the higher CAR group (ORR: 96.9% vs 80.0%, P =0.035; CRR: 63.6% vs 32.5%, P =0.008). With a median follow-up of 24 months, patients with lower CAR had significantly longer median PFS and OS than those with higher CAR (median PFS: not reached vs 67 months, P =0.0026; median OS: not reached vs 67 months, P =0.002), while there was no statistical difference in PFS (P =0.11) and OS (P =0.11) in patients with GPS of 0, 1, and 2. Multivariate Cox regression analysis indicated that only sex (male) and IPI score >2 were independent risk factors for both PFS and OS.
Conclusion: CAR is significantly correlated with disease progression and survival in DLBCL patients; And compared with GPS, CAR has more advantages in predicting disease outcomes in DLBCL patients.
题目: 比较C反应蛋白与白蛋白比值和格拉斯哥预后评分对弥漫大B细胞淋巴瘤患者预后判断的价值.
目的: 探讨并比较基于C反应蛋白（CRP）和白蛋白（ALB）的两种预测模型——CRP与ALB比值（CAR）和格拉斯哥预后评分（GPS）——在新诊断弥漫大B细胞淋巴瘤（DLBCL）患者预后评估中的价值。.
方法: 回顾性分析本院2014年5月至2022年1月收治的初诊DLBCL患者的资料，纳入至少完成4个周期R-CHOP或R-CHOP样方案化疗，且临床、实验室检查数据以及随访资料均完整的111例患者。根据患者治疗前CAR及随访截止时的生存状态绘制受试者工作特征曲线（ROC），初步判断CAR对疾病进展和生存结局的预测价值，进一步分析CAR与患者基线临床、实验室特征的关系，并比较不同CAR、GPS分组患者的无进展生存期（PFS）和总生存期（OS），采用单因素和多因素COX风险比例回归模型分析影响疾病预后的因素。.
结果: ROC曲线分析结果显示，CAR预测DLBCL患者PFS、OS的曲线下面积（AUC）分别为0.687（P =0.002），0.695（P =0.005），最佳截断值均为0.11；相较于低CAR（<0.11）组，高CAR（≥0.11）组患者具有更多的临床高危因素，包括年龄>60岁（P =0.025）、ECOG评分≥2分（P =0.004）、Lugano分期III-IV期（ P < 0.001）、non-GCB亚型（P =0.035）、乳酸脱氢酶升高（ P < 0.001）、结外病变数>1处（P =0.004）及IPI评分>2分（ P < 0.001）。对患者进行中期疗效评估，低CAR组患者的总有效率（ORR）及完全缓解率（CRR）均明显优于高CAR组（ORR: 96.9% vs 80.0%, P =0.035；CRR: 63.6% vs 32.5%, P =0.008）。中位随访24个月，低CAR组患者的中位PFS及OS均明显优于高CAR组（中位PFS：未达到 vs 67个月，P =0.0026；中位OS：未达到 vs 67个月，P ＝0.002），而GPS 0分、1分和2分3组患者的PFS（P =0.11）和OS（P =0.11）差异均不具有统计学意义。多因素分析显示，仅有性别为男性和IPI评分>2分是影响患者PFS和OS的独立危险因素。.
结论: CAR与DLBCL患者的疾病进展和生存显著相关；与GPS相比，CAR对患者的预后判断价值更大。.
Keywords: diffuse large B-cell lymphoma; C-reactive protein to albumin ratio; Glasgow prognostic score; prognosis.