Green Tina M, Jensen Andreas K, Holst René, Falgreen Steffen, Bøgsted Martin, de Stricker Karin, Plesner Torben, Mourits-Andersen Torben, Frederiksen Mikael, Johnsen Hans E, Pedersen Lars M, Møller Michael B
Department of Pathology, Odense University Hospital, Odense, Denmark.
Institute of Clinical Research, University of Southern Denmark, Odense, Denmark.
Br J Haematol. 2016 Sep;174(6):876-86. doi: 10.1111/bjh.14138. Epub 2016 May 16.
We present a multiplex analysis for genes known to have prognostic value in an attempt to design a clinically useful classification model in patients with diffuse large B-cell lymphoma (DLBCL). Real-time polymerase chain reaction was used to measure transcript levels of 28 relevant genes in 194 de novo DLBCL patients treated with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone). Including International Prognostic Index (IPI) as a variable in a penalized Cox regression, we investigated the association with disease progression for single genes or gene combinations in four models. The best model was validated in data from an online available R-CHOP treated cohort. With progression-free survival (PFS) as primary endpoint, the best performing IPI independent model incorporated the LMO2 and HLADQA1 as well as gene interactions for GCSAMxMIB1, GCSAMxCTGF and FOXP1xPDE4B. This model assigned 33% of patients (n = 60) to poor outcome with an estimated 3-year PFS of 40% vs. 87% for low risk (n = 61) and intermediate (n = 60) risk groups (P < 0·001). However, a simpler, IPI independent model incorporated LMO2 and BCL2 and assigned 33% of the patients with a 3-year PFS of 35% vs. 82% for low risk group (P < 0·001). We have documented the impact of a few single genes added to IPI for assignment in new drug trials.
我们对已知具有预后价值的基因进行了多重分析，试图为弥漫性大B细胞淋巴瘤（DLBCL）患者设计一种具有临床实用性的分类模型。采用实时聚合酶链反应来测量194例接受R-CHOP（利妥昔单抗、环磷酰胺、阿霉素、长春新碱、泼尼松）治疗的初治DLBCL患者中28个相关基因的转录水平。将国际预后指数（IPI）作为一个变量纳入惩罚性Cox回归，我们在四个模型中研究了单个基因或基因组合与疾病进展的关联。最佳模型在一个在线可用的接受R-CHOP治疗队列的数据中得到验证。以无进展生存期（PFS）作为主要终点，表现最佳的独立于IPI的模型纳入了LMO2和HLADQA1以及GCSAMxMIB1、GCSAMxCTGF和FOXP1xPDE4B的基因相互作用。该模型将33%的患者（n = 60）归为预后不良，估计3年PFS为40%，而低风险组（n = 61）和中风险组（n = 60）为87%（P < 0·001）。然而，一个更简单的、独立于IPI的模型纳入了LMO2和BCL2，将33%的患者归为该组，其3年PFS为35%，而低风险组为82%（P < 0·001）。我们已经记录了在新的药物试验中，在IPI基础上增加一些单个基因对分类的影响。
