Chao Chun, Silverberg Michael J, Chen Lie-Hong, Xu Lanfang, Martínez-Maza Otoniel, Abrams Donald I, Zha Hongbin D, Haque Reina, Said Jonathan
a Department of Research and Evaluation , Kaiser Permanente Southern California , Pasadena , CA , USA.
b Division of Research , Kaiser Permanente Northern California , Oakland , CA , USA.
Leuk Lymphoma. 2018 Feb;59(2):321-329. doi: 10.1080/10428194.2017.1334121. Epub 2017 Jun 13.
Existing prognostic tools for HIV + diffuse large B-cell lymphoma (DLBCL) fail to accurately predict patient outcomes. To develop a novel prognostic algorithm incorporating molecular tumor characteristics and HIV disease factors, we included 80 patients with HIV-related DLBCL diagnosed between 1996 and 2007. Immunohistochemistry staining was used to analyze the expression of 26 tumor markers. Clinical data were collected from medical records. Logistic regression and bootstrapping were used to select and assess stability of the prognostic model, respectively. Of the tumor markers examined, expression of cMYC, Ki 67, CD44, EBV, SKP2, BCL6, p53, CD20 and IgM were associated with two-year mortality. The final prognostic model, confirmed in bootstrapped samples, included IPI, circulating CD4 cell count, history of clinical AIDS, and expression of CD44, p53, IgM and EBV. This model incorporating HIV disease history and tumor markers, achieved better prediction for two-year mortality [AUC = 0.87, 95% CI: 0.78-0.96] compared with IPI alone [AUC = 0.63 (0.51-0.75)].
现有的针对HIV阳性弥漫性大B细胞淋巴瘤（DLBCL）的预后评估工具无法准确预测患者的预后情况。为了开发一种纳入分子肿瘤特征和HIV疾病因素的新型预后算法，我们纳入了80例在1996年至2007年间被诊断为HIV相关DLBCL的患者。采用免疫组织化学染色分析26种肿瘤标志物的表达情况。从病历中收集临床数据。分别使用逻辑回归和自抽样法来选择和评估预后模型的稳定性。在所检测的肿瘤标志物中，cMYC、Ki 67、CD44、EBV、SKP2、BCL6、p53、CD20和IgM的表达与两年死亡率相关。在自抽样样本中得到验证的最终预后模型包括国际预后指数（IPI）、循环CD4细胞计数、临床艾滋病病史以及CD44、p53、IgM和EBV的表达。与单独使用IPI相比，这个纳入了HIV疾病史和肿瘤标志物的模型对两年死亡率的预测效果更好 [曲线下面积（AUC）=0.87，95%可信区间（CI）：0.78 - 0.96]，而单独使用IPI时AUC为0.63（0.51 - 0.75）。
