Prediction of survival in multiple myeloma based on gene expression profiles reveals cell cycle and chromosomal instability signatures in high-risk patients and hyperdiploid signatures in low-risk patients:: A study of the intergroupe francophone du myelome
Prediction of survival in multiple myeloma based on gene expression profiles reveals cell cycle and chromosomal instability signatures in high-risk patients and hyperdiploid signatures in low-risk patients:: A study of the intergroupe francophone du myelome
复制标题
DOI：
10.1200/jco.2007.13.8545
复制
发表时间：
2008-10-10
影响因子：
45.3
通讯作者：
Minvielle, Stephane
中科院分区：
文献类型：
--
作者：
Decaux, Olivier;Lode, Laurence;Minvielle, Stephane
多发性骨髓瘤患者的生存期具有高度异质性，从几周到10年以上不等。我们使用诊断时获得的骨髓瘤细胞的基因表达谱来确定广泛适用的预后标志物。患者和方法在182例患者的训练集中，我们使用监督方法来识别与生存时间相关的个体基因。一个生存模型就是根据这些基因建立起来的。我们的模型的有效性在我们的68例患者的测试集和包括853例多发性骨髓瘤患者的三个独立队列中进行了评估。结果将与生存期相关的15个最强基因进行风险评分，并将患者分为低危组和高危组。在训练组和测试组以及外部验证组中，生存预测评分与生存显著相关。Kaplan-Meier估计3年生存率分别为90.5% （95% CI， 85.6%至95.3%）和47.4% (95% CI， 33.5%至60.1%)，在我们的低风险或高风险独立于传统预后因素的患者中。高风险患者是一个同质的生物实体，其特征是参与细胞周期进展及其监测的基因过表达，而低风险患者是异质的，表现出超二倍体特征。结论基于基因表达的高危患者生存预测和相关分子特征可能有助于开发预后标志物，并为开发新的靶向抗有丝分裂药物治疗高危患者提供依据。
Purpose Survival of patients with multiple myeloma is highly heterogeneous, from periods of a few weeks to more than 10 years. We used gene expression profiles of myeloma cells obtained at diagnosis to identify broadly applicable prognostic markers. Patients and Methods In a training set of 182 patients, we used supervised methods to identify individual genes associated with length of survival. A survival model was built from these genes. The validity of our model was assessed in our test set of 68 patients and in three independent cohorts comprising 853 patients with multiple myeloma. Results The 15 strongest genes associated with the length of survival were used to calculate a risk score and to stratify patients into low-risk and high-risk groups. The survival-predictor score was significantly associated with survival in both the training and test sets and in the external validation cohorts. The Kaplan-Meier estimates of rates of survival at 3 years were 90.5% (95% CI, 85.6% to 95.3%) and 47.4% (95% CI, 33.5% to 60.1%), respectively, in our patients having a low risk or high risk independently of traditional prognostic factors. High-risk patients constituted a homogeneous biologic entity characterized by the overexpression of genes involved in cell cycle progression and its surveillance, whereas low-risk patients were heterogeneous and displayed hyperdiploid signatures. Conclusion Gene expression-based survival prediction and molecular features associated with high-risk patients may be useful for developing prognostic markers and may provide basis to treat these patients with new targeted antimitotics.