Pharmacogenomic modeling in pancreatic cancer—response.
Pharmacogenomic modeling in pancreatic cancer—response.
复制标题
胰腺癌反应的药物基因组学模型。
DOI：
10.1158/1078-0432.ccr-15-0058
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Yu,KennethH
中科院分区：
文献类型：
--
作者：
Yu,KennethH
我们感谢Avan和他的同事(1)对我们的循环肿瘤和侵袭细胞(CTIC)的药物基因组学(PGx)研究的兴趣，以预测胰腺导管腺癌(PDAC)的有效治疗。2)和他们深思熟虑的评论。关于第一点，正如最初的文章(2)中所明确的那样，这是一项验证临床前模型的先导性研究。我们明确指出，正在进行的和计划中的研究正在进行中，以进一步验证所描述的结果，特别是将PGx模型应用于CTICs可以用于预测PDAC患者的有效药物治疗。招募更多的接受不同治疗的患者，特别是新药，如纳布紫杉醇，对于测试我们的PGx方法的有效性至关重要。我们还计划进行一项研究，利用我们的PGx模型所做的预测来指导治疗。关于第二点，由于Avan和他的同事所描述的对过度拟合的担忧(1)，没有进行多变量分析。我们确实单独研究了一些变量来确定失衡，正如最初的文章所描述的那样，通过单变量分析唯一具有统计学意义的变量是年龄。为了进一步探讨这一点，我们现在进行了年龄调整的无进展生存期(PFS)和总生存期(OS)的分析，即使考虑到年龄，敏感组和耐药组的PFS和OS差异仍然显著。正如最初的文章(2)所述，负责计算PFS和OS的研究人员对PGx预测视而不见，反之亦然。Avan和他的同事提出的最后一点与我们开发原始模型所使用的NCI60细胞系方法有关。我们认为，使用跨越多种肿瘤类型的细胞系实际上是我们方法的一个优点。药物治疗的未来将不仅仅是根据疾病的起源部位来治疗疾病，而是基于遗传易感性。与对奥沙利铂耐药的胰腺肿瘤相比，对奥沙利铂敏感的胰腺肿瘤可能与奥沙利铂敏感的结肠癌有更多的遗传相似性，使用NCI60细胞系资源可以捕捉到这一点。此外，正如在最初的文章中详细描述的那样，在进行预期的临床试验之前，我们实际上在一些患者来源的小鼠异种移植中验证了我们的方法。我们期待着在不久的将来报告我们正在进行的对这种有前景的临床工具的研究结果。
We thank Avan and colleagues (1) for their interest in our pharmacogenomic (PGx) study of circulating tumor and invasive cells (CTIC) for predicting effective therapy in pancreatic ductal adenocarcinoma (PDAC; ref. 2) and for their thoughtful comments. Regarding the first point, as is made clear in the original article (2), this is a pilot study validating a preclinical model. We make clear that ongoing and planned studies are under way to further validate the findings described, specifically that applying a PGx model to CTICs can be used to predict effective drug therapies for patients with PDAC. Enrolling a larger number of patients receiving diverse treatments, specifically newer drugs such as nabpaclitaxel, is critical to testing the validity of our PGx approach. We also plan a study that will utilize the prediction made by our PGx model to guide therapy. Regarding the second point, because of the concern for overfitting described by Avan and colleagues (1), a multivariable analysis was not performed. We did look at a number of variables individually to identify imbalances, and, as described in the original article, the only variable that was statistically significant by univariate analysis was age. To further explore this, we have now performed an age adjusted analysis of progressionfree survival (PFS) and overall survival (OS), and even accounting for age, the PFS and OS differences seen in the sensitive and resistant groups remain significant. As stated in the original article (2), the research personnel responsible for calculating PFS and OS were blinded to the PGx prediction, and vice versa. The final point made by Avan and colleagues relates to the NCI60 cell line approach used to develop our original models. We feel that using cell lines across a wide variety of tumor types is actually a strength of our approach. The future of drug therapy will not be treating diseases based solely on their site of origin, but rather, based on the genetic susceptibilities. Pancreatic tumors with susceptibility to oxaliplatin may have more genetic similarities to oxaliplatin-sensitive colon tumors, compared with oxaliplatin-resistant pancreatic tumors, and using the NCI60 cell line resource can capture this. Furthermore, as described in detail in the original article, we did in fact go on to validate our approach in a number of patient-derived mouse xenografts before conducting our prospective clinical trial. We look forward to reporting results from our ongoing studies of this promising clinical tool in the very near future.