Eunjin Lee

Synthetic lethal screening reveals FGFR as one of the combinatorial targets to overcome resistance to Met-targeted therapy

Met is a receptor tyrosine kinase that promotes cancer progression. In addition, Met has been implicated in resistance of tumors to various targeted therapies such as epidermal growth factor receptor inhibitors in lung cancers, and has been prioritized as a key molecular target for cancer therapy. However, the underlying mechanism of resistance to Met-targeting drugs is poorly understood. Here, we describe screening of 1310 genes to search for key regulators related to drug resistance to an anti-Met therapeutic antibody (SAIT301) by using a small interfering RNA-based synthetic lethal screening method. We found that knockdown of 69 genes in Met-amplified MKN45 cells sensitized the antitumor activity of SAIT301. Pathway analysis of these 69 genes implicated fibroblast growth factor receptor (FGFR) as a key regulator for antiproliferative effects of Met-targeting drugs. Inhibition of FGFR3 increased target cell apoptosis through the suppression of Bcl-xL expression, followed by reduced cancer cell growth in the presence of Met-targeting drugs. Treatment of cells with the FGFR inhibitors substantially restored the efficacy of SAIT301 in SAIT301-resistant cells and enhanced the efficacy in SAIT301-sensitive cells. In addition to FGFR3, integrin β3 is another potential target for combination treatment with SAIT301. Suppression of integrin β3 decreased AKT phosphorylation in SAIT301-resistant cells and restored SAIT301 responsiveness in HCC1954 cells, which are resistant to SAIT301. Gene expression analysis using CCLE database shows that cancer cells with high levels of FGFR and integrin β3 are resistant to crizotinib treatment, suggesting that FGFR and integrin β3 could be used as predictive markers for Met-targeted therapy and provide a potential therapeutic option to overcome acquired and innate resistance for the Met-targeting drugs.

Practical approach to determine sample size for building logistic prediction models using high-throughput data

An empirical method of sample size determination for building prediction models was proposed recently. Permutation method which is used in this procedure is a commonly used method to address the problem of overfitting during cross-validation while evaluating the performance of prediction models constructed from microarray data. But major drawback of such methods which include bootstrapping and full permutations is prohibitively high cost of computation required for calculating the sample size. In this paper, we propose that a single representative null distribution can be used instead of a full permutation by using both simulated and real data sets. During simulation, we have used a dataset with zero effect size and confirmed that the empirical type I error approaches to 0.05. Hence this method can be confidently applied to reduce overfitting problem during cross-validation. We have observed that pilot data set generated by random sampling from real data could be successfully used for sample size determination. We present our results using an experiment that was repeated for 300 times while producing results comparable to that of full permutation method. Since we eliminate full permutation, sample size estimation time is not a function of pilot data size. In our experiment we have observed that this process takes around 30min. With the increasing number of clinical studies, developing efficient sample size determination methods for building prediction models is critical. But empirical methods using bootstrap and permutation usually involve high computing costs. In this study, we propose a method that can reduce required computing time drastically by using representative null distribution of permutations. We use data from pilot experiments to apply this method for designing clinical studies efficiently for high throughput data.