To identify variables that could be used in differentiating pathologic responders from nonresponders, logistic regression analysis was conducted

To identify variables that could be used in differentiating pathologic responders from nonresponders, logistic regression analysis was conducted. analysis. Using multivariate analysis, kurtosis was found to Cortisone acetate be impartial predictor. In TKI group, intensity variability and size-zone variability were significantly decreased in pathologic responder group. Intensity variability was found to be an independent predictor for pathologic response on multivariate analysis. Conclusions Quantitative CT variables including histogram or texture analysis have potential as a predictive tool for response evaluation, and it may better reflect treatment response than standard response criteria based on size changes. Introduction Locally advanced non-small cell lung malignancy (NSCLC) has a dismal prognosis with a median overall survival (OS) of 25C35 months despite multimodal treatment including radiation therapy (RT), chemotherapy and surgery [1], [2]. Induction concurrent chemoradiation therapy (CCRT) is known to result in short-term gross tumor volume reduction, with aggressive locoregional control. Previous studies demonstrated variable responses with volume reduction averaging 38% to 73% [3], with improved survival compared with the treatment of Cortisone acetate surgery alone [4]. This may be explained by the potent local control effect of irradiation. Therefore, there is a need for studies directed toward predicting treatment benefit versus risk of treatment failure. Clinically, such predictors would allow further individualization of treatment during radiotherapy [5]. On the other hand, during the last decade, several molecular-targeted agentsfor example, epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) such as erlotinib and gefitinibhave emerged for treatment of NSCLC [6], as well as in the neoadjuvant setting, which has also shown to be effective in a subset of patients with NSCLC [7]C[10]. Furthermore, although TKI brokers are most active in patients with an EGFR mutation, patients without documented mutation still showed survival benefit compared with placebo [9]. With the advancement in imaging techniques and their increasing application to oncology practice, imaging-based tumor volume regression rate evaluated at mid-RT has been shown to predict local control rate and disease-free survival (DFS) after RT or CCRT [11], [12]. In addition, quantitative measurement of tumor regression rate becomes more realistic with the use of imaging particularly during therapy when the morphologic changes remain delicate and hard to assess by clinical examination [12], [13]. Evaluation of treatment response to TKI brokers is also challenging. It is well established that the conventional RECIST underestimates response rates than the proportion of patients who Cortisone acetate actually experience clinically effective disease control [7], [10], [14]C[20]. Since TKI brokers aim for inhibition of tumor cell growth, but not necessarily tumor cell death, tumor response may not emerge as early decrease in tumor size [19]. Very recently, histogram analysis or texture analysis is receiving attention as a method for quantifying tumor heterogeneity and evaluating treatment response [14], [21], [22]. Here, the question remains which is more effective to predict treatment response in the various imaging-based quantitative assessment methods, and whether or not each method functions differently depending on therapeutic regimen. Given the need for clinical validation of any updated analysis tool, our main objectives were to identify differences in serial changes of various CT-based parameters in patients with lung adenocarcinoma scheduled to undergo surgical resection after neoadjuvant therapy, and to correlate those changes with pathologic responses, focused on their relationship with different neoadjuvant therapeutic options, particularly of EGFR-TKI and concurrent chemoradiation therapy (CCRT) settings. Materials and Methods The institutional review table of Samsung Medical Center (SMC IRB) approved this retrospective study with a waiver of informed consent. Patients From September 2005 through December 2011, 398 patients with stage IIIA NSCLC underwent curative surgical resection of lung malignancy at our institution, after neoadjuvant treatment (chemotherapy, radiation therapy, or both). Among these patients, patients with pathologically confirmed adenocarcinoma were only included, while other histologic subtypes (such as Rabbit polyclonal to IGF1R.InsR a receptor tyrosine kinase that binds insulin and key mediator of the metabolic effects of insulin.Binding to insulin stimulates association of the receptor with downstream mediators including IRS1 and phosphatidylinositol 3′-kinase (PI3K). squamous cell carcinoma, large cell carcinoma, small cell lung malignancy, neuroendocrine malignancy, etc.) were excluded. We subdivided the patients into three groups, depending on different neoadjuvant treatment options: chemotherapy with TKI brokers, chemotherapy with standard brokers, and CCRT. Since our interest lay in comparing the imaging parameter changes for treatment response prediction between novel TKI and CCRT as a neoadjuvant option, patients who underwent neoadjuvant chemotherapy with standard agents were excluded. As a result, two groups of patients were enrolled in our study: patients who underwent neoadjuvant chemotherapy with TKI brokers, and those who underwent CCRT. Patients for neoadjuvant TKI agent were put together from a selected population fulfilling two of the following features: female, adenocarcinoma, nonsmoker, and Asian. Regimen of TKI group comprised 1 tablet of 150 mg of erlotinib daily for 3 weeks. Neoadjuvant CCRT included chemotherapy and concurrent thoracic radiotherapy (TRT). The chemotherapy regimen was weekly paclitaxel (50 mg/m2 per week IV) plus.