Predicting pathologic response to neoadjuvant chemoradiation in resectable stage III non-small cell lung cancer patients using computed tomography radiomic features

Highlights

• Tumoral heterogeneity is associated with more aggressive tumor phenotype.

• Stage III N2 non-small cell lung cancer is a heterogeneous disease.

• Major pathologic response is associated with amended overall survival, following NAC.

• Compared to clinicopathologic variables, texture features are associated with MPR.

• Predictive features to MPR are also associated with survival and time to recurrence.

Abstract

Objective

The use of a neoadjuvant chemoradiation followed by surgery in patients with stage IIIA NSCLC is controversial and the benefit of surgery is limited. There are currently no clinically validated biomarkers to select patients for such an approach. In this study we evaluate computed tomography (CT) derived intratumoral and peritumoral texture and nodule shape features in their ability to predict major pathological response (MPR). MPR being defined as ≤10% of residual viable tumor, assessed at the time of surgery.

Material and methods

Ninety patients with stage III NSCLC treated with chemoradiation prior to surgical resection were selected. The patients were divided randomly into two equal sets, one for training and one for independent testing. The radiomic texture and shape features were extracted from within the nodule (intra) and from the parenchymal regions immediately surrounding the nodule (peritumoral). A univariate regression analysis was performed on the image and clinicopathologic variables and then included into a multivariable logistic regression (MLR) for binary outcome prediction of MPR. The radiomic signature risk-score was generated by using a multivariate Cox regression model and association of the signature with OS and DFS was also evaluated.

Results

Thirteen stable and predictive intratumoral and peritumoral radiomic texture features were found to be predictive of MPR. The MLR classifier yielded an AUC of 0.90 ± 0.025 within the training set and a corresponding AUC = 0.86 in prediction of MPR within the test set. The radiomic signature was also significantly associated with OS (HR = 11.18, 95% CI = 3.17, 44.1; p-value = 0.008) and DFS (HR = 2.78, 95% CI = 1.11, 4.12; p-value = 0.0042) in the testing set.

Conclusion

Texture features extracted within and around the lung tumor on CT images appears to be associated with the likelihood of MPR, OS and DFS to chemoradiation.

Introduction

The treatment of stage III NSCLC involves a multi-disciplinary approach and careful patient selection to determine which resectable patients might benefit from a trimodality treatment [1,2]. Neoadjuvant chemotherapy administered prior to surgery can reduce tumor extent and metastases, thereby improving resectability. The role of surgery in stage IIIA patients still remains controversial. Survival benefit of surgery in this setting has long been debated and has proven difficult to demonstrate in multi-institute trials compared to definitive chemoradiation [1,3]. Yet, ad hoc subgroup analyses have provided data to suggest that resection (lobectomy) confers a clear survival benefit in this setting [1]. In current clinical practice the selection of patients for trimodality therapy is largely based on relatively limited lymph node burden, single lobe involvement, and patient fitness for this aggressive approach with little attention paid to specific markers of response.

Several studies have shown that pathologic response to chemoradiation is highly predictive of disease free survival and overall survival [[4], [5], [6]]. The degree of down staging seems to correlate with survival with the greatest benefit associated with ≤10% residual tumor noted on resected specimens (defined as major pathologic response (MPR)) [5]. Unfortunately, there are no extant clinically validated and approved biomarkers to predict MPR to chemoradiation.

Use of radiomics, or computerized feature analysis of radiographic scans, to capture quantitative phenotypic attributes of the tumor has emerged as an important prospect for survival prediction [[7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19]]. These approaches have been used to capture and associate quantitative measurements of intratumoral heterogeneity and tumor shape from CT scans to predict response to first line chemotherapy and neoadjuvant chemoradiation in patients with NSCLC [7,20,21]. However, apart from radiomic assessment of intratumoral heterogeneity patterns, there is increasing evidence that heterogeneity patterns associated with the peritumoral region—the area immediately surrounding the tumor mass—might harbor valuable disease specific prognostic cues. For instance, signatures of immune response such as presence of peritumoral lymphocytes have been shown to be associated with disease specific survival [22]. Also, vascular invasion and neovascularization within the peritumoral region have been shown to be associated with an increased likelihood of tumor recurrence as well as reduced overall survival [23].

In this work we sought to explore whether textural patterns of the peritumoral space coupled with measurements of lesion shape and intratumoral texture patterns might identify which stage IIIA NSCLC patients are likely to have MPR after chemoradiation. The association of these radiomic features with overall survival and disease free survival was also evaluated. The intra- and peritumoral texture features were extracted from baseline CT images and only a limited number of features that found to be strongly associated with response were validated in the testing set. Additionally, the prognostic ability of the radiomic features against tumor shape and clinicopathologic variables was also evaluated.