Pretreatment¹⁸F-FDG PET Textural Features in Locally Advanced Non–Small Cell Lung Cancer: Secondary Analysis of ACRIN 6668/RTOG 0235

Abstract

In a secondary analysis of American College of Radiology Imaging Network (ACRIN) 6668/RTOG 0235, high pretreatment metabolic tumor volume (MTV) on 18F-FDG PET was found to be a poor prognostic factor for patients treated with chemoradiotherapy for locally advanced non–small cell lung cancer (NSCLC). Here we utilize the same dataset to explore whether heterogeneity metrics based on PET textural features can provide additional prognostic information. Methods Patients with locally advanced NSCLC underwent 18F-FDG PET prior to treatment. A gradient-based segmentation tool was used to contour each patient’s primary tumor. MTV, maximum SUV, and 43 textural features were extracted for each tumor. To address over-fitting and high collinearity among PET features, the least absolute shrinkage and selection operator (LASSO) method was applied to identify features that were independent predictors of overall survival (OS) after adjusting for MTV. Recursive binary partitioning in a conditional inference framework was utilized to identify optimal thresholds. Kaplan–Meier curves and log-rank testing were used to compare outcomes among patient groups. Results Two hundred one patients met inclusion criteria. The LASSO procedure identified 1 textural feature (SumMean) as an independent predictor of OS. The optimal cutpoint for MTV was 93.3 cm3, and the optimal Sum-Mean cutpoint for tumors above 93.3 cm3 was 0.018. This grouped patients into three categories: low tumor MTV (n = 155; median OS, 22.6 mo), high tumor MTV and high SumMean (n = 23; median OS, 20.0 mo), and high tumor MTV and low SumMean (n = 23; median OS, 6.2 mo; log-rank P < 0.001). Conclusion We have described an appropriate methodology to evaluate the prognostic value of textural PET features in the context of established prognostic factors. We have also identified a promising feature that may have prognostic value in locally advanced NSCLC patients with large tumors who are treated with chemoradiotherapy. Validation studies are warranted.

Characterization of PET/CT images using texture analysis: the past, the present… any future?

“…Although a number of studies have included only between 20 and 70 patients [50, 71, 86–92], some of the most recent studies have included between 80 and more than 200 patients: 88 patients with oropharyngeal squamous cell carcinoma [93], 103 with bone and soft tissue lesions [94], 101 with early-stage NSCLC [95], 112 with oesophageal cancer and 101 with NSCLC [60], 113 with glioma [36], 107 and 217 with oesophageal cancer [96, 97], 132 with lymph node involvement in lung cancer [98], 116, 195 and 201 with NSCLC [99–101], 137 with pancreatic lesions [102], and 188 lesions in lymphoma patients [103]. Some of the most recent studies have also used more robust statistical analysis, compared to these recently reviewed [28], several of them using a machine-learning method, e.g.…”

“…Some of the most recent studies have also used more robust statistical analysis, compared to these recently reviewed [28], several of them using a machine-learning method, e.g. neural networks [96], support vector machines [94, 98, 103] or the least absolute shrinkage and selection operator (LASSO) [95, 101]. The majority of these recent studies have concluded that TA can provide useful quantitative metrics regarding patient management (prognosis, response to therapy, distant metastasis prediction) in different cancer models except one that showed more mixed results [104], whereas another concluded that the improvement, although significant, may not be sufficient to have a clinical impact [97].…”

Machine and deep learning methods for radiomics

“…Aerts et al built a radiomic signature consisting of a combination of four features in a retrospective lung cancer cohort, which was predictive for survival in head and neck and NSCLC independent cohorts. One textural feature calculated from GLCM, SumMean, was identified using the LASSO procedure as an independent predictor of overall survival that complements metabolic tumor volume (MTV) in decision tree . A radiomic signature was built from PET‐CT for survival after SBRT for lung cancer .…”

A multiobjective Bayesian networks approach for joint prediction of tumor local control and radiation pneumonitis in nonsmall‐cell lung cancer (NSCLC) for response‐adapted radiotherapy

“…43 PET tumor radiomics can help predict response in NSCLC. 44 Yang et al found that GLSZM-LZLGE exhibited significant temporal changes in partial metabolic responders. 45 Cheng et al showed that texture parameters (GLSZM-ZSV) can help predict the survival in NSCLC patients.…”