To identify imaging biomarkers of primary tumors and lymph nodes in patients with stage III-IV non-small cell lung cancer (NSCLC) and assess their predictive ability for treatment response (response vs. non-response) to immune checkpoint inhibitors (ICIs) after 6 months.

Retrospective analysis of 83 NSCLC patients treated with ICIs. Quantitative imaging features of the maximum primary lung tumors and lymph nodes on contrast-enhanced CT imaging were extracted at baseline (time point 0, TP0) and after 2-3 cycles of immunotherapy (time point 1, TP1). Delta-radiomics features (delta-RFs) were defined as the net changes in radiomics features (RFs) between TP0 and TP1. Interobserver interclass coefficient (ICC) and Pearson correlation analyses were applied for feature selection, and logistic regression (LR) was used to build a model for predicting treatment response.

Four and five important delta-RFs were selected to construct the nodal and tumor models, respectively. Δ Tumor diameter was used for constructing the clinical prediction model. The predictive efficacy of the nodal model for the treatment response status was higher than that of the tumor and clinical models. In the training set, the AUC values for the three models were 0.96 (95% CI = 0.90-1.00), 0.86 (95% CI = 0.76-0.95), and 0.82 (95% CI = 0.71-0.93), respectively. In the validation set, the AUC values were 0.94 (95% CI = 0.85-1.00), 0.77 (95% CI = 0.56-0.98), and 0.74 (95% CI = 0.48-1.00), respectively.

The nodal model based on delta-RFs performed well in distinguishing responders from non-responders and could identify patients more likely to benefit from immunotherapy. Finally, the nodal model exhibited a higher classification performance than the tumor model.

For many decades, the management of cancer has utilized radiation therapy, which continues to evolve with technology to improve patient outcomes. However, despite the standardization of treatment plans and the establishment of best clinical practices based on prospective, randomized trials and adherence to National Comprehensive Cancer Network (NCCN) guidelines, the outcomes from radiation therapy are highly variable and dependent on a number of factors, including patient demographics, tumor characteristics/histology, and treatment parameters. In this study, we attempt to use available patient data and treatment parameters at the time of radiation therapy to predict future outcomes using artificial intelligence (AI).

Six thousand five hundred ninety-five cases of patients who completed radiation treatment were selected retrospectively and used to train artificial neural networks (ANNs) and baseline models (i.e., logistic regression, random forest, support vector machines [SVMs], gradient boosting [XGBoost]) for binary classification of mortality at multiple time points ranging from six months to five years post-treatment. A hyperparameter grid search was used to identify the optimal network architecture for each time point, using sensitivity as the primary outcome metric.

The median age was 75 years (range: 2-102 years). There were 63.8% females and 36.1% males. The results indicate that ANNs were able to successfully perform binary mortality prediction with an accuracy greater than random chance and greater sensitivity than baseline models used. The best-performing algorithm was the ANN, which achieved a sensitivity of 83.00% ± 4.89% for five-year mortality.

The neural network was able to achieve higher sensitivity than Logistic Regression, SVM Random Forest, and XGBoost across all output target variables, demonstrating the utility of a neural network model for mortality prediction on the provided dataset.

Objectives: Radiomics and machine learning are innovative approaches to improve the clinical management of NSCLC. However, there is less information about the additive value of FDG PET-based radiomics compared with clinical and imaging variables. Methods: This retrospective study included 320 NSCLC patients who underwent PET/CT with FDG at initial staging. VOIs were placed on primary tumors only. We included a total of 94 variables, including 87 textural features extracted from PET studies, SUVmax, MTV, TLG, TNM stage, histology, age, and gender. We used the least absolute shrinkage and selection operator (LASSO) regression to select variables with the highest predictive value. Although several radiomics variables are available, the added value of these predictors compared with clinical and imaging variables is still under evaluation. Three hundred and twenty NSCLC patients were included in this retrospective study and underwent 18F-FDG PET/CT at initial staging. In this study, we evaluated 94 variables, including 87 textural features, SUVmax, MTV, TLG, TNM stage, histology, age, and gender. Image-based predictors were extracted from a volume of interest (VOI) positioned on the primary tumor. The least absolute shrinkage and selection operator (LASSO) Cox regression was used to reduce the number of variables and select only those with the highest predictive value. The predictive model implemented with the variables selected using the LASSO analysis was compared with a reference model using only a tumor stage and SUVmax. Results: NGTDM coarseness, SUVmax, and TNM stage survived the LASSO analysis and were used for the radiomic model. The AUCs obtained from the reference and radiomic models were 80.82 (95%CI, 69.01-92.63) and 81.02 (95%CI, 69.07-92.97), respectively (p = 0.98). The median OS in the reference model was 17.0 months in high-risk patients (95%CI, 11-21) and 113 months in low-risk patients (HR 7.47, p < 0.001). In the radiomic model, the median OS was 16.5 months (95%CI, 11-20) and 113 months in high- and low-risk groups, respectively (HR 9.64, p < 0.001). Conclusions: Our results indicate that a radiomic model composed using the tumor stage, SUVmax, and a selected radiomic feature (NGTDM_Coarseness) predicts survival in NSCLC patients similarly to a reference model composed only by the tumor stage and SUVmax. Replication of these preliminary results is necessary.

The goal of this study was to investigate whether the combined PET/CT radiomic features of the primary tumor and lymph node could predict lymph node metastasis (LNM) of resectable non-small cell lung cancer (NSCLC) in stage T2-4.

This retrospective study included 192 NSCLC patients who underwent tumor and node dissection between August 2016 and December 2017 and underwent ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT scanning 1-3 weeks before surgery. In total, 192 primary tumors (> 3 cm) and 462 lymph nodes (LN > 0.5 cm) were analyzed. The pretreatment clinical features of these patients were recorded, and the radiomic features of their primary tumor and lymph node were extracted from PET/CT imaging. The Spearman's relevance combined with the least absolute shrinkage and selection operator was used for radiomic feature selection. Five independent machine learning models (multi-layer perceptron, extreme Gradient Boosting, light gradient boosting machine, gradient boosting decision tree, and support vector machine) were tested as classifiers for model development. We developed the following three models to predict LNM: tumor PET/CT-clinical (TPC), lymph PET/CT-clinical (LPC), and tumor and lymph PET/CT-clinical (TLPC). The performance of the models and the clinical node (cN) staging was evaluated using the ROC curve and confusion matrix analysis.

The ROC analysis showed that among the three models, the TLPC model had better predictive clinical utility and efficiency in predicting LNM of NSCLC (AUC = 0.93, accuracy = 85%; sensitivity = 0.93; specificity = 0.75) than both the TPC model (AUC = 0.54, accuracy = 50%; specificity = 0.38; sensitivity = 0.59) and the LPC model (AUC = 0.82, accuracy = 70%; specificity = 0.41; sensitivity = 0.92). The TLPC model also exhibited great potential in predicting the N2 stage in NSCLC (AUC = 0.94, accuracy = 79%; specificity = 0.64; sensitivity = 0.91).

The combination of CT and PET radiomic features of the primary tumor and lymph node showed great potential for predicting LNM of resectable T2-4 NSCLC. The TLPC model can non-invasively predict lymph node metastasis in NSCLC, which may be helpful for clinicians to develop more rational therapeutic strategies.

Lung cancer is the second most common cancer and the leading cause of cancer-related death worldwide. Molecular imaging using [¹⁸F]fluorodeoxyglucose Positron Emission Tomography and/or Computed Tomography ([¹⁸F]FDG-PET/CT) plays an essential role in the diagnosis, evaluation of response to treatment, and prediction of outcomes. The images are evaluated using qualitative and conventional quantitative indices. However, there is far more information embedded in the images, which can be extracted by sophisticated algorithms. Recently, the concept of uncovering and analyzing the invisible data extracted from medical images, called radiomics, is gaining more attention. Currently, [¹⁸F]FDG-PET/CT radiomics is growingly evaluated in lung cancer to discover if it enhances the diagnostic performance or implication of [¹⁸F]FDG-PET/CT in the management of lung cancer. In this review, we provide a short overview of the technical aspects, as they are discussed in different articles of this special issue. We mainly focus on the diagnostic performance of the [¹⁸F]FDG-PET/CT-based radiomics and the role of artificial intelligence in non-small cell lung cancer, impacting the early detection, staging, prediction of tumor subtypes, biomarkers, and patient's outcomes.

Ideally, radiomics features and radiomics signatures can be used as imaging biomarkers for diagnosis, staging, prognosis, and prediction of tumor response. Thus, the number of published radiomics studies is increasing exponentially, leading to a myriad of new radiomics-based evidence for lung cancer. Consequently, it is challenging for radiologists to keep up with the development of radiomics features and their clinical applications. In this article, we review the basics to advanced radiomics in lung cancer to guide young researchers who are eager to start exploring radiomics investigations. In addition, we also include technical issues of radiomics, because knowledge of the technical aspects of radiomics supports a well-informed interpretation of the use of radiomics in lung cancer.

We explain the rationale for metabolically adaptive radiation dose escalation in stage III non-small-cell lung cancer and describe the design of a Canadian phase II randomized trial investigating this approach. In the trial, patients are randomized to either conventional chemoradiation treatment (60 Gy in 30 fractions) or metabolically adaptive chemoradiation, where fluorodeoxyglucose-avid tumor sub-volumes receive an integrated boost dose to a maximum of 85 Gy in 30 fractions. The trial sample size is 78 patients, and the target population is patients with newly diagnosed, inoperable stage III non-small-cell lung cancer treated with radical intent chemoradiation. The primary objective of the trial is to determine if dose escalation to metabolically active sub-volumes will reduce 2-year local-regional failure rate from 42.3% to 22.3%, when compared with standard treatment. The secondary objectives are to determine the effect of dose escalation on overall survival, progression-free survival, quality of life, and rate of grade 3 to 5 toxicities.

Lymph node stage prior to treatment is strongly related to disease progression and poor prognosis in non-small cell lung cancer (NSCLC). However, few studies have investigated metabolic imaging features derived from pre-radiotherapy 18F-fluorodeoxyglucose (FDG) positron-emission tomography (PET) of metastatic hilar/mediastinal lymph nodes (LNs). We hypothesized that these would provide complementary prognostic information to FDG-PET descriptors to only the primary tumor (tumor).

Two independent cohorts of 262 and 50 node-positive NSCLC patients were used for model development and validation. Image features (i.e. Radiomics) including shape and size, first order statistics, texture, and intensity-volume histograms (IVH) (http://www.radiomics.io/) were evaluated by univariable Cox regression on the development cohort. Prognostic modeling was conducted with a 10-fold cross-validated least absolute shrinkage and selection operator (LASSO), automatically selecting amongst FDG-PET-Radiomics descriptors from (1) tumor, (2) LNs or (3) both structures. Performance was assessed with the concordance-index. Development data are publicly available at www.cancerdata.org and Dryad (doi:10.5061/dryad.752153b).

Common SUV descriptors (maximum, peak, and mean) were significantly related to overall survival when extracted from LNs, as were LN volume and tumor load (summed tumor and LNs' volumes), though this was not true for either SUV metrics or tumor's volume. Feature selection exclusively from imaging information based on FDG-PET-Radiomics, exhibited performances of (1) 0.53 -external 0.54, when derived from the tumor, (2) 0.62 -external 0.56 from LNs, and (3) 0.62 -external 0.59 from both structures, including at least one feature from each sub-category, except IVH.

Combining imaging information based on FDG-PET-Radiomics features from tumors and LNs is desirable to achieve a higher prognostic discriminative power for NSCLC.

Noninvasive biomarkers that capture the total tumor burden could provide important complementary information for precision medicine to aid clinical decision making. We investigated the value of radiomic data extracted from pretreatment computed tomography images of the primary tumor and lymph nodes in predicting pathological response after neoadjuvant chemoradiation before surgery.

A total of 85 patients with resectable locally advanced (stage II-III) NSCLC (median age 60.3 years, 65% female) treated from 2003 to 2013 were included in this institutional review board-approved study. Radiomics analysis was performed on 85 primary tumors and 178 lymph nodes to discriminate between pathological complete response (pCR) and gross residual disease (GRD). Twenty nonredundant and stable features (10 from each site) were evaluated by using the area under the curve (AUC) (all p values were corrected for multiple hypothesis testing). Classification performance of each feature set was evaluated by random forest and nested cross validation.

Three radiomic features (describing primary tumor sphericity and lymph node homogeneity) were significantly predictive of pCR with similar performances (all AUC = 0.67, p < 0.05). Two features (quantifying lymph node homogeneity) were predictive of GRD (AUC range 0.72-0.75, p < 0.05) and performed significantly better than the primary features (AUC = 0.62). Multivariate analysis showed that for pCR, the radiomic features set alone had the best-performing classification (median AUC = 0.68). Furthermore, for GRD classification, the combination of radiomic and clinical data significantly outperformed all other feature sets (median AUC = 0.73).

Lymph node phenotypic information was significantly predictive for pathological response and showed higher classification performance than radiomic features obtained from the primary tumor.