Lung cancer diagnosis: radiologic imaging, histology, and genetics

Identifying the histologic subtypes of non-small cell lung cancer with computed tomography imaging: a comparative study of capsule net, convolutional neural network, and radiomics

BACKGROUND

Discriminating the subtypes of non-small cell lung cancer (NSCLC) based on computed tomography (CT) images is a challenging task for radiologists. Although several machine learning methods such as radiomics, and deep learning methods such as convolutional neural networks (CNNs) have been proposed to explore the problem, large sample sizes are required for effective training, and this may not be easily achieved in single-center datasets.

METHODS

In this study, an automated subtype recognition model with capsule net (CapsNet) was developed for the subtype discrimination of NSCLC. CapsNet utilizes an activity vector to record the relative spatial relationship of image elements that can subsequently better delineate the global image characteristics. CT images of 72 adenocarcinoma (AC) and 54 squamous cell carcinoma (SCC) patients were retrospectively collected from a single clinical center. The cancer region on the CT image was manually segmented for every subject by an experienced radiologist, and CapsNet, CNN, and four radiomics models were then used to construct the recognition model.

RESULTS

The study demonstrated that CapsNet achieved the best discriminative performance (accuracy 81.3%, specificity 80.7%, sensitivity 82.2%) although its area under the curve was just marginally better than that of the optimal random forest (RF) based radiomics model. Not surprisingly, the performance of the CNN was only comparable to the other radiomics models.

CONCLUSIONS

This study demonstrated that CapsNet is a viable potential framework for discriminating the subtypes of NSCLC, and its use could be extended to the recognition of other diseases especially in limited single-center datasets.

Is there any correlation between spectral CT imaging parameters and PD-L1 expression of lung adenocarcinoma?

Background

The aim of this study was to explore whether spectral computed tomography (CT) imaging parameters are associated with PD‐L1 expression of lung adenocarcinoma.

Methods

Spectral CT imaging parameters (iodine concentrations [IC] of lesion in arterial phase [ICLa] and venous phase [ICLv], normalized IC [NICa/NICv]‐normalized to the IC in the aorta, slope of the spectral HU curve [λHUa/λHUv] and enhanced monochromatic CT number [CT40keVa/v, CT70keVa/v] on 40 and 70 keV images) were analyzed in 34 prospectively enrolled lung adenocarcinoma patients with common molecular pathological markers including PD‐L1 expression detected with immunohistochemistry. Patients were divided into two groups: positive PD‐L1 expression and negative PD‐L1 expression groups. Two‐sample Mann‐Whitney U test was used to test the difference of spectral CT imaging parameters between the two groups.

Results

The CT40keVa (127.03 ± 37.92 vs. −54.69 ± 262.04), CT40keVv (124.39 ± 34.71 vs. −45.73 ± 238.97), CT70keVa (49.56 ± 11.76 vs. −136.51 ± 237.08) and CT70keVv (46.13 ± 15.81 vs. −133.10 ± 230.72) parameters in the positive PD‐L1 expression group of lung adenocarcinoma were significantly higher than the negative PD‐L1 expression group (all P < 0.05). There was no difference detected in IC, NIC and λHU of the arterial and venous phases between both groups (all P > 0.05).

Conclusion

CT40keVa, CT40keVv, CT70keVa and CT70keVv were increased in positive PD‐L1 expression. These parameters may be used to distinguish the PD‐L1 expression state of lung adenocarcinoma.

A continuous-flow acoustofluidic cytometer for single-cell mechanotyping

The biophysical properties of cells such as their compressibility have been found to be closely related to disease progression such as cancer development and metastasis. As cancer cells are heterogeneous, rapid and high-throughput evaluation of cell biophysical properties at single-cell resolution is needed to assess their potential as biomarkers for cancer staging and prognosis. Acoustofluidics has shown promise as a contactless method for accurately measuring cell biophysical properties; however, previously reported methods had relatively low throughput due to their requirement of no-flow conditions. This work presents a high-throughput continuous flow-based acoustofluidic cell mechanotyping method at single-cell resolution that retains the advantage of simplicity and low-cost.

Score for lung adenocarcinoma in China with EGFR mutation of exon 19: Combination of clinical and radiological characteristics analysis

BACKGROUD

The biopsy samples might be the only tumor material available for testing the EGFR mutation status in some cases, but these samples are often composed of variable ratios of tumor to normal cells. In this study, we sought to build a scoring system to predict Epidermal growth factor receptor (EGFR) exon 19 mutation in lung adenocarcinoma by clinical and radiological features.

METHODS

Enrolled in this study were 601 patients with lung adenocarcinoma. Qualitative evaluation of the clinical and radiological features included 25 aspects. Statistical analysis was used to assess the association of these features between the EGFR wild type and exon 19 mutation, based on a clinical scoring system built by the statistical model and the experience of the radiologists.

RESULTS

EGRF-exon-19-mutation was associated with the female gender [odds ratios (OR), 2.573; 95% confidence intervals (CI), 1.689-3.920], tumor maximum diameter (OR, 0.357; 95% CI, 0.235-0.542), the absence of emphysema (OR, 0.202; 95% CI, 0.110-0.368), the absence of fibrosis (OR, 0.168; 95% CI, 0.083-0.339), and pleural retraction (OR, 2.170; 95% CI, 1.434-3.285). The clinical scoring model assigned 3 points to the female gender, 2 points to small tumor maximum diameter (≤34.5 mm), 2 to the absence of emphysema, 2 to the absence of fibrosis, and 1 to the presence of pleural retraction.

CONCLUSIONS

The scoring system based on the statistical analysis of clinical and radiological features may be a new alternative to the prediction of EGFR mutation subtypes.

Imaging genotyping of functional signaling pathways in lung squamous cell carcinoma using a radiomics approach

Imaging features can be useful for identifying distinct genomic differences and have predictive power for certain phenotypes attributed to genomic mutations. We aimed to identify predictive imaging biomarkers that underpin genomic alterations and clinical outcomes in lung squamous cell carcinoma (SQCC) using a radiomics approach. In 57 patients with lung SQCC who underwent preoperative computed tomography (CT) and whole-exome DNA sequencing, 63 quantitative imaging features were extracted from CT and 73 clinicoradiological features including imaging features were classified into 8 categories: clinical, global, histogram-based, lung cancer-specific, shape, local, regional, and emphysema. Mutational profiles for core signaling pathways of lung SQCC were classified into five categories: redox stress, apoptosis, proliferation, differentiation, and chromatin remodelers. Range and right lung volume was significantly associated with alternation of apoptosis and proliferation pathway (p = 0.03, and p = 0.03). Energy was associated with the redox stress pathway (p = 0.06). None of the clinicoradiological features showed any significant association with the alteration of differentiation and chromatin remodelers pathway. This study showed that radiomic features indicating five different functional pathways of lung SQCC were different form one another. Radiomics approaches to lung SQCC have the potential to noninvasively predict alterations in core signaling pathways and clinical outcome.

Radiological and Clinical Features associated with Epidermal Growth Factor Receptor Mutation Status of Exon 19 and 21 in Lung Adenocarcinoma

The exon 19 and 21 in Epidermal Growth Factor Receptor (EGFR) mutation are the most common subtype of lung adenocarcinoma, and the strongest predictive biomarker for progression-free survival and tumor response. Although some studies have shown differences in radiological features between cases with and without EFGR mutations, they lacked necessary stratification. This article is to evaluate the association of CT features between the wild type and the subtype (exon 19 and 21) of EGFR mutations in patients with lung adenocarcinoma. Of the 721 finally included patients, 132 were positive for EGFR mutation in exon 19, 140 were positive for EGFR mutation in exon 21, and 449 were EGFR wild type. EGFR mutation in exon 19 was associated with a small-maximum diameter (28.51 ± 14.07) (p < 0.0001); sex (p < 0.0001); pleural retraction (p = 0.0034); and the absence of fibrosis (p < 0.0001), while spiculated margins (p = 0.0095), subsolid density (p < 0.0001) and no smoking (p < 0.0001) were associated with EGFR mutation in exon 21. Receiver Operating Characteristic (ROC) curves suggested that the maximum Area Under the Curve (AUC) was related to the female gender (AUC = 0.636) and the absence of smoking (AUC = 0.681). This study demonstrated the radiological and clinical features could be used to prognosticate EGFR mutation subtypes in exon 19 and 21.

Longitudinal evolution of incidentally detected solitary pure ground-glass nodules on CT: relation to clinical metrics

PURPOSE

We aimed to assess the relation between basic clinical parameters and evolution of solitary pure ground-glass nodules (pGGN) in the lungs.

METHODS

Baseline and follow-up computed tomography (CT) of patients with solitary pGGN were selected and two radiologists independently reviewed CTs for nodule characterization. CT features of solitary pGGN were manually measured maximum diameter (D1) and its orthogonal diameter (D2), mean diameter (mD), D1 to D2 ratio as surrogate of roundness, and location according to lobar anatomy. Longitudinal changes were assessed and solitary pGGNs were classified as resolved or persisting. Persisting nodules were further classified as stable or grown according to an increase in mD of ≥2 mm or appearance of solid component. Baseline CT features of solitary pGGNs and clinical metrics of patients were compared between resolved and persisting nodules and, thereafter, between stable and grown lesions.

RESULTS

A total of 95 subjects with solitary pGGN were included. After a median 16-month follow-up, 20 nodules resolved, while 75 persisted. Among persisting nodules, 18 were grown and 57 were stable. Grown nodules showed larger D1 and mD compared with stable pGGNs (P < 0.001). Subjects with grown nodules were older (P = 0.021). Logistic regression analyses showed higher likelihood of growth for nodules ≥10 mm (odds ratio [OR], 8.355; P = 0.001) and subjects older than 67 years (OR, 3.656; P = 0.034).

CONCLUSION

Nodules ≥10 mm in subjects older than 67 years showed higher likelihood of growth. These data could contribute to a more individual approach to the management of solitary pGGN.

Controversies in lung cancer screening

There remains an extensive debate over lung cancer screening, with lobbying for and against screening for very compelling reasons. The National Lung Screening Trial, International Early Lung Cancer Program, and other major screening studies favor screening with low-dose CT scans and have shown a reduction in lung cancer--specific mortality. The increasing incidence of lung cancer and the dismal survival rate for advanced disease despite improved multimodality therapy have sparked an interest in the implementation of national lung cancer screening. Concerns over imaging workflow, radiation dose, management of small nodules, overdiagnosis bias, lead-time and length-time bias, emerging new technologies, and cost-effectiveness continue to be debated. The authors address each of these issues as they relate to radiologic practice.