Artificial intelligence: A critical review of applications for lung nodule and lung cancer.

Artificial intelligence (AI) is a broad concept that usually refers to computer programs that can learn from data and perform certain specific tasks. In the recent years, the growth of deep learning, a successful technique for computer vision tasks that does not require explicit programming, coupled with the availability of large imaging databases fostered the development of multiple applications in the medical imaging field, especially for lung nodules and lung cancer, mostly through convolutional neural networks (CNN). Some of the first applications of AI is this field were dedicated to automated detection of lung nodules on X-ray and computed tomography (CT) examinations, with performances now reaching or exceeding those of radiologists. For lung nodule segmentation, CNN-based algorithms applied to CT images show excellent spatial overlap index with manual segmentation, even for irregular and ground glass nodules. A third application of AI is the classification of lung nodules between malignant and benign, which could limit the number of follow-up CT examinations for less suspicious lesions. Several algorithms have demonstrated excellent capabilities for the prediction of the malignancy risk when a nodule is discovered. These different applications of AI for lung nodules are particularly appealing in the context of lung cancer screening. In the field of lung cancer, AI tools applied to lung imaging have been investigated for distinct aims. First, they could play a role for the non-invasive characterization of tumors, especially for histological subtype and somatic mutation predictions, with a potential therapeutic impact. Additionally, they could help predict the patient prognosis, in combination to clinical data. Despite these encouraging perspectives, clinical implementation of AI tools is only beginning because of the lack of generalizability of published studies, of an inner obscure working and because of limited data about the impact of such tools on the radiologists' decision and on the patient outcome. Radiologists must be active participants in the process of evaluating AI tools, as such tools could support their daily work and offer them more time for high added value tasks.

Interstitial lung diseases after hematopoietic stem cell transplantation: New pattern of lung chronic graft-versus-host disease?

Bronchiolitis obliterans syndrome (BOS) after allogeneic HSCT is the only formally recognized manifestation of lung chronic graft-versus-host disease (GVHD). Other lung complications were reported, including interstitial lung diseases (ILDs). Whether ILDs belong to the spectrum of lung cGVHD remains unknown. We compared characteristics and specific risk factors for both ILD and BOS. Data collected from consecutive patients diagnosed with ILD or BOS from 1981-2019 were analyzed. The strength of the association between patient characteristics and ILD occurrence was measured via odds ratios estimated from univariable logistic models. Multivariable models allowed us to handle potential confounding variables. Overall survival (OS) was estimated using the Kaplan-Meier method. 238 patients were included: 79 with ILD and 159 with BOS. At diagnosis, FEV1 was lower in patients with BOS compared to patients with ILD, while DLCO was lower in ILD. 84% of ILD patients received systemic corticosteroids, leading to improved CT scans and pulmonary function, whereas most BOS patients were treated by inhaled corticosteroids, with lung-function stabilization. In the multivariable analysis, prior thoracic irradiation and absence of prior treatment with prednisone were associated with ILD. OS was similar, even if hematological relapse was more frequent in the ILD group. Both complications occurred mainly in patients with GVHD history.

Artificial intelligence in lung cancer: current applications and perspectives.

Artificial intelligence (AI) has been a very active research topic over the last years and thoracic imaging has particularly benefited from the development of AI and in particular deep learning. We have now entered a phase of adopting AI into clinical practice. The objective of this article was to review the current applications and perspectives of AI in thoracic oncology. For pulmonary nodule detection, computer-aided detection (CADe) tools have been commercially available since the early 2000s. The more recent rise of deep learning and the availability of large annotated lung nodule datasets have allowed the development of new CADe tools with fewer false-positive results per examination. Classical machine learning and deep-learning methods were also used for pulmonary nodule segmentation allowing nodule volumetry and pulmonary nodule characterization. For pulmonary nodule characterization, radiomics and deep-learning approaches were used. Data from the National Lung Cancer Screening Trial (NLST) allowed the development of several computer-aided diagnostic (CADx) tools for diagnosing lung cancer on chest computed tomography. Finally, AI has been used as a means to perform virtual biopsies and to predict response to treatment or survival. Thus, many detection, characterization and stratification tools have been proposed, some of which are commercially available.

Feasibility of Outpatient Transradial Prostatic Artery Embolization and Safety of a Shortened Deflation Protocol for Hemostasis.

BACKGROUND: to evaluate the safety and feasibility of a shorter time to hemostasis applied to outpatient transradial (TR) Prostatic Artery Embolization (PAE). METHODS: a retrospective bi-institutional study was conducted between July 2018 and April 2022 on 300 patients treated by outpatient TR PAE. Indications included lower urinary tract symptoms, acute urinary retention, and hematuria. Mean patient height was 176 ± 6.3 (158-192) cm. The primary endpoint was safety of a 45 min deflation protocol for hemostasis. The secondary endpoint was the feasibility of PAE using TR access. RESULTS: technical success was 98.7% (296/300). There was one failure due to patient height. Mean DAP/fluoroscopy times were 16,225 ± 12,126.3 (2959-81,608) µGy·m2/35 ± 14.7 (11-97) min, and mean time to discharge was 80 ± 6 (75-90) min. All access site and embolization-related adverse events were minor. Mild hematoma occurred in 10% (30/300), radial artery occlusion (RAO) in 10/300 (3.3%) cases, and history of smoking was a predictor for RAO. There was no major event. CONCLUSION: the safety of TR PAE using a 45 min time to hemostasis was confirmed, and TR PAE is feasible in most cases. Radial artery occlusion was still observed and may be favored by smoking.

Deep Learning for the Detection, Localization, and Characterization of Focal Liver Lesions on Abdominal US Images.

Purpose: To train and assess the performance of a deep learning-based network designed to detect, localize, and characterize focal liver lesions (FLLs) in the liver parenchyma on abdominal US images. Materials and Methods: In this retrospective, multicenter, institutional review board-approved study, two object detectors, Faster region-based convolutional neural network (Faster R-CNN) and Detection Transformer (DETR), were fine-tuned on a dataset of 1026 patients (n = 2551 B-mode abdominal US images obtained between 2014 and 2018). Performance of the networks was analyzed on a test set of 48 additional patients (n = 155 B-mode abdominal US images obtained in 2019) and compared with the performance of three caregivers (one nonexpert and two experts) blinded to the clinical history. The sign test was used to compare accuracy, specificity, sensitivity, and positive predictive value among all raters. Results: DETR achieved a specificity of 90% (95% CI: 75, 100) and a sensitivity of 97% (95% CI: 97, 97) for the detection of FLLs. The performance of DETR met or exceeded that of the three caregivers for this task. DETR correctly localized 80% of the lesions, and it achieved a specificity of 81% (95% CI: 67, 91) and a sensitivity of 82% (95% CI: 62, 100) for FLL characterization (benign vs malignant) among lesions localized by all raters. The performance of DETR met or exceeded that of two experts and Faster R-CNN for these tasks. Conclusion: DETR demonstrated high specificity for detection, localization, and characterization of FLLs on abdominal US images. Supplemental material is available for this article. RSNA, 2022Keywords: Computer-aided Diagnosis (CAD), Ultrasound, Abdomen/GI, Liver, Tissue Characterization, Supervised Learning, Transfer Learning, Convolutional Neural Network (CNN).

Long-term outcomes of adult pulmonary Langerhans cell histiocytosis: a prospective cohort.

BACKGROUND: The long-term outcomes of adult pulmonary Langerhans cell histiocytosis (PLCH), particularly survival, are largely unknown. Two earlier retrospective studies reported a high rate of mortality, which contrasts with our clinical experience. METHODS: To address this issue, all patients with newly diagnosed PLCH referred to the French national reference centre for histiocytoses between 2004 and 2018 were eligible for inclusion. The primary outcome was survival, which was defined as the time from inclusion to lung transplantation or death from any cause. Secondary outcomes included the cumulative incidences of chronic respiratory failure (CRF), pulmonary hypertension (PH), malignant diseases and extrapulmonary involvement in initially isolated PLCH. Survival was estimated using the Kaplan-Meier method. RESULTS: 206 patients (mean age 39±13 years, 60% female, 95% current smokers) were prospectively followed for a median duration of 5.1 years (IQR 3.2-7.6 years). Of these, 12 patients (6%) died. The estimated rate of survival at 10 years was 93% (95% CI 89-97%). The cumulative incidences of CRF and/or PH were <5% at both 5 and 10 years, and 58% of these patients died. 27 malignancies were observed in 23 patients. The estimated standardised incidence ratio of lung carcinoma was 17.0 (95% CI 7.45-38.7) compared to an age- and sex-matched French population. Eight (5.1%) of the 157 patients with isolated PLCH developed extrapulmonary involvement. CONCLUSION: The long-term prognosis of PLCH is significantly more favourable than has previously been reported. Patients must be closely monitored after diagnosis to detect severe complications early.

Nonpalpable breast lesions: impact of a second-opinion review at a breast unit on BI-RADS classification.

OBJECTIVE: To compare BI-RADS classification, management, and outcome of nonpalpable breast lesions assessed both by community practices and by a multidisciplinary tumor board (MTB) at a breast unit. METHODS: All nonpalpable lesions that were first assigned a BI-RADS score by community practices and then reassessed by an MTB at a single breast unit from 2009 to 2017 were retrospectively reviewed. Inter-review agreement was assessed with Cohen's kappa statistic. Changes in biopsy recommendation were calculated. The percentage of additional tumor lesions detected by the MTB was obtained. The sensitivity, AUC, and cancer rates for BI-RADS category 3, 4, and 5 lesions were computed for both reviews. RESULTS: A total of 1909 nonpalpable lesions in 1732 patients were included. For BI-RADS scores in the whole cohort, a fair agreement was found (κ = 0.40 [0.36-0.45]) between the two reviews. Agreement was higher when considering only mammography combined with ultrasound (κ = 0.53 [0.44-0.62]), masses (κ = 0.50 [0.44-0.56]), and architectural distortion (κ = 0.44 [0.11-0.78]). Changes in biopsy recommendation occurred in 589 cases (31%). Ninety of 345 additional biopsies revealed high-risk or malignant lesions. Overall, the MTB identified 27% additional high-risk and malignant lesions compared to community practices. The BI-RADS classification AUCs for detecting malignant lesions were 0.66 (0.63-0.69) for community practices and 0.76 (0.75-0.78) for the MTB (p < 0.001). CONCLUSION: Agreement between community practices and MTB reviews for BI-RADS classification in nonpalpable lesions is only fair. MTB review improves diagnostic performances of breast imaging and patient management. KEY POINTS: • The inter-review agreement for BI-RADS classification between community practices and the multidisciplinary board was only fair (κ = 0.40). • Disagreements resulted in changes of biopsy recommendation in 31% of the lesions. • The multidisciplinary board identified 27% additional high-risk and malignant lesions compared to community practices.

Impact of New Systemic Treatment and Radiotherapy in Melanoma Patients with Leptomeningeal Metastases.

IMPORTANCE: Few data are available on patients with leptomeningeal disease (LM) from melanoma treated with new systemic therapies. OBJECTIVE: To gain a better understanding of patients, disease characteristics, and therapeutic interventions in melanoma patients with LM in the era of new systemic treatment. DESIGN: Clinical characteristics, treatments, and survival of melanoma patients diagnosed with LM, isolated or associated with brain metastases, were collected. The Cox regression model assessed the influence of patient and melanoma characteristics on survival. SETTING: Monocentric, retrospective, real-life cohort of patients with LM from melanoma. PARTICIPANTS: All patients followed up at Saint-Louis University Hospital and diagnosed with LM between December 2013 and February 2020 were included. For each patient identified, a central review by dermato-oncologist and neuro-oncologist experts was performed to confirm the diagnosis of LM. EXPOSURE: Impact of new systemic therapies and radiotherapy. RESULTS: Among the 452 advanced melanoma patients followed at St Louis Hospital between 2013 and 2020, 41 patients with LM from melanoma were identified. Among them, 29 patients with a diagnosis of LM "confirmed" or "probable" after central neuro-oncologists reviewing were included. Nineteen patients had known melanoma brain metastases at LM diagnosis. Among the 27 patients treated with systemic therapy, 17 patients were treated with immunotherapy, 5 patients received targeted therapy, 1 was treated with chemotherapy, and 4 patients were treated with anti-PD-1 in combination with BRAF inhibitor. The median overall survival (OS) from LM diagnosis was 5.1 months. Median OS was 7.1 months for the 9 patients receiving systemic therapy combined with radiotherapy, and 3.2 months for the 20 patients not receiving combined radiotherapy. Elevated serum lactate dehydrogenase (LDH) (HR 1.44, 95% CI 1.09-1.90, p < 0.01) and presence of neurological symptoms at LM diagnosis (HR 2.96, 95% CI 1.25-6.99, p = 0.01) were associated with poor survival. At the time of data analysis, five patients were still alive with a median follow-up of 47.4 months and had persistent complete response. CONCLUSION: Targeted therapy and immunotherapy are promising new treatment options in LM from melanoma that can increase overall survival, and may induce long lasting remission in some patients.

Assessing invasiveness of subsolid lung adenocarcinomas with combined attenuation and geometric feature models.

The aim of this study was to develop and test multiclass predictive models for assessing the invasiveness of individual lung adenocarcinomas presenting as subsolid nodules on computed tomography (CT). 227 lung adenocarcinomas were included: 31 atypical adenomatous hyperplasia and adenocarcinomas in situ (class H1), 64 minimally invasive adenocarcinomas (class H2) and 132 invasive adenocarcinomas (class H3). Nodules were segmented, and geometric and CT attenuation features including functional principal component analysis features (FPC1 and FPC2) were extracted. After a feature selection step, two predictive models were built with ordinal regression: Model 1 based on volume (log) (logarithm of the nodule volume) and FPC1, and Model 2 based on volume (log) and Q.875 (CT attenuation value at the 87.5% percentile). Using the 200-repeats Monte-Carlo cross-validation method, these models provided a multiclass classification of invasiveness with discriminative power AUCs of 0.83 to 0.87 and predicted the class probabilities with less than a 10% average error. The predictive modelling approach adopted in this paper provides a detailed insight on how the value of the main predictors contribute to the probability of nodule invasiveness and underlines the role of nodule CT attenuation features in the nodule invasiveness classification.

The Growth Rate of Subsolid Lung Adenocarcinoma Nodules at Chest CT.

Background Confirming that subsolid adenocarcinomas show exponential growth is important because it would justify using volume doubling time to assess their growth. Purpose To test whether the growth of lung adenocarcinomas manifesting as subsolid nodules at chest CT is accurately represented by an exponential model. Materials and Methods Patients with lung adenocarcinomas manifesting as subsolid nodules surgically resected between January 2005 and May 2018, with three or more longitudinal CT examinations before resection, were retrospectively included. Overall volume (for all nodules) and solid component volume (for part-solid nodules) were measured over time. A linear mixed-effects model was used to identify the growth pattern (linear, exponential, quadratic, or power law) that best represented growth. The interactions between nodule growth and clinical, CT morphologic, and pathologic parameters were studied. Results Sixty-nine patients (mean age, 70 years ± 9 [standard deviation]; 48 women) with 74 lung adenocarcinomas were evaluated. Overall growth and solid component growth were better represented by an exponential model (adjusted R2 = 0.89 and 0.95, respectively) than by a quadratic model (r2 = 0.88 and 0.93, respectively), a linear model (r2 = 0.87 and 0.92, respectively), or a power law model (r2 = 0.82 and 0.93, respectively). Faster overall volume growth was associated with a history of lung cancer (P < .001), a baseline nodule volume less than 500 mm3 (P = .03), and histologic findings of invasive adenocarcinoma (P < .001). The median volume doubling time of noninvasive adenocarcinoma was significantly longer than that of invasive adenocarcinoma (939 days [interquartile range, 588-1563 days] vs 678 days [interquartile range, 392-916 days], respectively; P = .01). Conclusion The overall volume growth of adenocarcinomas manifesting as subsolid nodules at chest CT was best represented by an exponential model compared with the other tested models. This justifies the use of volume doubling time for the growth assessment of these nodules. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Kuriyama and Yanagawa in this issue.

Preoperative bronchial cytology for the assessment of tumor spread through air spaces in lung adenocarcinoma resection specimens.

BACKGROUND: Tumor spread through air spaces (STAS), a significant prognostic indicator, has been described recently as a pattern of invasion in pulmonary carcinomas. However, questions remain regarding preoperative identification of STAS and whether it represents an in vivo phenomenon versus an ex vivo artifact. METHODS: We retrospectively reviewed 67 paired preoperative bronchoalveolar lavage (BAL) or bronchial washing (BW) cytology specimens with the subsequent lung adenocarcinoma surgical resection specimen to determine whether preoperative cytology could predict STAS. Other clinical, radiologic, and pathologic features of the resected lesions were also correlated with preoperative bronchial cytology results. RESULTS: Positive bronchial cytology was observed in 28 cases (41.8%), 24 of which had STAS (85.7%); however, negative BAL/BW cytology was observed in 39 cases (58.2%), 29 of which had STAS (74.4%) (x2  = 1.27, P = .26, not significant). High-STAS burden was observed in 44 cases (83.0%), 21 (47.7%) with negative BAL/BW and 23 (52.3%) with positive BAL/BW. Low-STAS burden was observed in 9 cases (17.0%), 8 (88.9%) with negative BAL/BW and only 1 (11.1%) with positive BAL/BW (x2  = 5.11, P = .024, significant). For tumors with STAS, a statistically significant difference was identified in the maximal STAS distance from the main tumor edge between BAL/BW-positive and BAL/BW-negative groups (P = .007). Of the remaining clinicopathologic and radiologic features, only visceral pleural invasion was significantly associated with BAL/BW positivity. CONCLUSION: Presurgical bronchial cytology alone cannot adequately predict tumor STAS; however, it may provide useful information regarding the extent and overall burden of STAS on the subsequent resection specimen.

Growth Assessment of Pulmonary Adenocarcinomas Manifesting as Subsolid Nodules on CT: Comparison of Diameter-Based and Volume Measurements.

RATIONALE AND OBJECTIVES: To analyze the performances of diameter-based measurements, either using diameters, or by calculating diameter-based volumes, as compared to volume measurements in assessing growth of pulmonary adenocarcinomas manifesting as subsolid nodules on CT. MATERIALS AND METHODS: In this IRB-approved, retrospective study, 74 pulmonary adenocarcinomas presenting as subsolid nodules and resected in 69 patients (21 men, 48 women, mean age 70 ± 9 years) were included. Three CTs were available for each patient. Nodule size on each CT was assessed with diameter measurements, calculated volume based on diameter measurements, and measured volume. Nodule growth was defined as an increase of measured volume ≥25% between two sequential CTs. Sensitivity, specificity, accuracy, positive and negative predictive values of diameter-based measurements for growth assessment were calculated. Nodule characteristics were compared with nonparametric tests and analysis of variance. RESULTS: There were fewer growing nodules during CT1-CT2 interval (n = 22, 30%) than during CT2-CT3 interval (n = 33, 45%, p =.060). Specificity and negative predictive value of diameter-based measurements for growth assessment ranged respectively from 52 to 77% and 81 to 83% between CT1 and CT2, and from 66 to 76% and 79 to 90% between CT2 and CT3. Nongrowing nodules tended to be larger, regardless how size was measured, and some of these differences in size were statistically significant (p =.002 to .046). CONCLUSION: For pulmonary adenocarcinomas presenting as subsolid nodules on CT, diameter-based assessment of nodule volume is reasonably accurate at confirming a lack of nodule growth but may overestimate actual growth, as compared to growth assessment based on measured volume.

Management and outcomes of pneumothorax in adult patients with Langerhans cell Histiocytosis.

BACKGROUND: Pneumothorax may recur during pulmonary Langerhans cell histiocytosis (PLCH) patients' follow-up and its management is not standardised. The factors associated with pneumothorax recurrence are unknown. METHODS: In this retrospective study, PLCH patients who experienced a pneumothorax and were followed for at least 6 months after the first episode were eligible. The objectives were to describe the treatment of the initial episode and pneumothorax recurrences during follow-up. We also searched for factors associated with pneumothorax recurrence and evaluated the effect on lung function outcome. Time to recurrence was estimated by the Kaplan Meier method and the cumulative hazard of recurrence handling all recurrent events was estimated. Univariate Cox models and Andersen-Gill counting process were used for statistical analyses. RESULTS: Fourty-three patients (median age 26.5 years [interquartile range (IQR), 22.9-35.4]; 26 men, 39 current smokers) were included and followed for median time of 49 months. Chest tube drainage was the main management of the initial pneumothorax, which resolved in 70% of cases. Pneumothorax recurred in 23 (53%) patients, and overall 96 pneumothoraces were observed during the study period. In the subgroup of patients who experienced pneumothorax recurrence, the median number of episodes per patient was 3 [IQR, 2-4]. All but one recurrence occurred within 2 years after the first episode. Thoracic surgery neither delayed the time of occurrence of the first ipsilateral recurrence nor reduced the overall number of recurrences during the study period, although the rate of recurrence was lower after thoracotomy than following video-assisted thoracic surgery (p = 0.03). At the time of the first pneumothorax, the presence of air trapping on lung function testing was associated with increased risk of recurrence (hazard ratio = 5.08; 95% confidence interval [1.18, 21.8]; p = 0.03). Pneumothorax recurrence did not predict subsequent lung function decline (p = 0.058). CONCLUSIONS: Our results show that pneumothorax recurrences occur during an "active" phase of PLCH. In this observational study, the time of occurrence of the first ipsilateral recurrence and the overall number of pneumothorax recurrences were similar after conservative and thoracic surgical treatments. Further studies are needed to determine the best management to reduce the risk of pneumothorax recurrence in PLCH patients.

Lower respiratory tract amyloidosis: Presentation, survival and prognostic factors. A multicenter consecutive case series.

Lower-respiratory-tract (LRT) amyloidosis has rarely been investigated. Our study presents characteristics, outcomes and survival of LRT amyloidosis. This multicenter retrospective study, from 1995 to 2017, included 73 patients with amyloidosis and LRT involvement. Respiratory patterns were: tracheobronchial (n = 17), nodular (n = 10), interstitial (n = 14) or composite (several respiratory involvements, n = 32). Interstitial and composite patterns were associated with multi-organ amyloidosis (n = 37, 80%) while tracheobronchial and nodular patterns were associated with organ-limited amyloidosis (n = 21, 78%). Amyloid light chain (AL) amyloidosis was diagnosed in 43 patients (59%), mainly of lambda type (n = 33, 77%). Smokers' proportion was higher in tracheobronchial (71%) and nodular (90%) patterns than in interstitial (14%) and composite (34%) patterns. The B-cell neoplasms involved 15 patients (21%), solid neoplasms 8 (11%), connective tissue diseases 8 (11%) and multiple myeloma 6 (8%). The B-cell and solid neoplasms were most prevalent in nodular pattern. Median follow-up was 4.4 years (2.2-8.9). Twenty-four patients died, mostly from respiratory infection. Survival at 1, 5, 10 years was respectively 88%, 70% and 54% for multi-organ amyloidosis, 96%, 89% and 69% for organ-limited amyloidosis (P = .125). Tracheobronchial and nodular patterns survival was better than in other respiratory patterns (P = .039). Death risk factors (multivariate analysis) were: cardiac localization (hazard-ratio [HR] 4.3 [95% confidence interval 1.6-11.5]; P = .004), age (HR 2.1 [1.2-3.7]; P = .008) and dyspnea at diagnosis (HR 4.0 [1.3-12.3]; P = .014). Various LRT amyloidosis patterns depend on smoking habits, organ-limited or multi-organ extension and comorbidities. They are associated with a different survival, which is also predicted by age, cardiac localization and dyspnea at presentation.

Visceral Pleural Invasion in Pulmonary Adenocarcinoma: Differences in CT Patterns between Solid and Subsolid Cancers.

PURPOSE: To analyze the incidence and CT patterns of visceral pleural invasion (VPI) in adenocarcinomas on the basis of their CT presentation as solid or subsolid nodules. MATERIALS AND METHODS: A total of 286 adenocarcinomas in direct contact with a pleural surface, resected at an institution between 2005 and 2016, were included in this retrospective, institutional review board-approved study. CT size and longest contact length with a pleural surface were measured and their ratios computed. Pleural deviation, pleural thickening, spiculations, different pleural tag types, pleural effusion, and the CT appearance of transgression into an adjacent lobe or infiltration of surrounding tissue were evaluated. Fisher exact tests and simple and multiple logistic regression models were used. RESULTS: Of the 286 nodules, 179 of 286 (62.6%) were solid and 107 of 286 (37.4%) were subsolid. VPI was present in 49 of 286 (17.1%) nodules and was significantly more frequent in solid (44 of 179; 24.6%) than in subsolid nodules (five of 107; 4.7%; P < .001). In solid nodules, multiple regression analysis showed an association of higher contact length-to-size ratio (adjusted odds ratio [OR], 1.02; P = .007) and the presence of multiple pleural tag types (adjusted OR, 5.88; P = .002) with VPI. In subsolid nodules, longer pleural contact length of the solid nodular component (adjusted OR, 1.27; P = .017) and the CT appearance of transgression or infiltration (adjusted OR, 10.75; P = .037) were associated with VPI. CONCLUSION: During preoperative evaluation of adenocarcinomas for the likelihood of VPI, whether a tumor manifests as a solid or a subsolid nodule is important to consider because the incidence of VPI is significantly higher in solid than in subsolid nodules. In addition, this study showed that the CT patterns associated with VPI differ between solid and subsolid nodules.© RSNA, 2019Supplemental material is available for this article.See also the commentary by Elicker in this issue.