Stable and discriminating radiomic predictor of recurrence in early stage non-small cell lung cancer: Multi-site study

OBJECTIVES

To evaluate whether combining stability and discriminability criteria in building radiomic classifiers will improve the prognosis of cancer recurrence in early stage non-small cell lung cancer on non-contrast computer tomography (CT).

MATERIALS AND METHODS

CT scans of 610 patients with early stage (IA, IB, IIA) NSCLC from four independent cohorts were evaluated. A total of 350 patients from Cleveland Clinic Foundation and University of Pennsylvania were divided into two equal sets for training (D₁) and validation set (D₂). 80 patients from The Cancer Genome Atlas Lung Adenocarcinoma and Squamous Cell Carcinoma and 195 patients from The Cancer Imaging Archive, were used as independent second (D₃) and third (D₄) validation sets. A linear discriminant analysis (LDA) classifier was built based on the most stable and discriminate features. In addition, a radiomic risk score (RRS) was generated by using least absolute shrinkage and selection operator, Cox regression model to predict time to progression (TTP) following surgery.

RESULTS

A feature selection strategy focusing on both feature discriminability and stability resulted in the classifier having a higher discriminability on validation datasets compared to the discriminability alone criteria in discriminating cancer recurrence (D₂, AUC of 0.75 vs. 0.65; D₃, 0.74 vs. 0.62; D₄, 0.76 vs. 0.63). The RRS generated by most stable-discriminating features was significantly associated with TTP compared to discriminating alone criteria (HR = 1.66, C-index of 0.72 vs. HR = 1.04, C-index of 0.62).

CONCLUSION

Accounting for both stability and discriminability yielded a more generalizable classifier for predicting cancer recurrence and TTP in early stage NSCLC.

Prediction of Heparin Induced Thrombocytopenia (HIT) Using a Combination of 4Ts Score and Screening Immune Assays

Clinical assessment (4Ts) followed by testing for Heparin/platelet factor 4 (HPF4) antibody in intermediate and high risk patients is the standard algorithm of pretest for Heparin induced thrombocytopenia (HIT), and the diagnosis is confirmed by serotonin releasing assay (SRA) in those who have positive antibodies. We conducted a retrospective analysis in a cohort of patients treated in a community hospital who had HIT antibody test by either ELISA or a rapid Particle Immunofiltration Assay (PIFA), regardless of their 4Ts scores. Among 224 patients, 17 had HIT. The PPV for those with a 4 T score ≥4 was 10.4%, which misdianosed 3 patients with HIT who tested positive for antibodies. Combining 4 T score ≥4 AND positive HIT antibody showed a PPV of 20.3% and a sensitivity of 70.6%, misdiagnosing 5 HIT patients. Using 4Ts ≥4 OR positive HIT antibody showed 100% sensitivity and 100% negative predictive value (NPV). The ELISA test had 100% sensitivity and 100% NPV, while the PIFA test missed 2 HIT patients, with sensitivity of 60% and NPV of 96.7%. Our results suggest that SRA testing should be conducted if a patient presents with a 4 T score ≥4 OR a positive HIT antibody, and antibody tests should be conducted for every patient suspected of HIT.

Changes in CT Radiomic Features Associated with Lymphocyte Distribution Predict Overall Survival and Response to Immunotherapy in Non-Small Cell Lung Cancer

No predictive biomarkers can robustly identify patients with non-small cell lung cancer (NSCLC) who will benefit from immune checkpoint inhibitor (ICI) therapies. Here, in a machine learning setting, we compared changes ("delta") in the radiomic texture (DelRADx) of CT patterns both within and outside tumor nodules before and after two to three cycles of ICI therapy. We found that DelRADx patterns could predict response to ICI therapy and overall survival (OS) for patients with NSCLC. We retrospectively analyzed data acquired from 139 patients with NSCLC at two institutions, who were divided into a discovery set (D1 = 50) and two independent validation sets (D2 = 62, D3 = 27). Intranodular and perinodular texture descriptors were extracted, and the relative differences were computed. A linear discriminant analysis (LDA) classifier was trained with 8 DelRADx features to predict RECIST-derived response. Association of delta-radiomic risk score (DRS) with OS was determined. The association of DelRADx features with tumor-infiltrating lymphocyte (TIL) density on the diagnostic biopsies (n = 36) was also evaluated. The LDA classifier yielded an AUC of 0.88 ± 0.08 in distinguishing responders from nonresponders in D1, and 0.85 and 0.81 in D2 and D3 DRS was associated with OS [HR: 1.64; 95% confidence interval (CI), 1.22-2.21; P = 0.0011; C-index = 0.72). Peritumoral Gabor features were associated with the density of TILs on diagnostic biopsy samples. Our results show that DelRADx could be used to identify early functional responses in patients with NSCLC.

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

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.

Combination of Peri- and Intratumoral Radiomic Features on Baseline CT Scans Predicts Response to Chemotherapy in Lung Adenocarcinoma

Purpose

To identify the role of radiomics texture features both within and outside the nodule in predicting (a) time to progression (TTP) and overall survival (OS) as well as (b) response to chemotherapy in patients with non-small cell lung cancer (NSCLC).

Materials and Methods

Data in a total of 125 patients who had been treated with pemetrexed-based platinum doublet chemotherapy at Cleveland Clinic were retrospectively analyzed. The patients were divided randomly into two sets with the constraint that there were an equal number of responders and nonresponders in the training set. The training set comprised 53 patients with NSCLC, and the validation set comprised 72 patients. A machine learning classifier trained with radiomic texture features extracted from intra- and peritumoral regions of non-contrast-enhanced CT images was used to predict response to chemotherapy. The radiomic risk-score signature was generated by using least absolute shrinkage and selection operator with the Cox regression model; association of the radiomic signature with TTP and OS was also evaluated.

Results

A combination of radiomic features in conjunction with a quadratic discriminant analysis classifier yielded a mean maximum area under the receiver operating characteristic curve (AUC) of 0.82 ± 0.09 (standard deviation) in the training set and a corresponding AUC of 0.77 in the independent testing set. The radiomics signature was also significantly associated with TTP (hazard ratio [HR], 2.8; 95% confidence interval [CI]: 1.95, 4.00; P < .0001) and OS (HR, 2.35; 95% CI: 1.41, 3.94; P = .0011). Additionally, decision curve analysis demonstrated that in terms of clinical usefulness, the radiomics signature had a higher overall net benefit in prediction of high-risk patients to receive treatment than the clinicopathologic measurements.

Conclusion

This study suggests that radiomic texture features extracted from within and around the nodule on baseline CT scans are (a) predictive of response to chemotherapy and (b) associated with TTP and OS for patients with NSCLC.© RSNA, 2019Supplemental material is available for this article.

Perinodular and Intranodular Radiomic Features on Lung CT Images Distinguish Adenocarcinomas from Granulomas

Purpose To evaluate ability of radiomic (computer-extracted imaging) features to distinguish non-small cell lung cancer adenocarcinomas from granulomas at noncontrast CT. Materials and Methods For this retrospective study, screening or standard diagnostic noncontrast CT images were collected for 290 patients (mean age, 68 years; range, 18-92 years; 125 men [mean age, 67 years; range, 18-90 years] and 165 women [mean age, 68 years; range, 33-92 years]) from two institutions between 2007 and 2013. Histopathologic analysis was available for one nodule per patient. Corresponding nodule of interest was identified on axial CT images by a radiologist with manual annotation. Nodule shape, wavelet (Gabor), and texture-based (Haralick and Laws energy) features were extracted from intra- and perinodular regions. Features were pruned to train machine learning classifiers with 145 patients. In a test set of 145 patients, classifier results were compared against a convolutional neural network (CNN) and diagnostic readings of two radiologists. Results Support vector machine classifier with intranodular radiomic features achieved an area under the receiver operating characteristic curve (AUC) of 0.75 on the test set. Combining radiomics of intranodular with perinodular regions improved the AUC to 0.80. On the same test set, CNN resulted in an AUC of 0.76. Radiologist readers achieved AUCs of 0.61 and 0.60, respectively. Conclusion Radiomic features from intranodular and perinodular regions of nodules can distinguish non-small cell lung cancer adenocarcinomas from benign granulomas at noncontrast CT. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Nishino in this issue.

Isolated Pulmonic Valve Endocarditis

BACKGROUND Infective endocarditis (IE) has a high mortality rate, even when treated with appropriate antibiotic therapy and surgical intervention. Right-sided endocarditis is in itself rare, with some studies reporting an incidence of 5-10%. The majority of these cases involve the tricuspid valve, and isolated pulmonary valve endocarditis (PVE) is an extremely rare entity affecting less than 2% of patients with infective endocarditis. Identification and early management are crucial to prevent long-term complications and reduce mortality. CASE REPORT We present a patient with a history of essential hypertension and no underlying valvular disease, who underwent dental cleaning and subsequently developed low-grade fever, myalgia, and malaise. This occurred during the flu season, and was initially diagnosed and treated as flu, without any improvement. The patient was later found to be bacteremic with S. mitis, with no identifiable source, and a normal transthoracic echocardiogram (TTE). He was later hospitalized, had a transesophageal echocardiogram, and was found to have a large pulmonic valve vegetation. CONCLUSIONS This case presents an interesting and rare finding of endocarditis, isolated to the pulmonic valve, in an otherwise healthy individual with no predisposing risk factors. The lack of peripheral stigmata, as well as an unremarkable initial outpatient TTE, made the diagnosis more difficult. It should also be noted that current guidelines do not specifically address right-sided endocarditis, and do not specify the role of surgical intervention.

Mass Effect Deformation Heterogeneity (MEDH) on Gadolinium-contrast T1-weighted MRI is associated with decreased survival in patients with right cerebral hemisphere Glioblastoma: A feasibility study

Subtle tissue deformations caused by mass-effect in Glioblastoma (GBM) are often not visually evident, and may cause neurological deficits, impacting survival. Radiomic features provide sub-visual quantitative measures to uncover disease characteristics. We present a new radiomic feature to capture mass effect-induced deformations in the brain on Gadolinium-contrast (Gd-C) T1w-MRI, and their impact on survival. Our rationale is that larger variations in deformation within functionally eloquent areas of the contralateral hemisphere are likely related to decreased survival. Displacements in the cortical and subcortical structures were measured by aligning the Gd-C T1w-MRI to a healthy atlas. The variance of deformation magnitudes was measured and defined as Mass Effect Deformation Heterogeneity (MEDH) within the brain structures. MEDH values were then correlated with overall-survival of 89 subjects on the discovery cohort, with tumors on the right (n = 41) and left (n = 48) cerebral hemispheres, and evaluated on a hold-out cohort (n = 49 subjects). On both cohorts, decreased survival time was found to be associated with increased MEDH in areas of language comprehension, social cognition, visual perception, emotion, somato-sensory, cognitive and motor-control functions, particularly in the memory areas in the left-hemisphere. Our results suggest that higher MEDH in functionally eloquent areas of the left-hemisphere due to GBM in the right-hemisphere may be associated with poor-survival.

Shape Features of the Lesion Habitat to Differentiate Brain Tumor Progression from Pseudoprogression on Routine Multiparametric MRI: A Multisite Study

BACKGROUND AND PURPOSE

Differentiating pseudoprogression, a radiation-induced treatment effect, from tumor progression on imaging is a substantial challenge in glioblastoma management. Unfortunately, guidelines set by the Response Assessment in Neuro-Oncology criteria are based solely on bidirectional diametric measurements of enhancement observed on T1WI and T2WI/FLAIR scans. We hypothesized that quantitative 3D shape features of the enhancing lesion on T1WI, and T2WI/FLAIR hyperintensities (together called the lesion habitat) can more comprehensively capture pathophysiologic differences across pseudoprogression and tumor recurrence, not appreciable on diametric measurements alone.

MATERIALS AND METHODS

A total of 105 glioblastoma studies from 2 institutions were analyzed, consisting of a training (n = 59) and an independent test (n = 46) cohort. For every study, expert delineation of the lesion habitat (T1WI enhancing lesion and T2WI/FLAIR hyperintense perilesional region) was obtained, followed by extraction of 30 shape features capturing 14 "global" contour characteristics and 16 "local" curvature measures for every habitat region. Feature selection was used to identify most discriminative features on the training cohort, which were evaluated on the test cohort using a support vector machine classifier.

RESULTS

The top 2 most discriminative features were identified as local features capturing total curvature of the enhancing lesion and curvedness of the T2WI/FLAIR hyperintense perilesional region. Using top features from the training cohort (training accuracy = 91.5%), we obtained an accuracy of 90.2% on the test set in distinguishing pseudoprogression from tumor progression.

CONCLUSIONS

Our preliminary results suggest that 3D shape attributes from the lesion habitat can differentially express across pseudoprogression and tumor progression and could be used to distinguish these radiographically similar pathologies.

Tracheomediastinal Fistula: Rare Complication of Treatment with Bevacizumab

Tracheomediastinal fistula is a rare condition caused by multiple etiologies. We present a case of a patient of lung carcinoma receiving chemotherapy. A 63-year-old woman presented to the emergency room with a two-month history of worsening cough and shortness of breath. She was being treated with pemetrexed and bevacizumab for Stage IV non-small cell lung carcinoma. Chest X-ray showed a mass in the lung with mediastinal adenopathy. Computed tomography (CT) scan showed a perforation, confirmed with bronchoscopy. She had a secondary infection and she was started on intravenous antibiotics. The patient decided to continue care in a hospice. We present a rare complication of bevacizumab which has been only reported once in literature. Bevacizumab is known to cause tracheal fistulas when coupled with like invasive procedures. In our case, the patient developed a fistula without any invasive interventions. We advise that physicians using bevacizumab should be aware of the possibility of having such fistulas.

Combination of computer extracted shape and texture features enables discrimination of granulomas from adenocarcinoma on chest computed tomography

Differentiation between benign and malignant nodules is a problem encountered by radiologists when visualizing computed tomography (CT) scans. Adenocarcinomas and granulomas have a characteristic spiculated appearance and may be fluorodeoxyglucose avid, making them difficult to distinguish for human readers. In this retrospective study, we aimed to evaluate whether a combination of radiomic texture and shape features from noncontrast CT scans can enable discrimination between granulomas and adenocarcinomas. Our study is composed of CT scans of 195 patients from two institutions, one cohort for training ([Formula: see text]) and the other ([Formula: see text]) for independent validation. A set of 645 three-dimensional texture and 24 shape features were extracted from CT scans in the training cohort. Feature selection was employed to identify the most informative features using this set. The top ranked features were also assessed in terms of their stability and reproducibility across the training and testing cohorts and between scans of different slice thickness. Three different classifiers were constructed using the top ranked features identified from the training set. These classifiers were then validated on the test set and the best classifier (support vector machine) yielded an area under the receiver operating characteristic curve of 77.8%.

Abstract P4-09-12: Quantitative image features of nuclear and tubule architecture distinguish high and low oncotype DX risk categories of ductal carcinoma in situ from H&E tissue images

Background: Ductal Carcinoma in Situ (DCIS) is a pre-invasive stage of breast cancer, where malignant cells line the duct but have not spread into other parts of the breast. Oncotype DX (ODX) is a genomic test, which divide patients into three risk of recurrence categories (Low, Intermediate, and High) to help physicians determine if patients require adjuvant therapy. However, ODX is expensive, tissue destructive, and has a turnaround time of 7-10 days. There has been an interest in the use of image analysis of routine H&E histopathology slides to predict the course of the disease; the rationale being that the analytics are able to unearth subtle sub-visual cues regarding disease morphology that may escape visual examination. In this work, we evaluate the role of computer extracted features of nuclear morphology and the necrotic regions from surgically resected specimens to predict ODX categories in patients with DCIS.

Methods: H&E slides from breast tissue of 37 patients who were diagnosed with DCIS and underwent a lumpectomy were acquired. Nine of the 37 had high ODX score (higher than 54), while the rest had a low score (lower than 39). All the slides were digitized on a Philips slide scanner. For each image, a watershed algorithm segmented the individual nuclei, which were used to generate 230 nuclei features including nuclear architecture, nuclear shape and nuclear texture features within each candidate breast duct. In addition, we captured the area of necrosis and empty lumen region inside breast ducts to generate features pertaining to tubule packing.

The average feature values for each patient were calculated across all the breast ducts in each slide. A 3-fold cross validation scheme with 50 repetitions was used with the Support Vector Machine (SVM) classifier to predict the ODX risk category for each patient. We used a covariance algorithm to select the top 4 features that were independent of each other but relevant to the ODX class label.

Results: The top ranked features included features from three categories: nuclei architectural features (standard deviation of triangle area in Delaunay graph, skewness of edge length in Cell Cluster Graph), nuclear texture (standard deviation of Haralick matrix intensity) possibly reflecting chromatin patterns in the cell, and the Tubule Packing Ratio, a measure of the ratio of necrosis area and empty lumen area inside the breast ducts compared to the whole breast duct area. The SVM in conjunction with these 4 features yielded a mean area under receive operator characteristic curve (AUC) of 0.95 in correctly predicting high and low ODX risk categories.

Conclusion: We found that our histomorphometry features pertaining to nuclear arrangement, nuclear texture and necrosis could differentiate between DCIS patients with high and low ODX risk categories. Additional independent validation of the approach is needed to confirm the preliminary findings presented here.

Citation Format: Li H, Whitney J, Thawani R, Gilmore H, Badve S, Madabhushi A. Quantitative image features of nuclear and tubule architecture distinguish high and low oncotype DX risk categories of ductal carcinoma in situ from H&E tissue images [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-09-12.

Prediction of recurrence in early stage non-small cell lung cancer using computer extracted nuclear features from digital H&E images

Identification of patients with early stage non-small cell lung cancer (NSCLC) with high risk of recurrence could help identify patients who would receive additional benefit from adjuvant therapy. In this work, we present a computational histomorphometric image classifier using nuclear orientation, texture, shape, and tumor architecture to predict disease recurrence in early stage NSCLC from digitized H&E tissue microarray (TMA) slides. Using a retrospective cohort of early stage NSCLC patients (Cohort #1, n = 70), we constructed a supervised classification model involving the most predictive features associated with disease recurrence. This model was then validated on two independent sets of early stage NSCLC patients, Cohort #2 (n = 119) and Cohort #3 (n = 116). The model yielded an accuracy of 81% for prediction of recurrence in the training Cohort #1, 82% and 75% in the validation Cohorts #2 and #3 respectively. A multivariable Cox proportional hazard model of Cohort #2, incorporating gender and traditional prognostic variables such as nodal status and stage indicated that the computer extracted histomorphometric score was an independent prognostic factor (hazard ratio = 20.81, 95% CI: 6.42–67.52, P < 0.001).

Changes in computer extracted features of vessel tortuosity on CT scans post-treatment in responders compared to non-responders for non-small cell lung cancer on immunotherapy

11518

Background: Immune-checkpoint blockade treatments demonstrate promising clinical efficacy in patients with non-small cell lung cancer (NSCLC). Nivolumab is a PD-1 inhibitor that is FDA approved for treatment of patients with chemotherapy refractory advanced NSCLC. The current standard clinical approach to evaluating tumor response is sub-optimal in defining clinical benefit from immunotherapy drugs. We sought to evaluate whether computer extracted measurements of vessel tortuosity significantly and differentially change post treatment between NSCLC patients who do and do not respond to immunotherapy. Methods: A total of 50 NSCLC patients including pre- and post- treatment CT scans were included in this study. The patients were either responders or non-responders to Nivolumab. Patients who did not receive Nivolumab after 2 cycles due to lack of response or progression as per RECIST were classified as ‘non-responders’. A total of 35 tortuosity features of the vessels around the lung nodules were investigated. In the training cohort (N = 25), the features were ranked based on the degree of change between pre- and post- treatment CT. The top 4 features were used for training a Support Vector Machine (SVM) classifier to identify which patients did and did not respond to immunotherapy on a validation cohort of N = 25 patients. Results: The top features identified were the ones associated with the curvature of the vessel branches. The AUC for the SVM classifier was 0.75 for the training and 0.79 for the test set. Conclusions: Changes in specific vessel tortuosity features between baseline and post-treatment CT scans following nivolumab were different between NSCLC patients who did and did not respond. Multi-site validation of the vessel tortuosity features is needed to establish it as a predictive biomarker for NSCLC patients treated with immunotherapy.

Evolution of radiomic features on serial CT scans as an imaging based biomarker for evaluating response in patients with non-small cell lung cancer treated with nivolumab

e14534

Background: Nivolumab is a PD-1 inhibitor that is FDA approved for treatment of chemotherapy refractory advanced NSCLC. The current standard clinical approach to evaluating tumor response is sub-optimal in evaluating clinical benefit from immunotherapy drugs. Our study aims to explore whether changes in radiomic features of the tumor between baseline and 2-week post-treatment CT scans can predict treatment response. Methods: 41 NSCLC patients treated with nivolumab were included in this study. 22 patients with pre- and post-nivolumab CT scans were used as a learning set and the remaining 19 for independent testing. Patients who did not receive nivolumab after 2 cycles due to lack of response or progression as per RECIST were classified as ‘non-responders’, and patients who had radiological response as per RECIST, or stable disease as per RECIST and clinical improvement were classified as ‘responders’. Lung nodules on pre-treatment CT scans were annotated with 3D SLICER software by a radiologist. 312 texture features of lung nodules were extracted and investigated in the study. The percent difference of each extracted feature was calculated based on the baseline and 2 week post-therapy CT scan. In the learning set, the six features that most significantly changed between baseline and post-treatment scans and also maximally differentially expressed between responders and non-responders were identified. Unsupervised clustering was applied on the set of 6 features for the 19 patients in the test set to predict which patients did and did not respond. Results: The top 6 features predictive of response corresponded to the Haralick, Gabor and Laws texture families. Unsupervised clustering yielded an accuracy of 78.95%. Conclusions: Our results suggest that changes in certain radiomic texture features between baseline and post-treatment CT scans following nivolumab could identify early clinical response to treatment. Additional validation of these novel quantitative imaging based approaches is warranted to accurately define clinical benefit from immunotherapy.

Computer extracted measurements of intra-tumoral heterogeneity on H&E stained tissue images to distinguish short term and long term survivors in patients with non-small cell lung carcinoma

e20052

Background: Molecular and morphologic heterogeneity is an important characteristic of cancer. This spatial and temporal tumor heterogeneity has important implications on tumor behavior and response to therapies. This study aims to evaluate the role of computer extracted features of intra-tumoral heterogeneity (ITH) from digitized whole slide H&E stained images of early stage NSCLC patients treated with surgery as a prognostic marker for survival. Methods: A cohort of 89 early stage NSCLC patients treated with surgery with long term survival data were identified. 28 patients had OS> 3 years from the date of definitive surgery and were defined as long term survivors and 61 patients had OS < 3 years, and were defined as short term survivors. Corresponding H&E stained whole mounted lung tissue images was digitally scanned and a thoracic pathologist marked the primary tumor margins on these images. Our computational approach involved determining the variance in measurements relating to nuclear size, shape, and texture across the tissue section; Each feature was then assigned a morphologic diversity score (MDS) based off the variance; the top predictive MDS features were identified via Wilcoxon Rank Sum Test and then evaluated via a quadratic classifier using 3-fold cross validation. Kaplan-Meier (KM) survival analysis was performed for the ITH features, as well as T- and N-stage. Results: The top ranked MDS features yielded a mean area under the receiver operating characteristic curve (AUC) of 0.66 in discriminating short term from long term survivors. A p=0.00657 (see Table) was obtained for KM-analysis of the ITH features. Conclusions: Computer extracted image features of ITH enabled differentiation of NSCLC patients with short and longer term survival. Large scale multi-site validation will need to be done to establish ITH measurements as a prognostic biomarker for NSCLC patients. [Table: see text]

Computer extracted features of gland morphology on H&E surgically resected tissue images as predictive of biochemical recurrence and rate of expression in African American compared to Caucasian American men

e16559

Background: Prostate cancer (PCa) is 1.7 times more prevalent and 2.3 times more lethal in African American (AA) men than in Caucasian American (CA) men. Additionally there is a higher incidence (when controlling for relative population size) of more aggressive PCa in the AA compared to the CA population which has led a number of groups to investigate whether there are molecular differences in the disease phenotype between the two groups. In this work we investigate whether computer extracted features of glandular morphology from digitized pathology images of surgical specimens identified as predictive of biochemical recurrence are also differentially expressed between AA and CA patients. Methods: Two cohorts were gathered. Cohort 1 (C1) contained digitized whole mount Gleason 7 radical prostatectomy specimens (RPS) from 72 patients (46 CA, 26 AA). Cohort 2 (C2) comprising quarter and whole mount RPS from 145 patients (135 CA, 10 AA), who either had biochemical recurrence in 5 years (47) or did not (98). A pathologist annotated each image in both cohorts for a representative cancerous region. Glands were automatically segmented and 216 features describing gland arrangement, shape, and disorder were extracted from every image. Features were tested for significance by Wilcoxon rank sum test at p < .05. Results: 8 features of gland morphology were significantly different between AA and CA groups and also predictive of biochemical recurrence (Table). Conclusions: We identified 8 computer extracted features of gland morphology that were both differentially expressed between AA and CA men and predictive of biochemical recurrence, suggesting that there are histomorphometric differences in prostate cancer appearance between AA and CA men. [Table: see text]

Computer extracted measurements of vessel tortuosity on baseline CT scans to predict response to nivolumab immunotherapy for non-small cell lung cancer

11566

Background: Immune-checkpoint blockade treatments, particularly drugs targeting the programmed death-1 (PD-1) receptor, demonstrate promising clinical efficacy in patients with non-small cell lung cancer (NSCLC). We sought to evaluate whether computer extracted measurements of tortuosity of vessels in lung nodules on baseline CT scans in NSCLC patients(pts) treated with a PD-1 inhibitor, nivolumab could distinguish responders and non-responders. Methods: A total of 61 NSCLC pts who underwent treatment with nivolumab were included in this study. Pts who did not receive nivolumab after 2 cycles due to lack of response or progression per RECIST were classified as ‘non-responders’, patients who had radiological response per RECIST or had clinical benefit (defined as stable disease >10 cycles) were classified as ‘responders’. A total of 35 quantitative tortuosity features of the vessels associated with lung nodule were investigated. In the training cohort (N=33), the features were ranked in their ability to identify responders to nivolumab using a support vector machine (SVM) classifier. The three most informative features were then used for training the SVM, which was then validated on a cohort of N=28 pts. Results: The maximum curvature ( f1), standard deviation of the torsion ( f2) and mean curvature ( f3) were identified as the most discriminating features. The area under Receiver operating characteristic (ROC) curve (AUC) of the SVM was 0.84 for the training and 0.72 for the validation cohort. Conclusions: Vessel tortuosity features were able to distinguish responders from non-responders for patients with NSCLC treated with nivolumab. Large scale multi-site validation will need to be done to establish vessel tortuosity as a predictive biomarker for immunotherapy. [Table: see text]

Computer extracted shape features of prostate capsule from MRI to predict biochemical recurrence of prostate cancer post-treatment

e16579

Background: Pre-treatment identification of biochemical recurrence (BCR) from MRI may enable the use of aggressive neo-adjuvant therapies for prostate cancer patients to improve prognosis. BCR is often associated with aggressive cancer growth and/or extra prostatic extension resulting in an irregular bulge and focal capsular retraction. This may induce differences in the shape of the prostate capsule between BCR positive (BCR+) and BCR negative (BCR-) patients as observed on MRI. In this work, we show that computer extracted shape features of the prostate capsule on MRI can identify patients that are at a risk of BCR post-treatment Methods: In a single centre IRB approved study, from a registry of 874 patients, availability of complete image datasets (T1w, T2w and ADC map); no treatment for PCa before MRI; presence of clinically localised PCa; availability of Gleason score; and data available for post-treatment PSA and follow-up for at least 3 years in patients without BCR were used as inclusion criteria to select 80 patients. The prostate capsule was manually segmented on T2w MRI by an experienced radiologist. Two atlases A+ and A- were created for BCR+ and BCR- patients respectively with similar Gleason score (6 to 9), similar numbers in each cohort (25 each) and similar tumor stage (T2 to T3). A t-test based analysis corrected for multiple comparison revealed statistically significant prostate shape differences between A+ and A- in surface of interest (SOI). Curvature features (magnitude and surface normal orientations) were extracted from SOI of the two cohorts. A random forest classifier was trained using the 50 training images (from A+ and A-) and validated using a hold-out validation set of 30 patients. Results: The inter-quartile range, variance, skewness and kurtosis of curvature magnitude and normal orientations were found to be predictive of BCR. The RF classifier trained using these features could predict BCR with an accuracy of 78% and an AUC of 0.71 in the validation set. Conclusions: Curvature magnitude and orientation features of the prostate capsule from the SOI may be predictive of BCR. In future a multi centre independent datasets will be used to validate the findings.

Co-registration of pre-operative CT with ex vivo surgically excised ground glass nodules to define spatial extent of invasive adenocarcinoma on in vivo imaging: a proof-of-concept study

OBJECTIVE

To develop an approach for radiology-pathology fusion of ex vivo histology of surgically excised pulmonary nodules with pre-operative CT, to radiologically map spatial extent of the invasive adenocarcinomatous component of the nodule.

METHODS

Six subjects (age: 75 ± 11 years) with pre-operative CT and surgically excised ground-glass nodules (size: 22.5 ± 5.1 mm) with a significant invasive adenocarcinomatous component (>5 mm) were included. The pathologist outlined disease extent on digitized histology specimens; two radiologists and a pulmonary critical care physician delineated the entire nodule on CT (in-plane resolution: <0.8 mm, inter-slice distance: 1-5 mm). We introduced a novel reconstruction approach to localize histology slices in 3D relative to each other while using CT scan as spatial constraint. This enabled the spatial mapping of the extent of tumour invasion from histology onto CT.

RESULTS

Good overlap of the 3D reconstructed histology and the nodule outlined on CT was observed (65.9 ± 5.2%). Reduction in 3D misalignment of corresponding anatomical landmarks on histology and CT was observed (1.97 ± 0.42 mm). Moreover, the CT attenuation (HU) distributions were different when comparing invasive and in situ regions.

CONCLUSION

This proof-of-concept study suggests that our fusion method can enable the spatial mapping of the invasive adenocarcinomatous component from 2D histology slices onto in vivo CT.

KEY POINTS

• 3D reconstructions are generated from 2D histology specimens of ground glass nodules. • The reconstruction methodology used pre-operative in vivo CT as 3D spatial constraint. • The methodology maps adenocarcinoma extent from digitized histology onto in vivo CT. • The methodology potentially facilitates the discovery of CT signature of invasive adenocarcinoma.