Grading of lung adenocarcinomas with simultaneous segmentation by artificial intelligence (GLASS-AI)

Preclinical genetically engineered mouse models (GEMMs) of lung adenocarcinoma are invaluable for investigating molecular drivers of tumor formation, progression, and therapeutic resistance. However, histological analysis of these GEMMs requires significant time and training to ensure accuracy and consistency. To achieve a more objective and standardized analysis, we used machine learning to create GLASS-AI, a histological image analysis tool that the broader cancer research community can utilize to grade, segment, and analyze tumors in preclinical models of lung adenocarcinoma. GLASS-AI demonstrates strong agreement with expert human raters while uncovering a significant degree of unreported intratumor heterogeneity. Integrating immunohistochemical staining with high-resolution grade analysis by GLASS-AI identified dysregulation of Mapk/Erk signaling in high-grade lung adenocarcinomas and locally advanced tumor regions. Our work demonstrates the benefit of employing GLASS-AI in preclinical lung adenocarcinoma models and the power of integrating machine learning and molecular biology techniques for studying the molecular pathways that underlie cancer progression.

NIMG-01. PREDICTING POST-STEREOTACTIC RADIOTHERAPY MAGNETIC RESONANCE IMAGE OUTCOMES OF BREAST CANCER METASTASES TO THE BRAIN

BACKGROUND

Stereotactic radiosurgery (SRS) is a cornerstone in the management of Breast Cancer Metastases to the Brain (BCMB). While control rates are high following SRS, radiation necrosis is a rare but potentially devastating long-term toxicity. There is a clinical need for automated/semi-automated methods to assess tumor response and optimize the RT plans for local control with minimal long-term toxicity. Multiparametric MRI (mpMRI), particularly Apparent Diffusion Coefficient of water (ADC) maps, contain information that is mechanistically relatable to voxel-level tumor response to RT. We report a deep learning-based approach to predict post-SRS ADC maps, FLAIR, T2-weighted (T2W), T1-weighted unenhanced (T1W) and contrast-enhanced (T1WCE) images, from pre-SRS T1W, T1WCE, T2W and FLAIR images, ADC maps, and the delivered RT dose map. These “forward models” will enable the radiation oncologist to simulate radiologic outcomes and iteratively optimize RT plans for local control with minimal toxicity.

METHODS

We trained a variant of the pix2pix Generative Adversarial Network (GAN) on MRI and RT dose map data from 18 BCMB patients treated with stereotactic radiation with confirmed controlled and locally recurrent metastases. Patients were treated with stereotactic radiation dose of 1-40 Gy between 2013-2019.

RESULTS

On test data from 6 BCMB patients, the trained forward model predicted post-SRS ADC values within the Gross Tumor Volume (GTV) that were broadly in agreement with ground truth post-SRS ADC maps. In agreement with expectations, the forward model also predicts increasing post-RT ADC within the GTV with increasing simulated RT doses in the range of 1-71 Gy. We have also explored an inverse model to predict the RT dose map required to produce “prescribed” post-SRS ADC values within the GTV.

CONCLUSIONS

We envision that the forward models will assist the radiation oncologist in initial RT dose plan optimization, while the inverse model may be useful for daily RT plan optimization.

Abstract 5886: Pre-treatment adiposity measured by computed tomography and survival of women with high-grade serous ovarian cancer

The association of body mass index (BMI) with survival of women with ovarian cancer remains unclear due to mixed epidemiological evidence. This may be due, in part, to the fact that BMI is an imperfect measure of body fat as BMI does not distinguish weight from lean muscle versus adipose tissue. Here, we investigated the association of adiposity measured by computed tomography (CT) with survival among the most common histotype of ovarian cancer, high-grade serous ovarian cancer (HGSOC). The present study included 383 women diagnosed with HGSOC from 2008 to 2019 who were evaluated at H. Lee Moffitt Cancer Center and Research Institute and had pre-treatment computed tomography scans available for analysis. The sliceOmatic v5.0 rev13 (Tomovision, Magog, Canada) medical image analysis software and accompanying ABACS module for segmentation was used to quantify subcutaneous (SAT), visceral (VAT), and intermuscular adipose tissue (IMAT) from the third lumbar (L3) axial slice including the transverse processes. We used Cox proportional hazard regression to estimate hazard ratios (HR) and 95% confidence intervals (CIs) for the association of each measure of adiposity with overall survival (OS) and recurrence-free survival (RFS) while adjusting for age at diagnosis, stage, race and ethnicity, and first-line treatment. The degree of ascites was included in the VAT models as ascites fluid density can mask VAT. We also assessed these associations within first-line treatment groups (upfront chemotherapy [n=147], upfront surgery [n=236]). In the overall study population, we observed a positive but not statistically significant association with OS and RFS for the highest vs. lowest tertile of IMAT (HR= 1.18, 95% CI=0.83, 1.67 and HR=1.16, 95% CI=0.85, 1.58, respectively). Among women who received upfront surgery, the highest tertile of IMAT was associated with a 57% increased risk of recurrence compared to the lowest tertile (HR=1.57, 95% CI=1.04, 2.37), while the association between IMAT and OS was similar to the findings in the overall population (HR=1.14, 95% CI=0.73, 1.78). No association was observed between IMAT and OS or RFS among women who received upfront chemotherapy. No associations with OS or RFS were observed for SAT or VAT overall or within first-line treatment groups. In summary, we observed inferior RFS among HGSOC patients with higher IMAT. These findings suggest that IMAT measured from standard-of-care imaging may represent a biomarker of recurrence among HGSOC patients, and incorporating lifestyle and behavioral changes (e.g., diet, exercise) to decrease IMAT may be warranted for this patient population.

Citation Format: Christelle Colin-Leitzinger, Daniel Jeong, Mahmoud Abdalah, Rikki Cannioto, Jing-Yi Chern, Evan Davis, Robert Gillies, Melissa McGettigan, Jaileene Perez-Morales, Natarajan Raghunand, Sweta Sinha, Olya Stringfield, Rajwantee Tirbene, Matthew Schabath, Lauren C. Peres. Pre-treatment adiposity measured by computed tomography and survival of women with high-grade serous ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5886.

Quantitative Changes in Intratumoral Habitats on MRI Correlate With Pathologic Response in Early-stage ER/PR+ HER2- Breast Cancer Treated With Preoperative Stereotactic Ablative Body Radiotherapy

Objective

To quantitatively evaluate intratumoral habitats on dynamic contrast-enhanced (DCE) breast MRI to predict pathologic breast cancer response to stereotactic ablative body radiotherapy (SABR).

Methods

Participants underwent SABR treatment (28.5 Gy x3), baseline and post-SABR MRI, and breast-conserving surgery for ER/PR+ HER2- breast cancer. MRI analysis was performed on DCE T1-weighted images. MRI voxels were assigned eight habitats based on high (H) or low (L) maximum enhancement and the sequentially numbered dynamic sequence of maximum enhancement (H1-4, L1-4). MRI response was analyzed by percent tumor volume remaining (%VR = volume post-SABR/volume pre-SABR), and percent habitat makeup (%HM of habitat X = habitat X voxels/total voxels in the segmented volume). These were correlated with percent tumor bed cellularity (%TC) for pathologic response.

Results

Sixteen patients completed the trial. The %TC ranged 20%-80%. MRI %VR demonstrated strong correlations with %TC (Pearson R = 0.7-0.89). Pre-SABR tumor %HMs differed significantly from whole breasts (P = 0.005 to <0.00001). Post-SABR %HM of tumor habitat H4 demonstrated the largest change, increasing 13% (P = 0.039). Conversely, combined %HM for H1-3 decreased 17% (P = 0.006). This change correlated with %TC (P < 0.00001) and distinguished pathologic partial responders (≤70 %TC) from nonresponders with 94% accuracy, 93% sensitivity, 100% specificity, 100% positive predictive value, and 67% negative predictive value.

Conclusion

In patients undergoing preoperative SABR treatment for ER/PR+ HER2- breast cancer, quantitative MRI habitat analysis of %VR and %HM change correlates with pathologic response.

MRI Response to Pre-operative Stereotactic Ablative Body Radiotherapy (SABR) in Early Stage ER/PR+ HER2- Breast Cancer correlates with Surgical Pathology Tumor Bed Cellularity

OBJECTIVE

This study evaluates breast MRI response of ER/PR+ HER2- breast tumors to pre-operative SABR with pathologic response correlation.

METHODS

Women enrolled in a phase 2 single institution trial of SABR for ER/PR+ HER2- breast cancer were retrospectively evaluated for radiologic-pathologic correlation of tumor response. These patients underwent baseline breast MRI, SABR (28.5 Gy in 3 fractions), follow-up MRI 5 to 6 weeks post-SABR, and lumpectomy. Tumor size and BI-RADS descriptors on pre and post-SABR breast MRIs were compared to determine correlation with surgical specimen % tumor cellularity (%TC). Reported MRI tumor dimensions were used to calculate percent cubic volume remaining (%VR). Partial MRI response was defined as a BI-RADs descriptor change or %VR ≤ 70%, while partial pathologic response (pPR) was defined as %TC ≤ 70%.

RESULTS

Nineteen patients completed the trial, and %TC ranged 10% to 80%. For BI-RADS descriptor analysis, 12 of 19 (63%) showed change in lesion or kinetic enhancement descriptors post-SABR. This was associated with lower %TC (29% vs. 47%, P = .042). BI-RADS descriptor change analysis also demonstrated high PPV (100%) and specificity (100%) for predicting pPR to treatment (sensitivity 71%, accuracy 74%), but low NPV (29%). MRI %VR demonstrated strong linear correlation with %TC (R = 0.70, P < .001, Pearson's Correlation) and high accuracy (89%) for predicting pPR (sensitivity 88%, specificity 100%, PPV 100%, and NPV 50%).

CONCLUSION

Evaluating breast cancer response on MRI using %VR after pre-operative SABR treatment can help identify patients benefiting the most from neoadjuvant radiation treatment of their ER/PR+ HER2- tumors, a group in which pCR to neoadjuvant therapy is rare.

Standardization in Quantitative Imaging: A Multicenter Comparison of Radiomic Features from Different Software Packages on Digital Reference Objects and Patient Data Sets

Radiomic features are being increasingly studied for clinical applications. We aimed to assess the agreement among radiomic features when computed by several groups by using different software packages under very tightly controlled conditions, which included standardized feature definitions and common image data sets. Ten sites (9 from the NCI's Quantitative Imaging Network] positron emission tomography–computed tomography working group plus one site from outside that group) participated in this project. Nine common quantitative imaging features were selected for comparison including features that describe morphology, intensity, shape, and texture. The common image data sets were: three 3D digital reference objects (DROs) and 10 patient image scans from the Lung Image Database Consortium data set using a specific lesion in each scan. Each object (DRO or lesion) was accompanied by an already-defined volume of interest, from which the features were calculated. Feature values for each object (DRO or lesion) were reported. The coefficient of variation (CV), expressed as a percentage, was calculated across software packages for each feature on each object. Thirteen sets of results were obtained for the DROs and patient data sets. Five of the 9 features showed excellent agreement with CV < 1%; 1 feature had moderate agreement (CV < 10%), and 3 features had larger variations (CV ≥ 10%) even after attempts at harmonization of feature calculations. This work highlights the value of feature definition standardization as well as the need to further clarify definitions for some features.

Simulation-based evaluation of school reopening strategies during COVID-19: A case study of São Paulo, Brazil

During the coronavirus disease 2019 (COVID-19) pandemic, many countries opted for strict public health measures, including closing schools. After some time, they have started relaxing some of those restrictions. To avoid overwhelming health systems, predictions for the number of new COVID-19 cases need to be considered when choosing a school reopening strategy. Using a computer simulation based on a stochastic compartmental model that includes a heterogeneous and dynamic network, we analyse different strategies to reopen schools in the São Paulo Metropolitan Area, including one similar to the official reopening plan. Our model allows us to describe different types of relations between people, each type with a different infectiousness. Based on our simulations and model assumptions, our results indicate that reopening schools with all students at once has a big impact on the number of new COVID-19 cases, which could cause a collapse of the health system. On the other hand, our results also show that a controlled school reopening could possibly avoid the collapse of the health system, depending on how people follow sanitary measures. We estimate that postponing the schools' return date for after a vaccine becomes available may save tens of thousands of lives just in the São Paulo Metropolitan Area compared to a controlled reopening considering a worst-case scenario. We also discuss our model constraints and the uncertainty of its parameters.

Abstract PO-082: Automated tumor segmentation, grading, and analysis of tumor heterogeneity in preclinical models of lung adenocarcinoma

Preclinical mouse models of lung adenocarcinoma are invaluable for the discovery of molecular drivers of tumor formation, progression, and therapeutic resistance. Histological analyses of these preclinical models require significant investments of time and training to ensure accuracy and consistency. Analysis by a clinical pathologist is the gold standard in this approach, but may be difficult to obtain due to the cost and availability of their services. As an alternative we have developed a digital pathology tool to identify, segment, grade, and analyze tumors in mouse models of lung adenocarcinoma. This convolutional neural network (CNN) model, based on ResNet18, was trained to classify normal lung tissue, normal airways, and the different grades (1 – 4) of lung adenocarcinoma from 100,000 224 × 224 pixel image patches (~16,000 patches per class). Our training dataset was constructed from whole slide images of hematoxylin and eosin stained lung sections from 4 different mouse models of lung adenocarcinoma with oncogenic Kras (KrasG12D/+), in combination with oncogenic p53 mutations (KrasG12D/+; p53R172H/+ and KrasG12D/+;p53R270H/+), or with the loss of the tumor suppressive TAp73 (KrasG12D/+;TAp7fltd/fltd). Our CNN demonstrated a strong correspondence with human pathologists on our holdout dataset, achieving a micro-F1 score of 0.81 on a pixel-by-pixel basis. As a test of our CNN, we analyzed two mouse models to better understand the role of TAp73 in lung adenocarcinoma: KrasG12D/+ (“K”) and KrasG12D/+;TAp73fltd/fltd (“TK”). Both human raters and our CNN reported a significant increase in the tumor burden of the compound mutant “TK” mice compared to the single mutant “K” mice. According to our CNN, this increased tumor burden was driven primarily by an increase in tumor size and not an increased number of tumors in “TK” mice. Because our CNN can assign different grades to regions within the same image patch and tumor, we also uncovered a high degree of intratumor heterogeneity that was not reported by the human pathologists, who are trained to assign one grade to a single tumor with a bias for the highest grade present in a given tumor. The finer grading resolution allowed our CNN to uncover the increased tumor size observed in the “TK” mice was due to expansion of Grade 2 regions (characterized by enlarged nuclei without irregular shape) within tumors that would be considered a higher grade by pathologists. Our CNN also provides a detailed map of tumor grades overlaid on the H&E images used for analysis, allowing for precise targeting of regions within tumors with other assays. We are currently utilizing these outputs in conjunction with other assays, such as spatial transcriptomic analysis and immunohistochemistry, to investigate the molecular mechanisms that underlie the expansion of Grade 2 tumor regions in “TK” mice. Future work will expand this tool into a multidimensional digital pathology pipeline that can accelerate current investigations and reveal new therapeutic targets and prognostic markers.

Citation Format: John H. Lockhart, Hayley D. Ackerman, Kyubum Lee, Mahmoud Abdalah, Andrew Davis, Nicole Montey, Theresa Boyle, James Saller, Ayensur Keske, Kay Hänggi, Brian Ruffell, Olya Stringfield, Aik Choon Tan, Elsa R. Flores. Automated tumor segmentation, grading, and analysis of tumor heterogeneity in preclinical models of lung adenocarcinoma [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-082.

Novel clinical and radiomic predictors of rapid disease progression phenotypes among lung cancer patients treated with immunotherapy: An early report

OBJECTIVES

Immune-checkpoint blockades have exhibited durable responses and improved long-term survival in a subset of advanced non-small cell lung cancer (NSCLC) patients. However, highly predictive markers of positive and negative responses to immunotherapy are a significant unmet clinical need. The objective of this study was to identify clinical and computational image-based predictors of rapid disease progression phenotypes in NSCLC patients treated with immune-checkpoint blockades.

MATERIALS AND METHODS

Using time-to-progression (TTP) and/or tumor growth rates, rapid disease progression phenotypes were developed including hyperprogressive disease. The pre-treatment baseline predictors that were used to identify these phenotypes included patient demographics, clinical data, driver mutations, hematology data, and computational image-based features (radiomics) that were extracted from pre-treatment computed tomography scans. Synthetic Minority Oversampling Technique (SMOTE) was used to subsample minority groups to eliminate classification bias. Patient-level probabilities were calculated from the final clinical-radiomic models to subgroup patients by progression-free survival (PFS).

RESULTS

Among 228 NSCLC patients treated with single agent or double agent immunotherapy, we identified parsimonious clinical-radiomic models with modest to high ability to predict rapid disease progression phenotypes with area under the receiver-operator characteristics ranging from 0.804 to 0.865. Patients who had TTP < 2 months or hyperprogressive disease were classified with 73.41% and 82.28% accuracy after SMOTE subsampling, respectively. When the patient subgroups based on patient-level probabilities were analyzed for survival outcomes, patients with higher probability scores had significantly worse PFS.

CONCLUSIONS

The models found in this study have potential important translational implications to identify highly vulnerable NSCLC patients treated with immunotherapy that experience rapid disease progression and poor survival outcomes.

Predicting clinically significant prostate cancer using DCE-MRI habitat descriptors

Prostate cancer diagnosis and treatment continues to be a major public health challenge. The heterogeneity of the disease is one of the major factors leading to imprecise diagnosis and suboptimal disease management. The improved resolution of functional multi-parametric magnetic resonance imaging (mpMRI) has shown promise to improve detection and characterization of the disease. Regions that subdivide the tumor based on Dynamic Contrast Enhancement (DCE) of mpMRI are referred to as DCE-Habitats in this study. The DCE defined perfusion curve patterns on the identified tumor habitat region are used to assess clinical significance. These perfusion curves were systematically quantified using seven features in association with the patient biopsy outcome and classifier models were built to find the best discriminating characteristics between clinically significant and insignificant prostate lesions defined by Gleason score (GS). Multivariable analysis was performed independently on one institution and validated on the other, using a multi-parametric feature model, based on DCE characteristics and ADC features. The models had an intra institution Area under the Receiver Operating Characteristic (AUC) of 0.82. Trained on Institution I and validated on the cohort from Institution II, the AUC was also 0.82 (sensitivity 0.68, specificity 0.95).

Comparison of sparse domain approaches for 4D SPECT dynamic image reconstruction

PURPOSE

Dynamic imaging (DI) provides additional diagnostic information in emission tomography in comparison to conventional static imaging at the cost of being computationally more challenging. Dynamic single photon emission computed tomography (SPECT) reconstruction is particularly difficult because of the limitations in the sampling geometry present in most existing scanners. We have developed an algorithm Spline Initialized Factor Analysis of Dynamic Structures (SIFADS) that is a matrix factorization method for reconstructing the dynamics of tracers in tissues and blood directly from the projections in dynamic cardiac SPECT, without first resorting to any 3D reconstruction.

METHODS

SIFADS is different from "pure" factor analysis in dynamic structures (FADS) in that it employs a dedicated spline-based pre-initialization. In this paper, we analyze the convergence properties of SIFADS and FADS using multiple metrics. The performances of the two approaches are evaluated for numerically simulated data and for real dynamic SPECT data from canine and human subjects.

RESULTS

For SIFADS, metrics analyzed for reconstruction algorithm convergence show better features of the metric curves vs iterations. In addition, SIAFDS provides better tissue segmentations than that from pure FADS. Measured computational times are also typically better for SIFADS implementations than those with pure FADS.

CONCLUSION

The analysis supports the utility of the pre-initialization of a factorization algorithm for better dynamic SPECT image reconstruction.

Reconstruction of 4-D dynamic SPECT images from inconsistent projections using a Spline initialized FADS algorithm (SIFADS)

In this paper, we propose and validate an algorithm of extracting voxel-by-voxel time activity curves directly from inconsistent projections applied in dynamic cardiac SPECT. The algorithm was derived based on factor analysis of dynamic structures (FADS) approach and imposes prior information by applying several regularization functions with adaptively changing relative weighting. The anatomical information of the imaged subject was used to apply the proposed regularization functions adaptively in the spatial domain. The algorithm performance is validated by reconstructing dynamic datasets simulated using the NCAT phantom with a range of different input tissue time-activity curves. The results are compared to the spline-based and FADS methods. The validated algorithm is then applied to reconstruct pre-clinical cardiac SPECT data from canine and murine subjects. Images, generated from both simulated and experimentally acquired data confirm the ability of the new algorithm to solve the inverse problem of dynamic SPECT with slow gantry rotation.