Virtual Library
Start Your Search
Andre Moreira
Author of
-
+
MS05 - Diagnostic Dilemma in Lung Cancer (ID 784)
- Event: WCLC 2018
- Type: Mini Symposium
- Track: Pathology
- Presentations: 1
- Moderators:
- Coordinates: 9/24/2018, 13:30 - 15:00, Room 201 BD
-
+
MS05.04 - Diagnosis and Classification in Biopsies (ID 11422)
14:30 - 14:50 | Presenting Author(s): Andre Moreira
- Abstract
- Presentation
Abstract not provided
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
OA03 - Advances in Lung Cancer Pathology (ID 897)
- Event: WCLC 2018
- Type: Oral Abstract Session
- Track: Pathology
- Presentations: 1
- Moderators:
- Coordinates: 9/24/2018, 10:30 - 12:00, Room 205 BD
-
+
OA03.03 - Phase 2B of Blueprint PD-L1 Immunohistochemistry Assay Comparability Study (ID 14530)
10:50 - 11:00 | Author(s): Andre Moreira
- Abstract
- Presentation
Background
PD-L1 immunohistochemistry (IHC) has been established as companion or complementary diagnostic assays, each developed as predictive biomarker for specific anti PD1/PD-L1 immunotherapies. The Blueprint (BP) phase 1 comparability study demonstrated that three PD-L1 assays (28-8, 22C3, SP263) showed comparable analytical performance for assessment of PD-L1 expression on tumor cells (TPS), while the SP-142 PD-L1 assay appeared to stain a lower percentage of tumor cells when compared to the other assays. The first part of BP phase 2 (BP2A) re-affirmed these findings in a larger cohort of ‘real life’ specimens scored by 24 experienced pulmonary pathologists, and also showed that the 73-10 assay developed for avelumab showed greater sensitivity than all other assays to detect PD-L1 on tumour cells. BP2A also demonstrated generally excellent inter-observer agreement for tumor cell PD-L1 scoring using both glass slides and digital images, with slightly lesser agreement for the cytology samples included in the study cohort. Inter-observer agreement for immune cell scoring on glass or digital slides was poor. Phase 2B of Blueprint (BP2B) aimed to compare PD-L1 scoring on triplet samples representing large tumor resection blocks, small biopsy samples and fine needle aspirate cell blocks prepared from the same tumor. a9ded1e5ce5d75814730bb4caaf49419 Method
Triplet samples of large resected tumor block, small biopsy sample and fine needle aspirate cell block (the latter two taken from the resected tumour specimen) were gathered from 31 resected primary lung cancers (17 adenocarcinomas, 12 squamous cell carcinomas, and 2 large cell carcinomas). Sections from all 93 blocks were stained with the pharmDx 28-8 and 22C3, the FDA-approved SP142 and SP263, or clinical trial associated 73-10 PD-L1 assays, in a CLIA-approved immunohistochemistry laboratory. All H&E and PD-L1 IHC slides were scanned and digital images were used to score all cases by the same 24 pathologists involved in BP2A. As before, tumor cells PD-L1 staining were scored as continuous variable and into 7 cut-off-defined categories, as used in various immune checkpoint inhibitor trials. Immune cells were not scored. 4c3880bb027f159e801041b1021e88e8 Result
The data reaffirm the relative comparability of 28-8, 22C3 and SP263 assays across the range of scores; SP142 assay scores were lower, those for 73-10 higher. Inter-observer agreement between readers ranged from moderate to near perfect (Kappa-Fleiss (K-F) scores generally >0.7); best overall agreement was on aspirates. Overall, the agreement between scores on the different sample types from the same tumor was good (most K-F scores >0.7); aspirates showed no significant difference from biopsy samples or whole surgical blocks. In contrast to biopsies and surgical blocks, scores could, however, not be rendered in about 14% of aspirate sections. 8eea62084ca7e541d918e823422bd82e Conclusion
The results of BP2B confirms earlier results and also demonstrate comparable performance for fine needle aspirates in those cases where TPS scores were possible. 6f8b794f3246b0c1e1780bb4d4d5dc53Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
P1.09 - Pathology (Not CME Accredited Session) (ID 941)
- Event: WCLC 2018
- Type: Poster Viewing in the Exhibit Hall
- Track:
- Presentations: 1
- Moderators:
- Coordinates: 9/24/2018, 16:45 - 18:00, Exhibit Hall
-
+
P1.09-32 - Classification and Mutation Prediction from Non-Small Cell Lung Cancer Histopathology Images Using Deep Learning (ID 12631)
16:45 - 18:00 | Author(s): Andre Moreira
- Abstract
Background
Visual inspection of histopathology slides of lung tissues is one of the primary methods used by pathologists to assess pathological stage and classifications of lung cancer. Adenocarcinoma and squamous cell carcinoma are the two most prevalent types of non-small cell lung cancer (NSCLC), but their distinction can be challenging and time-consuming even for the expert eye. Furthermore, the increasing prevalence of new treatment modalities relies on classification of NSCLC mutational analyses of the tumor to guide targeted therapy and provide prognostic information.
a9ded1e5ce5d75814730bb4caaf49419 Method
In this study, we trained a deep convolutional neural network (CNN) model (inception v3) on histology images obtained from The Cancer Genome Atlas (TCGA) to accurately and automatically classify whole-slide images into adenocarcinoma, squamous cell carcinoma or normal lung tissue. Furthermore, we trained the neural network to predict the ten most commonly mutated genes in lung adenocarcinoma. Our model was then validated on independent datasets of images obtained from both frozen and formalin-fixed paraffin-embedded tissue, including resection and biopsy specimens.
4c3880bb027f159e801041b1021e88e8 Result
We found that the performance of our method is comparable, in terms of sensitivity and specificity, to that of pathologists, with a 0.97 average Area Under the Curve (AUC) on a held-out population of whole-slide images for the sub-type classification on frozen sections. Additionally, we found that six of the ten most commonly mutated genes – STK11, EGFR, FAT1, SETBP1, KRAS and TP53 – can be predicted by the algorithm solely using information from histologic images with an accuracy ranging from 0.733 to 0.856, as measured by the AUC on the held-out population.
8eea62084ca7e541d918e823422bd82e Conclusion
These findings suggest that deep learning models can offer both pathologists and patients a fast, accurate, and inexpensive classification of NSCLC or gene mutations, and thus have a significant impact on cancer treatment. Our computational approach can be applied to any cancer type and the code is available as free open-source software at https://github.com/ncoudray/DeepPATH.
6f8b794f3246b0c1e1780bb4d4d5dc53
