Start Your Search
JICC01 - Joint IASLC-CAALC-CSCO Session: The Truth and Myth of Oral Anti-VEGFR Inhibitors for Advance NSCLC (ID 276)
- Event: WCLC 2020
- Type: Workshop
- Track: N.A.
- Presentations: 1
P38 - Pathology - Pathology/Staging (ID 108)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
P38.06 - Evaluation of Non-invasive Transcutaneous Bioconductance Measurement - a Risk-Stratification Biomarker for Suspicious Pulmonary Lesions (ID 2263)
00:00 - 00:00 | Presenting Author(s): Dawei Yang
By conducting low dose CT screening program in the high-risk population, a large number of pulmonary nodules have been discovered by the radiology image. There is an urgent need for risk stratification biomarkers for pre-interventional assessment of growing numbers of Indeterminate Pulmonary Lesions (IPL). In the previous study, we found the bioconductance through normal lung parenchyma differs from pathologic lesions, including inflammation and malignant tumors. Using non-invasive multi-point and multiparametric measurements of transcutaneous bioconductance, the previous single-center open-labeled feasibility trial (JTO 2012;7:681) developed an algorithm with high-performance AUC/accuracy 90%. Subsequent prospective multi-center trials in US (NCT01566682) and China had lower performances (60% accuracy) attributed to data overfitting and inconsistent human performances. In this abstract, we presented the latest multi-center prospective trial (NCT0156668; China reg20170226) evaluating revised Prolung test protocol.Methods
Institutions IRB approved the recruitment of subjects undergoing multidisciplinary evaluation of IPLs, undergo single Prolung Test measurement before biopsies. Protocol modified from previous studies; briefly only 20 surface points interrogated instead of 62, 3X measurement plus within-patient calibration improved measurement consistency. Operator training and criteria for subjects’ enrollment also modified. Collected bioconductance data is analyzed on a modified algorithm, and binary malignancy prediction score (increased/decreased) generated and locked prior to follow-up interventions. In this non-interventional study, the result of prediction is withheld from clinical investigators prior to the final diagnosis.Results
418 of 486(86%) enrolled subjects evaluable for the outcome. Demographic distribution: male:female 44%/56%. Age range 3rd-9th decades with significant clustering around the 5th-6th decades (64%). Tissue diagnostic modalities are by surgery (193=46%) and bronchoscopy (84=20%), non-diagnostic biopsies followed by radiology(141=34%). Based on the pathological diagnosis, it includes 221(52.9%) cancers and 197(47.1%) benign. Cancer cell type predominantly adenocarcinoma (84.7%), squamous ca (9.1%), other cell types (6.2%). Test performance (Figure 1) demonstrates Sensitivity 84.2%, Specificity 73.6%, overall accuracy 79.2%, PPV 78.2%, NPV 80.6%. Focusing on the 153/193(79.3%) surgical group with pathologic staging, 128 stage I patients had a test sensitivity of 83.6%. Accuracy of test stable across size ranges 82%(4-8mm), 77.2%(8-30mm), 80%(30-50mm).
In this blinded prospective validation trial, the current iteration of the Prolung Test(v2.0) has significantly improved sensitivity (84.2% vs 70-75%) and specificity (73.6% vs 47-50%) over the previous version, based on the streamlined protocol of human operations and minor algorithm development. Overall performance achieved without subgroup selection based on risk score and without degradation in smaller stage I cancers. Efficacy as a follow-up tool and further benefits from hardware modifications await studies.
Only Active 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 or select "Add to Cart" and proceed to checkout.