Virtual Library
Start Your Search
Lancia Guo
Author of
-
+
P1.11 - Screening and Early Detection (Not CME Accredited Session) (ID 943)
- Event: WCLC 2018
- Type: Poster Viewing in the Exhibit Hall
- Track:
- Presentations: 1
- Moderators:
- Coordinates: 9/24/2018, 16:45 - 18:00, Exhibit Hall
-
+
P1.11-15 - Application of Lung-RADS vs. PAN-CAN Nodule Risk Calculation in the Alberta Lung Cancer Screening Study (ID 13052)
16:45 - 18:00 | Author(s): Lancia Guo
- Abstract
Background
False positive or negative examinations and high early recall rates are important factors in the performance of lung cancer screening programs. How low-dose chest tomography (LDCT) scans are interpreted and classified may impact these metrics.
a9ded1e5ce5d75814730bb4caaf49419 Method
LDCT examinations for participants in the Alberta Lung Cancer Screening Study (ALCSS) were interpreted by chest radiologist with information entered in a synoptic report. Baseline scans were classified according to highest risk of malignancy nodule as per the PAN-CAN nodule risk calculator (NRC) and according to the Lung-RADS scheme. A positive scan was any baseline LDCT requiring any intervention beyond an annual screening examination (NRC nodule with ≥5% malignancy risk; Lung-RADS category ≥3). In the calculation of sensitivity, false negative scans could include reader error or classification errors (NRC <5% or Lung-RADS <3 but cancer present regardless of perceived appropriateness of resulting management).
4c3880bb027f159e801041b1021e88e8 Result
Seven hundred and seventy-six participants in the ALCSS underwent LDCT screening and had no prior chest CT imaging on file. Median follow-up was 572 days (+/-205) with lung cancer confirmed in 16 (2.1%) participants. The early recall rate was 9.0% for NRC and 11.2% for Lung-RADS (p=0.044), with fair concordance between each approach (kappa 0.554). Sensitivity for malignancy was 87.5% vs. 87.5% (difference 0%, 95%CI -0.44%-0.44%) and specificity 92.6% vs. 90.4% (difference 2.2%, 95%CI 0.2%-4.3%) for NRC and Lung-RADS respectively. False negative screens were due to reader error (same case in both systems); and classification error (one different case for each system).
8eea62084ca7e541d918e823422bd82e ConclusionCancer +
Cancer -
Total
NRC +
14
56
70
NRC -
2
704
706
Lung-RADS +
14
73
87
Lung-RADS -
2
687
689
Total
16
760
776
Performance of both the NRC and Lung-RADS in the ALCSS was very good, with NRC resulting in a lower early recall rate. Application of the NRC demonstrated increased specificity over Lung-RADS without a change in sensitivity for lung cancer detection. Lung cancer program performance may be improved with the use of the PAN-CAN NRC classification.
6f8b794f3246b0c1e1780bb4d4d5dc53