P1.02-006 Interlaboratory Variation in Molecular Testing (EGFR, KRAS and ALK) in Stage IV Non-Squamous Non-Small Cell Lung Cancer in the Netherlands in 2013

Abstract

Adequate testing for molecular changes in non-small cell lung cancer (NSCLC) is necessary to ensure the best possible treatment. However, it is unknown how well molecular testing is performed in daily practice. Therefore we aimed to assess the performance of testing for EGFR, KRAS mutation and ALK translocation in metastatic NSCLC on a nationwide basis. Using the Netherlands Cancer Registry, all stage IV non-squamous NSCLC from 2013 were identified and matched to the Dutch Pathology Registry (PALGA). Data on molecular testing for EGFR, KRAS and ALK were extracted from excerpts of pathology reports. Proportions of tested and positive cases were determined and interlaboratory variation was assessed. Finally, degree of concordance between ALK immunohistochemistry (IHC) and fluorescent in situ hybridization (FISH) results was evaluated. In total, 3393 stage IV non-squamous NSCLCs were identified, and 3183 (93.8%) were matched to PALGA. Fifty-two tumors were excluded as pathology reports described a lung tumor other than non-squamous NSCLC or a tumor with another origin, leaving 3131 tumors. All 48 laboratories had access to molecular testing, either in house or via outsourcing. The table shows the nationwide proportions of cases tested and positive for EGFR, KRAS and ALK, as well as the interlaboratory variation. EGFR and KRAS mutations occurred together in 8 patients, ALK translocation occurred together with EGFR mutation in 3 patients and with KRAS mutation in 2 patients. In 272 cases, ALK had been tested using both IHC and FISH, and the methods were conclusive in 253 cases. IHC and FISH were concordant in 239 cases (94.5%; Kappa 0.728, p=0.069), 5 discordant cases were IHC+/FISH- and 9 were IHC-/FISH+.Tabled 1Nationwide proportions of tested and positive cases and interlaboratory variation.TotalTested cases; n (%)Range in tested cases between LaboratoriesPositive cases; n (%)Range in positive cases between laboratoriesEGFR31312237 (71.4)33.7% to 93.5%243 (10.9)6.1% to 21.6%KRAS31312292 (73.2)20.9% to 93.6%845 (36.9)27.0% to 48.1%ALK3131905 (28.9)7.0% to 72.6%51 (5.6)0% to 13.6%ALK (in case of EGFR and KRAS wildtype)1227685 (55.8)19.4% to 100% Open table in a new tab These results suggest that in 2013 molecular testing was suboptimal in the Netherlands, especially for ALK. To determine whether molecular testing has improved, 2015 data will be analyzed in the near future as well.