TY - JOUR
T1 - Performance comparison of different analytic methods in proficiency testing for mutations in the BRAF, EGFR, and KRAS genes
T2 - A study of the college of American pathologists molecular oncology committee
AU - Moncur, Joel T.
AU - Bartley, Angela N.
AU - Bridge, Julia A.
AU - Kamel-Reid, Suzanne
AU - Lazar, Alexander J.
AU - Lindeman, Neal I.
AU - Long, Thomas A.
AU - Merker, Jason D.
AU - Rai, Alex J.
AU - Rimm, David L.
AU - Rothberg, Paul G.
AU - Vasalos, Patricia
AU - Kim, Annette S.
N1 - Publisher Copyright:
© 2019 College of American Pathologists. All rights reserved.
PY - 2019
Y1 - 2019
N2 - Context. - The performance of laboratory testing has recently come under increased scrutiny as part of important and ongoing debates on regulation and reimbursement. To address this critical issue, this study compares the performance of assay methods, using either commercial kits or assays designed and implemented by single laboratories (''home brews''), including next-generation sequencing methods, on proficiency testing provided by the College of American Pathologists Molecular Oncology Committee. Objective. - To compare the performance of different assay methods on College of American Pathologists proficiency testing for variant analysis of 3 common oncology analytes: BRAF, EGFR, and KRAS. Design. - There were 6897 total responses across 35 different proficiency testing samples interrogating 13 different variants as well as wild-type sequences for BRAF, EGFR, and KRAS. Performance was analyzed by test method, kit manufacturer, variants tested, and preanalytic and postanalytic practices. Results. - Of 26 reported commercial kits, 23 achieved greater than 95% accuracy. Laboratory-developed tests with no kit specified demonstrated 96.8% or greater accuracy across all 3 analytes (1123 [96.8%] acceptable of 1160 total responses for BRAF; 848 [97.5%] acceptable of 870 total responses for EGFR; 942 [97.0%] acceptable of 971 total responses for KRAS). Next-generation sequencing platforms (summed across all analytes and 2 platforms) demonstrated 99.4% accuracy for these analytes (165 [99.4%] acceptable of 166 total next-generation sequencing responses). Slight differences in performance were noted among select commercial assays, dependent upon the particular design and specificity of the assay. Wide differences were noted in the lower limits of neoplastic cellularity laboratories accepted for testing. Conclusions. - These data demonstrate the high degree of accuracy and comparable performance across all laboratories, regardless of methodology. However, care must be taken in understanding the diagnostic specificity and reported analytic sensitivity of individual methods.
AB - Context. - The performance of laboratory testing has recently come under increased scrutiny as part of important and ongoing debates on regulation and reimbursement. To address this critical issue, this study compares the performance of assay methods, using either commercial kits or assays designed and implemented by single laboratories (''home brews''), including next-generation sequencing methods, on proficiency testing provided by the College of American Pathologists Molecular Oncology Committee. Objective. - To compare the performance of different assay methods on College of American Pathologists proficiency testing for variant analysis of 3 common oncology analytes: BRAF, EGFR, and KRAS. Design. - There were 6897 total responses across 35 different proficiency testing samples interrogating 13 different variants as well as wild-type sequences for BRAF, EGFR, and KRAS. Performance was analyzed by test method, kit manufacturer, variants tested, and preanalytic and postanalytic practices. Results. - Of 26 reported commercial kits, 23 achieved greater than 95% accuracy. Laboratory-developed tests with no kit specified demonstrated 96.8% or greater accuracy across all 3 analytes (1123 [96.8%] acceptable of 1160 total responses for BRAF; 848 [97.5%] acceptable of 870 total responses for EGFR; 942 [97.0%] acceptable of 971 total responses for KRAS). Next-generation sequencing platforms (summed across all analytes and 2 platforms) demonstrated 99.4% accuracy for these analytes (165 [99.4%] acceptable of 166 total next-generation sequencing responses). Slight differences in performance were noted among select commercial assays, dependent upon the particular design and specificity of the assay. Wide differences were noted in the lower limits of neoplastic cellularity laboratories accepted for testing. Conclusions. - These data demonstrate the high degree of accuracy and comparable performance across all laboratories, regardless of methodology. However, care must be taken in understanding the diagnostic specificity and reported analytic sensitivity of individual methods.
UR - http://www.scopus.com/inward/record.url?scp=85070549778&partnerID=8YFLogxK
U2 - 10.5858/arpa.2018-0396-CP
DO - 10.5858/arpa.2018-0396-CP
M3 - Article
C2 - 30969158
AN - SCOPUS:85070549778
SN - 0003-9985
VL - 143
SP - 1203
EP - 1211
JO - Archives of Pathology and Laboratory Medicine
JF - Archives of Pathology and Laboratory Medicine
IS - 10
ER -