APTIMA COMBO 2

Description of Event or Problem

FINAL REPORT: A (B)(6) PREVALENCE STUDY WAS INITIATED BY THE (B)(6) IN CONJUNCTION WITH THE (B)(6). ROUTINE SAMPLES THAT TESTED [?]EQUIVOCAL' FOR (B)(6) BY AC2 WERE RE-TESTED BY THE (B)(6) WHICH USES A DIFFERENT DETECTION TARGET (16S RRNA). SAMPLES THAT TESTED ACT (B)(6) WERE FORWARDED TO NMLFOR SEQUENCING TO EXAMINE IF MUTATION(S) IN THE 23S RRNA REGION PRODUCED (B)(6) VARIANT STRAINS. NML IDENTIFIED THREE (B)(6) VARIANT SAMPLES IN OVER 100,000 APTIMA COMBO2 SAMPLES SCREENED; 2 SAMPLES CARRIED THE C1514T MUTATION AND 1 SAMPLE CARRIED THE G1523A MUTATION. THESE MUTATIONS WERE PREVIOUSLY IDENTIFIED IN EUROPE AT VERY LOW PREVALENCE (ROBERTS ET AL., EUROSURVEILLANCE, 2019). THE NUMBER OF AC2 SAMPLES SCREENED WAS NEEDED TO ESTIMATE THE PREVALENCE OF THESE 3 (B)(6) VARIANTS IN (B)(6). DUE TO THE DIFFICULTY EXTRACTING THE NUMBERS FROM EACH LABORATORY'S INFORMATION SYSTEM, HOLOGIC USED CUSTOMER ORDERING DATA (DURING THE (B)(6) SURVEILLANCE PERIOD (JUNE 2019 TO OCTOBER 2019)) TO DETERMINE A VOLUME OF APPROXIMATELY 115,000 SCREENED AC2 SAMPLES. THE ESTIMATED PREVALENCE OF (B)(6) VARIANTS IDENTIFIED THUS FAR IN (B)(6) IS 0.0026%. THIS PREVALENCE RATE IS SIMILAR TO THE 0.003% PREVALENCE SEEN IN THE UK, BASED ON A LARGE SURVEILLANCE STUDY (>250,000 SAMPLES) CONDUCTED BY PUBLIC HEALTH ENGLAND (PHE) (ROBERTS ET AL., 2019). THIS OBSERVED PREVALENCE OF CT VARIANTS IS WELL BELOW THE ACCEPTABILITY THRESHOLD (FALSE NEGATIVE RATE) OF 0.1% SET BY PUBLIC HEALTH (B)(6) FOR ACCEPTABLE ASSAY PERFORMANCE. MUTATIONS IN THE (B)(6) 23S RRNA REGION DETECTED BY THE APTIMA COMBO (AC2) ASSAY MAY AFFECT PERFORMANCE. AS A RESULT, A SAMPLE CONTAINING A CT VARIANT STRAIN WITH A MUTATION IN THE 23S RRNA REGION MAY PRODUCE AN EQUIVOCAL OR FALSE NEGATIVE RESULT WHEN USING THE AC2 ASSAY. THE AC2 PACKAGE INSERT (502487-IFU-PI, REV. 004) STATES: "RESULTS FROM THE APTIMA COMBO 2 ASSAY SHOULD BE INTERPRETED IN CONJUNCTION WITH OTHER LABORATORY AND CLINICAL DATA AVAILABLE TO THE CLINICIAN." THE PROBABILITY OF HARM DUE TO AN AC2 FALSE NEGATIVE CT RESULT IS REMOTE, GIVEN THE LOW FREQUENCY OF MUTATIONS OBSERVED. MOREOVER, BASED ON SURVEILLANCE PERFORMED IN VARIOUS EUROPEAN COUNTRIES, THERE HAS BEEN NO DECLINE IN (B)(6) POSITIVITY RATES IN REGIONS THAT HAVE IDENTIFIED CT VARIANTS (HOKYNAR ET AL., 2019; ROBERTS ET AL., 2019; JOHANSEN ET AL., 2019). INTERNAL TESTING OF IN VITRO TRANSCRIPTS (IVTS) CONTAINING THE C1514T AND THE G1523A MUTATIONS AT HOLOGIC DEMONSTRATED THAT THESE MUTATIONS CAN IMPACT AC2 ASSAY PERFORMANCE. OF NOTE, A THREE-YEAR COMPLAINT TRENDING DATA ANALYSIS SHOWED NO INCREASE IN CUSTOMER COMPLAINTS ASSOCIATED WITH INCORRECT RESULTS. IN ADDITION, SEVERAL RECENT PUBLICATIONS HAVE DEMONSTRATED COMPARABLE PERFORMANCE OF THE AC2 ASSAY WITH OTHER MANUFACTURERS' ASSAYS. THE ROOT CAUSE OF THE (B)(6) RESULTS IS DUE TO MUTATION(S) IN THE 23S RRNA REGION OF THE (B)(6) GENOME WHICH REDUCE THE PROBE BINDING ABILITY OF THE AC2 ASSAY. HOLOGIC IS IN PROCESS OF UPDATING THE AC2 ASSAY SUCH THAT THESE VARIANTS WILL BE DETECTED. THIS ASSAY IS ANTICIPATED TO BE READY FOR SUBMISSION SPRING 2020.