A customer in the us reported the generation of discrepant g719x mutation results for a patient's tissue sample tested with the cobas egfr mutation test v2).The patient's sample was initially tested on (b)(6) 2021 using cobas egfr mutation test v2 and generated a "mutation detected" result for g719x.The original prepared eluate sample and re-extracted eluate from the same ffpe block but a different section were both tested/re-tested twice on (b)(6) 2022 and (b)(6) 2022 using cobas egfr mutation test v2 and generated a "mutation not detected" result for g719x.Finally, a new sample was collected and sequenced using next generation sequencing (unknown method) which noted braf mutation, but no egfr mutation in patient sample- findings were confirmed by nano-strain testing.Per the customer, a lobectomy was performed in (b)(6) 2021 based on the alleged erroneous egfr mutation detected result.An investigation was conducted to evaluate the customer issue.Examples of additional information requested but not provided at this time include: what exact ngs method was used for the retest and more information regarding the nano-strain testing that was also performed as part of the retest.If any other test was performed that supported the decision to pursue a lobectomy.Any treatment the patient received, specifically anti-egfr treatment administered related to the event.
|
The ngs method used to test the sample is unknown and may be less sensitive than the cobas ® egfr mutation test v2.It¿s unknown what nano-strain testing is and how sensitive the technology is compared with cobas test.If the test is nano-string , it¿s a gene expression profiling.The mechanism of how they report out mutations and what kind of sensitivity their technology possesses is also unknown.The ngs and "nano-strain" test were performed on a recollection section and not the original section that generated the positive result.Based upon data review, the original sample was observed with g719x mutation at or below the assay's limit of detection.Based upon the original and retest results, tumor heterogeneity may be a factor in the observed results.This is further supported because although the second section did not detect g719x mutation, there was signal generated in the g719x channel for the retest.Controls were valid and within their respective ranges showing the reactions were successful.As far as the sensitivity of the roche test, g719x was validated down to 1.25% allelic frequency against next generation sequencing(sensitive to 1% allelic frequency).Per the instructions for use, to achieve 95% mutation detected for g719x a 2.5 to 5.6% was observed to be needed however, it can detect lower amounts at a lower confidence level.In summary, the results obtained by the customer are expected by a sample that seems to be lower than lod as well as tumor heterogeneity.Roche customer site ((b)(6)) performed the test.A sponsor site, which is conducting the trial, collected the tissue.As a result, only limited information was able to be obtained in regards to the patient.Based on the limited information available at this time, a medical assessment was performed and concluded that it is not a standard medical practice in the united states to pursue lobectomy based on molecular testing.Such pulmonary resections are driven by cancer diagnosis (based on histopathological assessment of the tissue using h&e and/or ihc).According to nccn guidelines ( nccn v1.2022), if a preoperative tissue diagnosis is not obtained, intraoperative diagnosis (i.E wedge resection, needle biopsy) is necessary before lobectomy.Currently molecular biomarkers are not part of the lung cancer diagnosis as per who nomenclature (who 4th edition) and molecular testing for predictive biomarkers is done as part of the workup for non-surgical therapy selection.Furthermore, cobas egfr v2 test is intended to identify patients, who were previously diagnosed with non-small cell lung cancer, eligible for companion diagnostic therapies (as per the assay's instructions for use).(b)(4).
|