Unknown

Dataset Information

0

ACTR-68. CONCORDANCE OF EGFR AND MGMT ANALYSES BETWEEN LOCAL AND CENTRAL LABORATORIES: IMPLICATIONS FOR CLINICAL TRIAL DESIGN AND PRECISION MEDICINE FOR DEPATUXIZUMAB-MAFODOTIN (ABT-414) IN GLIOBLASTOMA (GBM)


ABSTRACT: Abstract BACKGROUND Depatuxizumab-mafodotin is an EGFR-directed monoclonal antibody that targets activated EGFR (wild type or EGFRvIII) conjugated to a microtubule inhibitor. EGFRamp correlated with recurrent GBM response in a phase I trial (M12-356, NCT01800695), thus becoming a centrally determined inclusion criterion in an ensuing phase II/III randomized study for newly diagnosed disease now accruing (RTOG 3508, M13-813, NCT02573324). Centrally detected EGFRvIII mutation and MGMT promoter methylation are also stratification factors in RTOG 3508 to balance randomization for biomarkers of potential predictive or prognostic significance. Allowing use of locally determined molecular profiling may reduce accrual barriers by avoiding added time of tissue shipping and repeat central testing. However, discordant results between local and central tests could confound interpretation of clinical trial results. Therefore, we compared local vs. central biomarker results among patients screened for M12-356 or RTOG 3508 at Columbia University Medical Center (CUMC). METHODS We retrospectively tabulated local and central EGFRamp, EGFRvIII, and MGMT results. EGFRamp was analyzed by FISH/CISH and EGFRvIII and MGMT by PCR, although the assays for each differed slightly between laboratories. RESULTS 109 GBMs were molecularly profiled from 73 patients who underwent 1–4 resection(s) and screened for M12-356 (n=44) or RTOG 3508 (n=29) at CUMC. EGFRamp, EGFRvIII, and MGMT promoter methylation each was observed in 54, 36, and 27% of tumors tested locally vs. 68, 33, and 43% centrally, respectively. Concordance between laboratories was observed 88%, 86%, and 61%. CONCLUSIONS EGFRamp and EGFRvIII results were mainly concordant (over 85%) between local and central laboratories. EGFRamp frequency was higher centrally, likely biased by investigators who preferentially sent cases known to harbor EGFRamp locally for central testing. Of concern, MGMT results were discordant in 39%. Allowing local MGMT testing for stratification could unacceptably imbalance randomization if confirmed in a larger dataset.

SUBMITTER: Lassman A 

PROVIDER: S-EPMC5692946 | biostudies-literature | 2017 Nov

REPOSITORIES: biostudies-literature

Similar Datasets

| S-EPMC5686264 | biostudies-literature
| S-EPMC5969257 | biostudies-literature
| S-EPMC6216179 | biostudies-literature
| S-EPMC6847153 | biostudies-literature
| S-EPMC5782648 | biostudies-literature
| S-EPMC6216489 | biostudies-literature
| S-EPMC9925712 | biostudies-literature
| S-EPMC5692444 | biostudies-literature
| S-EPMC8645742 | biostudies-literature
| S-EPMC3017549 | biostudies-literature