Abstract
Legally marketed comparator devices (LMCD) are required by many regulatory bodies in as a control for thrombogenicity testing when evaluating new devices. It is assumed by both the medical device manufacturing industry and regulatory bodies that these LMCD’s have good clinical history and should perform with no to minimal thrombus accumulation and thereby serve as valid negative controls for the assay. APS regularly performs these assays for many medical device manufacturers, all of whom select a predicate comparator device (required by FDA to be an LMCD), for both the in vivo Non-Anticoagulated Venous Implant (NAVI) assay as well as a custom in vitro blood loop AVI. In this retrospective analysis, we have compiled thrombogenicity scores of control/predicate devices (limited to assays which used LMCD’s), both the discrete score from the classification standard scoring scheme and the continuous values obtained from the percent surface area associated with thrombus. We have compared results from 37 NAVI studies and 22 in vitro blood loop studies. These compiled results show ∼25% of LMCDs score ≥ 3 (> 50% of the surface covered in thrombus) in the NAVI model while < 5% of LMCDs score ≥ 3 (> 50% thrombus) in the Blood-Loop assay. In addition, the median score and mean % thrombus for LMCD in the blood loop assay is substantially lower than the median and mean scores for LMCD in the NAVI assay. This retrospective assessment highlights a high proportion of false-positive scores for LMCD in a large number of NAVI assays.
