Austin, Christopher

Introduction: The prevalence of cardiac implantable electronic devices (CIED) is steadily increasing. Accordingly, the frequency of CIED extraction has also increased. Occasionally, safe retrieval of all CIED components is not feasabile and lead fragments are left behind. Little is known about the safety and diagnostic accuracy of MRI in patients with retained CIED fragments. Methods: Two hundred consecutive patients were identified as having undergone orthotopic heart transplantation (OHT) at our institution between April 2006 and December 2014. Chest x-rays immediately prior to OHT were reviewed for presence of CIED and chest x-rays immediately following OHT were reviewed for retained CIED fragments. MRI performed after OHT in patients with retained CIED fragments were inspected for image quality and procedural safety. The most recent chest x-ray in patients with retained CIED fragments subjected to MRI was reviewed for change in fragment position. Results: One hundred seventy-eight of 200 patients (89%) had CIED prior to OHT. All patients with CIED had surgical device extraction at the time of OHT. Of these, 29 patients (12%) had retained CIED fragments after OHT, the majority of which (23/29) were ICD coil fragments located in the superior vena cava, brachiocephalic or subclavian vein. Ten patients with retained CIED fragments underwent a total of 28 MRI after OHT. No patient safety issues were identified during MRI acquisition, and all MRI images were of diagnostic quality. There were no changes in the position of CIED fragments following MRI. Conclusions: Diagnostic MRI can be safely performed in patients with retained CIED fragments after incomplete device extraction. Adherence of the ICD coil to the vessel wall due to fibrosis has been well described and is a common impediment to complete device extraction. Retrieval of these fragments prior to MRI does not appear warranted given the demonstrated safety and preserved MRI image quality in this population.

Abstract Introduction: Recent reports have utilized von Willebrand factor (VWF) multimerquantitation in the assessment of native cardiac lesions and prosthetic valve dysfunction. However there is no standardized method for quantitation. We compared 3 methods of assessment which utilized a normal plasma control. Methods: We analyzed 476 samples and their control plasma from 395 patients including normal subjects, patients with aortic stenosis, hypertrophic cardiomyopathy, mitral or aortic regurgitation, normal and abnormal left cardiac valve prostheses, and left ventricular assist device therapy (LVAD). VWF multimers were assessed as normal or abnormal, or as normalized (patient / normal plasma) VWF multimerratios of gel bands >15/2-15 (NMR 15) or gel bands > 10/2-10 (NMR 10). We tested whether one technique was more strongly associated with hemodynamic severity of the cardiac lesion, with a history of acquired bleeding, and compared the results to a separate test of VWF function, platelet function analyzer 15 00 (PFA). Results: Two hundred seventeen patient samples reflected normal or mildly abnormal hemodynamics, 111 moderate, and 113 severely abnormal hemodynamics. Abnormal multimers were present in only 19% of samples with normal or mildly abnormal hemodynamics compared to 79% abnormal in the presence of moderate to severe hemodynamic abnormality. The distinction between moderate and severe hemodynamic abnormality was less striking, 65% versus 88% (figure, left). Quantitative NMR values declined with increasing cardiac lesion severity (figure, middle). The two quantitative multimermethods showed similar relationships to lesion severity by ANOVA, r=0.40 (NMR 15) versus r=0.36 (NMR 10), and by Spearman correlation, r=0.68 for NMR 15, and r=0.66 for NMR 10, and PFA results were comparable (all p< 0.001). Bleeding rates were associated with NMR 15, 10 and PFA (Mann-Whitney p<0.001, figure, right). Conclusion: The dichotomous normal / abnormal VWF multimer method appears to distinguish ≥ moderate hemodynamic disruption, whereas there is a continuous relationship between hemodynamic severity and quantitative multimer methods, which may be clinically useful in circumstances in which clinical estimation of lesion severity is challenging, such as with dysfunctional prosthetic valves. Progressive loss of VWF multimersappears to be associated with bleeding, suggesting that quantitative measures may provide incremental value over the dichotomous normal / abnormal designation. Figure Figure. Disclosures Blackshear: Baxalta, Inc: Consultancy, Research Funding.