Dear Editor,
We would like the thank Drs. Berndt, Zimmer, Kaesmacher, and Boeckh-Behrens for their interest in our study titled “Clot permeability and histopathology: Is a clot’s perviousness on CT imaging correlated with its histologic composition?” We read their letter with interest. The authors have been pioneers in stroke clot analysis and we greatly respect their academic rigor and expertise.
While we agree that there are certainly some methodological differences between our two studies, we do not believe that these are to blame for the differences in results. Rather, we feel that the observed differences in results between our studies could be due to differences in our patient populations.
Our group has previously shown that there is indeed a correlation between clot composition and etiology. In a recently published article in Stroke we found that large artery atherosclerosis clots were more likely to be platelet rich than those of a cardioembolic origin.1 To date, however, we have yet to find any definite correlation between etiology and RBC density or fibrin density, and we think it is too early to make any definite conclusions on the association between clot composition and etiology.
We agree that the association between perviousness, clot composition, etiology and clinical outcome is not conclusively clarified yet, and hence warrants further research, especially in a larger patient group in a multi-centric setting. Currently, our gr...
Dear Editor,
We would like the thank Drs. Berndt, Zimmer, Kaesmacher, and Boeckh-Behrens for their interest in our study titled “Clot permeability and histopathology: Is a clot’s perviousness on CT imaging correlated with its histologic composition?” We read their letter with interest. The authors have been pioneers in stroke clot analysis and we greatly respect their academic rigor and expertise.
While we agree that there are certainly some methodological differences between our two studies, we do not believe that these are to blame for the differences in results. Rather, we feel that the observed differences in results between our studies could be due to differences in our patient populations.
Our group has previously shown that there is indeed a correlation between clot composition and etiology. In a recently published article in Stroke we found that large artery atherosclerosis clots were more likely to be platelet rich than those of a cardioembolic origin.1 To date, however, we have yet to find any definite correlation between etiology and RBC density or fibrin density, and we think it is too early to make any definite conclusions on the association between clot composition and etiology.
We agree that the association between perviousness, clot composition, etiology and clinical outcome is not conclusively clarified yet, and hence warrants further research, especially in a larger patient group in a multi-centric setting. Currently, our group is collaborating with multiple centers across the United States and Canada in the Stroke Thromboembolism Registry of Imaging and Pathology and are in the process of performing histological and imaging analysis of over 1200 collected clots from stroke patients. Combining our findings with those of other ongoing multicenter studies may ultimately help us reach meaningful conclusions regarding the association between clot composition and imaging, outcomes and etiology.
References:
[1] Fitzgerald S, Dai D, Wang S, et al. Platelet-rich emboli in cerebral large vessel occlusion are associated with a large artery atherosclerosis source. Stroke 2019;50(7):1907-1910.
The authors describe two endovascular techniques for delivering IA chemotherapy to retinoblastoma patients. Technique A where a 1.2 Fr or 1.5 Fr micro catheter with continuous verapamil flush is advanced without a guide and technique B where a1.5 Fr or 1.7 Fr micro catheter is advanced within a 4 Fr catheter, through a 4 Fr sheath. We usually use a no sheath technique, using a 4Fr diagnostic catheter as a guide catheter for neonatal and pediatric cases. Most importantly, we do not use Echelon or Marathon micro catheters in neonatal and pediatric patients because their use is contraindicated per "instructions for use".
An increasing number of reports highlight polymer coating embolism as an iatrogenic complication of intravascular medical devices [1-3]. Autopsies, histologic evaluations of thrombectomy specimens, samples of captured debris, resected or biopsied tissues, are available methods used to study polymer emboli post investigative catherizations or interventional procedures. Reported data highlight the prevalence of this phenomenon and/or its clinicopathologic impacts, however, fall short of identifying higher-risk polymer emboli interventional devices. Consequently, an optimal approach for future investigations related to polymer coating embolism is required.
Mehta et. al investigate the histologic frequency of polymer emboli among patients who underwent endovascular thrombectomy for treatment of acute ischemic stroke due to large vessel occlusion by retrospectively evaluating thrombectomy specimens [2]. In this study, the reported frequency of polymer emboli includes the use of various devices and techniques among selected cases. However, literature highlights polymer coating embolism is device specific and dependent on coating integrity measured by particulates released [4]. Thus, the use of alternate devices with higher or lower particulate release for a given procedure may result in a large variation in incidence rates from reported results. Also, as mentioned by the authors, subsequent statistical correlations unless appropriately powered provide limited informatio...
An increasing number of reports highlight polymer coating embolism as an iatrogenic complication of intravascular medical devices [1-3]. Autopsies, histologic evaluations of thrombectomy specimens, samples of captured debris, resected or biopsied tissues, are available methods used to study polymer emboli post investigative catherizations or interventional procedures. Reported data highlight the prevalence of this phenomenon and/or its clinicopathologic impacts, however, fall short of identifying higher-risk polymer emboli interventional devices. Consequently, an optimal approach for future investigations related to polymer coating embolism is required.
Mehta et. al investigate the histologic frequency of polymer emboli among patients who underwent endovascular thrombectomy for treatment of acute ischemic stroke due to large vessel occlusion by retrospectively evaluating thrombectomy specimens [2]. In this study, the reported frequency of polymer emboli includes the use of various devices and techniques among selected cases. However, literature highlights polymer coating embolism is device specific and dependent on coating integrity measured by particulates released [4]. Thus, the use of alternate devices with higher or lower particulate release for a given procedure may result in a large variation in incidence rates from reported results. Also, as mentioned by the authors, subsequent statistical correlations unless appropriately powered provide limited information on higher-risk or culprit devices. Consequently, in addition to providing limited value, use of histologic evaluations of thrombi (or autopsies) to exclusively evaluate incidence rates may be time-consuming and have a high resource burden. Notably, this may not be the intent of the authors of this study who elude to an article series.
Few studies related to polymer coating embolism have included controls over procedures and specific devices used. For example, histologic evaluations of captured debris within cerebral protection devices during a mitral valve repair procedure highlighted polymer coating type material in 12 of 14 cases [5]. The source of polymer emboli was speculated to be from the mitral valve repair catheter and/or adjunct devices. In another study, intracranial polymer emboli incidence rates were determined for a branded hydrophilic coated guide sheath used for carotid and iliac stenting procedures respectively in Yucatan miniswine [6]. Since the experimental stents and stent delivery systems lacked any coating, impacts of polymer emboli were isolated to the coated guide sheaths, guide catheters and guidewires used during the procedures. These studies with controls over procedures and devices may be leveraged to understand an incidence rate from a specific set of interventional devices repetitively used during a given procedure. Notably however, even this approach is unable to provide device specific information or identify higher-risk interventional devices as the origin of emboli are difficult to determine.
Device type, coating composition-device substrate material combinations, coating application processes, coating thickness and degree of coating coverage are variables that impact polymer coating integrity on a device [4,7]. Particulate generation testing – the determination of a count, shape and size of particulates released from a device when used in an in-vitro vessel model – is the industry standard for evaluating polymer coating integrity from an intravascular device [8]. An intuitive correlation exists between particulates released in-vitro and the polymer emboli incidence rates [4]. Thus, particulate generation testing may be used to compare particulates released and understand relative incidence rates among intravascular devices. In-vitro particulate testing is an effective and efficient approach to determining higher-risk particulate release devices and may assist in identifying culprit interventional devices.
Based on the aforementioned study methodologies and outcomes, the following approach may be outlined for future investigations related to polymer coating embolism: a) Autopsy based studies that typically lack device information or controls over prior procedures should be limited to determining clinicopathologic impacts of polymer emboli. For these studies, estimating a particulate burden to correlate impacts with the quantity of polymer emboli for each area of assessment is relevant; b) Histologic evaluations of thrombi, captured debris, resected or biopsied tissues are also effective in determining localized disease processes. These methods may be used to determine incidence rates for a procedure if devices are consistently repeated for selected cases. For these studies (and postmortem investigations) combining incidence rates with particulate test data from devices will assist in categorizing embolic risk and determine higher risk devices; c) When clinical presentations preclude the consistent use of devices for a procedure, animal studies may be used to determine incidence rates. Extrapolating clinical impacts from healthy animals maybe acceptable, however, investigations that include relevant disease conditions are preferred; d) In-vitro particulate generation testing may be the optimal method to rank embolic risk and determine higher risk polymer emboli devices.
Inclusion and exclusion criteria for polymer emboli related investigations are essential for meaningful results. Patient history should be carefully evaluated as prior procedures may impact incidence rates. For this reason, patients with a history of multiple interventional procedures should be excluded from studies that attempt to determine polymer emboli frequencies. This patient subset may be included for studies determining potential impacts from polymer emboli for a worst-case scenario assessment. All investigations should include device specific information such as type, dimensions (e.g. length, diameter), brand, coating types (e.g. hydrophilic and/or hydrophobic) and coated dimensions. Other important parameters include patient baseline characteristics, procedural length, device in-dwell time (if available) and procedural outcomes.
Comparing particulate data among devices or reporting the magnitude of embolic burden may require assumptions to characterize total particulate volume. For the aforementioned studies, use of an efficient light obscuration methodology for particulate assessments that provides an equivalent circular diameter for particle areas may be used [9]. Combined with a uniform 1-micron height representing the typical lamellar nature of polymer particulates [3], a cylindrical volume calculation for each particle may be optimal. Summation of particle volumes may provide a total particulate burden per device or affected area of inspection.
The FDA continues to work with stakeholders to create tools which permit the standardization of particulate test methods and enable comparisons among devices [10]. Till these standards become available, investigations should include assumptions used to generate particulate data. In the future, an accumulation of procedural and device particulate data with associated incidence rates and clinicopathologic impacts may provide actionable input for regulators for setting device particulate limits. Given the sparse literature on this subject, more studies with controls over procedural parameters and devices used are required. Procedures with devices exposed to larger frictional forces (e.g. chronic total occlusions, aortic repair, or atherectomy), or aqueous environments for longer durations (e.g. percutaneous mechanical circulatory support) should be prioritized for future studies.
Acknowledgements, Funding Sources, Disclosures & Author Contributions
Acknowledgements: None. No persons other than the listed authors have made contributions to this manuscript.
Funding Sources: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosures: None. Authors declare no current relationship with industry and no conflicts of interest.
Author Contributions: All authors have contributed to this manuscript.
References:
1. Chopra AM, Mehta M, Bismuth J, et al. Polymer coating embolism from intravascular medical devices - a clinical literature review. Cardiovasc Pathol 2017;30:45-54.
2. Mehta RI, Rai AT, Vos JA, Solis OE, Mehta RI. Intrathrombus polymer coating deposition: a pilot study of 91 patients undergoing endovascular therapy for acute large vessel stroke. Part I: Histologic frequency. J Neurointerv Surg. Epub ahead of print 18 May 2019. doi: 10.1136/neurintsurg-2018-014684.
3. Hickey TB, Honig A, Ostry AJ, et al. Iatrogenic embolization following cardiac intervention: postmortem analysis of 110 cases. Cardiovasc Pathol 2019;40:12-18.
4. Babcock DE, Hergenrother RW, Craig DA, Kolodgie FD, Virmani R. In Vivo Distribution of Particulate Matter from Coated Angioplasty Balloon Catheters. Biomaterials 2013;34(13):3196-205.
5. Frerker C, Schlüter M, Sanchez OD, et al. Cerebral Protection During MitraClip Implantation: Initial Experience at 2 Centers. JACC Cardiovasc Interv. 2016;9(2):171-9.
6. Stanley JR, Tzafriri AR, Regan K, et al. Particulates from Hydrophilic-Coated Guiding Sheaths Embolize to the Brain. EuroIntervention 2016;11(12):1435-41.
7. Work JW. Technical white paper: considerations for hydrophilic surface coatings on medical devices [internet]. Biocoat, Inc. Horsham, PA, USA. 2016. Available from www.biocoat.com. Accessed 22 Apr 2016.
8. Center for Devices and Radiological Health Recognized Consensus Standards. Recognition number 3-99: AAMI TIR42:2010 Evaluation of Particulates Associated with Vascular Medical Devices (cardiovascular) [Internet]. 2010. Available from: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfStandards/detail.cf.... Accessed Jan. 13, 2019.
9. AAMI TIR42:2010. Evaluation of Particulates Associated with Vascular Medical Devices. Arlington, VA: Association for the Advancement of Medical Instrumentation, 2010.
10. ASTM WK60510. New Test Method for Simulated Use Testing of Neurointerventional Device in Tortuous Vasculature. West Conshohocken, PA: American Society for Testing and Materials International, 2018.
Proprietary nature of intravascular medical device coatings limits safety testing
Dear Dr. Albuquerque:
We are glad that our work has generated interest and discussion in the field [1]. Four years have elapsed since a need for updated device coating testing was officially announced [2], however complexities on the matter and persistent knowledge gaps limit safety studies of devices currently on the market for clinical intravascular use [3,4]. Standardized in vitro particulate generation testing is needed. However, available literature shows that preclinical device testing is not fully predictive of clinical response. Therefore, in vitro and animal studies cannot replace investigation in humans. Currently, lack of consensus on the following prevent meaningful testing in humans: I) optimal clinical testing methods; ii) definitions of permissible risk; iii) adverse cellular, organ, and temporal-specific effects of distinct coating biomaterials; and iv) effects of pre-existing comorbid conditions. Nevertheless, in vitro testing that does not incorporate clinical data has limited utility for safety guidance. Likewise, in vivo studies that do not incorporate biomaterial factors are incomplete. Thus, the proprietary nature of intravascular device coatings remains a significant limitation to clinical device testing and safety assurances. Growing data [2-6] suggest that it may be time for this to be addressed.
1. Chopra AM, Hu YC, Cruz JP. The Device Specific...
Proprietary nature of intravascular medical device coatings limits safety testing
Dear Dr. Albuquerque:
We are glad that our work has generated interest and discussion in the field [1]. Four years have elapsed since a need for updated device coating testing was officially announced [2], however complexities on the matter and persistent knowledge gaps limit safety studies of devices currently on the market for clinical intravascular use [3,4]. Standardized in vitro particulate generation testing is needed. However, available literature shows that preclinical device testing is not fully predictive of clinical response. Therefore, in vitro and animal studies cannot replace investigation in humans. Currently, lack of consensus on the following prevent meaningful testing in humans: I) optimal clinical testing methods; ii) definitions of permissible risk; iii) adverse cellular, organ, and temporal-specific effects of distinct coating biomaterials; and iv) effects of pre-existing comorbid conditions. Nevertheless, in vitro testing that does not incorporate clinical data has limited utility for safety guidance. Likewise, in vivo studies that do not incorporate biomaterial factors are incomplete. Thus, the proprietary nature of intravascular device coatings remains a significant limitation to clinical device testing and safety assurances. Growing data [2-6] suggest that it may be time for this to be addressed.
1. Chopra AM, Hu YC, Cruz JP. The Device Specific Nature of Polymer Coating Emboli: An Optimal Approach For Future Investigations Related to Polymer Embolism. Journal of Neurointerventional Surgery.
2. U.S. Food and Drug Administration Lubricious Coating Separation From Intravascular Medical Devices FDA Safety Communication. Silver Spring MD: FDA; 2015. Available at: https://wayback.archive-it.org/7993/20161022044037/http://www.fda.gov/Me.... Accessed September 11, 2019.
3. Mehta RI, Rai AT, Vos JA, et al. Intrathrombus polymer coating deposition: a pilot study of 91 patients undergoing endovascular therapy for acute large vessel stroke. Part I: Histologic frequency. Journal of NeuroInterventional Surgery Published Online First: 18 May 2019. doi: 10.1136/neurintsurg-2018-014684
4. Mehta RI , Mehta RI. Hydrophilic polymer embolism: implications for manufacturing, regulation, and postmarketsurveillance of coated intravascular medical devices. J Patient Saf 2018 [Epub ahead of print 19 Mar 2019].doi:10.1097/PTS.0000000000000473
5. Mehta RI , Mehta RI , Solis OE , et al. Hydrophilic polymer emboli: an under-recognized iatrogenic cause of ischemia and infarct. Mod Pathol 2010;23:921–30.doi:10.1038/modpathol.2010.74
6. Mehta RI , Mehta RI. Hydrophilic polymer embolism: an update for physicians. Am J Med 2017;130:e287–90.doi:10.1016/j.amjmed.2017.01.032
Sincerely,
Rashi I. Mehta, MD
West Virginia University
Department of Neuroradiology
Ansaar T. Rai, MD
West Virginia University
Department of Neuroradiology
James W. Simpkins, PhD
West Virginia University
Department of Physiology and Pharmacology
Center for Basic and Translational Stroke Research
Rockefeller Neuroscience Institute
Rupal I. Mehta, MD
University of Rochester
Center for Translational Neuromedicine
Acknowledgments: RIM (Rashi I Mehta) is supported by a grant from the National Institute of General Medical Sciences of the National Institutes of Health (5U54GM104942-03). RIM (Rupal I Mehta) is supported by a grant from the National Institute of Neurological Disorders and Stroke (K08NS089830).
Competing interests: ATR serves as a consultant for Stryker Corporation.
We read with interest the case series by Manning’s et al.[1] using a surface modified flow-diverter stent (Pipeline Flex with Shield Technology, Medtronic Neurovascular, Irvine, California, USA). In this retrospective series, 14 ruptured intracranial aneurysms have been treated in the acute phase after Sub-Arachnoid Hemorrhage (SAH) with the Pipeline shield device under a Single Anti-Platelet Therapy (SAPT). The article concluded the PED-Shield to be safe to use in the acute treatment of ruptured intracranial aneurysms with SAPT.
However, in this small series, the authors reported one case of total stent occlusion and two cases of platelet aggregation noted on the PED-Shield device requiring to switch from single to dual antiplatelet treatment. Considering those three patients, thrombotic complications have been observed in 21.4 % of cases (3/14) in the acute period. Furthermore, in two cases (14.3 %), the authors reported rebleeding of the culprit aneurysm leading to patient death, pointing out the fact that flow-diverter devices may not immediately prevent the risk of aneurysm rerupture.
Anti-thrombogenic coating might have an added value in case of very specific aneurysms cases requiring the placement of a stent in the acute phase after rupture. Those specific cases are mainly dissecting or blister aneurysms for which endovascular or even surgical approach are difficult and carry a high risk of morbi-mortality[2][3]. In case of endovascular treatment, rece...
We read with interest the case series by Manning’s et al.[1] using a surface modified flow-diverter stent (Pipeline Flex with Shield Technology, Medtronic Neurovascular, Irvine, California, USA). In this retrospective series, 14 ruptured intracranial aneurysms have been treated in the acute phase after Sub-Arachnoid Hemorrhage (SAH) with the Pipeline shield device under a Single Anti-Platelet Therapy (SAPT). The article concluded the PED-Shield to be safe to use in the acute treatment of ruptured intracranial aneurysms with SAPT.
However, in this small series, the authors reported one case of total stent occlusion and two cases of platelet aggregation noted on the PED-Shield device requiring to switch from single to dual antiplatelet treatment. Considering those three patients, thrombotic complications have been observed in 21.4 % of cases (3/14) in the acute period. Furthermore, in two cases (14.3 %), the authors reported rebleeding of the culprit aneurysm leading to patient death, pointing out the fact that flow-diverter devices may not immediately prevent the risk of aneurysm rerupture.
Anti-thrombogenic coating might have an added value in case of very specific aneurysms cases requiring the placement of a stent in the acute phase after rupture. Those specific cases are mainly dissecting or blister aneurysms for which endovascular or even surgical approach are difficult and carry a high risk of morbi-mortality[2][3]. In case of endovascular treatment, recent meta-analyses have demonstrated the superiority of reconstructive techniques with flow diverter against deconstructive techniques despite the need for dual antiplatelet treatment with standard flow-diverter[4]. However, those specific aneurysms represent a tiny minority of the endovascularly treated aneurysms. They account for only 0.3% to 1% of intracranial aneurysms and 0.9% to 6.5% of ruptured aneurysms[5–7]. Most of others ruptured aneurysms are suitable for surgical or endovascular treatments without any need for use of a stent in the acute phase. Also, in case of implantation of stent for unruptured aneurysms, the need for dual antiplatelet treatment does not carry a high ischemic risk[8] and new antiplatelet treatments such as ticagrelor or prasugrel [9–14] recently helped to reduce the occurrence of this complication.
Beyond the potential indication of using PED shield with SAPT for ruptured aneurysms, we assume that PED-shield might be used with a SAPT in the setting of unruptured aneurysms. We recently followed this strategy in the setting of a 47-year-old woman allergic to aspirin with a medical history of severe angioedema and harboring an unruptured 11 mm left ophthalmic aneurysm. This patient had been premedicated with ticagrelor (Brilique 90 mg twice a day) and a PED-shield had been implanted without any trouble with perfect wall apposition improved with intra-stent balloon angioplasty. As routinely performed in our center, an MRI has been performed 24 hours after the treatment showing a normal patency of the flow-diverter without any ischemic complication on the DWI sequence. The patient has been discharged home on day 3 but suffered on day 6 of transient right arm palsy and right eye ptosis. An MRI has been performed in emergency depicting minimal ischemic lesions in the right MCA territory with an acute thrombosis of the stent. This patient improved with no deficit thanks to efficient supply from the Willis polygon. However, this observation illustrates the risk of using SAPT in the setting of intracranial stent implantation for the treatment of aneurysms even with the PED-shield device which has been presumed to carry low thrombogenicity.
In light of this observation, we would recommend further randomized clinical trial in order to assess the security of using SAPT treatment with the PED-Shield flow-diverter device.
REFERENCES
1 Manning NW, Cheung A, Phillips TJ, et al. Pipeline shield with single antiplatelet therapy in aneurysmal subarachnoid haemorrhage: multicentre experience. Journal of NeuroInterventional Surgery 2019;11:694–8. doi:10.1136/neurintsurg-2018-014363
2 Gonzalez AM, Narata AP, Yilmaz H, et al. Blood blister-like aneurysms: Single center experience and systematic literature review. European Journal of Radiology 2014;83:197–205. doi:10.1016/j.ejrad.2013.09.017
3 Kaschner MG, Kraus B, Petridis A, et al. Endovascular treatment of intracranial ‘blister’ and dissecting aneurysms. The Neuroradiology Journal 2019;32:353–65. doi:10.1177/1971400919861406
4 Rouchaud A, Brinjikji W, Cloft HJ, et al. Endovascular Treatment of Ruptured Blister-Like Aneurysms: A Systematic Review and Meta-Analysis with Focus on Deconstructive versus Reconstructive and Flow-Diverter Treatments. American Journal of Neuroradiology 2015;36:2331–9. doi:10.3174/ajnr.A4438
5 Nakagawa F, Kobayashi S, Takemae T, et al. Aneurysms protruding from the dorsal wall of the internal carotid artery. Journal of Neurosurgery 1986;65:303–8. doi:10.3171/jns.1986.65.3.0303
6 McLaughlin N, Laroche M, Bojanowski MW. Blister-like aneurysms of the internal carotid artery – management considerations. Neurochirurgie 2012;58:170–7. doi:10.1016/j.neuchi.2012.02.025
7 Abe M, Tabuchi K, Yokoyama H, et al. Blood blisterlike aneurysms of the internal carotid artery. Journal of Neurosurgery 1998;89:419–24. doi:10.3171/jns.1998.89.3.0419
8 Pikis S, Mantziaris G, Mamalis V, et al. Diffusion weighted image documented cerebral ischemia in the postprocedural period following pipeline embolization device with shield technology treatment of unruptured intracranial aneurysms: a prospective, single center study. Journal of NeuroInterventional Surgery 2019;:neurintsurg-2019-015363. doi:10.1136/neurintsurg-2019-015363
9 Atallah E, Saad H, Bekelis K, et al. The use of alternatives to clopidogrel in flow-diversion treatment with the Pipeline embolization device. Journal of Neurosurgery 2018;129:1130–5. doi:10.3171/2017.5.JNS162663
10 Moore JM, Adeeb N, Shallwani H, et al. A Multicenter Cohort Comparison Study of the Safety, Efficacy, and Cost of Ticagrelor Compared to Clopidogrel in Aneurysm Flow Diverter Procedures. Neurosurgery Published Online First: 5 May 2017. doi:10.1093/neuros/nyx079
11 Choi HH, Lee JJ, Cho YD, et al. Antiplatelet Premedication for Stent-Assisted Coil Embolization of Intracranial Aneurysms: Low-Dose Prasugrel vs Clopidogrel. Neurosurgery 2018;83:981–8. doi:10.1093/neuros/nyx591
12 Soize S, Foussier C, Manceau P-F, et al. Comparison of two preventive dual antiplatelet regimens for unruptured intracranial aneurysm embolization with flow diverter/disrupter: A matched-cohort study comparing clopidogrel with ticagrelor. Journal of Neuroradiology Published Online First: February 2019. doi:10.1016/j.neurad.2019.01.094
13 Barra ME, Berger K, Tesoro EP, et al. Periprocedural Neuroendovascular Antiplatelet Strategies for Thrombosis Prevention in Clopidogrel‐Hyporesponsive Patients. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy 2019;39:317–34. doi:10.1002/phar.2228
14 Dmytriw AA, Phan K, Salem MM, et al. The Pipeline Embolization Device: Changes in Practice and Reduction of Complications in the Treatment of Anterior Circulation Aneurysms in a Multicenter Cohort. Neurosurgery Published Online First: 12 March 2019. doi:10.1093/neuros/nyz059
We thank the respondents for providing their case experience and allowing further discussion of this important topic. We would first like to draw attention to specific points in the described case, before discussing some of the more general issues raised.
The respondents report one of many scenarios in which it may be undesirable to use dual antiplatelet therapy in the elective treatment of intracranial aneurysms with flow diverters. In this case, a 47-year-old woman with an 11 mm left ophthalmic aneurysm harbours a significant aspirin allergy. A single Pipeline Shield device under cover of ticagrelor was used to treat the aneurysm. The patient was well at discharge on postoperative day three but then developed symptomatic stent thrombosis on day 6. We draw attention to three points:
1) The respondents state that the stent achieved “perfect wall apposition improved with intra-stent balloon angioplasty.” Setting aside the impossibility of improving “perfect wall apposition” with angioplasty, this does allude to the increasingly understood importance of flow diverter wall apposition. [1] However, digital subtraction angiography assess stent apposition poorly. [1] The use of angioplasty suggests that there may have been some initial concern. Moreover, angioplasty itself may contribute to thrombosis if it promotes activation of the extrinsic clotting pathway by disrupting the endothelial layer. The phosphocholine “Shield” layer reduces thrombosis and platelet...
We thank the respondents for providing their case experience and allowing further discussion of this important topic. We would first like to draw attention to specific points in the described case, before discussing some of the more general issues raised.
The respondents report one of many scenarios in which it may be undesirable to use dual antiplatelet therapy in the elective treatment of intracranial aneurysms with flow diverters. In this case, a 47-year-old woman with an 11 mm left ophthalmic aneurysm harbours a significant aspirin allergy. A single Pipeline Shield device under cover of ticagrelor was used to treat the aneurysm. The patient was well at discharge on postoperative day three but then developed symptomatic stent thrombosis on day 6. We draw attention to three points:
1) The respondents state that the stent achieved “perfect wall apposition improved with intra-stent balloon angioplasty.” Setting aside the impossibility of improving “perfect wall apposition” with angioplasty, this does allude to the increasingly understood importance of flow diverter wall apposition. [1] However, digital subtraction angiography assess stent apposition poorly. [1] The use of angioplasty suggests that there may have been some initial concern. Moreover, angioplasty itself may contribute to thrombosis if it promotes activation of the extrinsic clotting pathway by disrupting the endothelial layer. The phosphocholine “Shield” layer reduces thrombosis and platelet aggregation via retarding activation of clotting factor XII at the stent surface(intrinsic pathway). It will exert no effect on the extrinsic pathway should it be activated.
2) The respondents do not mention if platelet response to ticagrelor was tested. This is not a criticism, as it is common to use new generation P2Y12 antagonists without testing their efficacy. However, it is also essential to understand ticagrelor resistance is reported in up to 3% of the population. [2] More recently we have experienced stent occlusions using the aspirin only technique in patients who were subsequently found to be aspirin non-responders. Given the loss of redundancy inherent in a single antiplatelet therapy (SAPT) technique, anti-platelet response testing should be reconsidered.
3) The timing of the symptomatic stent occlusion is noteworthy as it occurred three days post-discharge from hospital. This is somewhat unusual, as typically SAPT stent occlusions occur more acutely. Ticagerol’s shorter half-life means its effect will be subtherapeutic if discontinued for three days. Therefore, given the suspicious timing of the patient's stent occlusion, issues with patient compliance should be considered.
The use of the Pipeline Shield with aspirin only is a technique reserved for extreme circumstances. In aneurysmal SAH patients who cannot be treated successfully without the use of a stent and in whom antiplatelet therapy is undesirable or contraindicated, this technique may be appropriate. This has been stated in the original manuscript and is apparent in the description of the aneurysms treated. Our study reported an asymptomatic thromboembolic complication rate of 14.3% and a symptomatic rate of 7.1%. [3] Given the complex and extremely challenging nature of these cases, a symptomatic thromboembolic complication rate of 7.1% may be acceptable. The respondents point out that the combined rate is 21.4%. Any complications are too many, however, bear in mind that the thromboembolic complication rate in CLARITY was 12.5%. [4] Given the latter patient population underwent routine aneurysm coiling and the former represents patients without conventional treatment options this difference seems acceptable. Moreover, the symptomatic thromboembolic complication rate in the SAPT study is not very different from that reported in either ASPIRe or IntrePED. [5,6]
We believe that the addition of the phosphocholine "Shield" surface modification may allow greater flexibility in the use of antiplatelet therapy. It is increasingly clear that antiplatelet therapy is central to complications with flow diverting stents. Recent meta-analysis demonstrates that a hypo-response to DAPT increases the risk of thromboembolic complications and that a hyper-response increases the risk of haemorrhagic complications. [7] In the setting of acute aneurysm rupture, antiplatelet therapy is even more critical. The use of DAPT in such patients has been shown to increase the risks of haemorrhagic complications dramatically. [8] Therefore, any technology that may reduce our reliance on antiplatelet therapy should be fully explored.
The use of SAPT in complex ruptured aneurysms has nuanced advantages that may not be apparent to all readers. Many operators choose to delay treatment of ruptured aneurysms requiring a flow diverter to avoid dual antiplatelet therapy (DAPT) acutely. The mean time from subarachnoid haemorrhage (SAH) to aneurysm treatment in our study was one day. In comparison to 7 days in the meta-analysis reported by Cagnazzo et al. [9] This is particularly important to bear in mind when comparing retrospective studies as patients who re-ruptured before treatment will be excluded from analysis, however, undoubtedly represent treatment failure. This point is highlighted in a recent report by ten Brick and colleagues. [10] In this study, 44 patients with aneurysmal subarachnoid haemorrhage were treated with flow diverting stents on DAPT in five European centres. A range of flow diverters were used. However, none had the "Shield" surface modification. Aneurysm and patients characteristics were similar to the SAPT series we reported. Aneurysms were treated acutely (mean of three days post-SAH), although not as early as in the SAPT study. [3] Procedure-related complications occurred in 44%. Permanent neurological deficit due to procedure-related complications occurred in 27%. Perhaps of most interest to this discussion, post-procedure intracranial bleeding occurred in 22.7% including five aneurysm re-bleeds, two intraparenchymal haemorrhages not related to ventriculostomy, two ventricular shunt-related haemorrhages and one extra-axial haemorrhage. With extracranial haemorrhagic complications reported in another 9.1%. [10] In the SAPT series, two aneurysm re-bleeds occurred with one patient making a full recovery. The other patient, who was already a poor WFNS grade SAH never regained consciousness from the initial ictus. Both patients were being treated with postoperative heparin infusion at the time of aneurysm re-rupture, a practice which was subsequently abandoned as described in the original manuscript. In spite of similar patient and aneurysm characteristics, the DAPT case series reported good clinical outcomes in 45.5% compared to 64.3% in the SAPT series. Procedure-related mortality occurred in 11.4% and 7.1% of the DAPT and SAPT series, respectively.
Finally, it is essential to clarify the use of the term single antiplatelet therapy (SAPT). We are forced to take considerable responsibility for any misuse of this term. One should not consider the use of only one agent as the goal of this work. Instead, the goal is to offer patients with complex ruptured aneurysms, treatment using flow diverting (or other) stents with as little suppression of platelet function as is possible to avoid thromboembolic complications. The term "single antiplatelet therapy" has begun to appear in the literature, regarding the use of agents such as ticagrelor and prasugrel. Such terminology is literally correct; however, it misses the real goal of this work. As compared to P2Y12 antagonists, aspirin has a much less pronounced effect on platelet function and naturally, less haemorrhagic complication risk. [11] In most patients, agents such as ticagrelor and prasugrel likely suppress platelet function to such a degree that they make the addition of aspirin mostly redundant. [12,13]
We agree that further investigation of this technique is required, as stated in the original article. However, we think a randomised control trial would be inappropriate at this stage. It is somewhat facile to call for a randomised control trial to solve all medical dilemmas. It is safe to say that no treatment option is established for patients harbouring such complex ruptured aneurysms and therefore deciding on a control group would be somewhat arbitrary.
Furthermore, it is likely to be challenging to detect a significant difference in clinical outcomes given the uncertain prognosis of such patients independent of how their aneurysms are secured. Therefore, the number of patients that would need to be randomised likely exceeds 1000. We have elected to take a different approach and will study the Pipeline Shield SAPT technique in a multicentre, single-arm, prospective study. We hope that the first patients will be enrolled before this letter is published. The study will begin in Australia; however, it is planned to expand to other countries, and we encourage experienced and skilled centres such as yours to be a part of this investigation.
Thank you once again for an engaging discussion. Warm regards,
Nathan Manning
References:
1 Rouchaud A, Ramana C, Brinjikji W, et al. Wall Apposition Is a Key Factor for Aneurysm Occlusion after Flow Diversion: A Histologic Evaluation in 41 Rabbits. Am J Neuroradiol 2016;37:2087–91. doi:10.3174/ajnr.a4848
2 Warlo EM, Arnesen H, Seljeflot I. A brief review on resistance to P2Y12 receptor antagonism in coronary artery disease. Thrombosis J 2019;17:11. doi:10.1186/s12959-019-0197-5
3 Manning NW, Cheung A, Phillips TJ, et al. Pipeline shield with single antiplatelet therapy in aneurysmal subarachnoid haemorrhage: multicentre experience. Journal of neurointerventional surgery 2019;11:694–8. doi:10.1136/neurintsurg-2018-014363
4 Pierot L, Cognard C, Anxionnat R, et al. Ruptured intracranial aneurysms: factors affecting the rate and outcome of endovascular treatment complications in a series of 782 patients (CLARITY study). Radiology 2010;256:916–23. doi:10.1148/radiol.10092209
5 Kallmes DF, Brinjikji W, Boccardi E, et al. Aneurysm Study of Pipeline in an Observational Registry (ASPIRe). Interventional Neurology 2016;5:89–99. doi:10.1159/000446503
6 Kallmes D, Hanel R, Lopes D, et al. International retrospective study of the pipeline embolization device: a multicenter aneurysm treatment study. American Journal of Neuroradiology 2015;36:108–15. doi:10.3174/ajnr.a4111
7 Ajadi E, Kabir S, Cook A, et al. Predictive value of platelet reactivity unit (PRU) value for thrombotic and hemorrhagic events during flow diversion procedures: a meta-analysis. J Neurointerv Surg 2019;11:1123. doi:10.1136/neurintsurg-2019-014765
8 Hudson JS, Prout BS, Nagahama Y, et al. External Ventricular Drain and Hemorrhage in Aneurysmal Subarachnoid Hemorrhage Patients on Dual Antiplatelet Therapy: A Retrospective Cohort Study. Neurosurgery Published Online First: 2018. doi:10.1093/neuros/nyy127
9 Cagnazzo F, di Carlo D, Cappucci M, et al. Acutely Ruptured Intracranial Aneurysms Treated with Flow-Diverter Stents: A Systematic Review and Meta-Analysis. American Journal of Neuroradiology 2018;39:1669–75. doi:10.3174/ajnr.a5730
10 ten Brinck MF, Jäger M, de Vries J, et al. Flow diversion treatment for acutely ruptured aneurysms. J Neurointerv Surg 2019;:neurintsurg-2019-015077. doi:10.1136/neurintsurg-2019-015077
11 Gresele P. Antiplatelet agents in clinical practice and their haemorrhagic risk. Blood Transfus Trasfusione Del Sangue 2013;11:349–56. doi:10.2450/2013.0248-12
12 KIRKBY N, LEADER P, CHAN M, et al. Antiplatelet effects of aspirin vary with level of P2Y12 receptor blockade supplied by either ticagrelor or prasugrel. J Thromb Haemost 2011;9:2103–5. doi:10.1111/j.1538-7836.2011.04453.x
13 Warner TD, Nylander S, Whatling C. Anti‐platelet therapy: cyclo‐oxygenase inhibition and the use of aspirin with particular regard to dual anti‐platelet therapy. Brit J Clin Pharmaco 2011;72:619–33. doi:10.1111/j.1365-2125.2011.03943.x
Thank you for your technical considerations regarding stent in stent placement without hooking the first stent.
Use of 3 D Roadmap may be helpful. Moreover, following passage of the microwire, reconstructions of a second flat panel angioCT with the microwire in place clearly outlines the relationship between the microwire and the struts of the first stent, especially if reconstructions perpendicular to the orientati...
Thank you for your technical considerations regarding stent in stent placement without hooking the first stent.
Use of 3 D Roadmap may be helpful. Moreover, following passage of the microwire, reconstructions of a second flat panel angioCT with the microwire in place clearly outlines the relationship between the microwire and the struts of the first stent, especially if reconstructions perpendicular to the orientation of the first stent are performed. This may avoid the passage of the stent with a DAC (Distal Access Catheter) which may demage or displace the first stent, especially at it's proximal end.
Using this technique, even passage through a deformed and fragmented stent can be performed, and the location of the microwire within the central axis of the first stent can be confirmed before introduction of the second stent or flow diverter.
Mordasini P, Al-Senani F, Gralla J, Do D, Schroth G: The use of flat Panel angioCT (DynaCT) for Navigation through a deformed and fractured carotid stent. Neuroradiology 52; 2010:629-632.
Prof. Gerhard Schroth Neurologist and Radiologist gerhard.schroth@insel.ch University of Bern Senior Consultant of the Institute for Diagnostic and Interventional Neuroradiology
I read with interest the article "Onyx extrusion through the scalp
after embolization of dural arteriovenous fistula" by SIngla et al. The
objective of the article is mainly to reinforce in our pre-treatment
discussions the need to include as much as possible the outcomes arising
from the use of Onyx that include, but are not exclusively related to,
micro catheter retention or rupture, unintended ves...
I read with interest the article "Onyx extrusion through the scalp
after embolization of dural arteriovenous fistula" by SIngla et al. The
objective of the article is mainly to reinforce in our pre-treatment
discussions the need to include as much as possible the outcomes arising
from the use of Onyx that include, but are not exclusively related to,
micro catheter retention or rupture, unintended vessel occlusion and the
possibility of stroke.
The use of the occipital artery is a valid means of occluding
fistulas of the transverse sinus and on occasion the torcula however there
is the risk of skin necrosis especially if reflux is obtained to the
degree seen here in this case, filling the contralateral occipital artery.
We have used the middle meningeal artery to occlude such superior
sagittal sinus, other transverse sinus and higher grade dural fistulas
with detachable tip micro catheters such as Apollo (eV3) and Sonic (BALT)
and dual lumen remodelling balloons (Sceptre C, Microvention)1.
Perhaps use of the middle meningeal artery feeders would have pre-
empted surgery and completed the treatment endovascularly as well as
avoiding possible skin necrosis from use of the occipital arteries and
potential skin extrusion of the embolic agent. We understand that a
complete endovascular cure may not be achievable without venous balloon
remodelling using the Copernic balloon ( BALT) but the same
clinical/angiographic result can be obtained with careful monitoring of
the sinus lumen on an anteroposterior view. It may be that both middle
meningeal arteries arose from the ophthalmic arteries, thus precluding
their use. This would be very unusual however.
Thus in summary, use of the middle meningeal arteries in this case
may have precluded the surgery and having to expose the occipital artery
territories and scalp to the liquid embolic agent. Thus possible skin
necrosis and the observed material extrusion could be avoided.
Reference:
1. Preliminary experience with the liquid embolic material agent PHIL
(Precipitating Hydrophobic Injectable Liquid) in treating cranial and
spinal dural arteriovenous fistulas: technical note
Joe J Leyon, Swarupsinh Chavda, Allan Thomas and Saleh Lamin
J NeuroIntervent Surg published online May 20, 2015
It is with great interest that we read the study of Alawieh et al(1), in which they developed a machine learning algorithm, called ‘SPOT’, to select stroke patients older than 80 years for endovascular therapy (EVT). Prediction modeling to optimize patient selection for EVT is an emerging topic of interest and we agree that predicting individual patient outcomes is increasingly important for decision making in medicine. However, we were surprised by the strong conclusions that were drawn by the authors, considering some serious limitations of the study.
First, the size of the training set is insufficient to develop a complex model with twelve predictor variables and many correlations. Only 22 patients had a good functional outcome, which means that the number of events per tested predictor variable is less than two. For the development of a reliable model, a sample size of at least ten events per variable is needed to minimize the risk of overfitting(2, 3). It has been suggested that even far more events per variable are needed to achieve stable predictions with machine learning techniques(4). Especially complex models developed on small sample sizes have a high risk of overfitting, resulting in unstable predictions and too optimistic model performance measures. The reported AUC of 0.92 is therefore very likely to be an overestimation.
Second, the SPOT algorithm provides a treatment advice based on the predicted outcome after treatment, without providing the...
It is with great interest that we read the study of Alawieh et al(1), in which they developed a machine learning algorithm, called ‘SPOT’, to select stroke patients older than 80 years for endovascular therapy (EVT). Prediction modeling to optimize patient selection for EVT is an emerging topic of interest and we agree that predicting individual patient outcomes is increasingly important for decision making in medicine. However, we were surprised by the strong conclusions that were drawn by the authors, considering some serious limitations of the study.
First, the size of the training set is insufficient to develop a complex model with twelve predictor variables and many correlations. Only 22 patients had a good functional outcome, which means that the number of events per tested predictor variable is less than two. For the development of a reliable model, a sample size of at least ten events per variable is needed to minimize the risk of overfitting(2, 3). It has been suggested that even far more events per variable are needed to achieve stable predictions with machine learning techniques(4). Especially complex models developed on small sample sizes have a high risk of overfitting, resulting in unstable predictions and too optimistic model performance measures. The reported AUC of 0.92 is therefore very likely to be an overestimation.
Second, the SPOT algorithm provides a treatment advice based on the predicted outcome after treatment, without providing the absolute probability of good functional outcome or the treatment benefit. Rational treatment decisions should be based on expected outcome with treatment compared to the expected outcome without treatment. A low likelihood of good outcome does not imply absence of treatment benefit. Besides that, many octogenarians might not be able to achieve complete recovery to functional independence, but an improvement from an mRS score of 4-5 to a score of 3 as a result of EVT can still be very relevant in clinical practice, and such outcome is not covered by the SPOT algorithm.
Well-developed prediction models may guide us in the selection of patients that benefit from treatment, but external validation in a large validation set is always needed before these models are implemented in everyday clinical care. The SPOT algorithm does not yet fulfill the minimum requirements for a well-developed and validated decision support tool. This means that an effective treatment may be withheld from patients who could benefit from it. Therefore, the SPOT algorithm should not yet be implemented in clinical care.
References
1. Alawieh A, Zaraket F, Alawieh MB, Chatterjee AR, Spiotta A. Using machine learning to optimize selection of elderly patients for endovascular thrombectomy. J Neurointerv Surg. 2019.
2. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373-9.
3. Harrell FE, Jr., Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15(4):361-87.
4. van der Ploeg T, Austin PC, Steyerberg EW. Modern modelling techniques are data hungry: a simulation study for predicting dichotomous endpoints. BMC Med Res Methodol. 2014;14:137.
We read with interest the response to our manuscript on using machine learning to optimize elderly patient selection for endovascular thrombectomy (1). We acknowledge here, as the author reports, the limitation of SPOT being based on single center data, and the need for multicenter prospective validation of SPOT as next step in development. The author raises additional technical concerns that we do not necessarily view as applicable to this study.
First, we would like to stress the general limitations of artificial intelligence based techniques such as the overfitting and the data specific local optima problems. However, the specific comments brought by the author are not applicable in our case. First, studies on the number of events per predictor are applicable for logistic regressions (LRs) which is not used in the SPOT algorithm. In fact, our results show poor LR performance which is consistent with the rule of thumb of 1 to 10 referred to by the author. Hence, while serving as a good guidance for LR, the rule is not binding and more importantly it does not guarantee the generalization of the learned model. To further illustrate, classification models using convolutional neural networks have millions of parameters and are trained with datasets that, in most cases, do not have millions of samples in each group. However, these models have acceptable generalization capabilities and are tested using the data-split method. In SPOT, the model at its core is a regressi...
We read with interest the response to our manuscript on using machine learning to optimize elderly patient selection for endovascular thrombectomy (1). We acknowledge here, as the author reports, the limitation of SPOT being based on single center data, and the need for multicenter prospective validation of SPOT as next step in development. The author raises additional technical concerns that we do not necessarily view as applicable to this study.
First, we would like to stress the general limitations of artificial intelligence based techniques such as the overfitting and the data specific local optima problems. However, the specific comments brought by the author are not applicable in our case. First, studies on the number of events per predictor are applicable for logistic regressions (LRs) which is not used in the SPOT algorithm. In fact, our results show poor LR performance which is consistent with the rule of thumb of 1 to 10 referred to by the author. Hence, while serving as a good guidance for LR, the rule is not binding and more importantly it does not guarantee the generalization of the learned model. To further illustrate, classification models using convolutional neural networks have millions of parameters and are trained with datasets that, in most cases, do not have millions of samples in each group. However, these models have acceptable generalization capabilities and are tested using the data-split method. In SPOT, the model at its core is a regression model with continuous output. More importantly, while the overfitting concern is a valid one with the high area under a curve, SPOT was tested in an adequate fashion using data-splitting, a well-accepted validation scheme to test the generalization capability of the model, and thus detect overfitting. In fact, studies discussing events per variable rule for logistic regression use data-splitting as one of the validation method (2). The approach used for testing SPOT using a prospective data not part of the training is the most stringent approach to test a model. In fact, state of the art machine learning model evaluations use the data-splitting technique, and consider the performance of the testing data as the golden metric to judge upon the model’s generalization and over-fitting (3,4).
While we agree with the fact that machine learning are data hungry, the size of the dataset is highly dependent on the problem at hand. For example, in sparse regression models, the number of samples in the training dataset can be orders of magnitude smaller than that of the parameters. However, sparse regression models joined with proper training techniques are able to generalize to unseen data (5). Again, this ability is tested using the data-split method which was used to test SPOT.
Further, we emphasize that although SPOT returns a continuous output of mRS scores, the tool will specifically report grouped outcomes into mRS 0-2 and mRS 3-6. The decision to choose this dichotomy of outcomes was to be consistent with clinical trials that predominantly report functional independence as outcome measure to guide interventions even when elderly patients were included. In response to the concern about returning probability for outcomes, and as stated in the manuscript, when SPOT returns poor outcome prediction, its negative predictive value was 95.2% which represents the probability for a patient that screened negative to have a poor outcome and this probability is reported in the text. In the current form, SPOT does not return a probability for every mRS score.
Finally, while we agree that multicenter data is needed to additionally validate SPOT as a tool as stated in the manuscript, the current version of SPOT does meet technical requirements for a validated tool. We do stress again that “SPOT is designed to aid clinical decision of whether to undergo ET in elderly patients” (1), and not a stand-alone tool.
References
1. Alawieh A, Zaraket F, Alawieh MB, Chatterjee AR, Spiotta A. Using machine learning to optimize selection of elderly patients for endovascular thrombectomy. J Neurointerv Surg. 2019.
2. Harrell FE, Jr., Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15(4):361-87.
3. Kohavi, Ron. "A study of cross-validation and bootstrap for accuracy estimation and model selection." Ijcai. Vol. 14. No. 2. 1995.
4. Arlot, Sylvain, and Alain Celisse. "A survey of cross-validation procedures for model selection." Statistics surveys4 (2010): 40-79.
5. Donoho, David Leigh, et al. Sparse solution of underdetermined linear equations by stagewise orthogonal matching pursuit. Department of Statistics, Stanford University, 2006.
Dear Editor,
We would like the thank Drs. Berndt, Zimmer, Kaesmacher, and Boeckh-Behrens for their interest in our study titled “Clot permeability and histopathology: Is a clot’s perviousness on CT imaging correlated with its histologic composition?” We read their letter with interest. The authors have been pioneers in stroke clot analysis and we greatly respect their academic rigor and expertise.
While we agree that there are certainly some methodological differences between our two studies, we do not believe that these are to blame for the differences in results. Rather, we feel that the observed differences in results between our studies could be due to differences in our patient populations.
Our group has previously shown that there is indeed a correlation between clot composition and etiology. In a recently published article in Stroke we found that large artery atherosclerosis clots were more likely to be platelet rich than those of a cardioembolic origin.1 To date, however, we have yet to find any definite correlation between etiology and RBC density or fibrin density, and we think it is too early to make any definite conclusions on the association between clot composition and etiology.
We agree that the association between perviousness, clot composition, etiology and clinical outcome is not conclusively clarified yet, and hence warrants further research, especially in a larger patient group in a multi-centric setting. Currently, our gr...
Show MoreThe authors describe two endovascular techniques for delivering IA chemotherapy to retinoblastoma patients. Technique A where a 1.2 Fr or 1.5 Fr micro catheter with continuous verapamil flush is advanced without a guide and technique B where a1.5 Fr or 1.7 Fr micro catheter is advanced within a 4 Fr catheter, through a 4 Fr sheath. We usually use a no sheath technique, using a 4Fr diagnostic catheter as a guide catheter for neonatal and pediatric cases. Most importantly, we do not use Echelon or Marathon micro catheters in neonatal and pediatric patients because their use is contraindicated per "instructions for use".
An increasing number of reports highlight polymer coating embolism as an iatrogenic complication of intravascular medical devices [1-3]. Autopsies, histologic evaluations of thrombectomy specimens, samples of captured debris, resected or biopsied tissues, are available methods used to study polymer emboli post investigative catherizations or interventional procedures. Reported data highlight the prevalence of this phenomenon and/or its clinicopathologic impacts, however, fall short of identifying higher-risk polymer emboli interventional devices. Consequently, an optimal approach for future investigations related to polymer coating embolism is required.
Mehta et. al investigate the histologic frequency of polymer emboli among patients who underwent endovascular thrombectomy for treatment of acute ischemic stroke due to large vessel occlusion by retrospectively evaluating thrombectomy specimens [2]. In this study, the reported frequency of polymer emboli includes the use of various devices and techniques among selected cases. However, literature highlights polymer coating embolism is device specific and dependent on coating integrity measured by particulates released [4]. Thus, the use of alternate devices with higher or lower particulate release for a given procedure may result in a large variation in incidence rates from reported results. Also, as mentioned by the authors, subsequent statistical correlations unless appropriately powered provide limited informatio...
Show MoreProprietary nature of intravascular medical device coatings limits safety testing
Dear Dr. Albuquerque:
We are glad that our work has generated interest and discussion in the field [1]. Four years have elapsed since a need for updated device coating testing was officially announced [2], however complexities on the matter and persistent knowledge gaps limit safety studies of devices currently on the market for clinical intravascular use [3,4]. Standardized in vitro particulate generation testing is needed. However, available literature shows that preclinical device testing is not fully predictive of clinical response. Therefore, in vitro and animal studies cannot replace investigation in humans. Currently, lack of consensus on the following prevent meaningful testing in humans: I) optimal clinical testing methods; ii) definitions of permissible risk; iii) adverse cellular, organ, and temporal-specific effects of distinct coating biomaterials; and iv) effects of pre-existing comorbid conditions. Nevertheless, in vitro testing that does not incorporate clinical data has limited utility for safety guidance. Likewise, in vivo studies that do not incorporate biomaterial factors are incomplete. Thus, the proprietary nature of intravascular device coatings remains a significant limitation to clinical device testing and safety assurances. Growing data [2-6] suggest that it may be time for this to be addressed.
1. Chopra AM, Hu YC, Cruz JP. The Device Specific...
Show MoreWe read with interest the case series by Manning’s et al.[1] using a surface modified flow-diverter stent (Pipeline Flex with Shield Technology, Medtronic Neurovascular, Irvine, California, USA). In this retrospective series, 14 ruptured intracranial aneurysms have been treated in the acute phase after Sub-Arachnoid Hemorrhage (SAH) with the Pipeline shield device under a Single Anti-Platelet Therapy (SAPT). The article concluded the PED-Shield to be safe to use in the acute treatment of ruptured intracranial aneurysms with SAPT.
However, in this small series, the authors reported one case of total stent occlusion and two cases of platelet aggregation noted on the PED-Shield device requiring to switch from single to dual antiplatelet treatment. Considering those three patients, thrombotic complications have been observed in 21.4 % of cases (3/14) in the acute period. Furthermore, in two cases (14.3 %), the authors reported rebleeding of the culprit aneurysm leading to patient death, pointing out the fact that flow-diverter devices may not immediately prevent the risk of aneurysm rerupture.
Anti-thrombogenic coating might have an added value in case of very specific aneurysms cases requiring the placement of a stent in the acute phase after rupture. Those specific cases are mainly dissecting or blister aneurysms for which endovascular or even surgical approach are difficult and carry a high risk of morbi-mortality[2][3]. In case of endovascular treatment, rece...
Show MoreWe thank the respondents for providing their case experience and allowing further discussion of this important topic. We would first like to draw attention to specific points in the described case, before discussing some of the more general issues raised.
The respondents report one of many scenarios in which it may be undesirable to use dual antiplatelet therapy in the elective treatment of intracranial aneurysms with flow diverters. In this case, a 47-year-old woman with an 11 mm left ophthalmic aneurysm harbours a significant aspirin allergy. A single Pipeline Shield device under cover of ticagrelor was used to treat the aneurysm. The patient was well at discharge on postoperative day three but then developed symptomatic stent thrombosis on day 6. We draw attention to three points:
1) The respondents state that the stent achieved “perfect wall apposition improved with intra-stent balloon angioplasty.” Setting aside the impossibility of improving “perfect wall apposition” with angioplasty, this does allude to the increasingly understood importance of flow diverter wall apposition. [1] However, digital subtraction angiography assess stent apposition poorly. [1] The use of angioplasty suggests that there may have been some initial concern. Moreover, angioplasty itself may contribute to thrombosis if it promotes activation of the extrinsic clotting pathway by disrupting the endothelial layer. The phosphocholine “Shield” layer reduces thrombosis and platelet...
Show MoreThank you for your technical considerations regarding stent in stent placement without hooking the first stent.
Use of 3 D Roadmap may be helpful. Moreover, following passage of the microwire, reconstructions of a second flat panel angioCT with the microwire in place clearly outlines the relationship between the microwire and the struts of the first stent, especially if reconstructions perpendicular to the orientati...
Dear Sirs,
I read with interest the article "Onyx extrusion through the scalp after embolization of dural arteriovenous fistula" by SIngla et al. The objective of the article is mainly to reinforce in our pre-treatment discussions the need to include as much as possible the outcomes arising from the use of Onyx that include, but are not exclusively related to, micro catheter retention or rupture, unintended ves...
It is with great interest that we read the study of Alawieh et al(1), in which they developed a machine learning algorithm, called ‘SPOT’, to select stroke patients older than 80 years for endovascular therapy (EVT). Prediction modeling to optimize patient selection for EVT is an emerging topic of interest and we agree that predicting individual patient outcomes is increasingly important for decision making in medicine. However, we were surprised by the strong conclusions that were drawn by the authors, considering some serious limitations of the study.
First, the size of the training set is insufficient to develop a complex model with twelve predictor variables and many correlations. Only 22 patients had a good functional outcome, which means that the number of events per tested predictor variable is less than two. For the development of a reliable model, a sample size of at least ten events per variable is needed to minimize the risk of overfitting(2, 3). It has been suggested that even far more events per variable are needed to achieve stable predictions with machine learning techniques(4). Especially complex models developed on small sample sizes have a high risk of overfitting, resulting in unstable predictions and too optimistic model performance measures. The reported AUC of 0.92 is therefore very likely to be an overestimation.
Second, the SPOT algorithm provides a treatment advice based on the predicted outcome after treatment, without providing the...
Show MoreWe read with interest the response to our manuscript on using machine learning to optimize elderly patient selection for endovascular thrombectomy (1). We acknowledge here, as the author reports, the limitation of SPOT being based on single center data, and the need for multicenter prospective validation of SPOT as next step in development. The author raises additional technical concerns that we do not necessarily view as applicable to this study.
First, we would like to stress the general limitations of artificial intelligence based techniques such as the overfitting and the data specific local optima problems. However, the specific comments brought by the author are not applicable in our case. First, studies on the number of events per predictor are applicable for logistic regressions (LRs) which is not used in the SPOT algorithm. In fact, our results show poor LR performance which is consistent with the rule of thumb of 1 to 10 referred to by the author. Hence, while serving as a good guidance for LR, the rule is not binding and more importantly it does not guarantee the generalization of the learned model. To further illustrate, classification models using convolutional neural networks have millions of parameters and are trained with datasets that, in most cases, do not have millions of samples in each group. However, these models have acceptable generalization capabilities and are tested using the data-split method. In SPOT, the model at its core is a regressi...
Show MorePages