Predicting the Safety and Effectiveness of Inferior Vena Cava Filters Study: Design of a unique safety and effectiveness study of inferior vena cava filters in clinical practice

Predicting the Safety and Effectiveness of Inferior Vena Cava Filters Study: Design of a unique safety and effectiveness study of inferior vena cava filters in clinical practice David L. Gillespie, MD, RVT, FACS,a James B. Spies, MD,b F. Sandra Siami, MPH,c John E. Rectenwald, MD,d Rodney A. White, MD,e and Matthew S. Johnson, MD,f Dartmouth and Watertown, Mass; Washington, D.C.; Ann Arbor, Mich; Long Beach, Calif; and Indianapolis, Ind

ABSTRACT Background: Death from venous thromboembolism remains a significant cause of death worldwide. Although anticoagulation is the cornerstone of treatment in patients at risk for venous thromboembolism, inferior vena cava (IVC) filter use has increased exponentially over the last decade driven predominantly by the prophylactic use in patients at risk for venous thromboembolism despite limited evidence supporting this practice. The Predicting the Safety and Effectiveness of Inferior Vena Cava Filters (PRESERVE) Study is being implemented by the Society for Vascular Surgery, Society of Interventional Radiology, U.S. Food and Drug Administration, and several IVC filter manufactures to better understand the safety, effectiveness, and current patterns of real-world use of IVC filters. Methods: The PRESERVE Study includes IVC filters from seven manufacturers: ALN (ALN 6 hook), Argon (Option Elite), B. Braun (LP, Vena Tech Convertible), CR Bard (Denali), Cook (Gunther Tulip), Cordis (OptEase, TrapEase), and Philips Volcano (Crux). The indications for filter placement, filter brand, complications, stability, frequency and success of retrieval, and clinical effectiveness of each filter will be recorded. Approximately 2100 patients (300 for each filter brand included in the study) are intended to be enrolled at 60 U.S. centers. Results: Men and women age 18 years or older requiring IVC filters for prevention of venous thromboembolism will be included in the study if no contrast allergy is present and they are willing to commit to the prescribed study follow-up. Participants will be evaluated at discharge, 3, 6, 12, 18, and 24 months after filter placement and/or 1 month after retrieval, which ever occurs first. Intravascular ultrasound examination or venography will be done before and after IVC filter placement, with abdominal plain film at 3 months, and contrast enhanced computed tomography scans at 12 and 24 months to evaluate filter stability. The primary safety end point is a composite of clinical end points, including freedom from perforation, embolization, thrombosis, recurrent DVT, and defined serious adverse events. Secondary end points include mechanical stability and procedure related complications at 3 months, major adverse events at 6, 12, 18, and 24 months, and filter tilt of more than 15
at any point. Conclusions: The PRESERVE Study represents the largest prospective study ever undertaken to investigate real-world outcomes with contemporary use of IVC filters. The investigators await results with the hope that it can improve patient care. (J Vasc Surg: Venous and Lym Dis 2019;-:1-8.) Keywords: Inferior vena cava; Filter; Venous thromboembolism; Pulmonary embolism

Pulmonary embolism (PE), the most severe consequence of venous thromboembolism, represents one of the greatest sources of morbidity among hospitalized patients around the world, with an annual incidence of

400,000 and resulting in 250,000 deaths annually in the United States alone.1 Anticoagulation has historically been the first-line therapy for venous thrombosis and prophylactic prevention of PE. However, in patients in

From the Department of Vascular and Endovascular Surgery, Southcoast

(a Becton Dickinson Company)., Cordis Corporation, (a Cardinal Health Com-

Health System, Dartmoutha; the Department of Radiology, MedStar Georgetown University Hospital, Washington, D.C.b; the HealthCore-NERI, Watertownc; the Section of Vascular Surgery, Department of Surgery, University of

pany), and Philips Volcano. Author conflict of interest: D.L.G is a consultant for Cook, Inc. M.S.J. is a consultant for Argon Medical.

Michigan, Ann Arbord; the Department of Radiology & Imaging Sciences,

Additional material for this article may be found online at www.jvsvenous.org.

Heart and Vascular Institute Long Beach Memorial Care, Long Beache; and

Correspondence: David L. Gillespie, MD, RVT, FACS, Department of Vascular and

the Department of Vascular Surgery, Indiana University School of Medicine,

Endovascular Surgery, Southcoast Health System, 300 A Faunce Corner Rd,

Indianapolis.f

Dartmouth, MA (e-mail: [email protected]).

Sponsored by the Inferior Vena Cava Filter Study Group Foundation (IVCFSGF),

The editors and reviewers of this article have no relevant financial relationships to

a joint effort between the Society for Vascular Surgery (SVS) and the Society

disclose per the Journal policy that requires reviewers to decline review of any

of Interventional Radiology. Funding for this study has been provided to the IVCFSGF by ALN Implant Chirurgicaux, Argon Medical Devices Inc., B. Braun Interventional Systems Inc., Cook Inc, C.R. Bard Peripheral Vascular Inc

manuscript for which they may have a conflict of interest. 2213-333X Copyright Ó 2019 by the Society for Vascular Surgery. Published by Elsevier Inc. https://doi.org/10.1016/j.jvsv.2019.07.009

1

2

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders ---

whom anticoagulation is contraindicated or in whom it has failed, inferior vena cava (IVC) filters are commonly placed to prevent death from PE, though the evidence supporting their use is limited.2 The growth in the use of vena cava filters may be due in part to altered perceptions around the risk/benefit ratio, owing to the availability of retrievable devices. Despite relatively little evidence of effectiveness in purely prophylactic settings, IVC filters have been placed in trauma patients, bariatric surgery patients, and those who have limited pulmonary reserve.3-5 As an example of the marked growth in their use, one study showed that between 1999 and 2008, IVC filter placement procedures increased by 111.5% (from 30,756 to 65,041) in the Medicare population alone.6 Beyond whether IVC filters are indicated in particular any clinical setting, the safety of some retrievable IVC filters remains a major question. The marked growth in their use is a source of concerns among clinicians and regulators, as reports of filter fracture, filter embolization, and vena cava penetration of filter parts have emerged. In August 2010 the U.S. Food and Drug Administration (FDA) issued a warning letter concerning a perceived increase in IVC filter-related complications.7 This warning was followed by an updated FDA Safety Communication in May 2014 based on newly published research and postmarket surveillance for these devices, which reported device migration, filter fracture, embolization, perforation of the IVC, as well as difficulty removing the device.8 In addition, the FDA was concerned that the retrievable IVC filters were not always being removed once the risk of PE had subsided. In response the FDA formally recommended that physicians consider removing retrievable IVC filters as soon as protection from PE was no longer needed. This recommendation was codified in a decision analysis publication by the agency.9 This concern regarding safety and effectiveness of the IVC filters led to the collaboration of the Society of Interventional Radiology, the Society for Vascular Surgery, the FDA, and seven manufacturers of IVC filters (ALN [Ghisonaccia, France], Argon [Frisco, Tex], B. Braun [Melsungen, Germany], C.R. Bard [Tempe, AZ], Cook [Bloomington, Ind], Cordis [Milpitas, Calif], and Philips Volcano [San Diego, Calif]) to study the current patterns of use and safety of these devices in clinical practice.

METHODS The Predicting the Safety and Effectiveness of Inferior Vena Cava Filters (PRESERVE) Study is a prospective, multicenter, open-label, nonrandomized, investigational study to evaluate the safety and effectiveness of commercially available IVC filters from participating manufacturers placed in patients for the prevention of death from fatal or symptomatic PE. The overall goal of the study is to characterize the current practice of IVC

2019

ARTICLE HIGHLIGHTS d

d

d

Type of Research: Prospective, multicenter, openlabel, nonrandomized, investigational study Key Findings: Clinical equipoise regarding safety and efficacy of seven of the available U.S. Food and Drug Administration-approved inferior vena cava filters are investigated using an investigational device exemption (IDE) alternative to U.S. Food and Drug Administration 522 Postmarket inferior vena cava filter study. Take Home Message: The Predicting the Safety and Effectiveness of Inferior Vena Cava Filters study is a collaborative research study designed to investigate several different filters across seven manufacturers for a proof-of-principle IDE in an effort to maximize the generalizability and relevance of the study. The protocol is designed to collect real-world evidence and contemporary standard of care procedures in a pragmatic study design.

filter placement, including the indications, filter brand, the safety of placement, the mechanical stability of the device, frequency and success of filter removal, and their clinical effectiveness in preventing PE and the incidence of recurrent deep vein thrombosis. The PRESERVE Study holds an investigational device exemption (IDE) with the FDA in lieu of mandating a Postmarket Surveillance Study under Section 522 of the Federal Food, Drug, and Cosmetic Act, as required for other manufacturers not participating in PRESERVE. The PRESERVE IDE is sponsored by the IVC Filter Study Group Foundation (IVCFSGF), the 501(c) (3) not-for-profit entity of this joint collaboration between the Society of Interventional Radiology and the Society for Vascular Surgery. The study is currently funded by seven participating manufacturers including ALN Implants Chirurgicaux (ALN), Argon Medical Devices Inc (Argon, manufactured by Rex Medical), B. Braun Interventional Systems Inc (B. Braun), C.R. Bard Peripheral Vascular (Bard), Cook Inc (Cook), Cordis Corporation (Cordis), and Philips Volcano Corporation (Crux). The study is being executed by the PRESERVE Steering Committee and New England Research Institutes, Inc (now acquired by HealthCore, a subsidiary of Anthem, Inc). Filter manufacturers have no involvement in the execution of the study or analysis of the data. The foundation has the right to publish study data or other findings relating to the study, but filter manufacturers will be provided the opportunity to suggest input for PRESERVE study publications. The study intends to enroll 2100 patients in aggregate, 300 patients for each filter, at approximately 60 U.S. centers. The current filters being used in PRESERVE include ALN, with and without hook (ALN), Option Elite Retrievable (Argon), VenaTech LP and VenaTech Convertible (B. Braun), DENALI (Bard), Günther-Tulip (Cook), Crux

Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume

-,

Number

Gillespie et al

3

-

(Philips Volcano), and OPTEASE Retrievable and TRAPEASE Permanent (Cordis). The study is conducted in compliance with the protocol, International Conference for Harmonization Good Clinical Practice Guidelines (ICH e6), 21 CFR 50 Protection of Human Subjects, 21 CFR 812 IDE, the Declaration of Helsinki, and the Health Insurance Portability and Accountability Act. The experimental protocol and informed consent were approved by a site or central institutional review board and all patients gave informed consent. Eligibility criteria. Given the desire to collect real-world evidence based on current clinical practice, the PRESERVE Study has broad eligibility criteria. Inclusion is open to men and women age 18 years or older requiring an IVC filter for prevention of PE. The patient, or legally authorized representative, must provide written consent using an institutional review board-approved informed consent for the study and be willing to comply with the follow-up evaluation schedule. Patients are excluded if they have a known sensitivity to contrast or are unable to participate in the pretreatment and post-treatment evaluations. It is up to the treating investigator to determine which of the participating filters is appropriate for a particular patient. Follow-up schedule and imaging. After placement, patients are evaluated at discharge, 3 months, 6 months (phone), 12 months, 18 months (phone), and 24 months or until 1 month after filter retrieval, whichever occurs first. The assessments at baseline and follow-up are shown in Table I. The radiologic imaging protocol follows standard of care guidelines. During the filter placement, a cavagram or correlative intravascular ultrasound examination before placement; and abdominal radiographs, digital subtraction angiography runs, or intravascular ultrasound examination immediately after placement. Additional imaging is also required: an abdominal radiographs at 3 month after placement and a computed tomography scan with contrast at 12 and 24 months after placement if the filter has not yet been retrieved. At retrieval, abdominal radiographs, on-table digital radiographs, or early (noncontrast) images from digital subtraction angiography runs may be obtained both before and after retrieval. The imaging follow-up is designed to determine the stability of the filter placement (lack of migration or filter embolization), the physical integrity of the device, and secondary complications related to caval penetration or thrombosis of the filter. The protocol was developed to reflect current practice as closely as possible while achieving the aims of the study. Hence, primary safety and effectiveness end points will be evaluated based on site data, but, because imaging after insertion is not standard of care, especially in

asymptomatic patients, the baseline, 3-month, 12-month, and 24-month images will be evaluated by a core imaging laboratory at the University of Virginia, Department of Radiology and Medical Imaging. This strategy will provide standardized reads at set timepoints, in accordance with protocol definitions, as well as data and timelines for asymptomatic events. The Triad software, provided by the American College of Radiology, will be used to upload and de-identify images before submission to the core laboratory. End points and statistical considerations. The primary safety end point is a composite of clinical end points (confirmed or documented by imaging) after successful filter placement at 12 months, that includes freedom from clinically significant vena cava perforation, embolization of the filter or components of the filter, caval thrombotic occlusion, new deep vein thrombosis, and all serious adverse events within the perioperative period (30 days after the procedure). The definitions of these clinical end points are provided in Table II. The primary effectiveness end point is a composite end point at 12 months in situ or 1 month after retrieval (whichever occurs first) that includes procedural and technical success and freedom from clinically significant PE. Definitions of these end points are shown in Table III. Secondary end points include mechanical stability, procedure-related complications at 3 months, major adverse events (composite and individual components) at 6 months, 12 months, 18 months and 24 months, filter tilting of more than 15
at any timepoint, and filter retrieval at any time. Details of these secondary end points are shown in Table IV. In addition to the aggregate primary and secondary end points above, a secondary effectiveness end point that includes freedom from PE at 12 months will be calculated for each individual filter brand (ie, not in aggregate). This secondary effectiveness end point was developed in collaboration with the FDA to allow for a potential labeling modification for individual IVC filter brands. Statistical analysis and sample size calculation. The objective of the statistical design for this study is, within the constraints of single-arm study, to demonstrate that IVC filters in patients with clinical need for mechanical prophylaxis of PE meet safety and effectiveness performance criteria, based on performance goals. The null hypothesis tested is that the population proportion is at, or below, a performance goal rate, versus the alternative hypothesis that the population proportion is above the performance goal rate. The estimated rates for this study are based on premarket studies and published medical literature.9 Primary safety and effectiveness hypotheses, as well as hypotheses for secondary end points, will be formulated for all IVC filters in aggregate (ie, all patients). For the purposes of labeling

4

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders ---

2019

Table I. Time and event schedule of measurements Assessment

Baseline/ Procedure

D/C

3 months 6 15 days

X

X

X

X

6 monthsa6 30 days

Informed consent

X

Eligibility

Xc

Medical history

Xc

Physical examination

Xc

Procedure and filter information

X

Clinical usefulness measures

Xd

Anticoagulant and antiplatelet medication

Xc

Laboratory tests: CBCþplt, Cr, Coag panel

Xe

Imaging: Venogram

Xf

X

X

Imaging: Contrast-enhanced abdominal CT scan Imaging: radiographs (AP and lateral) AE assessmentg

12 months 6 30 days

c

X

X X

X X

X

X

X

AE, Adverse event; AP and LAT, anteroposterior and lateral; CBCþplt, complete blood count and platelet; Cr, serum creatinine; Coag, coagulation; CT, computed tomography; D/C, discharge. After placement, patients are evaluated at discharge, 3 months, 6 months (phone), 12 months, 18 months (phone), and 24 months or until 1 month after filter retrieval, whichever occurs first. The assessments at baseline and follow-up. a The 6-month and 18-month visits are telephone visits. b Patients who do not have a filter placed at time of procedure will also be followed for 1-month after the procedure. c Before any study-related procedure and within 6 weeks before index procedure. d Clinical usefulness measures include length of (1) hospital stay, (2) intensive care unit stay, and (3) index procedure time. e Laboratory tests must be performed within 1 month before the procedure. f For patients having an inferior vena cava filter placed at bedside, an intravascular ultrasound examination may be performed instead of a venogram. g It is expected that all device-related AEs will have appropriate imaging regardless of imaging schedule.

modification, secondary effectiveness hypotheses will be formulated for each IVC filter type (ie, not in aggregate). All analyses will be based on the intent-to-treat population (enrolled at time of consent) and using all available data; no imputation will be performed. Clustering of filter type within enrollment sites is expected, because institutions have preferred devices, and will be accounted for during the data analysis. The primary safety end point will be based on aggregate data for patients with the IVC filter in situ at 12 months. For the primary composite safety end point (defined as freedom from the various complications outlined in Table II), the null hypothesis tested is that the population proportion is at, or below, the performance goal of 80%, versus the alternative hypothesis that the population proportion is above the performance goal. A rejection of the null hypothesis means that filters in situ at 12 months after the procedure have, in aggregate, a relatively low rate of key safety issues, thus, meeting the performance goal. A sample size of 619 patients will provide 90% power at a type I error rate of 0.025. The primary effectiveness end point will be based on aggregate data for patients with the IVC filter in situ at 12 months after placement or 1 month after retrieval, whichever occurs first. Similar testing as for the primary safety end point will be used except the performance

goal is 90%. A sample size of 292 patients will provide 90% power at a type I error rate of 0.025. If both the primary safety and primary effectiveness end points have been met, the secondary effectiveness end point hypothesis will be tested based on aggregate data for all patients. This end point is based on the null hypothesis that the proportion of patients with PE at 12 months after the procedure is greater than or equal to a performance goal of 2%. A sample size of 1539 patients will provide 90% power at a type I error rate of 0.025. Only if this null hypotheses for secondary effectiveness for all filters in aggregate is rejected and a 3% performance goal is met for each IVC filter type individually, will this mean each IVC filter type at 12 months after the procedure has a relatively low rate of key safety issues and may be eligible to apply for labeling modification. Study organization and oversight. The study is sponsored by the IVC Filter Study Group Foundation as previously described. The PRESERVE Steering Committee, comprised of experts in interventional radiology and vascular surgery and New England Research Institutes, oversee the day-to-day conduct of the study, with regular reports to the foundation, participating manufacturers, FDA, and Sites. The data and safety monitoring board has the oversight for the safety of the study. The

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume

-,

Number

5

-

Table I. Continued. 18 monthsa 6 30 days

24 months 6 30 days

Retrieval

1 month after retrievalb 6 15 days from retrieval

X

X X

X

X

X

X

X X X X

X

clinical end points committee consists of two vascular surgeons and an interventional cardiologist who are considered experts in IVC filters and their use, who are not participating in the trial, and who have no actual or potential conflicts. They centrally adjudicate relevant events based on clearly defined protocol and charter definitions. The core laboratory centrally reads specified images obtained during the course of the study.

X

X

Site evaluation and selection. To try to ensure some equitable distribution in enrollment, participating manufacturers will be queried to obtain contacts for sites and investigators, such that six to eight sites participating in PRESERVE would be using their filter. A site questionnaire will be established and sent to sites or investigators with different specialties. The PRESERVE Steering Committee will review each survey and evaluate sites based upon

Table II. Primary composite safety end points Clinical end points: Freedom from

Definition

Clinically significant vena cava perforation

Protrusion of the filter legs through the wall of the IVC causing hemorrhage, hematoma, touching, impressing, or perforating another organ such a liver, bowel, aorta, psoas muscle, vertebral body, or lymph nodes or that triggers the decision to remove the IVC filter, resulting in an attempt to remove the IVC filter, or requiring other intervention, within the first 12 months

Embolization of the filter or components

Movement of the filter or its components to a distant anatomic site completely out of the target zone after successful filter placement

Caval thrombotic occlusion

Presence of an asymptomatic or symptomatic occluding thrombus in the IVC

New DVT

Lower extremity DVT that had not been present previously and occurs after the placement of the IVC filter

SAEs within the perioperative period

SAEs within the first 30 days after filter insertion

DVT, Deep vein thrombosis IVC, inferior vena cava; SAE, serious adverse event. The primary safety end point is a composite of clinical end points (confirmed or documented by imaging) after successful filter placement at 12 months that includes freedom from clinically significant vena cava perforation, embolization of the filter or components of the filter, caval thrombotic occlusion, new deep vein thrombosis, and serious adverse events within the perioperative period.

6

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders ---

2019

Table III. Primary effectiveness end point definitions End points

Definition

Procedural and technical success

Deployment of the initial filter such that the filter is judged suitable for mechanical protection against PE, and placement of second filter to address any anatomic variation without clinically significant perforation, filter embolization, or insertion problems. Insertion problems are defined as incomplete filter opening, clinically significant filter tilt >15
from the IVC axis (eg, nonself-centering filters), misplacement of filter outside the infrarenal IVC when the operator’s intent is to place the filter in the infrarenal IVC (eg, when a portion of the filter is within 1 iliac vein), prolapse of filter components, or filter malposition requiring surgical/endovascular removal.

Freedom from clinically significant PE

Freedom from new symptomatic PE confirmed by appropriate imaging

IVC, Inferior vena cava; PE, pulmonary embolism. The primary effectiveness end point is a composite end point at 12 months in situ or 1 month after retrieval (whichever occurs first) that includes procedural and technical success and freedom from clinically significant PE.

volume of filter insertion. In addition, the PRESERVE Steering Committee desires an equitable distribution of filter use, site principal investigators by specialty, and geographic distribution. The trial sites currently participating are listed in the Appendix (online only).

DISCUSSION Vena caval interruption through percutaneous IVC filter insertion to prevent PE has become widely used, not only in patients with venous thromboembolism, but also in some patient populations as primary prophylaxis against PE. The accepted indications for placement of an IVC

filter are in a patient with either a pulmonary embolus or iliocaval or femoropopliteal deep vein thrombosis and one of more of the following: contraindication to anticoagulation, anticoagulation complication or failure, inability to maintain adequate anticoagulation, or progressive thromboembolic disease despite adequate anticoagulation.10 There is less unanimity on the use of filters in the setting of massive PE, free-floating iliofemoral or IVC thrombus, or in patients with limited cardiopulmonary reserve in the presence of otherwise uncomplicated deep vein thrombosis. Prophylactic use in the absence of deep vein thrombosis has also been suggested in high-

Table IV. Secondary effectiveness end point definitions Secondary end point Mechanical stability

Description Absence of the following at the time of retrieval or at each follow-up: Migration: evidence of cephalad or caudal movement of the filter >20 mm relative to fixed anatomic landmarks compared to the time of placement as determined by radiographs Perforation: >5 mm outside apparent cava wall as determined by CT or perforation of adjacent viscera or major vessel Filter fracture: any loss of a filter’s structural integrity (ie, breakage or separation) documented by imaging or autopsy Filter embolization: postdeployment movement of the filter or its components to a distant anatomic site completely out of the target zone

Procedure related Judgment of the principal investigator complications Major adverse events

Death, PE, caval thrombotic occlusion, DVT, clinically significant perforation, retroperitoneal hematoma, or adjacent organ penetration (eg, bowel, spinal cord, aorta)

Filter tilting >15


As determined by appropriate imaging

Filter retrieval

Attempted retrieval Successful retrieval Failed retrieval Percentage of retrieval success Complications associated with filter retrieval Reasons for failed retrieval

CT, Computed tomography; DVT, deep venous thrombosis; PE, pulmonary embolism. Secondary end points include mechanical stability, procedure-related complications at 3 months, major adverse events (composite and individual components) at 6 months, 12 months, 18 months, and 24 months, filter tilting >15
at any timepoint and filter retrieval at any time.

Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume

-,

Number

Gillespie et al

7

-

risk patients, such as those with closed head injury, spinal cord injury, or multiple fractures. However, in the past decade there has been growing evidence that retrievable IVC filters have their own welldefined risks. These include fracture of the filter itself, with or without embolization of all or part of the filter to the heart or lungs, perforation of the IVC by the legs of the device, IVC thrombosis, and tilting and/or migration of the filter either cephalad or caudally.11 The frequency of these complications appears to vary widely depending on filter type, but the evaluation of the risk of each is greatly limited by the small size of many of the studies. Despite the limited data sets available to accurately estimate risk, the FDA was sufficiently concerned to issue a warning in 2014 regarding the importance of removal of these filters once their clinical need had passed.8 It also led the FDA to work with the Society of Interventional Radiology and the Society of Vascular Surgery to develop the PRESERVE Study with the collaboration of the manufacturers of these devices. Recent publications suggest that IVC filter retrieval rates are increasing. An analysis of data from an administrative claims database, has reported an increase in retrievals from 14% in 2010 to 24% in 2014.12 In the Medicare population, an analysis of the Physician/Supplier Procedure Summary Master Files in Medicare fee-forservice beneficiaries found retrieval rates increased from 12.0% to 17.7% between 2012 and 2015,13 and an analysis of Medicare claims data showed an increase in retrievals from 6.9% to 22.1% between 2012 and 2016.14 A literature review to identify interventions that increase IVC retrieval rates15 concluded that tracking patients typically lost to follow-up after IVC placement is the most effective strategy to increase retrieval rates. Reasons for nonretrieval have also been studied, with one systematic literature review covering publications from 1984 to 2016 showing that two-thirds of retrieval filters were not retrieved, even though 85% were intended to be temporary.16 Technical, system, and patient factors have all been identified as contributing to the failure to retrieve IVC filters in a recent review article,17 but both publications noted that there is a subset of patients with a need for a long-term IVC filter placement. There are several unique features to PRESERVE. First, this study includes many different types of filters across seven manufacturers, for a proof-of-principle IDE to maximize the generalizability and relevance of the study. Second, the protocol is designed to collect real-world evidence and contemporary standard-of-care procedures in a pragmatic study design, with minor exceptions for standardization. Third, the study is a joint venture of two medical professional societies working together in an unusual, paradigm-shifting initiative that brings together researchers, multidisciplinary clinicians, regulatory agencies, and multiple manufacturers, focused on a significant public health issue. The important clinical

goal of preventing potentially life-threatening pulmonary embolus must be balanced with the risks associated with prevention, as well as effectiveness. The studies currently available do not provide enough reliable data on either the safety, or the effectiveness, of IVC filters to answer the question definitively. Randomized trials in this area are few, because they have been very difficult to implement. The French-led Prevention du Risque d’Embolie Pulmonaire par Interruption Cave (PREPIC) trial remains the only level 1 clinical trial that has been performed evaluating the long-term effectiveness of IVC filters in preventing PE with 400 patients enrolled.18 The study demonstrated that, even in a population in which all subjects were anticoagulated, those with vena cava filters have a decreased incidence of pulmonary embolus, but no improvement in survival and, with an increased incidence of deep vein thrombosis, this leaves uncertainty as to their ultimate usefulness. The PREPIC2 trial19 was a randomized, open-label trial of similar size, but with only a 6-month follow-up and filter retrieval planned at 3 months from placement. Hospitalized patients with acute, symptomatic PE associated with lower limb vein thrombosis were studied and similarly assigned to retrievable IVC filter implantation plus anticoagulation (n ¼ 200) or anticoagulation alone with no filter implantation (n ¼ 199). The authors concluded that the use of a retrievable IVC filter plus anticoagulation did not reduce the risk of symptomatic recurrent PE at 3 months, in these patients, compared with anticoagulation alone. Because anticoagulation is the first-line approach in the United States and patients who are able to be anticoagulated would not also be considered for filter placement, owing to the added risk, these trials were not conducted using a design that would be accepted in this country. In addition, the studies were too small to answer key questions about the safety of the devices. With a planned enrollment of 2100 patients, the PRESERVE study includes devices from multiple manufacturers in subjects requiring primary and/or secondary prophylaxis and is the first focused primarily on the safety of these devices. PRESERVE represents the collective best effort of the IVC Filter Study Group Foundation, the FDA, and industry to answer key open questions, as safely as possible, in a clinically relevant population. It also will provide an indication of the effectiveness of the devices in preventing clinically important pulmonary emboli and deep vein thrombosis. A further benefit of the PRESERVE Study is as a demonstration of a new type of collaboration to study an important, large-scale clinical issue, with the collaboration of medical professional societies, a regulatory agency and manufacturers. As similar questions arise regarding other medical devices or treatments, it will serve as a useful template for future studies.

8

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders ---

CONCLUSIONS PE represents a major public health challenge based upon the impact of its morbidity and mortality in the United States. IVC filters represent a viable method of preventing fatal pulmonary emboli in patients with deep venous thrombosis who are not candidates for anticoagulation, but more information is needed on their safety and effectiveness. The PRESERVE Study represents the largest prospective study ever undertaken to investigate this with the hope that it can improve patient care. More information on the PRESERVE study can be found at www.preservetrial.com or through ClinicalTrials.gov under NCT02381509.

AUTHOR CONTRIBUTIONS Conception and design: DG, JS, FS, JR, RW, MJ Analysis and interpretation: DG, JS, FS, JR, RW, MJ Data collection: DG, JS, FS, JR, RW, MJ Writing the article: DG, JS, FS, JR, RW, MJ Critical revision of the article: DG, JS, FS, JR, RW, MJ Final approval of the article: DG, JS, FS, JR, RW, MJ Statistical analysis: DG, JS, FS, JR, RW, MJ Obtained funding: DG, JS, FS, JR, RW, MJ Overall responsibility: DG

REFERENCES 1. Grassi CJ, Swan TL, Cardella JF, Meranze SG, Oglevie SB, Omary RA, et al. Quality improvement guidelines for percutaneous permanent inferior vena cava filter placement for the prevention of pulmonary embolism [Practice Guideline]. J Vasc Interv Radiol 2003;14:271-5. 2. Bikdeli B, Chatterjee S, Desai NR, Kirtane AJ, Desai MM, Bracken MB, et al. Inferior vena cava filters to prevent pulmonary embolism: systematic review and meta-analysis. J Am Coll Cardiol 2017;70:1587-97. 3. Sarosiek S, Crowther M, Sloan JM. Indications, complications, and management of inferior vena cava filters: the experience in 952 patients at an academic hospital with a level I trauma center. JAMA Intern Med 2013;173: 513-7. 4. Rowland SP, Dharmarajah B, Moore HM, Lane TR, Cousins J, Ahmed AR, et al. Inferior vena cava filters for prevention of venous thromboembolism in obese patients undergoing bariatric surgery: a systematic review. Ann Surg 2015;261: 35-45. 5. DeYoung E, Minocha J. Inferior vena cava filters: guidelines, best practice, and expanding indications. Semin Intervent Radiol 2016;33:65-70. 6. Duszak R Jr, Parker L, Levin DC, Rao VM. Placement and removal of inferior vena cava filters: national trends in the Medicare population. J Am Coll Radiol 2011;8:483-9.

2019

7. Removing retrievable inferior vena cava filters: initial communication. 2010. Available at: www.fda.gov/ MedicalDevices/Safety/AlertsandNotices/ucm221676.htm. Accessed July 11, 2018. 8. Removing retrievable inferior vena cava filters: FDA safety communication. 2014. Available at: www.fda.gov/Medical Devices/Safety/AlertsandNotices/ucm396377.htm. Accessed July 11, 2018. 9. Morales JP, Li X, Irony TZ, Ibrahim NG, Moynahan M, Cavanaugh KJ Jr. Decision analysis of retrievable inferior vena cava filters in patients without pulmonary embolism. J Vasc Surg Venous Lymphat Disord 2013;1:376-84. 10. Caplin DM, Nikolic B, Kalva SP, Ganguli S, Saad WE, Zuckerman DA, et al. Quality improvement guidelines for the performance of inferior vena cava filter placement for the prevention of pulmonary embolism. J Vasc Interv Radiol 2011;22:1499-506. 11. Deso SE, Idakoji IA, Kuo WT. Evidence-based evaluation of inferior vena cava filter complications based on filter type. Semin Invernt Radiol 2016;33:93-100. 12. Brown JD, Raissi D, Han Q, Adams VR, Talbert JC. Vena cava filter retrieval rates and factors associated with retrieval in a large US cohort. J Am Heart Assoc 2017;6. 13. Morris E, Duszak R Jr, Sista AK, Hemingway J, Hughes DR. National trends in inferior vena cava filter placement and retrieval procedures. J Am Coll Radiol 2018;15:1080-6. 14. Ahmed O, Wadhwa V, Patel K, Patel MV, Turba UC, Arslan B. Rising retrieval rates of inferior vena cava filters in the United States: insights from the 2012 to 2016 Summary Medicare Claims Data. J Am Coll Radiol 2018;15:1533-57. 15. Goodin A, Han Q, Raissi D, Brown JD. A review of interventions to increase vena cava filter retrieval rates. Ann Vasc Surg 2018;51:284-97. 16. Jia Z, Fuller TA, McKinney JM, Paz-Fumagalli R, Frey GT, Sella DM, et al. Utility of retrievable inferior vena cava filters: a systematic literature review and analysis of the reasons for nonretrieval of filters with temporary indications. Cardiovasc Intervent Radiol 2018;41:675-82. 17. Crumley KD, Hyatt E, Kalva SP, Shah H. Factors affecting inferior vena cava filter retrieval: a review. Vasc Endovascular Surg 2019;53:224-9. 18. Decousus H, Leizorovicz A, Parent F. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. Prevention du Risque d’Embolie Pulmonaire par Interruption Cave Study Group. N Engl J Med 1998;338:409-15. 19. Mismetti P, Laporte S, Pellerin O, Ennezat PV, Couturaud F, Elias A, et al. Effect of a retrievable inferior vena cava filter plus anticoagulation vs anticoagulation alone on risk of recurrent pulmonary embolism: a randomized clinical trial. JAMA 2015;313:1627-35. Submitted Apr 7, 2019; accepted Jul 18, 2019.

Additional material for this article may be found online at www.jvsvenous.org.

Gillespie et al

Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume

-,

Number

-

APPENDIX (online only).

Participating sites

Participating sites

Albany Medical Center

University of Colorado

Boston Medical Center

University of Miami Jackson Memorial Hospital

Carle Heart & Vascular Institute

University of Michigan

Carolinas Medical Center

University of Minnesota

Cleveland Clinic

University of Oklahoma-Tulsa

Duke Medical Center

University of Pittsburgh Medical Center

Fairfield Medical Center

University of Texas Southwestern Medical Center

Florida Hospital

Veterans Administration Palo Alto Health Care

Hackensack University Medical Center

Wake Forest Baptist Health

Harbor e University of California Los Angeles Medical Center

Washington University

Holy Cross Hospital

William Beaumont Hospital

Hospital of the University of Pennsylvania

Yale New Haven Hospital

Indiana University Inova Fairfax Hospital Massachusetts General Hospital Mayo Clinic Rochester Medical College of Wisconsin MedStar Georgetown University Hospital Memorial Hermann Hospital Memorial Sloan Kettering Miami Valley Hospital/Wright State Physicians Health Center Mount Sinai Hospital New York-Presbyterian/Weill Cornell Medical Center Northshore University Manhasset Northwestern Memorial Hospital Oregon Health & Science University Overlook Medical Center Rhode Island Hosp/Miriam Rochester Regional Health Ronald Regan Medical Center Sarasota Memorial Hospital Southcoast Health Spartanburg Regional Medical Center St. Louis University St. Mary’s Medical Center Tallahassee Memorial Hospital The Heart Institute Largo University Health/Louisiana State University Health Shreveport School of Medicine University Hospital, State University of New York University of Arkansas for Medical Sciences University of California San Diego Medical Center University of California, San Francisco Medical Center

(Continued)

8.e1