CDU Contributing Editor

SAN ANTONIO, Texas – The interventional radiology products market, historically a secondary target for many medical device companies focusing on the larger coronary sector, is attracting increased attention with the advent of a number of new technology-driven opportunities. As shown in Table 1, the combined market for interventional radiology products in the U.S. is estimated at approximately $617 million for 2000, with strong growth forecast over the next few years. Key segments of the market contributing to that anticipated growth include peripheral stents and endovascular grafts, neuroradiology products, and new types of thrombectomy and embolectomy devices.

Table 1
U.S. Interventional Radiology Products Market Trends

Year Total Market Growth

1999 $504 million
2000 $617 million 22.5%
2001 $769 million 24.6%
2002 $968 million 25.9%
2003 $1.09 billion 12.6%
2004 $1.19 billion 9.8%
2005 $1.28 billion 7.5%
CAGR, 1999-2005 15.8%

Source: Cardiovascular Device Update

The annual meeting of the Society of Cardiovascular and Interventional Radiology (SCVIR, Fairfax, Virginia), held here in early March, provided a window on the status of the new technologies driving growth in interventional radiology, and also highlighted some of the important issues facing the discipline.

One major area of development focus for companies in the interventional radiology market is in products for carotid angioplasty and stenting, as well as embolic protection devices to help avoid procedural complications during carotid interventions. A number of stents are now in clinical trials, along with a variety of protection devices, and many suppliers view the carotid stent market as the next major opportunity after coronary stents. Other interventional approaches to stroke treatment, such as intra-arterial thrombolysis for treatment of acute occlusive stroke, are coming into more widespread use, and researchers continue to pursue development of alternative approaches such as mechanical thrombectomy, ultrasound-assisted thrombolysis, and balloon catheters for clearing cerebral occlusions in stroke patients. Endovascular grafts for aneurysm treatment represent a major new growth area, although significant issues remain regarding endoleaks, device migration, cost and patient selection. Other growing applications for peripheral vascular stents include treatment of stenosis in the renal and femoral arteries.

Another segment of the interventional radiology products market that is attracting renewed interest is vena caval filters. Vena caval filters have been in use since the 1970's for the prevention of pulmonary embolism in patients with deep vein thrombosis and other conditions predisposing them to clots. The devices have generally been implanted permanently, even though in at least some patients the need for protection against embolism is only short-term. A new permanent filter was launched by Cordis Endovascular (Warren, New Jersey) in mid-2000. Now, some suppliers are developing temporary filters for the U.S. market that can be removed once they are no longer needed, addressing concerns about migration and other adverse effects of filters implanted for many years. Temporary filters already are available in Europe, Canada and Asia, and clinical experience so far has been positive.

Neurointerventional technology advances

The neurointerventional products segment of the interventional radiology market has historically been one of the more dynamic segments of the market. In 2000, neurointerventional products, including microcatheters, microguidewires, embolization coils, other embolization agents and ancillary products, comprised about 17% of the total interventional radiology products market. The neurointerventional segment is poised for expansion as a number of products now in clinical trials begin to enter the market. Carotid stents are the main focus of development activity, coupled with embolic protection devices to help ensure the safety of the stenting procedure. Table 2 presents an update on the status of clinical trials for carotid stents.

Table 2: Status of Carotid Stent Trials and Registries

Clinical Trial

Device Under Study

Status

CREST (Carotid Revascularization Endarterectomy versus Stent Trial) AccuLink stent, AccuNet protection device; Guidant (Indianapolis, Indiana) 24 U.S. centers beginning lead-in credentialing phase of trial. Goal is 50 centers by June, randomization may begin by November.
SAPPHIRE (Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy) Precise stent, AngioGuard protection device; Cordis Endovascular (Warren, New Jersey) Involves high-risk cases; 500 patients.
SHELTER (Stenting of High-risk patients Extracranial Lesions Trialwith Emboli Removal) Carotid Wallstent; Boston Scientific; (Natick, Massachusetts); GuardWire Plus; Medtronic/PercuSurge, (Minneapolis, Minnesota) IDE granted January 2001. Trial to enroll 400 high surgical risk patients at 30 sites. Carotid Wallstent has CE mark in Europe; GuardWire used in more than 5,000 procedures since 1999 release in Europe.
CARESS (Carotid Revascularization with Endarterectomy or Stenting Systems) Carotid Wallstent; GuardWire Plus; Others also allowed to enter Registry of low-risk carotid stent and endarterectomy cases; 900 patients. Organized under an IDE by the International Society of Endovascular Specialists.

Note: A trial of the Memotherm stent from C.R. Bard (Billerica, Massachusetts) for carotid stenting was halted because use of protection devices was not part of the protocol.
Source: Cardiovascular Device Update

While the market opportunity is substantial given that 300,000 carotid endarterectomy surgeries are performed annually worldwide (including 150,000 in the U.S.) of which a significant portion could potentially be converted to stenting, there is an ongoing debate about the efficacy, safety and cost of carotid artery stenting relative to carotid endarterectomy.

Some initial studies, such as CAVATAS (Carotid Artery Vertebral Artery Transluminal Angioplasty Study), have shown no difference in 30-day death and disabling stroke rates for patients treated with carotid stents vs. surgery. Another study of carotid stenting initiated in the UK was halted after five strokes occurred in seven treated patients. Furthermore, a recent retrospective review of 40 carotid artery stenting procedures vs. 428 carotid endarterectomy procedures described at the SCVIR conference by Dr. A. Silas of Dartmouth Hitchcock Medical Center (Lebanon, New Hampshire) showed that stenting resulted in higher costs ($2,801 vs. $1,533) and no improvement in outcome. In fact, there were two deaths secondary to cerebral ischemia in the stent group, and none in the surgical group. Silas concluded that stenting does not offer safety or cost advantages for treatment of routine internal carotid artery disease, but that it may be indicated for patients with restenosis following surgery. However, others noted that the surgical team at Dartmouth Hitchcock typically achieves better-than-average results in terms of both cost and outcome, and that the stent patients in the group studied may in some cases have been ineligible for surgery due to the severity of their disease.

Overall, as indicated by the large number of clinical studies now being initiated with carotid stents, most clinicians believe that carotid stents hold promise, particularly as new devices now under development become available. Stents such as the Cordis Precise, which has replaced the Smart stent as the preferred Cordis device for carotid applications; the carotid Wallstent; the NexStent self-expanding nitinol stent from EndoTex (Cupertino, California), now entering Phase II trials; the AccuLink Carotid from Guidant; and a new stent under development by Medtronic AVE incorporate features intended to optimize their performance in carotid applications. Key features that are desirable for carotid stents include ease of deployment, conformability (preferably via use of a tapered design), plaque coverage/retention, absence of protruding struts, high radial force, and elasticity. Furthermore, the availability of cerebral protection devices is expected to greatly diminish problems with embolization due to plaque dislodged during stenting, resulting in reduced complications. As discussed by Takao Ohki, MD, of Montefiore Medical Center (New York), at the SCVIR conference, there are now a number of cerebral protection devices under development for use in carotid artery and other vascular procedures. Table 3 presents a listing of the various devices segmented by their mechanism of action.

Table 3: Cerebral Protection Devices
Type of Device/Mechanism of Action Device/Supplier

Distal Occlusion Medtronic/PercuSurge GuardWire
Theron balloon
Henry-Amor balloon

Distal Filter NeuroShield (MedNova; Galway, Ireland)
Filterwire (Embolic Protection/Boston Scientific; Natick, Massachusetts)
Angioguard (Cordis; Warren, New Jersey)
AccuNet (Guidant; Indianapolis, Indiana)
TRAP (Microvena; White Bear Lake, Minnesota)
E-Trap (Metamorphic Surgical Devices; Pittsburgh, Pennsylvania)
Bate floating filter (ArteriA; San Francisco, California)
Sci-Pro (Scion Cardio-Vascular; Miami, Florida)
Intraguard (Intratherapeutics; St. Paul, Minnesota)

Proximal Occlusion Parodi catheter (ArteriA)
Kachel balloon

Source: Takao Ohki, MD, Montefiore Medical Center, New York, presented at SCVIR annual meeting, San Antonio, Texas, March 3-8, 2001

The PercuSurge device is closest to market, according to Ohki. The PercuSurge device has already been evaluated for use in saphenous vein graft procedures in the SAFER trial, and improved outcomes were observed. Devices that employ vessel occlusion, such as the PercuSurge Guardwire and the ArteriA Parodi Catheter, rely on the principle that flow is required for downstream embolization. The Parodi device, for example, reverses blood flow in the vessel during the stent procedure and allows capture and removal of embolic particles. Such an approach may prove more effective than filtration, since transcranial Doppler measurements indicate that emboli can still migrate downstream past some filtration devices.

Stents are also being evaluated for treatment of cerebral aneurysms by a number of companies. Medtronic recently launched a device in Europe that is purported to be the first stent for the treatment of high-risk brain aneurysms. Studies have also been conducted with the Wallstent to treat internal carotid artery pseudoaneurysms. As described by Dr. J. Aulino of the University of Maryland School of Medicine (Baltimore, Maryland) at the SCVIR sessions, stenting can be combined with embolization therapy using coils such as the Guglielmi Detachable Coil from Boston Scientific/Target Therapeutics to successfully occlude aneurysms via transcatheter techniques. Such approaches expand the potential applications of carotid and cerebral stents, resulting in a potential market in excess of $300 million worldwide. Another factor driving the use of carotid stents is multi-specialty involvement. According to Randall Higashida, MD, of University of California, San Francisco, Medical Center (San Francisco, California), 60% of carotid stent procedures are now performed by cardiologists, 15% are performed by interventional radiologists, 10% by interventional neuroradiologists, 5% by neurosurgeons, 5% by vascular surgeons, and 2% by interventional neurologists.

While carotid stents and protection devices can address stroke prevention by treating stenosis in the cerebral vessels before they occlude, there is also a large opportunity for devices used to treat stroke once it occurs. There are now more than 750,000 new strokes each year in the U.S., resulting in over 150,000 deaths and in excess of 4.4 million stroke survivors with a disability. The cost to the U.S. health care system is quoted at between $30 billion and $45 billion annually. At present, therapy centers on the use of fibrinolytic agents such as Retevase from Genentech (South San Francisco, California) or alteplase (tissue plasminogen activator) to potentiate clot dissolution. The use of antiplatelet agents in combination with fibrinolytic agents is also under study. Such therapy is extremely time-critical. Treatment should preferably be initiated within three hours of onset of symptoms in order to provide benefit, although in some cases the window may be as long as six hours. A stroke registry called INSTOR is now being formed to begin to gather comprehensive data on fibrinolytic therapy for stroke, with a goal of learning more about patient selection to allow more precise identification of those patients who will benefit from therapy as well as those who will have adverse effects such as cerebral hemorrhage and to determine optimal dosage for various agents. The American Heart Association (Dallas, Texas) is initiating a public education program that will start in May to foster better awareness of the need for rapid response when a stroke occurs so that more patients will get to the hospital sooner.

New devices under development or recently introduced with applications in stroke treatment include the Neurojet catheter from Possis Medical (Minneapolis, Minnesota); the Sub-Microinfusion and RPM catheters from NeuroVASx (Maple Grove, Minnesota); and an ultrasound catheter used in conjunction with thrombolytic agents from Ekos (Everett, Washington). The Possis Neurojet is a 1.5 Fr device, small enough to fit through a microcatheter lumen, that uses miniaturized water jet technology to macerate clots and remove debris. The Possis device is approved for use in clearing dialysis access grafts and for coronary occlusions but remains investigational for cerebral applications. A pilot trial of the device was halted to address safety issues, according to Thomas Tomsick, MD, of the University of Cincinnati Medical Center (Cincinnati, Ohio), but the company plans to restart studies with the device in treating middle cerebral artery occlusions. The NeuroVASx microcatheter, while not employing a high-speed jet design, allows infusion of drugs in small distal vessels at pressures of up to 100 psi and has sufficient mechanical stiffness to allow penetration of clots. The catheter is cleared for use in infusion procedures in the U.S. NeuroVASx also is developing the RPM, which will provide clot disruption and maceration for re-establishment of flow. The Ekos catheter has been evaluated in a 10-patient study, which demonstrated clearance of a vascular branch in 15 minutes, vs. hours that may be needed to clear a clot using thrombolytic therapy alone. Shorter treatment times, and perhaps the ability to use lower drug dosages, are one advantage of devices employing mechanical techniques to eliminate clots. However, an even more important advantage may be improved patient outcome in cases where the protracted time needed to reperfuse brain tissue with thrombolytics results in added loss of neuronal cells.

Endovascular market may exceed expectations

Another major growth segment of the interventional radiology products market is endovascular grafts used in the treatment of aneurysms, arterial lesions, and TIPS procedures. The first two devices, the Ancure from Guidant and the AneuRx from Medtronic, were approved for use in aneurysm treatment by the FDA in the fall of 1999. The number of implants has since grown substantially, from about 500 placed in the U.S. clinical trial of the Ancure and an additional number placed abroad, to more than 6,000 Ancure devices as of March. However, Guidant vountarily recalled the Ancure in mid-March to correct regulatory deficiencies (see page 11).

At SCVIR, Renan Uflacker, MD, of the Medical University of South Carolina (Charleston, South Carolina), clinicians initially anticipated that only about 30% of patients who presented with abdominal aortic aneurysm would be candidates for endovascular graft treatment, but at some centers the proportion is now up to 70% to 80%. As discussed by Brian Stainken, MD, of Albany Medical College (Albany, New York), at the SCVIR gathering, the potential worldwide market at existing prices (around $10,000 per device) is over $500 million. A more conservative market model shown in Table 4 indicates that the U.S. endovascular graft market for all applications will approach $300 million by 2006. However, there are a number of issues that currently limit market expansion and have raised concerns among some in the physician community regarding more widespread use of the devices.

Table 4
U.S. Market Development Model for Endovascular Grafts

Year U.S. Sales Growth

1999 $5 million
2000 $34 million 538.4%
2001 $97 million 183.7%
2002 $202 million 107.6%
2003 $236 million 17.1%
2004 $262 million 10.7%
2005 $278 million 6.3%
2006 $292 million 5.1%
CAGR, 2000-2006 24.6%

Note: Applications include AAA and other aneurysm repair, repair of arterial lesions, TIPS procedures and treatment of certain types of arterial stenosis.
Source: Cardiovascular Device Update

Endoleaks remain an issue with most devices, resulting in continued exposure of the treated aneurysm to hydrodynamic pressures that could cause rupture. Device failures due to strut breakage, leaks in the covering material and migration have also occurred at relatively high rates (endoleaks occurred in 24% of patients receiving the implant in a study published in December 1999). Endovascular grafts represent a significant engineering challenge, since the deployed size of the device for AAA repair is considerably larger than for typical vascular stents, and both the metal framework as well as the graft covering must withstand compression to as small a diameter as possible to make percutaneous insertion practical. Furthermore, the configuration of the repaired vessel typically is altered as the aneurysm shrinks following graft deployment, such that the shape that gives maximal retention and optimal sealing initially may not be the same as that needed a few weeks post-implant. As more has been learned about graft performance in clinical trials, companies developing the devices have re-engineered their designs and introduced second and third generation grafts.

Examples of endovascular grafts that represent the most recent generation of designs include the Zenith graft from Cook (Bloomington, Indiana), a three-piece stainless steel/Dacron device that provides the flexibility to conform to a variety of anatomies; the Vanguard III from Boston Scientific, which features thicker wires and larger iliac branches; the Anaconda from Sulzer Vascutech (Renfrewshire, Scotland), which employs magnets to aid guidewire insertion into the contra-lateral gate; and the Talent LPS, with the CoilTrac Delivery System. In addition, Medtronic AVE has developed a new integrated delivery system for the AneuRx device. W.L. Gore (Flagstaff, Arizona) is continuing with development of the Excluder for AAA treatment and is also developing the Viatorr for TIPS applications and the Hemobahn and Biliary endovascular grafts. Gore is targeting market introduction of its endovascular graft products sometime in 2002. Edwards Life Sciences (Irvine, California) is developing the LifePath graft for AAA applications and is planning to re-start clinical studies before the end of 2001. The LifePath features a polyester covering over an Elgiloy metal mesh and both balloon-expandable and self-expanding components. Balloon expansion is used to place the main body of the device as well as the contra-lateral leg, while the other leg is self-expanding. Cordis also is pursuing development of an endovascular graft, but at present the device is being redesigned prior to continuing with trials. Other companies pursuing the development of endovascular grafts include Bard Vascular and Jomed NV (Ulestraten, the Netherlands).

Continued expansion of the endovascular graft market will hinge on the development of additional types of devices that are optimized for treatment of the varying range of anatomies encountered in the abdominal aorta, as well as devices for other types of procedures such as TIPS, arterial repair and reconstruction, and treatment of other aneurysms. As more patients are treated with existing devices, developers are learning more about design requirements, although many clinicians believe the performance envelope may be being pushed too far. Patients are clearly voting in favor of endovascular treatment over surgery, in response not only to the less invasive nature of endovascular therapy but also to statistics that have indicated a high mortality for surgically treated patients. Thus physicians often feel compelled to offer endovascular graft treatment for AAA, even though recent studies have shown that the procedure may not provide a major improvement in outcome or a cost savings. And results of the most recent endovascular graft trials indicate that more remains to be learned about device performance in the body. As discussed by Michael Soulen, MD, of the University of Pennsylvania Hospital (Philadelphia, Pennsylvania), there have been cases of aneurysm rupture even after successful graft placement with no confirmed presence of an endoleak.

Overall hospital costs for endovascular repair were quoted by Soulen at $21,000 vs. $12,000 for conventional surgical repair. Those costs do not include the cost of follow-up exams, which now are considered mandatory because of the number of long-term failures that have occurred. Consequently, at current graft prices endovascular repair may not be economically viable if open surgical repair is an option. For patients who are unable to undergo surgical repair, the long-term benefit of endovascular repair over medical therapy remains to be demonstrated. Nevertheless, based on straw polls taken at the SCVIR meeting, at least one-third of interventional radiologists are performing endovascular graft implants, and the growth trend in the number of implants indicates continuing rapid adoption. In the future, introduction of new designs that resolve many of the issues with current-generation devices such as endoleaks, device migration, and difficulty in placement will promote wider adoption, as will increased competition in the market and the resultant decline in costs.

New devices emerge for peripheral uses

A variety of other new devices described at the SCVIR meeting promise to expand the range of patients who can be treated with interventional radiology techniques. New stents for placement in the renal and superficial femoral arteries are attracting strong interest among physicians. Sulzer Intratherapeutics (St. Paul, Minnesota) exhibited the IntraCoil SE stent for placement in the superficial femoral and popliteal arteries, and expects FDA approval for the device soon. The IntraCoil SE will be the first peripheral vascular stent to be approved for other than biliary applications since 1992, according to the company. Intratherapeutics was acquired by Sulzer Medica (Houston, Texas) in February and is generating about $17 million in annual sales, primarily of peripheral vascular stents.

Another new stent, the aSPIRE covered stent, was exhibited by Vascular Architects (San Jose, California). The aSPIRE stent has a unique open-coil spiral design and employs a double spiral Nitinol body covered with ePTFE. The open-coil design promotes endothelialization, according to the company. An IDE has been granted for studies in the superficial femoral and iliac arteries, and for the treatment of venous stenosis. The Phase I study will enroll 30 patients and require six-month follow-up. At present, venture-funded Vascular Architects has no sales force and is focusing its efforts on obtaining regulatory approvals. According to James Benanati, MD, of Miami Cardiac and Vascular Institute (Miami, Florida), covered stents for infrainguinal interventions in vessels such as the SFA are now an area of intense investigation and development activity, although at present there are a number of unexplained failures of covered stents used in the SFA. Other suppliers developing covered stents include Jomed, Boston Scientific with the Wallgraft device, and W.L. Gore.

Renal artery stenting was the topic of a dedicated session at the SCVIR conference. Balloon angioplasty and stenting can potentially be applied to the treatment of a number of diseases affecting the renal arteries, including fibromuscular dysplasia, renal hypertension and renal stenosis. The use of interventional techniques, including stents, for the treatment of fibromuscular dysplasia is highly effective and safe, and well-accepted by physicians. However, the treatment of atherosclerotic renal vascular hypertension is more controversial. As discussed by Steve Schwab, MD, of Duke University Medical Center (Durham, North Carolina), a patient must have at least 50% stenosis of a renal artery, and evidence of significant increases in renin secretion from the affected kidney, in order to be a candidate for treatment. Renal artery stenosis is responsible for less than 1% of severe hypertension. However, the use of stenting has had a major impact on the treatment of patients with atherosclerotic renal vascular hypertension and has essentially replaced surgical treatment because of its lower (3% to 5%) complication rate.

The most controversial area, according to Schwab, is interventional treatment of renal ischemic disease, a surprisingly common condition which affects about 11% of patients undergoing cardiac catheterization based on a study at Duke, with 6% to 7% having significant disease and 1% to 2% having bilateral disease. About 25% of patients with renal ischemic disease show progression of renal artery stenosis at five years. Interventional therapy with PTA was not highly effective in previous studies due to restenosis, but the use of stents has provided a dramatic improvement in long-term patency, and has been shown to be safe and successful in eliminating renal stenosis. However, there have been no large randomized controlled trials to demonstrate a survival or morbidity benefit for renal artery stenting. Most studies have been too small to allow definitive conclusions to be drawn, and in many cases patients were not adequately screened for the presence of excessive pressure gradients in the renal arteries. The prevailing opinion at present, according to Schwab, is that renal artery stenting provides a modest benefit for a modest risk in most patients but can prove useful in patients with a clear-cut loss of renal function and progressive renal artery stenosis.

At present, there are no stents with FDA approval for use in the renal arteries. A number of stents, however, including the Palmaz stent from Cordis, have been used off-label for treating renal artery stenosis. Because of the high level of off-label use, there is little incentive for suppliers to conduct the randomized controlled trials needed to obtain specific approval for renal artery use. However, a trial has been initiated using the Palmaz stent, the Aspire II trial, which has so far enrolled 205 patients at 23 sites. Renal stenosis improved or stabilized in 92% of patients, but long-term data on hypertension is not yet available. The SOAR (Sub-optimal Renal Angioplasty Results) registry has also been established to track outcomes for patients treated with the Medtronic AVE Bridge stent, with a goal of enrolling a total of 225 subjects. One company pursuing development of dedicated renal stents is Jomed. The Jomed renal stent features increased wall thickness to provide high radial force in the ostial region, but the device is not cleared for marketing in the U.S. AngioDynamics (Queensbury, New York) has introduced a new biliary stent, the OmniFlex, which also is appropriate for use in the renal arteries due to its short length.

Another important segment of the interventional radiology products market is vena caval filters. Vena caval filters are used to help prevent pulmonary embolism in patients with coagulation disorders. Table 5 presents key statistics for the U.S. and worldwide market.

Table 5: Vena Caval Filter Market Parameters

Number of patients experiencing pulmonary embolism in the U.S. 600,000
Number of deaths per year in U.S. due to pulmonary embolism 60,000
Cases of deep venous thrombosis in the U.S. About 2 million
Number of vena caval filters placed in the U.S., 2000 79,000 (109,000 worldwide)
Percentage placed by interventional radiologists 60%
Typical device cost $1,100 to $1,200

Source: John Kaufman, MD, Dotter Interventional Institute, Portland, Oregon, presented at SCVIR annual meeting,San Antonio, Texas, March 3-8, 2001

Devices available in the U.S. market, all intended for permanent implant, include the Greenfield 12 Fr Titanium filter from Boston Scientific; the Cook Bird's Nest; the Vena Tech filter from B. Braun (Bethlehem, Pennsylvania), including a new Low Profile version; and the Simon Nitinol Filter from NMT Medical/Bard. Two new permanent filters have been introduced recently, including the TrapEase from Cordis and the Gunther Tulip Vena Cava Mreye Filter from Cook.

A number of suppliers have begun development of removable filters, which have the advantage of avoiding concerns about possible long-term adverse effects of filter placement. Some studies have suggested that the presence of a permanent filter increases the risk of developing recurrent deep venous thrombosis. The Gunther Tulip, while cleared only for permanent implant in the U.S., incorporates a hook on the distal tip that allows the device to be snared for removal. B. Braun is developing the Tempofilter II and the Lysofilter. The Tempofilter II can be implanted for up to six weeks, and includes a subcutaneous anchoring device. A 12 Fr sheath is required for introducing the device into the right jugular. The Lysofilter can be implanted for up to two weeks, and the tethering catheter supplied with the device (Lysocath) can be used for direct injection of thrombolytic drugs.

Outside the U.S., the Cook Gunther Tulip filter is marketed as a removable device that can be deployed for up to 14 days. Other temporary or removable filters marketed outside the U.S. include the Bard Protect catheter, the Cordis Prolyser, and the Antheor filter. Yet another filter, the Recovery from Nitinol Medical Technologies (Boston, Massachusetts), can be removed at up to 12 weeks post-implant but is only sold outside the U.S.

No Comments