Diagnostic Imaging
May 2003

INTERVENTION

Trials reveal need for changes in drug-eluting stent designs

Modified PV stent configurations address unexpectedly fast drug release and fractures in superficial femoral arteries

Drug-eluting stents may not live up to optimistic predictions that they would provide a panacea for restenosis. But with the right combination of pharmaceutical agent, coating, and stent configuration, they can stave off the reformation of intra-arterial plaque after angioplasty.

Physicians were astounded in 2001 and early 2002 when several preliminary trials, including the first trial of a drug-eluting stent for the superficial femoral arteries, found no in-stent restenosis six months after installation. The expected restenosis rate in the coronary arteries is 20% to 40%, and in the SFA it is close to 50%.

Drug-eluting stent proponents have settled back to earth following reports of patient experience with the stents up to 18 months after installation. Several sirolimus- and paclitaxel-eluting stents continue to perform well in clinical trials. The SIRIUS trial of the Cordis Cypher coronary stent, announced last September, found the devices reduced the in-stent restenosis rate after eight months to 3.2%, or 91% less than the rate observed in the study's controls.

Twelve-month data for Boston Scientific's Taxus paclitaxel-coated stent reveal that only one target vessel revascularization was required among 61 patients. Four follow-up interventions were reported for the control population.

Both devices were awarded CE Mark approvals and were introduced in Europe in 2002. Achieve and V-Flex Plus PTX, paclitaxel-eluting coronary stents manufactured by Cook, were also approved for European sales in 2002. The company is initiating a U.S. trial as part of the FDA approval process for the V-Flex Plus PTX. U.S. sales of the drug-eluting Achieve are not planned. As of March 15, Cordis was awaiting final word from the FDA about approval of its Cypher stent. An FDA advisory panel unanimously recommended approval in October 2002. The agency granted expedited review for Boston Scientific's Taxus stent in March.

Even the usually circumspect Centers for Medicare and Medicaid Services-the administrator of Medicare and Medicaid-appears to be on the bandwagon. Billing codes for drug-eluting coronary artery stents have been created, and reimbursement rates for the devices have been set. According to a Cordis reimbursement guide, Medicare will pay hospitals an extra $2000 for a drug-eluting stent installed after a heart attack or an extra $2100 in the absence of previous myocardial infarction. Business analysts predict that drug-eluting stents could eventually generate $5 billion in sales annually.

As good as all this looks, the past year has been marked by failures as well. Guidant was knocked out of the market in 2002 when tests of its actinomycin-D stent showed that the agent failed to prevent restenosis, but the company returned in January by acquiring the everolimus-eluting coronary stent program of Biosensors International. Six-month results of a preliminary trial of that stent have been encouraging.

Biocompatibles International of the U.K. suspended the pivotal clinical trial of its batimastat-eluting Bio divYsio stent in March 2002. Biocompatibles sold its cardiovascular stent business to Abbott Laboratories less than a month later.

The emergence of the first clear product winners and losers may have been the most important development for drug-eluting stent developers in the past year, according to Dr. Brian G. Firth, vice president of medical affairs and medical economics at Cordis.

"If you ask me for the greatest 'ah-hah' for the past year, it would be that some drugs look good, some drugs look really bad. Some look good early and not good later on, and some lack enough information to come to judgment," he said.

APPLES AND ORANGES

It has also become clear that conditions that govern the performance of drug-eluting stents in the coronary artery do not necessarily apply to devices designed for installation elsewhere in the body.

This finding was reinforced by the experience of Dr. Stephan Duda, principal investigator of the Sirocco preliminary trial of the Cordis Smart stent for the SFA. Although the Smart stent releases the same sirolimus agent as the Cordis Cypher stent for the coronary arteries, the similarities between the two devices end there. The balloon-expandable Cypher is composed of stainless steel mesh, while the self-expanding Smart stent is made of nitinol, and the two devices reside in radically different environments, Duda said. The SFAs are subjected to severe mechanical forces not found in the coronaries and are much wider and longer. Very slow flow or nearly no flow is observed in the SFA during diastole, while peak perfusion occurs during diastole in the coronary arteries.

In addition, stents designed for the peripheral vascular system have to address the long length of the occlusions typically found in these regions. PV stents must be overlapped to treat long segments of disease and should be able to handle the longitudinal stretching and flexion encountered in those vessels.

"You cannot easily transpose the knowledge gained by observing a stent in a muscular small-diameter artery in the heart to an artery like the SFA," Duda said.

Durability is major issue in the fabrication of drug-eluting PV stents, said Dr. Barry Katzen, medical director of the Miami Cardiac & Vascular Institute. Encouraging early trial results with the coated peripheral vascular stents in the coronary arteries do not necessarily apply to coated stents designed for the peripheral vascular system.

"Everybody is looking at their stent designs to make sure they can sustain durability testing with or without a coating," he said. "In the coronary arteries, if you make the six-month timetable, you are pretty home-free. The incidence of restenosis begins to drop. This is not true in periphery. The critical time period is in the nine-to-12-month range," he said.

FULL OF SURPRISES

The picture drawn by Duda after 18 months of testing as part of the SIROCCO trial differs from his view at six months into it. The 18-month results were full of surprises. He determined, for example, that either slow- or fast-eluting versions of the Cordis Smart stent were installed in the 14 patients who composed the experimental group. Analysis showed that 95% of the fast-eluting sirolimus embedded in the polymer coating was released three days after installation. The slow-eluting stent released 95% of its drug in three weeks.

Consequently, the experience of the five patients with the fast-release formula was nearly identical to that of the control subjects who received uncoated stents. The late loss, or buildup of neointimal tissue, after six months for fast-release subjects was 0.68 mm, compared with 0.8 mm for the controls, and zero for the slow-release formula. After 18 months, Duda calculated the binary restenosis rate for the slow-eluting, fast-eluting, and bare metal control groups was zero, 33%, and 30% respectively.

The fast-release stents were responsible for the three binary in-stent restenoses seen in the experimental group after 18 months. In comparison, four binary restenoses had already been observed in the bare-metal stent group after six months.

Duda continued to be surprised by the good performance of the bare-metal controls, however. After 18 months, only one total occlusion, accounting for 6% of the control group, was detected in the entire study cohort. The patient in that instance was hospitalized and underwent angioplasty.

"Obviously, the rate of restenosis is higher. It is 23.5% in the bare Smart group, but that is also an astoundingly good result after 18 months," he said.

Six instances of stent fractures in the coated-stent population tempered the overall positive nature of the 18-month results for SIROCCO. The patient in each instance had received three overlapping 8-cm-long stents to cover the length of a long lesion. In one case, a small ulceration was seen at the site of a strut fracture eight months after installation. In response, the attending physician placed a covered stent inside the Smart stent for prophylactic reasons, Duda said.

Two deaths among subjects in the sirolimus group were unrelated to the treatment, he said. Two other patients were successfully treated with antibiotics and anti-inflammatory drugs for cellulitis associated with the stents. The 18-month Sirocco trial results were presented in January at the International Symposium on Endovascular Therapy in Miami.

The design of phase II Sirocco trials that began in November was modified to address early problems with drug elution rates and fractures, Duda said. Fifty-seven additional subjects were recruited from five newly participating hospitals in Australia, Austria, Belgium, and Germany. Subjects assigned to the experimental group received slow-eluting sirolimus Smart stents. No lesion longer than 14.5 cm was treated. Phase II results will be presented in September at the Transcatheter Cardiovascular Therapeutics annual meeting.

Katzen will collaborate with Dr. Krishan Rocha-Singh, an interventional cardiologist with Prairie Cardiovascular Consultants in Springfield, IL, in another trial of the sirolimus-eluting Smart stent. The FDA phase II Sentry trial will attempt to learn why drug-eluting stents installed in the SFA result in a higher patency rate than uncoated stents. The number of sites and the date the trial will begin have not yet been determined.

Cook has developed a paclitaxel-eluting version of its uncoated Zilver nitinol stent originally designed for biliary indications. A U.S. clinical trial is planned, and once approved, the device will be called the Zilver PTX. It will be marketed for peripheral applications, according to a company spokesperson.