CRF/ACC: Xience is economically dominant to Taxus
ATLANTA—At one-year in the SPIRIT IV trial, the use of Abbott’s Xience V stent was an “economically dominant strategy” compared with Boston Scientific’s Taxus stent with improved clinical outcomes and lower overall medical costs, according to the interim results of an analysis presented Saturday during CRF's Optimizing PCI Outcomes Symposium, which ran concurrent to the American College of Cardiology (ACC) conference.

Lead author David J. Cohen, MD, from St. Luke’s Mid America Heart Institute in St. Louis, Mo., noted that previous studies have demonstrated that both Cordis’ Cypher sirolimus-eluting stent and the paclitaxel-eluting Taxus first-generation drug-eluting stent (DES) are reasonably cost-effective compared with bare-metal stents for patients with moderate to high risk of restenosis.

Recently, several second-generation DES have been introduce and found to be either noninferior (Endeavor) or superior (Xience V) to one or more first-generation DES designs, according to Cohen. He added that its “unknown” whether these second-generation DES offer “economic advantages over predicate devices.”

The researchers sought to compare the long-term costs of PCI using Xience V with those of PCI with Taxus; to examine the overall cost effectiveness of Xience V versus Taxus in patients undergoing PCI in a real-world setting (i.e., multiple lesions, no angiographic follow up); and to examine the impact of alternative device acquisition on the cost effectiveness of Xience V.

For the study, Cohen and his colleagues prospectively collected medical resource utilization for all patients for initial hospitalization and for one year after enrollment. They assessed services, such as cardiovascular hospitalizations, revascularization procedures, diagnostic cath and dual-antiplatelet therapy.

Cohen mentioned they did not collect outpatient testing, emergency room visits and other medications. He added that the costs were assessed from the perspective of the U.S. healthcare system in constant 2009 dollars.

Specifically, cath lab procedure costs were based on “bottom-up” cost methodology using measured resource utilization and current unit costs, and assumed equal stent costs ($2,000 per stent) for base case analysis. Cohen noted that all other inpatient costs for initial PCI procedures were calculated using a regression-based cost model derived from the MAHI cost accounting system. Finally, the cost of follow-up procedures and hospitalizations (including physician fees) were based on the average U.S. reimbursement rates.

For index procedure resources and hospital costs for both stents, the procedure duration was the same (43 minutes); the amount of study stents were the same (1.5); the number of non-study stents used were the same (0.01); the number of balloons used were the same (1.2); and the length of stay was the same (1.4 days). However, Cohen reported that the total initial in-hospital costs were $17,735 with Xience V and $17,758 with Taxus.

He also reported that the total costs with one-year follow up were $20,189 with Xience V and $20,355 with Taxus due to less repeat revascularization in the twelve-month period—representing $146 difference.

Additionally, the analysis found that the total costs at one year excluding non-target vessel revascularization events showed a great difference, with $18,610 for Xience V and $19,072 for Taxus.

“Depending on the specific analytic approach,” Cohen said, “total one-year cost savings ranged from $150-$450 per patient.” He explained that sensitivity analyses demonstrate that Xience V would remain “reasonably cost effective” by U.S. standards, even with higher stent prices of about $200-$300 per stent.

“Since the findings were derived from a multicenter, comparative trial without protocol-driven angiographic follow up, these results should be readily generalizable to most U.S. practice settings,” Cohen said.

He added that further study will be necessary to determine the impact of longer term follow-up on these results, and that the final two-year results should be available by TCT.10.