NCHR Comments on CMS’s Proposed Decision Regarding Percutaneous Transluminal Angioplasty (PTA) of the Carotid Artery

August 10, 2023

The National Center for Health Research (NCHR) is a nonprofit think tank that conducts, analyzes, and scrutinizes research on a range of health issues, with a particular focus on which prevention strategies and treatments are most effective for which patients and consumers. We do not accept funding from companies that make products that are the subject of our work, so we have no conflicts of interest.

We appreciate the opportunity to comment and support CMS’s mission of ensuring Medicare beneficiaries receive quality, evidence-based medical care. We agree that CMS has an obligation to enact policies that are in the best interest of the population it serves and are based on robust scientific evidence. Thus, we strongly oppose CMS’s proposed decision: “Percutaneous Transluminal Angioplasty (PTA) of the Carotid Artery Concurrent with Stenting.” In this comment we review the science that supports our recommendation.

CMS’s draft decision expands reimbursement to patient populations that have not been shown to benefit from carotid artery revascularization over optimal medical treatment (OMT). Proposing to extend coverage to patients with asymptomatic carotid artery stenosis of at least 70% instead of 80%, or patients with symptomatic stenosis of at least 50% instead of 70% is not evidence-based and will lead to unnecessary procedures. One of the largest, best-quality, randomized clinical trials, the European Carotid Surgery Trial, showed a statistically significant six-fold reduction in strokes at 3-years for asymptomatic patients with >80% CAS who underwent a carotid endarterectomy (CEA) compared to OMT. In addition, the North American Symptomatic Carotid Trial showed a statistically significant three-fold stroke reduction at 2-years in symptomatic patients with >70% stenosis who underwent CEA compared to OMT.  In contrast, a sub-analysis of this same study showed that symptomatic patients with 50-69% stenosis who underwent CEA had no significant benefit of stroke reduction at 5-years compared with OMT alone. Since expanding CMS coverage to include populations who are at lower risk of stroke is not supported by the most current, best-quality, evidence, we conclude that it will lead to unnecessary procedures that have long-term consequences and is likely to cause harm to patients.

Carotid artery stenting (CAS) has not been demonstrated as an equivalent, non-inferior, or superior treatment to CEA or OMT and has a high rate of periprocedural complications. The strongest, most-contemporary evidence comparing CAS to CEA or OMT does not suggest that CAS is equivalent, non-inferior, or superior to CEA or OMT. It is crucial to remember that if statistical significance is not reached when comparing two treatment modalities (i.e., CAS to CEA) it does not mean the options are equivalent or non-inferior, especially when the studies are underpowered. Non-inferiority is only determined when the difference between the two treatment modalities is insubstantial and a non-inferiority margin is referenced and clearly demonstrated. This is of critical importance when studies are underpowered (such as the SAPPHIRE, CREST, SPACE-2, ACT-1, and ASCT-2 trials) and reach an inconclusive outcome, as the actual difference might be statistically significant and clinically meaningful when adequately powered. The CREST-2 Trial may eventually provide evidence that better informs CMS’s decision on CAS as a noninferior treatment option to CEA. However, since the CREST-2 trial is ongoing, it would be irresponsible to expand coverage for an otherwise experimental procedure on a lower-risk stroke population using the preliminary data that are currently available.

The most current and highest-quality studies comparing CAS to CEA or OMT have major limitations and are not generalizable to the typical patient population with vascular disease. The major relevant studies comparing CAS to CEA or OMT use data from homogenous study populations where most are male and almost all are white (when race is described) and received treatment at well-resourced academic facilities. This homogeneity greatly limits the generalizability of study outcomes, and results should not be extrapolated to the typical patient population with vascular disease. Other serious flaws with the key relevant studies comparing CAS to CEA or OMT are worth noting. We have included a summary table of the highest-quality, contemporary, studies comparing CAS to CEA, or OMT and highlight the major findings, limitations, and complication rates identified in each study. We strongly advise CMS to review our summary table as we show there is truly a lack of evidence to support CAS as a non-inferior approach to CEA or OMT and further note the high complication rates of CAS. 

CAS is associated with a significantly higher 30-day periprocedural risk of major complications compared to CEA. Severe complications including stroke, myocardial infarction, and death occur at rates nearly double that of CEA. These are life-threatening, life-altering complications that require extensive medical resources to treat and place a tremendous emotional and economic burden on patients and their families. Higher 30-day periprocedural complication rates of stroke or death among CAS patients are highlighted in our summary table of contemporary evidence and have also been reaffirmed in the Vascular Quality Initiative database (CEA: 1.4%, CAS: 2.6%) and Medicare database (CEA: 3.5% CAS: 5.1%)— these databases show a near doubled rate of periprocedural stroke or death among patients undergoing CAS.

The evidence comparing CAS to CEA or OMT is inconclusive, which is reaffirmed by the United States Preventative Services Task Force, which reviewed the major evidence comparing treatment modalities among asymptomatic patients and rated it grade D (i.e., low-quality, contradictory, and/or inconclusive). A major reason these studies yield inconclusive results, despite being randomized controlled trials is that they were inadequately designed for establishing equivalence or noninferiority. A margin for clinical noninferiority of CAS against CEA or OMT has only been calculated in two trials (Long-Term Results of SAPPHIRE and ACT-1), both of which were substantially underpowered. Further, these trials were performed in patients at either a high-operative risk (SAPPHIRE) or with severe carotid artery stenosis (ACT-1), not among the lower-risk stroke population for whom CMS is proposing to expand coverage. While the ACST-2 trial data were based on an intention to treat analysis, it showed that CAS had a significantly higher rate of non-disabling strokes at 30-days compared to CEA; therefore, it should not be considered non-inferior. Moreover, no studies have ever evaluated a margin for relative efficacy when CAS is compared to a reference treatment (i.e., OMT) in patients with moderate stenosis—yet this is the exact population for whom CMS is proposing to expand coverage.

CMS proposes that physicians engage in formal shared decision-making using a validated tool prior to offering CAS; unfortunately, no such tool exists. There is no standardized or validated shared decision-making tool for the management of carotid artery revascularization, and so this objective proposed by CMS cannot currently be met. And, since not all physicians practice every treatment modality (CAS, CEA, and OMT), this will often undermine the transparency and objectivity of shared decision-making discussions. Expanding coverage without a validated shared decision-making tool or clear evidence-based guidelines in place creates a framework prone to extreme bias. Until a shared decision-making tool for carotid artery revascularization is validated and evidence-based guidelines on revascularization options are clearly adopted, we strongly urge CMA to revise its draft decision. Without such infrastructure in place, this decision jeopardizes patient autonomy, introduces bias, and potentially harms patient care.

In summary, the NCHR is committed to policies that are based on robust scientific evidence and limit harm, and so we strongly oppose CMS’s draft decision: “Percutaneous Transluminal Angioplasty (PTA) of the Carotid Artery Concurrent with Stenting.” The major limitations of the key randomized controlled trials comparing CAS to CEA or OMT were insufficiently addressed in CMS’s draft decision. These studies are grossly underpowered, have homogeneous study populations that are not representative of most Medicare patients, and conflate inconclusive results with equivalence and noninferiority. CAS may be substantially inferior to CEA or OMT in an adequately powered study with set noninferiority margins. And, CAS has been repeatedly shown to have a significantly higher risk of periprocedural complications compared to CEA.

We urge CMS to revise its draft decision regarding PTA reimbursement to better reflect current scientific evidence. There is no evidence that CAS is a beneficial treatment over optimal medical (non-surgical) care in patients with moderate carotid artery stenosis, nor enough evidence to show that CAS is a noninferior treatment to CEA in patients with high grade stenosis. However, there are real documented dangers to CAS that impact patient morbidity, mortality, and increase the cost of medical care. Enacting a policy that does not conform to evidence-based guidelines or rely on a validated shared decision-making model for carotid artery revascularization jeopardizes patient care and is likely to harm patients.