By James Cockburn, David Hildick-Smith, and Adam de Belder
Historically, studies found rotational atherectomy to be associated with both high restenosis and complication rates. However, it has recently made a resurgence due to a combination of drug-eluting stents, which have reduced the rate of restenosis, and lower burr:artery ratios/lower burr speeds which has led to a reduction in complication rates.
The aim of our study was to compare the rate of all-cause mortality in patients who had undergone PCI with adjunctive rotational atherectomy to the rate of all-cause mortality in patients who underwent standard PCI using contemporary data from the British Cardiovascular Intervention Society (BCIS) database.
Over a three-and-a-half-year period, 221,699 patients under went PCI and 2,152 (1%) of them underwent PCI with adjunctive rotational atherectomy. Comparison of baseline characteristics demonstrated a higher pre-morbid risk profile in patients undergoing rotational atherectomy. Procedural success was lower for rotational atherectomy patients (90.3% vs. 94.6% for patients undergoing standard PCI; p<0.001) and procedural complications were more frequent (9.7% vs 5.4%, respectively; p<0.001).
With regard to survival (mean follow up 2.4±1.4 years), unadjusted cox proportional hazard modelling demonstrated poorer survival for rotational atherectomy patients (HR 2.21, 95% CI 1.97–2.49, p<0.0001). After adjusting for univariables of significance, there remained significant evidence of poorer survival for rotational atherectomy patients despite a reduction in the estimated hazard ratio (HR 1.26, 95% CI 1.11–1.44, p=0.0004).
Propensity analysis was undertaken to compensate for the non-randomised design of the study and to achieve a balance between the two treatment groups due to the considerable difference in sample size. Reassuringly hazard ratios (both unadjusted and adjusted) from the propensity analysis were consistent.
Implications of study
Firstly, patients with calcific coronary disease tend towards a higher pre-morbid cardiovascular risk profile. Consequently, it is important to recognise this when interpreting results from rotational atherectomy trials as adverse events and outcomes may simply be a reflection of an individual’s comorbidity and the complexity of their coronary anatomy rather than of the technique itself. However, after correcting for those risk factors, the risk associated with rotational atherectomy remains significant.
Secondly, despite the high procedural success with rotational atherectomy in our study, the intervention was also associated with higher procedural complication rates. However, this rate was lower than in previous series. Critics of rotational atherectomy as an adjunct to PCI focus on high complication rates reported in older trials and registries, but it is important to remember that rotational atherectomy strategies at the time of these studies included high burr:artery ratios, high burr speeds, large calibre catheters, and the absence of dual antiplatelet therapy and drug-eluting stent technologies; therefore, making direct comparisons between the complication rates reported in these studies and the rates associated with current devices difficult. In this context, our study represents a significant improvement in clinical results, and is consistent with the most contemporary rotational atherectomy trial—ROTAXUS.
Thirdly, one interesting finding in our study was an increased use of rotational atherectomy for left main stem intervention (13.6% for PCI with rotational atherectomy vs. 2.3% standard PCI; p<0.001), which is likely to reflect a proportion of “surgical turn-down” patients considered too high risk for coronary artery bypass surgery. The results found a significant interaction in the multivariable model, suggesting improved survival for patients treated with rotational atherectomy with left main stem disease. It may be that this reflects the size of the artery (internal diameter generally >3.5mm), which offsets any potential complications seen as a result of the interaction between burr and plaque.
Finally, our study suggested no adverse effect on survival based on rotational atherectomy performed by the radial access site (HR 1.06 95% CI: 0.77–1.47), or any difference in outcome between of rotational atherectomy procedures performed in surgical and non-surgical centres (HR 1.17, 95% CI 0.83–1.63, p=0.037), offering further support to the development of programmes in centres without immediate access to on-site surgical back-up.
We believe rotational atherectomy remains a technique of significant value for patients with heavily calcified plaque burden, though mortality in this group during medium term follow-up remains significant (a proportion of which appears to be linked to the premorbid risk of the patient population rather than the technique itself).
James Cockburn, David Hildick-Smith, and Adam de Belder are all at Sussex Cardiac Centre, Royal Sussex County Hospital, Brighton, UK