Commentary: The Adluminal Origin of Restenosis in Peripheral Artery Interventions and Its Implications for the Development of Future Treatment Strategies

The incidence of restenosis after infrainguinal arterial intervention is high, with conventional balloon angioplasty (BA), reaching 40% to 75% on long-term follow-up.^1,2 Although stenting reduced restenosis compared to BA in randomized trials, the main mechanism has been preventing recoil and negative remodeling. Smooth muscle cells are not altered substantially with stenting. A heightened inflammatory state measured by high-sensitivity C-reactive protein, which is present at baseline in patients with peripheral artery disease (PAD), is significantly increased post intervention and correlates with restenosis. ³

Endovascular specialists routinely perform high pressure balloon inflation in the coronary arteries to achieve good stent apposition and expansion. This practice has been carried over to the peripheral vasculature. There are significant differences, however, between peripheral and coronary arteries, including plaque burden, vessel size, lesion length, flow characteristics, inflammatory milieu at baseline, the use of nitinol self-expanding rather than balloon-expandable stents, and the need for shorter and faster inflation/deflation time in the coronary arteries to reduce ischemic complications. Adopting a similar methodology in treating PAD may not be an ideal approach. Excessive vessel injury by high pressure barotrauma, deeper atherectomy into the vessel wall, constant vessel stretching by oversized self-expanding stents, rapid inflation/deflation of balloons may all precipitate deeper disruption in the tunica media and adventitia, triggering a significant restenotic process.^4–6 Understanding the mechanisms of restenosis in peripheral artery interventions is therefore critical to developing therapies and strategies to effectively treat this highly prevalent problem.

The article by Tarricone et al ⁷ is an important step toward understanding how the depth of injury during the treatment of PAD can be a strong predictor of restenosis. Although the disruption of the internal elastic lamina is known to correlate with smooth muscle cell proliferation, ⁴ their study demonstrates that deeper disruption at the medial and adventitial levels is a powerful predictor of restenosis. In fact, very few patients in this study escaped the restenotic process with directional atherectomy (DA) when medial and adventitial disruption were demonstrated by histopathology (60 of 62 patients). As expected, the overall restenosis in their study was 57% at 1 year, mostly accounted for by the deeper injury (97% with adventitial or medial injury vs 11% with no deep injury).

This important concept offers at least a partial explanation as to why target lesion revascularization (TLR) was significantly low in studies where deeper injury, particularly adventitial injury, was avoided, such as with orbital atherectomy (OA). In the CALCIUM 360 randomized trial, ⁸ for instance, adjunctive BA treatment following OA was performed at small increments of pressure (1 atmosphere every 10 seconds until full balloon expansion was achieved). The overall pressure needed for full balloon expansion was low with adjunctive dilation post OA compared with dilation alone (≤4 atm for 59.3% of OA lesions vs 8.8% for BA lesions; mean OA arm pressure 5.9±4.2 vs 9.4±3.8 atm in the BA arm). This translated into numerically less bailout stenting, dissection, and TLR (6.7% vs 20% in the OA vs BA arms, respectively).

Orbital atherectomy is known to cause differential sanding of hard calcium and plaque and is deflected from normal elastic tissue. Similarly, rotational atherectomy was shown to cause lumen enlargement by selectively ablating hard and calcific plaque with no elastic tissue disruption or arterial expansion or change in vessel volume (external elastic membrane), whereas adjunctive BA leads to arterial dissection and further vessel expansion.^9,10 Slow inflation and low balloon pressure may have avoided significant deeper balloon injury and resulted in overall favorable outcome. On the other hand, aggressive DA (multiple cuts in the same plane or cuts in a normal quadrant) may result in deeper injury, as seen in the study of Tarricone et al. ⁷ Although atherectomy may improve vessel compliance and reduce bailout stenting, ¹¹ deep injury may offset the benefit of adjunctive low pressure balloon inflation. Therefore, in atherectomy of de novo lesions, larger may not necessarily be better if it comes at the expense of tunica media or adventitia disruption and injury. Atherectomy, particularly directional, may be best performed less aggressively, primarily targeting plaque and avoiding normal tissue.

It is yet to be clarified in well powered and randomized trials if proliferation of the neointima or neomedia remains a problem if drug-coated balloons (DCB) are applied as an adjunctive to atherectomy. Based on the findings of Tarricone et al, ⁷ DCB should be able to deliver drugs deep enough into the tissue to inhibit adventitial myofibroblasts and neomedia proliferation. Although deeper cuts trigger this adluminal process of restenosis, they may also carry the advantage of allowing deeper penetration of a higher concentration of an antiproliferative drug that can inhibit the restenotic process.

Early preclinical data have shown that drug delivery post atherectomy in a vessel wall is safe. ¹² Reducing dissection and bailout stenting and facilitating a safe and effective antiproliferative drug concentration in a vessel wall provides a compelling rationale for adding DCB as an adjunctive treatment postatherectomy. In a small nonrandomized study, ¹³ 30 patients with heavily calcified lesions were treated with intravascular ultrasound–guided DA and DCB. Bailout stenting was 6.5% and the TLR rate was 10%. Also, the Definitive AR study randomized patients with femoropopliteal disease to DA with adjunctive DCB vs DCB alone. ¹⁴ The atherectomy arm yielded a higher primary patency (96.8%) for long (≥10 cm) lesions at 1-year follow-up compared with 85.9% of patients treated with DCB alone. Larger randomized trials will be eagerly awaited to conclusively test this hypothesis.

In conclusion, the study by Tarricone et al ⁷ presents an opportunity to evaluate our current approach in treating PAD. The goal of atherectomy should be to alter vessel compliance and improve on the acute procedure results, avoiding deep tissue injury in particular to reduce restenosis. The advent of DCB will likely reduce the restenotic process following atherectomy by possibly delivering an effective and safe drug concentration into the tissue. Coupling these 2 modalities will likely result in a higher acute success rate, less dissection and stenting, and an effective drug penetration that could result in an optimal strategy to reduce restenosis after peripheral artery interventions.