Commentary: Excimer Laser in Treating Femoropopliteal In-Stent Restenosis

In-stent restenosis (ISR) is a prevalent problem following stenting of the superficial femoral artery (SFA) and/or popliteal vessels. Restenosis or loss of patency measured angiographically (≥50% lesion severity) or by duplex ultrasound (peak systolic velocity ratio >2.4) is predicted by longer lesion length and the presence of total occlusions. Other important factors also include TransAtlantic Inter-Society Consensus (TASC) C/D lesions; the presence of critical limb ischemia, diabetes, and female gender; and a heightened baseline inflammatory state.^1–5

Change in stent design, such as covered stents (Viabahn; W.L. Gore & Associates, Flagstaff, AZ, USA), ⁶ interwoven nitinol wire stents (Supera; Abbott Vascular, Redwood City, CA, USA), ⁷ drug-coated stents (Zilver PTX; Cook Medical, Bloomington, IN, USA), ⁸ and, more recently, polymer-based drug-eluting stents (Eluvia; Boston Scientific Corporation, Marlborough, MA, USA) ⁹ have all shown promise in reducing the incidence of restenosis when a primary stenting strategy is used in treating de novo femoropopliteal lesions. A primary stent strategy appears to yield a higher patency rate and a lower target lesion revascularization (TLR) rate (with drug-coated stents or the Supera stent) than conventional balloon angioplasty (BA) when treating the femoropopliteal segment. It is unlikely, however, to significantly improve TLR rates with the traditional mesh-design nitinol self-expanding stents (when excluding intraprocedure bailout stenting as a TLR).^10–14

The long-term outcome (>3 years) with stenting, however, remains unclear. The fate of these stents after several years of compression, elongation, torsion, and twisting is unknown. Despite significant improvement in stent design leading to zero fracture rate at 1 year in the SUPERB trial using an interwoven stent design, ⁷ stent fractures (particularly with mesh-like design) tend to occur more frequently in complex and calcified lesions and their significance is yet to be clearly defined, although high grade fractures are associated with loss of patency. ¹⁵ In addition, treatment of femoropopliteal ISR is not benign and continues to be associated with higher rates of recurrent restenosis and TLR, particularly in long lesions and/or total occlusions. Furthermore, the risk of compromising surgical targets with stenting, particularly in common femoral or popliteal segments, is a real problem and should be avoided when possible.

The problems with stenting the femoropopliteal segment are well illustrated by Armstrong et al ¹⁶ in this issue of the JEVT. In this “real world” retrospective study, stent fractures at index were observed in 10% to 11% of patients. Also, treatment of ISR, though highly effective from a procedure standpoint, continued to carry a high rate of recurrent restenosis and need for TLR, particularly in total occlusions at 2 years (43%), which was not different from BA (48%) despite higher patency with the laser. This is in contrast to data from the EXCITE ISR trial, ¹⁷ which showed a significant advantage of the laser over BA in long and complex lesions in reducing TLR. The main differences between these 2 studies are the extent of debulking, which was more aggressive in EXCITE using the tandem catheter, and the overall shorter duration of follow-up that would favor a better outcome with the laser. Beyond the 6-month follow-up interval, a rapid decline in patency is seen with the laser following ISR treatment.^7,18,19 It is unclear whether the initial advantage of the laser in reducing TLR in complex lesions is lost at 2 years’ follow-up, and further data are needed to clarify this hypothesis. Kaplan-Meier plots in EXCITE appear to show an advantage for the laser over BA at 1-year follow-up in reducing TLR; however, the study does not extend its follow-up to >1 year. Regardless, there is consistency among all the studies,^7,18,19 including the article by Armstrong et al, ¹⁶ that a continued decline in patency and freedom from TLR is seen with the laser on follow-up, and adjunctive therapy with anti-restenotic drugs is needed to sustain this initial advantage over BA.

Armstrong and colleagues ¹⁶ have interestingly shown in their study that the TLR at 2 years in short lesions and nonocclusive disease favored the laser compared with BA (14% vs 44%, respectively) despite the significantly longer lesions in the laser cohort. This finding contrasts with the 6-month follow-up of EXCITE, and it could be related to chance although one cannot exclude the possibility that short lesions may to do better with the laser vs BA on long-term follow-up. This hypothesis is also interesting and may need to be further explored.

Treating femoropopliteal ISR was also associated with distal embolization (higher with laser than BA), not a surprising finding since total in-stent occlusions are generally stenotic-thrombotic in nature. In the DEEP EMBOLI study, ²⁰ 22.2% of filters in patients treated with the laser had significant macrodebris deemed large enough to have resulted in clinically significant embolization if not for the placement of a filter. The cost-effectiveness of the filter use in treating femoropopliteal ISR with the laser is unknown, but it is generally associated with less radiation exposure and low contrast use. ²¹ The use of filters with the laser is significant and seems to be common among operators, as seen in several studies.^16-18 A low threshold for filter use in treating femoropopliteal ISR is likely to continue, particularly in long lesions, total occlusions, or thrombotic-restenotic lesions.

In addition, in the study by Armstrong et al, ¹⁶ a high rate of bailout stenting was associated with both BA and the laser, but the authors argued that partial debulking with the predominant use of the Turbo Elite laser catheter (44/54 cases) may have led to this outcome. This was also seen with another retrospective study where the Turbo Elite catheter was the predominant device used, and bailout stenting was reported to be 50%. ¹⁸ More aggressive debulking with the tandem laser^17,19 has been associated with a low bailout stenting rate. Furthermore, the bailout stenting rate cannot be well defined in retrospective studies because of lack of prospective pre-specified endpoints as to when stenting should occur. When left to operator’s bias, the incidence of true bailout stenting cannot be accurately determined. It should be noted that the Turbo Elite catheter has been frequently used to create a pilot channel before the use of the Turbo-Tandem laser, particularly in severe or total femoropopliteal in-stent occlusions.^17,19 Although this strategy is important for a successful laser tandem application, it adds another level of expense to the procedure, which should be kept in mind. The cost of using 2 laser catheters and a filter to obtain optimal angiographic results and reduce embolization may be partially offset by reducing the cost of excessive bailout stenting (usually more than one stent in long lesions), a hypothesis that needs further testing.

The advent of drug-coated balloons (DCBs), however, is likely to change the landscape of femoropopliteal ISR treatment. Early clinical trials showed excellent patency and reduced TLR with DCBs (alone or as adjunctive treatment with atherectomy) when compared to BA in treating ISR.^22-24 The advantage of atherectomy prior to DCB or BA is the reduction of dissections and bailout stenting when optimal debulking is achieved. A low bailout stenting rate was also seen with the JetStream XC (Boston Scientific Corporation) off-label application in femoropopliteal ISR. In the JetStream ISR study, bailout stenting was reported at 5.3%. ²⁵ Debulking with the JetStream Navitus device has been shown to yield optimal results in porcine models using intravascular ultrasound and quantitative vascular angiography. ²⁶ Therefore, minimizing stenting with an initial debulking strategy coupled with a reduction of smooth muscle cell proliferation with DCBs may lead to improvement in clinical outcomes and the avoidance of concurrent stenting, a concept that applies to both the upfront treatment of de novo femoropopliteal lesions as well as the restenotic process when it occurs within or without a stent.

Armstrong et al ¹⁶ confirmed prior observations that operators appear to select the laser for longer (222 vs 114 mm) and more complex lesions (total occlusions 69% vs 20%) than BA, respectively. Shammas et al ²⁷ noted the same when laser was compared with SilverHawk (Medtronic CardioVascular, Santa Rosa, CA, USA) for treating femoropopliteal ISR. In this study, the laser was also selected by operators for more complex and long lesions. This bias in device selection is substantiated by data from the EXCITE ISR trial, in which the Turbo-Tandem laser was shown to be superior to BA in reducing TLR and improving patency in more complex lesions. The Turbo-Tandem laser yields a higher level of aggressive debulking, which is expected to delay the occurrence of TLR.

In conclusion, several important points emerge from the study of Armstrong and colleagues. ¹⁶ The long-term follow-up with the laser beyond 1 year in complex femoropopliteal ISR lesions continues to show a progressive loss of patency and lower freedom from TLR. Also, in these complex lesions and despite better patency with the laser, the advantage of reducing TLR with the laser is lost at 2 years. This observation, however, does not seem to apply to short, nonocclusive disease where the laser on short-term follow-up does not show significant advantage over BA, ¹⁷ but on longer follow-up the gap widens to the advantage of the laser. ¹⁶ In addition, suboptimal debulking leads to higher bailout stenting, as seen in the Armstrong study, and operators need to pursue an aggressive approach using the Turbo-Tandem catheter. Furthermore, distal embolization remains a problem with the laser in treating femoropopliteal ISR, and a low threshold for filter use is needed to avoid this complication, particularly in high risk complex lesions. Finally, adjunctive treatment with DCBs following the laser may emerge as an optimal strategy in treating femoropopliteal ISR, with the laser preparing the lesion prior to treatment with the DCB. The conclusions from the study by Armstrong et al ¹⁶ need to be considered hypotheses-generating, taking into account the inherent biases and the limitations in the design of this retrospective study.