Drug-coated balloon angioplasty versus balloon angioplasty for treating patients with in-stent restenosis in the femoropopliteal artery

Abstract Background: The introduction of endovascular surgery has led to frequent stent use, although in-stent restenosis (ISR) remains a challenging issue. Drug-coated balloon (DCB) and conventional balloon angioplasty (BA) are common endovascular procedures for addressing ISR in the femoropopliteal artery. However, there is controversy regarding which procedure provides the greatest benefit to patients. Methods: The PubMed, EMBASE, and Cochrane Central Register of Controlled Trials databases were searched for prospective controlled trials that compared DCB and BA for patients with ISR in the femoropopliteal artery. The study has been approved by Ethics Committee of Wuhan Central Hospital. Results: The meta-analysis included 6 prospective trials with 541 patients. We found that DCB use was associated with significant reductions in binary restenosis at 6 months (relative risk [RR]: 0.45, 95% confidence interval [CI]: 0.33–0.63; P < .00001), binary restenosis at 1 year (RR: 0.44, 95% CI: 0.34–0.57; P < .00001), target lesion revascularization (TLR) at 6 months (RR: 0.36, 95% CI: 0.20–0.65; P = .0006), and TLR at 1 year (RR: 0.38, 95% CI: 0.27–0.54; P < .00001). The DCB group also had significantly better clinical improvement (RR: 1.39, 95% CI: 1.13–1.71; P = .002), although we did not detect inter-group differences in terms of death, target vessel thrombosis, or ipsilateral amputation. The brand of DCB may a cause of heterogeneity. Conclusion: Relative to BA, DCB use increases the durability of treatment for ISR in the femoropopliteal artery, based on significant reductions in binary restenosis and TLR at 6–12 months after the procedure. Furthermore, DCB use was associated with better clinical improvement. However, additional randomized controlled trials are needed to validate these findings.


Introduction
The prevalence of peripheral artery disease (PAD) increases with age, and this disease affects an estimated 20% of people who are >70 years old. [1] Rapid improvements in endovascular instruments and physician experience have led to endovascular therapy being recommended as a first-line option for femoropopliteal artery disease. [2] However, percutaneous intervention still has a limited success rate, especially in patients with stenosis or occlusion of the femoropopliteal arteries, as percutaneous transluminal angioplasty (PTA) is associated with restenosis rates of up to 58% during the first 6 to 12 months. [3] Thus, instent restenosis (ISR) has become a challenging issue. [4] Drug-coated balloon (DCB) use has shown promising results in reducing restenosis of coronary stents, [5,6] although there is no standard treatment for ISR in the femoropopliteal artery. [7] Several recent randomized studies have evaluated the safety and efficacy of DCBs among patients with ISR in the femoropopliteal artery, [8][9][10] although the findings have been inconsistent. Thus, there is no consensus regarding the ideal strategy for treating ISR in the femoropopliteal artery, and we performed a meta-analysis to compare the outcomes of DCB use and balloon angioplasty (BA) in this setting.

Search strategy
The PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials databases were searched for English reports of studies that were published before October 2020. The searches were independently performed by two reviewers using the keywords "in-stent restenosis," "drug coated balloon," "drugeluting balloon," "paclitaxel-eluting balloon," and "femoralpopliteal arteries." The reference lists of identified reports and review articles were also examined to identify potentially relevant studies. This study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines.

Study selection
Titles and abstracts were independently screened by two investigators to identify prospective controlled trials that evaluated the outcomes of DCB and BA treatment for patients with ISR in the femoropopliteal artery. Any disagreements regarding a study's eligibility were resolved via re-reading and discussion. The inclusion criteria were: 1. patients with documented ISR in the femoral and/or popliteal arteries and 2. follow-up of ≥6 months.
Studies were not restricted based on the use of any specific type or brand of DCB. The exclusion criteria were: 1. target vessels that were not the femoropopliteal arteries, 2. non-BA and non-DCB treatment, and 3. review articles or duplicate studies.

Data collection and quality assessment
Two reviewers independently extracted and tabulated data from each article regarding baseline demographic characteristics, procedural variables, follow-up time, and primary and secondary endpoints. Data were extracted from the main text and tables of the published reports, as well as from online materials if available. The same authors also evaluated the quality of the included studies using the Cochrane Collaboration's tool for assessing the risk of bias. [11] Any disagreements between the reviewers were resolved via discussion.

Outcome variables
The outcome measures were defined according to previously published reporting standards and were analyzed on an intention-to-treat basis. [12] The primary outcome was defined as the likelihood of target lesion revascularization (TLR). TLR was performed if clinically indicated, and when a target lesion diameter stenosis of 50% was present. Secondary outcomes were defined as binary restenosis, clinical improvement, death, target vessel thrombosis, and ipsilateral amputation. Binary restenosis was defined as a >50% diameter stenosis (by angiography) or a peak systolic velocity ratio 2.5 (by duplex ultrasound). Clinical improvement was defined as at least 1 Rutherford-Becker category observed. Ipsilateral amputation defined as unplanned amputation of the target limb. Death defined as all-cause death.

Statistical analysis
Data were managed and analyzed using Review Manager software (version 5.2; the Cochrane Collaboration) and STATA software (version 12.0; STATA Corporation, College Station). Pooled risk ratios (RR) were calculated for dichotomous variables using the Mantel-Haenszel method and pooled mean differences were calculated for continuous variables. Outcomes were reported with the 95% confidence intervals (CIs) and results were considered statistically significant at P-values of <.05.
Heterogeneity was evaluated using the Cochrane Q test. [13] A fixed effects model was used if there was no significant heterogeneity (P > .1 and I 2 < 50%), while a random effects model was used if there was significant heterogeneity (P < .1 or I 2 > 50%). Sensitivity analyses were performed by excluding each study and re-analyzing the data. Various subgroup analysis was also been carried to address the heterogeneity. Publication bias was evaluated using funnel plots, Begg rank correlation, and the Egger linear regression test, [14] with significant publication bias considered present at P-values of <.05.

Eligible studies
The study selection flowchart is shown in Figure 1. The search identified 968 articles, although 825 articles were excluded after a review of the titles and abstracts. Full-text reviews were performed for the remaining 143 articles, and 6 studies with 541 patients ultimately fulfilled the inclusion criteria. Three different brands (Medtronic, FREEWAY, Medrad) of DCBs and 2 different paclitaxel doses (3 and 3.5 ug/mm 2 ) were used in the included studies. The main characteristics of the included studies are reported in Table 1. The patients' demographic characteristics and risk factors are shown in Table 2.

Risk of bias
The findings regarding risk of bias are summarized in Figure 2. There were low risks of selection bias and reporting bias for all trials, although 4 trials had high risks of performance bias, which was the most prominent quality issue. Two studies with >2 risk factors for bias were considered low-quality, although the other studies were considered high-quality and the results were considered applicable. [15][16][17][18][19][20] with 541 patients provided data regarding TLR at 6 months. A fixed effects model was used based on the absence of significant heterogeneity (P = .5, I 2 = 0%). The use of DCBs was associated with a significantly reduced risk of TLR at 6 months (RR: 0.36, 95% CI: 0.20-0.65; P = .0006) (Fig. 3A).

Publications bias and heterogeneity analysis
A visual inspection of the funnel plot did not reveal any clear asymmetry. Similarly, no significant publication bias was detected using the Egger and Begg tests (P = .22) (Fig. 6). Sensitivity analyses, which involved omitting one study at a time from the meta-analysis, failed to indicate that the results were influenced by a particular study. Subgroup analysis showed that study designs (RCT, and prospective study) and paclitaxel dose (3 and 3.5 ug/mm 2 ) were not the cause of heterogeneity. Brand (Medtronic, FREEWAY) of the DCB may be a heterogeneity cause in the binary restenosis result.

Discussion
The main disadvantage of PTA and stenting is that high rates of ISR can significantly affect the clinical outcomes of femoropopliteal artery stenting. [21][22][23] Furthermore, the 1-year rates of ISR range from 18% to 37%. [24] Thus, the increased use of endovascular therapy makes ISR and its treatment a challenging issue. [23] The most recent meta-analysis included 3 trials with 263 patients, and revealed that DCB use provided advantages (vs uncoated BA) for treating ISR in the superficial femoral artery based on rates of TLR, binary restenosis, and clinical improvement within 2 years after the procedure. However, the level of evidence was considered low, based on the small number of studies and potential risks of bias. [25] In addition, the evidence was considered insufficient in practice to confirm the superiority of DCB over BA for treating ISR, given the high cost of DCBs. [25] Our review included larger numbers of trials and patients, and the results revealed that the DCB treatment group had significantly better outcomes in terms of the 6-to 12-month rates of TLR, binary restenosis, and clinical improvement. We also evaluated the adverse events for each treatment, which revealed similar amputation and mortality rates in the DCB and BA groups. Unfortunately, few studies have provided data regarding the costs during the follow-up period, although previous meta-analyses also support our conclusion to some extent. [26][27][28] Katsanos et al [26] reported that DCBs provided a >50% reduction in the rates of restenosis (including ISR) and TLR in the femoropopliteal artery, and suggested that standard paclitaxel dose (3.0 and 3.5 ug/mm 2 ) DCBs were more effective compared with low paclitaxel dose (2.0 ug/mm 2 ) in reduce both restenosis and TLR. In our meta-analysis, low paclitaxel dose subgroup (3.0 ug/mm 2 ) and high paclitaxel dose subgroup (3.5 ug/mm 2 ) were both superior to BA in restenosis and TLR. The differences between the two subgroups were hard to analysis, because of the small number of the included studies. Anantha-Narayanan et al also reported that the TLR rates (including for ISR) in the femoropopliteal artery were 45% lower in the DCB group than in the BA group. Tepe et al [20] reported a recent randomized controlled trial, which has not been included in previous meta-analyses, which revealed that DCB use was associated with significantly less late lumen loss and fewer TLR procedures up to 24 months after treatment. To the best of our knowledge, ours is the most comprehensive metaanalysis to compare the outcomes of DCB and BA treatment for ISR in the femoropopliteal artery. There are limited data regarding DCB use for PAD, although the existing evidence seems to indicate that DCB use is associated with a significant benefit. [4,29,30] The first cohort study revealed a 92.1% primary patency rate at 12 months after treatment using DCBs, [9] while a more recent study of 53 patients with ISR in the femoropopliteal artery revealed a primary patency rate of 83.7% ± 5.0% and ∼90% freedom from TLR after 1 year. [31] The results of these 2 trials suggest that DCBs are a promising treatment option for patients with ISR in the femoropopliteal artery, which agrees with our findings, although it is important to note that both studies did not have a control group.
Various strategies have been used to treat ISR in the femoropopliteal artery. One randomized study revealed that, relative to DCB alone, a combination of laser debulking and DCB provided significantly better patency rates at 6 months (91.7% vs 58.3%; P = .01) and at 12 months (66.7% vs 37.5%; P = .01). [32] Bosiers et al compared the ISR rates after treatment using PTA and Viabahn ePTFE-covered stents, which revealed that the Viabahn stents provided better 1-year rates of primary patency (74.8% vs 28%; P < .001) and freedom from TLR (80% vs 42%; P < .001). [33] Silverhawk atherectomy is not superior to PTA, as it was associated with increased reoccurrence of intimal media hyperplasia. [34] Drug-eluting stents are a questionable treatment for ISR, as the scaffolding is not needed to manage the migration, proliferation, and collagen synthesis of smooth muscle cells. [31] Cutting balloons are also not superior to conventional PTA. [35] Our meta-analysis has some important limitations. First, the pooled analysis was based on study-level data, which could be confounded by inaccurate or incomplete data reporting. Second, the analysis only included a small number of studies with relatively short follow-up periods, which might be inadequate for detecting late adverse events, such as amputation, death, and very late thrombosis. Third, we only considered reports that were published in English, which is a potential source of bias.

Conclusion
In conclusion, our results indicate that DCB use was superior to BA for treating patients with ISR in the femoropopliteal artery, based on significantly better 6-to 12-month rates of binary restenosis, TLR, and clinical improvement. Furthermore, the rates of amputation and mortality were similar between the DCB and BA treatment groups. However, additional randomized controlled trials are needed to validate these findings.