Drug-eluting balloons for femoro-popliteal artery disease: emerging research.

Background

Endovascular therapy is generally preferred as the first choice for revascularisation of symptomatic patients with PAD due to lower risk of procedural complications. Stenting is generallly the preferred modality but which one is optimal in femoral-popliteal artery disease has yet to be determined.
Some good-quality randomised controlled trials in the last five years will be reviewed in a final chapter of a short story regarding DEB and ischemic arterial disease.

Outcome measures; results.

Coronary Artery

Neo-intimal formation in the coronary artery using drug-eluting balloons; positive results: Pre-clinical research was first carried out in the context of coronary artery disease in 2010 and results were positive. Indeed, the anti-mitotic drug paclitaxel was reported to have been attached to the surface of a dilatation balloon and brought within the inner 2 mm. of a pig's coronary artery within a 30-second inflation inflation time (1). The drug loaded on the balloon was retreived in the vessel wall of immediately excised specimens at only 9-17% (2); concentration then decayed to a quarter of that value 12 hours following delivery and declined dramatically over the 7 days thereafter (3).
Compared to controls, relevant and statistically significant histomorphometric reduction of neo-intimal formation and vessel stenosi was measurable as early as 2 weeks after treatment: up to half the stenosis was noted in the treated arm (4,5). 
The antiproliferative action of paclitaxel was accountable for these results. Indeed, its lipophilic properties favour tissue uptake, and its cytotoxic mechanism of action lead to cell death - despite short tissue persistence. This action, moreso than transient inhibition of cell growth and replication or cytostatitism, was determined at play in preventing restenosis.
The highest drug efficacy was observed optimal at 2 to 3 µg/mm2 and 30 to 60 seconds.

Revascularation in stable de-novo lesions comparing DEB and DES, negative results: Then, clinical testing was carried on to stable de-novo lesions, the preferred setting for coronary artery disease. Drug-eluting balloons - rather then conventional balloon angioplasty, reflecting that coronary procedures are usually stent-based with established antiproliferative therapy - were compared to drug eluting stents. Results were negative.
Indeed, the PEPCAD III trial randomized 637 patients to treatment with a DEB followed by bare metal stent or a sirolimus-eluting stent. A nearly four-fold increase in late luminal loss and twice-fold higher rates of target-lesion revascularisation were observed at study's end - after a 9-month follow-up (6). These results were presented at the American Heart Association Scientific Sessions in 2009.
Two more studies comparing a DEB followed by BMS with an everolimus-eluting stent in 225 patients had similarly negative results (7, 8).
Two trials then randomised DEB alone and paclitaxel stents for the treatment of small coronary vessels (lumen diameter < 2,8 mm) and yielded diverging results: one trial was prematurely interrupted for inferiority of DEB (9), while in the other, DEB outperformed DES in terms of 6 month late lumen loss and target vessel revascularisation (10).

Coronary restenosis in stable de-novo lesions comparing DEB and DES, positive results: Drug eluting balloons were determined non-inferior to DES and leading to better outcomes than plain balloon with established transcatheter therapies in 5 randomised controlled trials with a total of 871 patients (11, 12).

Femoro-popliteal artery

Immediate technical failure, restenosis, and target vessel revascularisation comparing routine and provisional stenting, mixed results: The best endovascular treatment of symptomatic femoro-popliteal artery disease has not been established according to the 2009 meta-analysis led by Kasapis (13). The results of ten randomised controlled trials enrolling 1,343 patients altogether examining routine stent implantation vs. percutaneous transluminal angioplasty in femoropopliteal artery disease were reviewed. Indeed, systematic stent implantation may result in lower restenosis compared to balloon with provisional stenting, but with no significant differences in target vessel revascularisation.
It should also be noted that 1) the rate of stent implantation in the balloon arm of the pooled trials was low: 0% in four and only 8-32% across the other six (mean 7,6%) 2) heterogeneity of endpoints however no potential publication bias was explored. One additional randomised trial was completed after the meta-analysis was performed comparing nitinol (a nickel-titanium alloy) BMS with balloon in 150 symptomatic long, complex superficial femoral artery lesions (12 cm. mean length, total occlusion 93% of cases). With a crossover rate as low as 5% in the balloon group, this trial ended with a similar 12-month restenosis rate and a non significant reduction in TVR in the stent arm (14).

Neo-intimal proliferation comparing DES and BMS, inconclusive results: Furthermore research on the possible added benefit of DES in limiting neo-intimal proliferation after femoro-popliteal intervention is scarce. 
A 93 patient randomised trial of a sirolimus-coated nitinol stent vs. the same, uncoated device in relatively long lesions (mean length 8.3 ± 4.5 cm) ended with high long-term efficacy of BMS: the two-year angiographic restenosis rate was 21% in both groups. Two-year target lesion revascularisation was very low overall and there was no significantly different restenosis between the BMS and the DES arm: 13% vs. 6% (15). Nevertheless, it could be expected that this figure was twice that of the BMS group, and the difference would have reached significant figures with a bigger sample size.

Polimer-free paclitaxel-coated DES implantation or balloon dilatation (PTA), debatable results: In a subsequent larger trial, 479 symptomatic patients were assigned to polimer-free paclitaxel-coated DES implantation or balloon dilatation (PTA) of femoro-popliteal stenoses with a mean length of 6.3 ± 4 cm. (16). Patients in the balloon arm were further randomised to BMS or DES during the intervention if balloon only did not achieve satisfactory angiographic results, which happened in 50% of cases. These patients were counted as having PTA failure in the 12-month outcome analysis, thereby generating a huge primary patency difference between the two groups. When provisional stent placement was no longer considered a failure, in a more correct clinical perspective, twelve-month outcomes were again in favour of DES, with a restenosis rate of 17% vs. 27% (p=0.01) and a clinically driven target lesion revascularisation rate of 9.5% vs. 17.5% (p=0.01), Notably, 25% of PTA patients were treated with a DES, thus making the comparison rather spurious; it may be argued that the restenosis and revascularisation rates would have been higher in the PTA arm if only BMS were used.
These two pioneer trials did not encourage further clinical studies on DES, and industry and researchers turned their attention and efforts to DEB.

Late-lumen loss, and others in DEB or uncoated balloon treatment; positive results: Six randomised controlled trials of DEB vs. uncoated balloon in the treatment of chronic, symptomatic femoro-popliteal disease have been published between 2008 and 2013, and included 535 patients (17-22). The trials have similar design and inclusion criteria: symptomatic patients with angiography-measured ≥ 70% diameter stenosis or occlusion, mean lesion length of 7 ± 4 cm. and at least one patent vessel below the knee were randomly assigned to either DEB or uncoated balloon treatment, with provisional bare-metal stenting in case of residual diameter stenosis > 30% or flow-limiting dissection. The proportion of stent implantation varied from 4% to 21% among the DEB arms and from 14 to 34% among the uncovered balloon arms in 4 of the 6 trials. In one of them  all the patients were intentionally treated with a nitinol stent, and in another, 56% of the patients received a stent in both study groups. Another important descriptive feature is the percentage of restenotic lesion treatment across the trials. Two trials had restenosis as an exclusion criterion, while in the other four trials, restenosis accounted for 11% to 34% of treated lesions.
The six trials are homogeneous and consistent as to what regards pre-specified end-points. All of them had late lumen loss – defined as the difference between vessel diameter at the end of the procedure and vessel diameter at follow-up – as their primary end-point, which was assessed with paired selective quantitative angiography or duplex echo scan. Since DEB are assumed to specifically inhibit tissue proliferation at the site of balloon dilatation, it was completely adequate to address the measure of the phenomenon as top priority in these studies. In five of six trials late lumen loss was evaluated at 6 months, in one trial it was measured at 12 months. Late lumen loss was assessed by an independent core-laboratory blinded to treatment assignment in two of the trials. The results of these two trials did not differ from those found in the others, with unblinded assessment. All six trials showed comparable and statistically significant reductions in late lumen loss in lesions treated with DEB compared to controls, with mean values of -0.05 to 0.86 mm. for DEB-treated lesions and mean values of 0.7 to 1.7 mm. for uncoated balloon-treated lesions (Table).
Binary restenosis, defined as > 50% vessel diameter reduction, was again consistently and significantly reduced by DEB, with similar findings among the six trials. Restenosis rates in DEB-treated groups varied between 8.6% and 19% and ranged from 29% to 47% in their balloon-treated counterparts.
These angiographic/echographic measures translated in improved clinical outcomes across the six trials. Target lesion revascularisation, evaluated at the same time point of vessel imaging (mostly six months after the intervention), was always and largely in favour of DEB, with prevalence between 4% and 17%, compared with 22% to 37% in uncovered balloon (Table). No safety issues related to DEB use (e.g. stent thrombosis, aneurysm formation) were reported at follow-up in either study.
Four of the six randomised studies were included in a meta-analysis published in 2012 (23). Restenosis rate, calculated on 233 patients with available data, gave an odds ratio of 0.26 (95% confidence interval 0.14 – 0.48) in favour of DEB, with number needed to treat of 4 and no heterogeneity across the trials. Of 360 patients with available data for target lesion revascularisation, an odds ratio of 0.23 (95% confidence interval 0.13 – 0.40) in favour of DEB was calculated, also with number needed to treat of 4 and low heterogeneity. Both odds ratios were statistically highly significant. The trials were scrutinised for biases due to small study effects and lack of blinding. No significant influence of small sample, blinding status, and publication status was found, nor were the results altered by the subtraction of each one of the trials.

Patency or stenosis comparing DEB and DES, mixed results: In a retrospective analysis of 228 patients with femoropopliteal lesions ≥10 cm the two drugs were found to perform equally well with either treatment (24).
Compared with DES in long infrapopliteal lesions, DEB were found to produce positive vessel remodeling however DES were related with lower residual post-procedure stenosis and vessel restenosis at 6 months. PCB may produce positive vessel remodeling (25). 

Conclusions

Drug-eluting balloons are to be held equally safe and more effective than uncoated balloons for the treatment of femoro-popliteal arterial disease.  
Preliminary studies comparing DEB and DES have produced mixed results, however head-to-head comparisons of are planned in many upcoming trials.
Once their cost (today thrice as high as uncoated balloons) begins to lower, their use will conceivably increase. 
Regarding paclitaxel, it has been the only antiproliferative agent used in DEB as a single-shot high-dose therapy delivered to atherosclerotic plaques.

Table: Study results comparing drug-eluting balloons and uncoated balloons in the treatment of chronic, symptomatic femoro-popliteal disease (23).

DEB: drug-eluting balloon
UCB: uncoated balloon
Values are expressed as mean ± standard deviation except *: median value 
THUNDER, Fem Pac, LEVANT I, PACIFIER and DEBELLUM: 6-month follow up comparisons
DEBATE-SFA: 12-month follow-up comparisons 
#: p<0.05 vs. UCB