Author + information
- Received August 6, 2015
- Revision received October 14, 2015
- Accepted October 22, 2015
- Published online February 8, 2016.
- Konstantinos Katsanos, MSc, MD, PhDa,b,∗ (, )
- Stavros Spiliopoulos, MD, PhDb,
- Athanasios Diamantopoulos, MD, PhDa,
- Dimitris Siablis, MD, PhDb,
- Dimitris Karnabatidis, MD, PhDb and
- Dierk Scheinert, MDc
- aDepartment of Interventional Radiology, Guy's and St. Thomas' Hospitals, NHS Foundation Trust, King's Health Partners, London, United Kingdom
- bDepartment of Interventional Radiology, Patras University Hospital, School of Medicine, Rion, Greece
- cDepartment of Angiology, Park Hospital and Universitätsklinikum, Leipzig, Germany
- ↵∗Reprint requests and correspondence:
Dr. Konstantinos Katsanos, Department of Interventional Radiology, Guy's and St. Thomas' Hospitals, NHS Foundation Trust, King's Health Partners, London SE1 7EH, United Kingdom.
Objectives The authors sought to report the wound healing outcomes, health-related quality-of-life changes and quality-adjusted life-years (QALYs) gain in the 2 treatment arms of the ACHILLES (Comparing Angioplasty and DES in the Treatment of Subjects With Ischemic Infrapopliteal Arterial Disease) multicenter randomized trial.
Background The ACHILLES randomized trial has previously shown that sirolimus-eluting stents (SES) may achieve lower vessel restenosis and higher event-free survival rates compared with plain balloon angioplasty (PTA) for infrapopliteal lesions.
Methods A total of 200 patients were randomly assigned between SES and PTA for the treatment of infrapopliteal arterial occlusive lesions. Progression of wound healing was serially assessed by digital photography. Health-related quality-of-life scores were assessed with the self-administered EQ-5D questionnaire up to 1 year from randomization. QALYs gained were calculated with a standard multiplicative model using distribution-free Bayesian modeling.
Results In total, 109 open wounds (n = 54 in SES; n = 55 in PTA) were documented at baseline. At 6 months, wound volume reduction (%) was significantly higher in the SES group (95% healing [95% confidence interval (CI): 80% to 99%] compared with 60% healing [95% CI: 13% to 90%] in the PTA group; p = 0.048). At 1 year, rates of complete wound closure were higher in the case of SES (72.9% vs. 55.6% closed wounds in PTA; p = 0.088). The recorded weighted EQ-5D score improved significantly up to 1 year in case of SES (p < 0.0001), but not in case of PTA. There was a trend of more QALYs gained with SES compared with PTA up to 1 year after randomization. Relative QALY gain was 0.10 (95% CI: −0.01 to 0.21; p = 0.08) in the whole study and 0.17 (95% CI: −0.03 to 0.35; p = 0.09) in the wound subgroups comparison.
Conclusions Infrapopliteal SES accelerates wound healing and may improve quality of life compared with PTA. (Comparing Angioplasty and DES in the Treatment of Subjects With Ischemic Infrapopliteal Arterial Disease [ACHILLES]; NCT00640770)
- balloon angioplasty
- critical limb ischemia
- sirolimus-eluting stent(s)
- wound healing
The incidence of critical limb ischemia (CLI) in the Western population is estimated between 500 to 1,000 per 1,000,000 persons per annum, but is anticipated to rise mainly due to the continuously increasing incidence of diabetes and progressive increase of life expectancy in the developed world countries (1,2). Ischemic foot ulcers typically develop in end-stage CLI, resulting in minor or major tissue loss and are the main cause of major amputations, especially if no revascularization is attempted (3). Infrapopliteal arterial occlusive disease is a leading cause of CLI, and recent developments in percutaneous endovascular treatment of CLI included 3 multicenter randomized controlled trials (RCT); the YUKON-BTX (YUKON-Drug-Eluting Stent Below the Knee-Randomised Double-Blind Study), the DESTINY (Drug Eluting Stents in the Critically Ischemic Lower Leg), and the ACHILLES (Comparing Angioplasty and DES in the Treatment of Subjects With Ischemic Infrapopliteal Arterial Disease) trials, which have provided Class 1, Level of Evidence: A, about the safety and clinical effectiveness of infrapopliteal placement of drug-eluting stents (4–7).
The ACHILLES prospective multicenter RCT was the only one among the aforementioned studies that compared the performance of drug-eluting stents (CYPHER SELECT sirolimus-eluting stent [SES]; Cordis, Johnson & Johnson, Bridgewater, New Jersey) with the well-established percutaneous transluminal balloon angioplasty (PTA) in infrapopliteal ischemic arterial disease (5). In total, 200 patients were randomized between SES and PTA for the treatment of infrapopliteal arterial lesions at 16 European institutions. The study was designed and powered using 1-year angiographic in-segment binary restenosis assessed by quantitative analysis as the primary endpoint. The previously published 1-year analysis reported significantly lower angiographic vascular restenosis (22.4% vs. 41.9%; p = 0.019) and higher vessel patency (75.0% vs. 57.1%; p = 0.025), but rates of index-limb amputation, patient death, target limb revascularization (TLR), and Rutherford class changes were largely similar in the SES and PTA groups at 1 year. Nonetheless, patients treated with infrapopliteal SES experienced a significantly improved overall event-free survival (p = 0.028 compared with PTA), which was a post-hoc composite survival endpoint defined as freedom from death, TLR, bypass/amputation, and Rutherford class ≥4) (5).
The ACHILLES protocol included also the documentation of further important clinical variables such as index-limb wound status and the recording of health-related quality-of-life (HRQOL) parameters. Progression of wound healing and serial changes of HRQOL outcomes were not reported in the initial report because these data were scheduled to be collated, analyzed, and published separately. The authors herein report the final 1-year results from the wound healing outcomes and HRQOL changes from the ACHILLES multicenter RCT to provide a more comprehensive consideration of the treatment effect of SES compared with PTA in the infrapopliteal arteries.
The ACHILLES multicenter RCT was registered in a public electronic database (NCT00640770), and its design, material and methods, and angiographic and clinical results have been published previously in great detail (5). Briefly, randomization was performed on a 1:1 basis using a numbered envelope system, and both severe intermittent claudication and CLI patients were recruited in the study (Rutherford-Becker class 3 to 5; ≤2 lesions with ≥70% stenosis, maximum lesion length ≤120 mm). In total, 99 patients (113 lesions) were assigned to SES placement, and 101 patients (115 lesions) to standard PTA treatment. Baseline demographics and lesions characteristics were evenly distributed between the 2 groups, and average treated lesion length was nearly 29 mm in both cases. Eight cases crossed over from PTA to SES treatment because of a suboptimal angioplasty result. After the index procedure, all participants received low-dose aspirin, and SES subjects were also prescribed standard dose thienopyridine treatment (clopidogrel or ticlopidine) for 6 months. After 1 year of follow-up, 74 patients were evaluable in the SES group and 80 patients in the PTA group.
Wound healing assessment
The assessment of wound healing was pre-specified in the original protocol. Index-limb wound status was assessed at baseline and at pre-scheduled follow-up visits at 6 weeks, 6 months, and 12 months following the index procedure. Digital photographs of the wounds were taken before the index procedure and on each follow-up visit with a dedicated 1-cm2 graded metallic frame to assess progress of wound healing. All wounds were reviewed and evaluated by a diabetic foot specialist acting as an independent reviewer, who was blinded to initial treatment allocation. Quantitative volumetric analysis of wound healing progress was performed to allow for objective comparison between the 2 study arms. Wound status reports included pre-determined wound codes for the anatomic location of the index wound on the target limb, assessment of local infection, and review of wound closure status at different time points.
All amputations were classified by the investigational sites and adjudicated by the clinical events committee with all clinical documentation available (e.g. operation report, photograph). A minor amputation was defined as any amputation involving removal of a toe or part of the foot (Level 1). A wound that was healed within 30 days following a minor amputation was considered “closed.” Once a wound was adjudicated as “closed” in one assessment, and wound healing information was missing at a later follow-up time point, the status “closed” was carried forward if there was no repeat revascularization between the 2 follow-up time points. Conversely, once a wound was adjudicated as “not closed,” and wound status information was missing at a later follow-up time point, then the wound status information was considered as missing at the later time point.
Health-related quality of life
HRQOL utility scores were assessed using a self-administered EuroQoL 5-D system (EQ-5D) with higher scores indicating improved health and well-being as subjectively perceived by each patient. All subjects were asked to provide HRQOL data up to 1 year following randomization regardless of the presence of tissue loss at baseline or amputation at follow-up. The EQ-5D questionnaire is designed to evaluate different health parameters, including mobility, self-care, everyday activities, pain or discomfort, and anxiety or depression (EuroQol health dimensions), using 3 ordinal levels (EuroQol health scoring system: “no problems,” “some problems,” or “extreme problems/unable to”) for each dimension. The EQ-5D scores were then adapted into a preference-based, single weighted, utility score (EQ-5D score) using the time trade-off method as described elsewhere (8,9). Aforementioned patient-reported outcome measures (PROMS) were recorded at screening (before the procedure) and at 6 weeks, 6 months, and 12 months after the index procedure.
Quality-adjusted life-years (QALYs) were calculated for each group at each time interval (0 to 6 weeks, 6 weeks to 6 months, 6 months to 12 months) with a standard multiplicative model using the calculated group-specific EQ-5D score relating to each time interval (8). Cumulative QALY scores of each study arm for every time point were calculated by summation of individual time-specific QALYs over the duration of the trial. Some PROMS data (individual domains and weighted EQ-5D score) were missing either because trial participants died or did not complete the questionnaires or were lost from follow-up. Overall, PROMS data were missing because of attrition in 17.6% of the cases (5.5% at screening, 13.5% at 6 weeks, 23.5% at 6 months, and 28.0% at 12 months). Because of nonnormal heavily right-sided skewed distributions of PROMS data in both groups at the various time points, all QALY calculations and pairwise comparisons were performed with noninformative distribution-free Markov Chain Monte Carlo simulation (MCMC with Gibbs sampling; WinBUGS version 1.4.3, MRC Biostatistics Unit, Cambridge, United Kingdom).
All data including patients’ demographics, procedural details, electronic photo files, wound status and HRQOL reports, were referred from all 16 centers to a contract research organization (Clinquest Europe, Leiden, the Netherlands) where data management and analysis was performed independently from the study sponsor (Cordis, Johnson & Johnson). Amassed data were provided by Cordis Corporation and analyzed independently from the company without any direct sponsorship or funding for the purposes of this analysis. Descriptive statistics were used for all baseline, lesion, and procedural variables. Categorical variables were expressed as counts (percentages), and continuous variables were given as medians with interquartile ranges (i.e., between the 25th and 75th percentiles) or as mean ± SD if they passed the Kolmogorov-Smirnov goodness-of-fit normality test. The unpaired Student t test was used to test normally distributed continuous variables; the Mann-Whitney U test was used for nonparametric continuous variables. The Kruskal-Wallis test was used for nonparametric testing of multiple group comparisons. Comparison of proportions was done by testing the null hypothesis that the proportions were equal and the chi-square test (or the Fisher exact test in case of single-digit event counts) was applied in case of cumulative analysis. Event-specific adjusted denominators were applied for reporting of cumulative event rates up to 360 days following the index procedure. All outcome measures were analyzed on both an intent-to-treat (ITT) and as-treated basis (10). The threshold of statistical significance was set at alpha <0.05. Statistical analysis was performed with the GraphPad Prism statistical software package (version 5, GraphPad Software, La Jolla, California). Bayesian modeling for distribution-free calculation and comparison of wound healing percentages and QALYs differences between SES and PTA groups was done with a Gibbs sampler and Bayesian p values and 95% credible intervals are reported (WinBUGS version 1.4.3; code available on request).
A total of 109 open wounds were observed at baseline and evaluated per protocol at all follow-up time points. At baseline, 54 open wounds were recorded in 41 subjects in the SES arm (41 of 99; 41.4%) and 55 open wounds in 37 subjects in the PTA arm (37 of 101; 36.6%). Baseline demographics of the open wound patient cohorts were well balanced and in line with the characteristics of the overall patient groups (Table 1). Discordance was noted only in case of hyperlipidemia (well matched in the whole RCT, but significantly higher in the SES wound subgroup compared with the PTA wound subgroup) and a trend in case of diabetes (numerically more frequent in the PTA wound subgroup compared with the SES wound subgroup) as outlined in detail in Table 1. Pre-procedure target lesion stenosis was significantly higher in the PTA group both when analyzing the whole randomized study and the wound subgroups separately. Up to 1 year, comparative analysis of the angiographic and clinical findings in the wound cohorts yielded similar results with the overall study; post-procedure residual stenosis was significantly less in the SES group (19.3 ± 9.49% compared with 28.9 ± 11.2% in the PTA group; p < 0.001), as well as binary restenosis of angiographically evaluable lesions at 1 year (7 of 25 [28.0%] in the SES arm compared with 15 of 23 [65.2%] in the PTA arm; p = 0.019). Minor and major amputations were largely similar in the wound subgroups analysis and represented the majority of the amputations recorded in the ACHILLES study (Table 1).
Wound healing outcomes
Rates of closed wounds increased consistently throughout the whole follow-up period in both treatment groups. When analyzed per ITT, patients treated with SES had 25.9% of wounds closed by 6 weeks following stent placement, 54.2% closed at 6 months, and 72.3% closed at 1 year follow-up. For comparison, wound closure rates in the PTA control arm were 20.0%, 46.9%, and 56.5%, respectively. None of these differences were statistically significant (p = 0.501, p = 0.544, p = 0.132), although a trend was noted in favor of the SES arm at 1 year. There were no differences in wound healing outcomes among different randomization canters (analysis of variance p = 0.29). At 1 year, the relative risk of complete wound healing was 1.29 (95% confidence interval [CI]: 0.83 to 2.00) between the SES and PTA arms when including the treatment center as a random effect to account for potential healthcare variations across different randomization sites (I2 heterogeneity: 15.2%). In the “as-treated” group comparison to account for crossovers, results were similar, but the difference at 1 year was more pronounced in favor of the SES, albeit also not statistically significant (72.9 vs. 55.6% closed wounds; p = 0.088). For more details, please see Figure 1.
Figure 2 demonstrates the ITT results of the volumetric quantitative analysis of wound healing progression that showed better results in case of the SES-allocated patients. Absolute reduction of wound volumes (cm3) was numerically higher both at 6 weeks and at 6 months in the SES arm compared with the PTA arm. Adjusted percentage reduction of wound volumes (% healing) demonstrated similar differences in favor of the SES group at 6 weeks and at 6 months; the latter being significantly better (95% healing [95% CI: 80% to 99%] in the SES-treated limbs compared with 60% healing [95% CI: 13% to 90%] in the PTA cases; p = 0.048). At 1 year, SES- and PTA-treated wounds showed very similar results in terms of progression of wound healing (Figure 2).
HRQOL outcomes were analyzed and compared between SES and PTA first for the whole-trial arms (N = 200) and second for the subgroups of patients with recorded wounds at baseline (n = 78; post-hoc wound cohort analysis). Overall, there was an improvement in most of the individual domains of the captured EQ-5D data. Recorded changes were more prominent in the domains of “mobility” and “pain,” and to a lesser extent in the domain of “anxiety,” whereas no notable differences were noted in the domains of “self-care” and “activity.” Most of the improvement in the quality-of-life variables was noted within the first 6 weeks following the revascularization procedure, but there were no significant differences when comparing individual PROMS domains between the 2 treatments under investigation (SES vs. PTA) (Figure 3). Calculated weighted EQ-5D scores showed some improvement over time in both groups, which again was more pronounced within the first 6 weeks after treatment. However, multiple group comparison showed that this pattern of improvement was highly statistically significant only in the SES group (p < 0.0001 on ITT analysis). Largely similar findings in terms of overall quality-of-life improvement were found when analyzing separately only the subgroups of patients with wounds at baseline (wound cohort subgroup analysis; p = 0.02 on ITT analysis) (Figure 4).
Compared with baseline, infrapopliteal SES placement produced statistically significant QALY gains at all 3 different time points (6 weeks, 6 months, and 12 months), whereas in the PTA group, QALY gains were significant only up to 6 months. Furthermore, the SES group exhibited a steady gain of QALYs throughout the 1-year follow-up period, compared with only minor gains in the PTA group up to 6 months. In the SES group, 0.13 (95% CI: 0.05 to 0.21) QALYs were gained at 1 year after randomization compared with 0.03 (95% CI: −0.06 to 0.11) in the PTA control group. Head-to-head comparisons between the SES and PTA groups demonstrated a relative gain of 0.10 (95% CI: −0.01 to 0.21) QALYs in favor of the SES group at 1 year (p = 0.08). The relative QALY gains were more pronounced in favor of the SES group when focusing on the wound subgroups. In the SES group, 0.22 (95% CI: 0.09 to 0.36) QALYs were gained at 1 year after randomization compared with 0.06 (95% CI: −0.08 to 0.19) in the PTA control group. Head-to-head subgroup comparisons demonstrated a relative gain of 0.17 (95% CI: −0.03 to 0.35) QALYs in favor of the SES wound cohort at 1 year with a trend towards significance (p = 0.09) (Figure 5).
Generic HRQOL tools are frequently used as secondary outcome measures in randomized trials to assess the perceived health improvement and to allow clinically relevant cost-utility analysis. Typically, quality of life is depressed in diabetes and further impaired by the presence of diabetic foot ulcers. On the other hand, successful wound healing is associated with improvements in HRQOL (11,12). Over the past decade, several studies including multicenter randomized trials provided evidence demonstrating the superiority of below-the-knee drug-eluting stenting compared with PTA or bare-metal stenting in terms of both angiographic and clinical endpoints (13). To our knowledge, this is the first report of wound healing outcomes and HRQOL outcomes following balloon angioplasty and drug-eluting stent placement in the infrapopliteal arteries.
The ACHILLES RCT has previously reported compelling evidence of reduced vascular restenosis and improved overall event-free survival in favor of sirolimus-eluting stents compared with PTA as the control group. The authors have herein further reported on the wound healing outcomes of patients with tissue loss at baseline and have shown a clear benefit of SES over PTA. Progression of wound healing was significantly accelerated in the SES allocated limbs (up to 6 months) and final complete wound closure rates were higher at 1-year follow-up. In order to attain precise and unbiased outcomes, wound healing assessment was performed and adjudicated by an independent diabetic foot specialist blinded to initial treatment allocation. In addition, wound volumes were quantified to allow a more robust comparison of the healing process. Results from the ACHILLES wound cohorts suggest that more than two-thirds of the patients treated with SES will experience complete wound healing within 1 year from the baseline procedure compared with around half of the patients treated with PTA. Of further interest, in case of infrapopliteal SES placement most of the wound healing progress will take place within the first 6 months from treatment compared with more delayed healing up to 1 year in case of PTA. Accelerated wound healing could be theoretically associated with reduced risk of infection and also reduced health care costs (e.g., shorter duration of wound care needed). Successful wound healing has been previously related with improved HRQOL ratings (10,14). On the other hand, low reported HRQOL has been found to correlate with future major amputations and death (15).
In the ACHILLES trial, HRQOL was measured using the self-administered EQ-5D questionnaire, which is an established, easy-to-use generic instrument for the clinical assessment of perceived health outcomes in various settings, as well as for health-economic evaluations enabling the transformation of outcomes in QALYs (8,15). Moreover, it has been widely used for the investigation of HRQOL in diabetic patients with foot ulcers (10,14,15). According to rigorous analysis of all collected EQ-5D data, both revascularization strategies (SES and PTA) improved HRQOL at all time-points compared with baseline, up to 1 year of follow-up. Interestingly, in line with the wound healing outcomes, a trend was noted when analyzing the quality-of-life metrics provided by the weighted score when combining all EQ-5D domains. Improvement of HRQOL outcome measures over time was markedly more pronounced in the SES group. Furthermore, in head-to-head comparisons of both the overall study arms and the wound cohort subgroups, there was a stronger numerical benefit of QALYs gained in the case of SES-treated subjects. The trend of improvement seemed to be even more pronounced in the case of participants with wounds at baseline with nearly twice as much a relative gain of QALYs compared with the complete dataset analysis. Up to 1 year, use of SES produced a relative gain of 0.10 QALYs (p = 0.08) when examining the whole randomized trial and 0.17 QALYs (p = 0.09) when comparing the wound subcohorts separately. Arguably, the trial was not a priori adequately powered for quality-of-life changes.
Still, our findings compare favorably with other HRQOL reports following peripheral angioplasty. For example, in the SUPER (SMART stent versus balloon angioplasty in long superficial femoral artery lesions) study that investigated stenting of the superficial femoral artery, each event of TLR related to a 0.17 reduction of the HRQOL utility score (16). Similarly, in the BASIL (Bypass versus Angioplasty in Severe Ischaemia of the Leg) trial that randomized femoropopliteal bypass versus angioplasty for severe leg ischemia, the HRQOL utility score improved by 0.30 in the angioplasty group and 0.34 in the bypass surgery group up to 1 year (8); by comparison, in the ACHILLES trial, the utility scores improved by 0.13 in the SES arm (0.22 in the SES wound subgroup) and only 0.03 in the PTA arm (0.06 in the PTA wound subgroup).
Although there is no analogous trial comparing wound healing outcomes and HRQOL metrics following different infrapopliteal revascularization strategies, these results are in accord with previously published data indicating that poor diabetic ulcer prognosis is associated with reduced HRQOL and that the self-assessment of quality of life score in patients with foot ulcers is lower compared with that reported from patients with healed foot ulcers (11,12,14). Therefore, future randomized trials should include PROMS, especially in the case of critical limb ischemia patients. Quality-of-life metrics and cost-utility analyses will be important in demonstrating a more pragmatic clinically relevant benefit in this very morbid patient population.
First, one limitation is the use of a single HRQOL instrument for the evaluation of perceived quality of life. However, the EQ-5D has been reported to provide reliable data regarding the health status of peripheral artery disease patients, whereas other analogous instruments such as the Short Form-36 (SF-36) do not permit calculations of QALYs for health-economic assessments. Furthermore, the use of additional tests would complicate the self-assessment process and could further increase attrition rates (8). Second, the ACHILLES trial was designed and powered using 1-year angiographic in-segment binary restenosis as the primary endpoint. Hence, the trial was not adequately powered for wound healing and quality-of life metrics, and this might explain why some of the results had a trend towards significance but did not cross the p < 0.05 threshold. Finally, the study did not include a priori collection of cost data, and therefore, a complete incremental cost-effectiveness analysis was not feasible.
Wound healing and HRQOL analysis from the ACHILLES prospective randomized controlled trial suggests that revascularization of infrapopliteal lesions with SES placement may improve wound healing and quality-of-life metrics compared with PTA.
WHAT IS KNOWN? The ACHILLES randomized trial has previously shown that sirolimus-eluting stents may achieve lower vessel restenosis and higher event-free survival rates compared with plain balloon angioplasty for infrapopliteal lesions.
WHAT IS NEW? We reported the HRQOL changes and quality-adjusted life-years gain in the 2 treatment arms of the ACHILLES multicenter randomized trial. We have demonstrated that infrapopliteal sirolimus-eluting stents accelerate wound healing and may improve quality-of-life metrics compared with plain balloon angioplasty.
WHAT IS NEXT? More appropriately powered randomized trials are necessary to further investigate the clinical utility and cost-effectiveness of below-knee drug-eluting stents for critical limb ischemia treatment.
Dr. Scheinert is a consultant for and/or member of scientific advisory board of Abbott, Biotronik, Boston Scientific, Cook Medical, Cordis, CR Bard, Gardia Medical, Medtronic/Covidien, TriReme Medical, Trivascular, and Upstream Peripheral Technologies. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- confidence interval
- critical limb ischemia
- EuroQoL EQ-5D
- 5-D system
- health-related quality of life
- patient-reported outcome measures
- plain balloon angioplasty/percutaneous transluminal angioplasty
- quality-adjusted life-year
- randomized controlled trial
- sirolimus-eluting stent(s)
- target lesion revascularization
- Received August 6, 2015.
- Revision received October 14, 2015.
- Accepted October 22, 2015.
- 2016 American College of Cardiology Foundation
- Scheinert D.,
- Katsanos K.,
- Zeller T.,
- et al.
- Siersma V.,
- Thorsen H.,
- Holstein P.E.,
- et al.