Abstract
Keywords
Introduction
Restenosis is a major limitation in superficial femoral artery (SFA) intervention, occurring in 40% to 60% of vessels 6 to 12 months after percutaneous transluminal angioplasty (PTA). 1 Drug-eluting balloons (DEB) are one potential solution to this limitation, as there is increasing evidence that this therapy reduces restenosis rates. DEBs are coated with the antiproliferative drug, paclitaxel, which is quickly released on balloon contact with the vessel wall. 2 The antiproliferative drug reduces neointimal hyperplasia, resulting in a lower incidence of restenosis and target lesion revascularization (TLR). 3 In randomized controlled trials comparing DEB therapy to PTA, the rate of TLR at 6 months was lower with DEB (4%–13%) as compared with uncoated balloons (22%–37%).3–7
Currently, solid data regarding predictors of the outcome and possible limitations of DEB therapy are lacking. Daily practice, nevertheless, seems to show that calcium may have a negative impact on DEB efficacy. Of the several prospective DEB studies, most excluded patients with severely calcified lesions.3–5 Few studies have examined the relationship between calcification and DEB efficacy. In the DEBELLUM trial, 6-month late lumen loss (LLL) increased with increasing severity of lesion calcification, suggesting calcium may either serve as a barrier to adequate drug uptake or result in subacute vessel recoil. 7 A recent DEB study reported higher LLL at 12 months with increasing severity of lesion calcification. 8
The purpose of this retrospective study was to analyze potential variables that may predict the outcome following DEB therapy in the SFA and popliteal artery.
Methods
Study Design
Peripheral artery disease (PAD) patients who received treatment with paclitaxel-coated DEBs at 3 different hospitals in Germany were included in this retrospective, core laboratory–adjudicated analysis. Patients with lesions up to 24 cm in length located in the SFA or popliteal artery treated with DEB were included in the study if they had evaluable angiograms obtained in 2 projections at baseline, postintervention, and at 6±1-month follow-up.
In the participating centers, endovascular therapy with DEBs was standardized independent of brand. Briefly, predilation was required in total occlusions only, inflation time of the drug-coated balloons was 1 to 2 minutes, and stents were implanted only if indicated after a second long dilation (5 minutes) with an uncoated balloon.
The objective of the study was to examine the association of patient baseline demographics, lesion characteristics, and procedure variables with LLL at 6 months postprocedure to determine predictors of outcome of DEB therapy. LLL was calculated as minimum lumen diameter (MLD) postintervention compared with MLD after 6 months and was evaluated by an independent angiography core laboratory (Core Lab Bad Krozingen, Bad Krozingen, Germany).
In addition, lesion calcification was assessed by the core laboratory based on 2 previously published grading systems (Table 1): the peripheral artery calcification scoring system (PACSS) 9 and an angiographic calcium grading system based on circumference (arc) and length of calcium. 10 In addition to circumference (unilateral or bilateral) and length (5-cm cut point), the PACSS added location of the calcium in the vessel wall (intimal, medial, or mixed). For this, the core laboratory determined calcification zones and patterns in angiographic frames displaying the target lesion without contrast to discriminate calcification patterns representing medial sclerotic from intimal type.
Calcification Scoring Systems.
May be on both sides of vessel but not at the same location.
Patient Sample
Ninety-one patients (mean age 72.0±8.62 years; 50 men) with Rutherford category 1 to 5 ischemia were included in this retrospective analysis. Patient characteristics are presented in Table 2 and lesion characteristics in Table 3. Nearly half (45.1%, 40/89) of the patients had a previous peripheral intervention. Lesions were located in the SFA (n=68) and popliteal artery (n=23); 41/91 (45.1%) were restenotic and a third were occlusions (33%, 30/91). More than a third of lesions (40%, 36/91) were graded as moderately severe or severely calcified using the angiographic calcium score. During the index procedure, 10 (11%) patients required stent implantation after DEB due to flow-limiting dissection or residual stenosis.
Demographics and Clinical Characteristics of 91 Study Patients. a
Continuous data are presented as the means ± standard deviation (range); categorical data are given as the counts/sample (percentage).
Vessel and Lesion Characteristics. a
Abbreviations: PACSS, peripheral arterial calcium scoring system; SFA, superficial femoral artery.
Continuous data are presented as the means ± standard deviation (range) or median and interquartile range (IQR); categorical data are given as the counts/sample (percentage).
Statistical Methods
Patient characteristics at baseline, postprocedure, and 6-month follow-up were summarized using counts and percentages for nominal data and median with interquartile range (IQR) or mean ± standard deviation as appropriate for continuous data.
The primary endpoint was LLL after 6 months. Subgroup analyses were performed to evaluate the association between calcification and LLL, which was first compared between groups with differing calcification severity, between groups of differing baseline risk factors (ie, smoking), and between groups of different baseline medical history (ie, history of cardiac disease). The influence of calcification on 2-year TLR was also explored in similar fashion. To account for nonnormally distributed data and small cell counts, these analyses were conducted using nonparametric tests (chi-square, Mann-Whitney U, or Kruskal-Wallis) or with nonparametric correlation analyses (Spearman). Additionally, multivariate regression analyses were performed for all variables attaining p<0.05 in the subgroup analysis. Even though the LLL data were not normally distributed, the data were not transformed because a normal distribution was attained (Kolmogorov-Smirnov test, p=0.056) after exclusion of an extreme outlier (deviation >3 box lengths).
The associations of risk factors, comorbid conditions, and lesion-specific findings (lesion length, prior stenosis, occlusion, length of the occlusion) with LLL after 6 months were also analyzed. Any influence of the DEB model, number of balloon catheters used, implantation of stents, or occurrence of dissections was sought.
Results
One angiogram at the 6-month follow-up was not readable by the core laboratory, so the LLL analysis was based on 90 patients. The median MLD was 0.95 mm (IQR 0.1, 1.71) at baseline, 3.4 mm (IQR 2.87, 3.96) postintervention, and 2.95 mm (IQR 2.21, 3.83) at 6-month follow-up. The median LLL at 6 months was 0.20 mm (IQR −0.50, 1.14). The rate of TLR was 20.9% (19/91) at 6 months.
Influence of Calcification
Because of low cell counts, calcification categories were collapsed for analysis. Table 4 shows stratified analysis of LLL at 6 months based on variable calcium lesion characteristics. Using the angiographic calcium score, LLL differed by calcification severity (p=0.042), with moderately severe/severe calcified lesions showing median LLL of 0.93 mm (IQR −0.24, 1.30), mild/moderate calcification showing median LLL of 0.17 mm (IQR −0.81, 1.32), and no calcification showing median LLL of 0.00 mm (IQR −0.56, 0.60). The difference in LLL was even more pronounced when comparing moderately severe/severe lesions vs none/mild/moderate lesions (p=0.014).
Comparison of Late Lumen Loss at 6 Months: Calcification Severity.
Abbreviation: PACSS, peripheral arterial calcium scoring system.
Data are presented as the median (interquartile range).
Kruskal-Wallis test.
Mann-Whitney U test.
The PACSS score showed tendencies but no significant differences between the groups. The length of calcification (p=0.258), as well as the location in the wall (intimal, medial, or mixed type), did not have a significant impact on the 6-month LLL (p=0.351). In contrast, there were strong tendencies toward a higher LLL in bilateral calcifications (0.72 mm, IQR −0.25, 1.17) compared to unilateral (0.17 mm, IQR −0.83, 1.39) or no calcification (0 mm, IQR −0.56, 0.60; p=0.072).
Within 2 years, 25 of 90 patients had a TLR. In general, the degree of calcium had an impact on the TLR in the same way. As an example, the 2-year TLR rate was 5/26 in noncalcified, 5/16 in unilaterally calcified, and 15/23 in bilaterally calcified arteries, respectively (p=0.030).
Influence of Patient and Lesion Characteristics
Significantly higher LLL values were observed in patients with type 2 diabetes (p=0.034), coronary artery disease (p=0.024), or prior intervention (p=0.013, Table 5). There was no correlation between 6-month LLL and the length of the lesion (r = −0.067), the degree of prior stenosis (r = −0.007), and/or the length of the occlusion (r = −0.009). Further analyses of the degree of calcification among the various concomitant diseases showed that the frequency distributions of the calcification were independent from those diseases.
Comparison of Late Lumen Loss at 6 Months: Baseline Demographics.
Data are presented as the median (interquartile range).
Kruskal-Wallis test.
Mann-Whitney U test.
Spearman correlation coefficient; no correlation to late lumen loss.
Influence of Procedure Variables
In 60.4% of cases, only 1 DEB was used, 2 balloon catheters were used in 34.1%, and few cases required 3 or 5 DEBs. Although there was numerical variation in LLL among the different models of DEBs, no statistical differences were observed in relation to the DEB model or number of the balloon catheters used. The degree of residual stenosis did not demonstrate statistically significant influence on LLL (r = −0.238; Figure 1).
Scatter diagram showing residual stenosis vs late lumen loss after 6 months (n=90, Spearman r = −0.238).
In 37.4% of the patients, predilation was performed prior to the DEB application. There was no impact of predilation on LLL. Provisional stent implantation had a significant association with LLL. In the 10 patients with stents, median LLL was 1.11 mm (IQR 0.13, 2.21), while in the patients without stents, the median LLL was 0.16 mm (IQR −0.52, 1.10; p=0.048). There were no correlations between calcification severity and stents used.
Dissections (type A-E) were reported in 61.5% (56/91) of all cases. There was no influence of dissection on LLL and no statistically significant difference between the dissection types and calcification severity. The dissection types stratified for risk factors and concomitant diseases also did not differ significantly.
Regression Analysis
The results for the unadjusted univariate analysis, with reference category “none to moderate” (angiographic calcium score version 2) or “mixed type” (PACCS version 3) or “none” (remaining predictors), are shown in Table 6. Significant factors from the univariate regression analysis where entered into the multivariate model. According to this model (Table 6), 2 predictors remained statistically significant. For patients with moderately severe or severe angiographic calcium score, the incidence of LLL at 6 months would be expected to be 1.8 times higher, adjusted for diabetes type 2, coronary artery disease, and previous intervention, than for patients with none, mild, or moderate angiographic calcium scores. In addition, patients with previous intervention would have a 1.8 times higher LLL at 6 months than patients without previous intervention.
Univariate and Multivariate Regression Analyses for Predictors of Late Lumen Loss After 6 Months.
Abbreviation: PACSS, peripheral arterial calcium scoring system.
Data are presented as the estimator, 95% confidence interval, and p value.
Discussion
With the wide adoption of DEB for treatment of femoropopliteal arterial disease, DEB efficacy in lesions with severe calcium burden remains questionable. This retrospective study used calcium scores to show that severe calcification is a predictor of inferior outcome of DEB therapy in femoropopliteal lesions based on LLL at 6 months. Selecting an appropriate calcification grading system and determining the relevant components of the score that correlated with the major procedure outcome (6-month LLL) was important in this analysis. This study demonstrated that assessment of baseline angiograms for calcification scoring is predictive of an increased LLL. A high calcification score was correlated with a less favorable LLL, specifically with circumferential calcification. Absence of calcium or unilateral presence was not associated with unfavorable outcome. Lesion length and calcium location did not correlate with the outcome and may not be considered as predictors of LLL.
Reasons for increased LLL in severely calcified lesions are 2-fold. First, biological efficacy of DEB treatment depends on adequate drug transfer and accumulation in the target artery. Calcium in large quantities, in particular if distributed circumferentially in the arterial wall, may be an impermeable barrier for the drug, thus diminishing efficacy of the therapy. Second, circumferential calcification in particular reduces vessel compliance and may result in subacute vessel recoil, triggering increased LLL. The calcification scoring system based on the angiogram assessment could be an important and easy-to-use tool for immediate optimization of the treatment strategy and device selection during the procedure. In cases of high-grade calcifications bilaterally on the vessel wall, the drug-coated balloons are more likely to fail, with reduced long-term patency.
Other imaging modalities may help with assessment of calcification but may not always be practical. As presented by Fanelli et al, 8 computed tomography angiography and intravascular ultrasound with virtual histology are very reliable tools for calcium identification in the peripheral vasculature. Combining these imaging tools could be a very sensitive approach for characterization of calcium in the lesion and play an important role in research. However, those additional diagnostic steps may not be feasible in clinical practice due to additional radiation exposure, time, and related cost. In addition, the study by Fanelli et al 8 had several limitations that do not apply to the approach suggested in this study. The assessment of LLL by Fanelli’s group was done using Doppler ultrasound, which may be difficult or sometimes impossible to perform in highly calcified lesions. Other limitations of the study 8 could be attributed to a limited number of patients in the highly calcified group (n=8), a single-center cohort, absence of a core laboratory, and no analyses of other variables besides calcium.
In the present study, evaluation of multiple variables demonstrated a correlation of LLL in patients with diabetes (type 2), coronary heart disease, or previous interventions of the target lesion, but further analyses showed that the degree of calcification was independent of these conditions. Renal insufficiency was not a predictor of the LLL, likely due to a small number of patients in this group. Interestingly, lesion length, degree of stenosis, and severity of dissection did not correlate with LLL at 6 months or with calcium distribution and severity and were not predictors of LLL. Thus, severity of lesion calcification remains a single independent predictor of LLL outcome after DEB treatment. Pretreatment of severely calcified lesions with atherectomy devices or plaque modification devices might be mandatory to maximize the therapeutic effect of DEB treatment.
Baseline lesion characteristics with impact on DEB therapy outcome can be divided in 2 groups: (1) cannot be changed prior to DEB (eg, lesion length, degree of occlusion, restenotic lesion, and comorbidities) and (2) modifiable prior to DEB (eg, severity of calcification). In our dataset, postprocedure severity of residual stenosis without calcification, even if high grade, did not play a significant role in DEB outcome. The treatment with DEBs cannot be compared with plain balloon angioplasty because the result of the intervention is not the final result with the DEB balloon. The Pacifier trial showing a negative LLL following In.Pact DEB therapy supports this. 6 This effect might differ from one DEB to another DEB and might be also dependent on the lesion characteristics. Further studies are mandatory in order to prove that no stent is needed in order to treat a residual stenosis if there is no flow limitation.
Limitations
This study was retrospective and included a relatively small number of patients. The large, ongoing, prospective, single arm, all-comers studies of DEB treatment might support our findings (eg, InPACT Global, Levant global). In addition, future prospective studies with specific focus on using DEB in patients with severe calcification may be of interest. Even though a dedicated analysis of the calcium scores was performed through our core laboratory, the description of the calcium location was difficult and might have been more precise if we would have also have used intravascular ultrasound. Nevertheless, the analysis was performed because we wanted to follow the PACSS and angiographic calcium scores.
Conclusion
Severe lesion calcification can be considered as a single independent variable and predictor of LLL at 6 months after treatment with DEB. Late lumen loss at 6 months is also associated with type 2 diabetes, coronary artery disease, and prior intervention. Calcification is modifiable prior to DEB, and partial calcium removal before DEB application may improve the outcome in patients with femoropopliteal disease.
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Gunnar Tepe consults for Abbott, Covidien, Medtronic, Medrad, and Cardiovascular Systems, Inc. Thomas Zeller serves on the advisory boards to Medtronic, W.L. Gore & Associates, and Covidien and has received consulting fees from C.R. Bard, Johnson & Johnson Cordis, Abbott Vascular, Boston Scientific, Cook, Straub Medical, and Biotronik.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Cardiovascular Systems, Inc, St Paul, MN, USA.