Clinical outcomes of directional atherectomy versus plain balloon angioplasty as vessel preparation prior to drug-coated balloon treatment for femoropopliteal occlusive disease

Abstract

This retrospective multicenter cohort study from the PROMISING registry compared directional atherectomy (DA) and plain old balloon angioplasty (POBA) as vessel preparation before drug-coated balloon (DCB) therapy for femoropopliteal occlusive disease. After propensity score matching, 147 DA-treated and 480 POBA-treated patients were analyzed with well-balanced baseline characteristics. DA markedly reduced the need for stent implantation (8.8% vs. 20.4%) and shortened total stent length, while allowing subsequent use of larger-diameter DCBs. Major peri-operative complications were rare in both groups, but minor complications were more frequent after DA. Postoperative ankle–brachial index and Rutherford classification improved significantly in both groups, without intergroup differences. During 24-month follow-up, Kaplan–Meier analysis showed no significant differences between DA and POBA in all-cause mortality, clinically driven target lesion revascularization, amputation-free survival, or major adverse limb events. Subgroup analyses revealed that patients with chronic total occlusion (CTO) derived greater benefit from DA, with a significant interaction (P for interaction = 0.002), whereas outcomes were consistent across other subgroups. Overall, DA before DCB therapy effectively reduces stent use and enhances vessel preparation without compromising long-term efficacy or safety, supporting its selective application in complex femoropopliteal lesions, particularly those with CTO.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-40423-8.

Introduction

Peripheral artery disease (PAD) now affects more than 230 million adults worldwide, a figure that has continued to rise in parallel with population ageing and the global spread of atherosclerotic risk factors such as diabetes, hypertension, and smoking.¹Femoropopliteal occlusive disease (FPOD) represents the most common localisation of symptomatic lower-extremity PAD and is increasingly treated by endovascular means.²The contemporary “leave nothing behind” paradigm, adequate vessel preparation, followed by the use of Drug-coated balloons (DCB), and if necessary, the use of stent implantation, has emerged as a routine strategy for the treatment of FPOD.³.

Directional atherectomy (DA) is a technique designed to remove atherosclerotic plaques. Theoretically, it can improve luminal gain and plaque modification, resulting in low rates of stent implantation, perforation, and dissection. Additionally, it potentially facilitates more homogeneous drug delivery to the vessel wall and enhances drug penetration.⁴DA combined with drug-coated balloon (DAART) holds promising clinical application prospects.

Nevertheless, the use of DA for vessel preparation remains a subject of considerable debate. Several studies have failed to demonstrate the superiority of DA over plain old balloon angioplasty (POBA) for vessel preparation.^5–8Additionally, researchers have highlighted the potential waste of medical resources associated with these costly devices in their meta-analyses.^6,9Concurrently, clinical evidence supporting DAART in real-world settings is also insufficient and necessitates further supplementation. Firstly, reports of real-world comparative outcomes with large sample sizes and long-term follow-up (exceeding 1 year) remain relatively scarce.⁷Secondly, there is a paucity of trials focusing on patient-oriented outcomes; instead, some studies are designed to focus on hemodynamic effects rather than overall patient benefit.⁸Thirdly, potential ethnic or regional variations in the use of DAART exist, with relatively few reports originating from the Chinese population.

We therefore conducted this multicentre, real-world study in Chinese patients with FPOD to comprehensively compare the clinical effectiveness and safety of DA versus POBA as vessel preparation before DCB angioplasty.

Methods.

Patients population

We retrospectively analyzed the cases from the PROMISING registry (A prospective, multicenter, real-world observational study evaluating the effectiveness and safety of DCB therapy for femoropopliteal lesions. NCT04826705). It includes consecutive cases of patients with FPOD treated with DCB angioplasty between 2021 and 2024. The registry aims to recruit 2500 patients, with an expected follow-up period of 4 years. Details of the PROMISING registry can be found at: https://clinicaltrials.gov/study/NCT04826705. Current study analyzed patients enrolled between January 2021 and December 2023.

The key inclusion criteria for the PROMISING Study encompassed Rutherford grades 2–5, with all patients undergoing treatment for FPOD using DCBs. Furthermore, the current analysis excluded all patients who underwent vessel preparation using specialty balloons. Detailed inclusion and exclusion criteria are provided in Table S1 of the Supplementary Appendix.

The protocol for the PROMISING registry was approved by the institutional review boards of the participating centers, all patients provided informed consent, and the study was approved by the local ethics committees. All methods were performed in accordance with the Declaration of Helsinki and relevant institutional and national guidelines and regulations.

Procedure and Follow-up procedure

All interventions were performed in a hybrid operating suite by experienced endovascular surgeons. Device selection, including vessel preparation technique, use of an embolic protection device (EPD), chosen DCB brand, and the need for stent implantation, was made at the operator’s discretion. The details of the device used in this study are provided in Supplementary Table 3.

In the POBA group, the stenosis was predilated by uncoated balloon catheters to avoid drug loss and then was dilated by DCB at least 180s (the same diameter as the target vessel, 10 mm longer than the lesion segment). If flow-limiting dissection or residual stenosis > 50% which is diagnosed by angiographic examination occurs, an uncoated balloon was used to dilate again (> 2 min). When the flow-limiting dissections or the recoil (residual stenosis is estimated to be > 50%) still exist, stents need to be implanted.

In the DA group, SilverHawk or TurboHawk plaque excision system (Medtronic, Minneapolis, MN) was used. Under the guidance of the guidewire, the target lesions were treated with the plaque excision system near-to-far, and the debulking speed and angle were controlled. Every time, according to the fixed direction, adjust the debulking angle 15°-30°, 4 to 6 times in total. For every 1–2 times of removal, the plaque debris in the collection tank should be cleaned after exiting the system so as to avoid falling off to the distal end and causing an embolism. After the removal of the plaques, the lesion site was predilated with an uncoated balloon and then was treated with DCB angioplasty. Similarly, an uncoated balloon was used for redilation of the flow-limiting dissections (> 2 min). Stents were implanted in cases of acute DCB failure, such as flow-limiting dissection or residual stenosis > 50%. The grading of the dissection patterns after balloon angioplasty was assessed based on the National Heart, Lung, and Blood Institute (NHLBI) classification.¹⁰.

After the procedure, 2 antiplatelet drugs (aspirin 100 mg/d, clopidogrel 75 mg/d) were recommended to be taken orally for 6 months, followed by a long-term oral administration of one antiplatelet drug.

After discharge, routine follow-up calls were made at 1, 3, 6, 12, 18 and 24 months.Systematic imaging-based assessment of vessel patency (e.g., duplex ultrasound or angiography) was not mandated by the study protocol and was therefore not routinely performed; however, imaging data were available and reviewed in patients who were rehospitalized or underwent further clinical evaluation, and were used to assist in endpoint adjudication when applicable. Baseline characteristics, procedural details, and follow-up outcomes were collected in the VASCBASE@ database (www.coachpvd.com) and anonymized for analysis.

Endpoint and definition

The primary effectiveness endpoints of this study were stent implantation rate, and maximum DCB dilation diameter. The primary safety endpoints included perioperative complications and major adverse limb-event (MALE)-free survival¹¹during the 2-year follow-up period. Decisions regarding stent implantation and the selection of maximum DCB diameter were made at the discretion of the treating operator.

Additional endpoints included the average stent length among stent recipients, postoperative ankle-brachial index (ABI), length of hospital stay, and a series of follow-up outcomes, including clinically-driven target lesion revascularization (CD-TLR) within 2 years, major amputation within 2 years, all-cause mortality within 2 years, and composite endpoints derived from these follow-up endpoints, specifically amputation-free survival (AFS).

CD-TLR was defined as any re-intervention within the target lesion(s) due to symptoms or ABI decrease of ≥ 20% or > 0.15 compared with post-index procedure baseline ABI.¹²Major amputation was defined as any procedure resulting in amputation at or above the ankle. Major adverse limb event (MALE) was a composite endpoint consisting of CD-TLR and major amputation.

Statistical analysis

All analyses were conducted with R 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria). Two-tailed P < 0.05 indicated statistical significance. Continuous variables are reported as mean ± standard deviation and categorical variables as counts and percentages. Baseline group differences were assessed with the Student’s t-test (or Welch’s t-test, as appropriate) for continuous variables and the χ² test or Fisher’s exact test for categorical variables.

Propensity-score matching

To control for confounding, a propensity score for receiving DA was generated for each patient using a multivariable logistic-regression model that included the 23 baseline covariates listed in Supplementary Table S3. Patients were matched 1:4 without replacement by nearest neighbour within a caliper width of 0.20 × the pooled standard deviation of the logit of the score. Covariate balance was evaluated with standardised mean differences (SMDs); an SMD < 0.10 denoted acceptable balance (Supplementary Figure S1).

Post-matching outcome comparisons

Difference tests were performed for continuous and categorical outcome variables using previously described methods. Time-to-event outcomes were described by Kaplan–Meier curves, compared with the log-rank test, and modelled with Cox proportional-hazards regression using pair-clustered robust standard errors.

Sensitivity analysis in the full cohort

In the entire analytic cohort, predictors of stent implantation were explored with multivariable logistic regression. Variables with P < 0.05 on univariable screening or deemed clinically relevant were entered into the model. Multicollinearity was checked with variance-inflation factors (VIF < 5 acceptable), and calibration with the Hosmer–Lemeshow test.

Prespecified subgroup analyses

The comparative performance of DA was further examined in four prespecified lesion subgroups: (i) chronic total occlusion vs. non-CTO; (ii) lesion length ≥ 10 cm vs. < 10 cm; (iii) moderate-to-severe vs. none/mild calcification; and (iv) involvement vs. non-involvement of popliteal segments P2/P3. Subgroup differences were evaluated with an interaction term; P for interaction < 0.05 indicated significant effect modification.

The present study adhered to the STROBE reporting guidelines, and the corresponding checklist is available in the Supplementary Table S4.

Results

Baseline characteristics of the study population

Among 2,367 screened patients in the PROMISING registry, 1,222 DCB-treated patients entered the analytic cohort (POBA = 1,070; DA = 152). By the reverse Kaplan–Meier method, the cohort’s median potential follow-up was 24.0 months (IQR 12.7–27.9), corresponding to 723 days (IQR 386–849). The enrollment and follow-up of patients are shown in Fig. 1.

Fig. 1.

Study flowchart. Flow diagram showing patient inclusion and exclusion from the PROMISING registry. Among 2,367 screened cases, 1,919 underwent endovascular therapy with drug-coated balloons (DCBs). After excluding cases with acute limb ischemia and incomplete lesion DCB coverage, 1,222 patients were included in the present analysis. Propensity-score matching (PSM) generated 147 patients treated with directional atherectomy (DA) and 480 patients treated with plain old balloon angioplasty (POBA). Abbreviations: DCB, drug-coated balloon; DA, directional atherectomy; POBA, plain old balloon angioplasty; PSM, propensity-score matching.

The baseline characteristics of the two groups before and after matching are shown in Table 1. Before matching (n = 1,222), DA patients were younger, had slightly higher BMI and better pre-operative VascuQoL, and more often had hyperlipidemia, coronary artery disease, and diabetes; they also presented with higher pre-op ABI and less impaired runoff. Lesion features showed fewer popliteal P2/P3 involvements and less in-stent restenosis in DA; arterial calcification distributions differed, while CTO prevalence and lesion length were comparable. (All P values as shown in Table 2) Procedurally, distal embolic-protection umbrellas were used in 99/152 DA cases (65.1%).

Table 1.

Baseline demographic and clinical characteristics before and after propensity score matching.

Variable	Unmatched cohort(n = 1222)			Matched cohort (n = 627)
	POBA (n = 1070)	DA (n = 152)	P	POBA (n = 480)	DA (n = 147)	P
Age, year	71.7 ± 9.5	68.9 ± 8.1	< 0.001	70.0 ± 8.9	68.9 ± 8.1	0.193
Male	777 (72.6)	110 (72.4)	0.949	358 (74.6)	106 (72.1)	0.550
Body-mass index, kg / m^2	23.2 ± 3.1	23.9 ± 3.9	0.011	23.6 ± 3.1	23.9 ± 3.9	0.349
Pre-operative VascuQoL score	2.6 ± 0.9	2.9 ± 1.0	< 0.001	2.8 ± 1.0	2.9 ± 0.9	0.866
Hypertension	786 (73.5)	114 (75.0)	0.686	370 (77.1)	109 (74.2)	0.464
Hyperlipidemia	224 (20.9)	56 (36.8)	< 0.001	130 (27.1)	51 (34.7)	0.075
Smoking History	384 (35.9)	50 (32.9)	0.471	191 (39.8)	47 (32.0)	0.087
Coronary Artery Disease	207 (19.4)	44 (29.0)	0.006	112 (23.3)	41 (27.9)	0.260
Atrial Fibrillation	39 (3.6)	3 (2.0)	0.290	10 (2.1)	2 (1.4)	0.829
Diabetes mellitus	624 (58.3)	105 (69.1)	0.011	318 (66.3)	101 (68.7)	0.580
Renal Dysfunction	107 (10.0)	15 (9.9)	0.960	57 (11.9)	14 (9.5)	0.431
Cerebral infarction	160 (15.0)	36 (23.7)	0.006	94 (19.6)	33 (22.5)	0.449
COPD	14 (1.3)	0 (0.0)	0.312	7 (1.5)	0 (0.0)	0.306
Previous Peripheral Revascularization	456 (42.6)	42 (27.6)	< 0.001	146 (30.4)	40 (27.2)	0.457
Pre-op ankle–brachial index	0.4 ± 0.2	0.5 ± 0.2	0.006	0.5 ± 0.3	0.5 ± 0.2	0.499

Values are presented as mean ± standard deviation (SD) or number (percentage). P values were calculated using Student’s t-test or χ² test as appropriate. P < 0.05 was considered statistically significant. Abbreviations: DA = directional atherectomy; POBA = plain old balloon angioplasty; COPD = chronic obstructive pulmonary disease.

Table 2.

Lesion and angiographic characteristics before and after propensity score matching.

Variable	Unmatched cohort(n = 1222)			Matched cohort (n = 627)
	POBA (n = 1070)	DA (n = 152)	P	POBA (n = 480)	DA (n = 147)	P
Rutherford Classification			0.094			0.997
0	1 (0.1)	0 (0.0)
1	17 (1.6)	1 (0.7)		5 (1.0)	1 (0.7)
2	169 (15.8)	25 (16.5)		78 (16.3)	25 (17.0)
3	436 (40.8)	80 (52.6)		238 (49.6)	75 (51.0)
4	111 (10.4)	11 (7.2)		42 (8.8)	11 (7.5)
5	309 (28.9)	31 (20.4)		103 (21.5)	31 (21.1)
6	27 (2.5)	4 (2.6)		14 (2.9)	4 (2.7)
Pedal-Plantar Loop Score			0.114			0.949
0	180 (16.8)	24 (15.8)		76 (15.8)	23 (15.7)
1	279 (26.1)	51 (33.6)		145 (30.2)	49 (33.3)
2	332 (31.0)	48 (31.6)		154 (32.1)	46 (31.3)
2.5	180 (16.8)	23 (15.1)		86 (17.9)	23 (15.7)
3	99 (9.3)	6 (4.0)		19 (4.0)	6 (4.1)
Run off totlScore	8.2 ± 5.1	6.7 ± 3.9	< 0.001	7.1 ± 4.6	6.8 ± 3.9	0.461
leasion length, cm	18.6 ± 10.8	18.7 ± 10.9	0.930	17.8 ± 10.8	19.0 ± 10.9	0.217
P2/P3 popliteal artery involvement	363 (33.9)	38 (25.0)	0.028	132 (27.5)	37 (25.2)	0.578
Chronic Total Occlusion	575 (53.7)	79 (52.0)	0.683	244 (50.8)	79 (53.7)	0.537
In-Stent Restenosis	191 (17.9)	9 (5.9)	< 0.001	36 (7.5)	9 (6.1)	0.571
Arterial Calcification			< 0.001			0.944
No	253 (23.6)	13 (8.6)		44 (9.2)	13 (8.8)
Mild	380 (35.5)	49 (32.2)		172 (35.8)	49 (33.3)
Moderate	284 (26.5)	61 (40.1)		179 (37.3)	58 (39.5)
Severe	153 (14.3)	29 (19.1)		85 (17.7)	27 (18.4)

Values are presented as mean ± standard deviation (SD) or number (percentage). P values were calculated using Student’s t-test or χ² test as appropriate. P < 0.05 was considered statistically significant. Abbreviations: DA = directional atherectomy; POBA = plain old balloon angioplasty; COPD = chronic obstructive pulmonary disease.

Propensity-score matching yielded 627 patients (POBA = 480; DA = 147). After matching, no statistically significant differences remained across demographics, comorbidities, hemodynamics, or lesion characteristics. (all P > 0.05). The two matched groups were highly comparable, as shown in Figure S1 and Figure S2.

Comparison of surgical outcomes between propensity-matched groups

Procedural success was high and comparable between groups (99.3% vs. 99.4%). DA was associated with significantly lower rates of stent implantation (8.8% vs. 20.4%, P = 0.001) and shorter stent length (7.6 ± 3.6 cm vs. 14.3 ± 8.0 cm, P < 0.001), while achieving a slightly larger DCB dilation diameter (5.3 ± 0.6 mm vs. 5.0 ± 0.6 mm, P < 0.001). Operative time and hospital stay were similar, but treatment costs were higher in the DA group (P < 0.001).(Table 3).

Table 3.

Comparison of surgical and in-hospital outcomes between the propensity-matched DA and POBA groups.

Variables	POBA (n = 480)	DA (n = 147)	P value
Stent implantation	98 (20.4)	13 (8.8)	0.001
Stent length, cm	14.3 ± 8.0	7.6 ± 3.6	< 0.001
Maximum DCB dilation diameter, mm	5.0 ± 0.6	5.3 ± 0.6	< 0.001
Operative time, minutes	104.6 ± 45.3	107.4 ± 47.6	0.505
Procedural success	475 (99.4)	145 (99.3)	1.000
Minor peri-operative complications	4 (0.8)	6 (4.1)	0.017
Major peri-operative complications	2 (0.4)	0 (0.0)	1.000
Endovascular therapy cost, USD	2005.8 ± 1569.7	2593.8 ± 1001.3	< 0.001
Total in-hospital cost, USD	2259.9 ± 1064.6	2913.1 ± 1042.4	< 0.001
Length of hospital stay, days	8.4 ± 7.6	7.9 ± 6.2	0.511

Legend: Comparison of intraoperative parameters, procedural outcomes, and hospitalization data between matched DA and POBA groups. DA resulted in a significantly lower rate and length of stent implantation, higher endovascular and total hospital costs, and comparable procedural success and complication rates. Values are presented as mean ± standard deviation (SD) or number (percentage). P values were obtained by Student’s t-test or Fisher’s exact test as appropriate. P < 0.05 was considered statistically significant. Abbreviations: DA = directional atherectomy; POBA = plain old balloon angioplasty; DCB = drug-coated balloon; USD = U.S. dollars.

Both groups showed marked improvements in ABI and Rutherford classification after intervention, without significant intergroup differences.(Fig. 2). Kaplan-Meier analysis up to two years revealed no differences between DA and POBA regarding all-cause death, CD-TLR, major amputation, or composite endpoints (all log-rank P > 0.05).(Fig. 3).

Fig. 2.

Comparison of ankle–brachial index and Rutherford classification before and after the procedure. (A) Changes in ankle–brachial index (ABI) before and after the intervention in the DA and POBA groups. Both groups showed significant improvement postoperatively (P < 0.001), with no between-group difference. (B) Distribution of Rutherford classification grades before and after treatment in each group, showing marked symptomatic improvement after endovascular therapy. Abbreviations: ABI, ankle–brachial index; DA, directional atherectomy; POBA, plain old balloon angioplasty.

Fig. 3.

Kaplan–Meier curves for survival analysis at each endpoint. Kaplan–Meier estimates comparing DA and POBA groups for (A) all-cause death-free survival, (B) freedom from clinically driven target lesion revascularization (CD-TLR), (C) freedom from major amputation, (D) amputation-free survival, and (E) freedom from major adverse events (MAE). No significant differences were observed between groups for any endpoint (all P > 0.05 by log-rank test). Abbreviations: DA, directional atherectomy; POBA, plain old balloon angioplasty; CD-TLR, clinically driven target lesion revascularization; MAE, major adverse events.

Minor peri-operative complications were more frequently observed in the DA group (4.1% vs. 0.8%, P = 0.017). In the DA group, minor complications included two intra-procedural distal embolization without clinical manifestations, an early thrombosis relieved by surgical or endovascular intervention without progression to severe ischemia, two puncture site complications, and a contrast-induced nephropathy with full recovery. In the POBA group, minor complications comprised early thrombosis relieved by conservative treatment, two cases of asymptomatic distal embolization, and one case of profunda femoris artery embolization. Major complications were rare in both groups. Two patients in the POBA group experienced major adverse events (one peri-operative cardiovascular/cerebrovascular event and one case of no-reflow after DCB dilation leading to failed tibial recanalization), whereas no major events occurred in the DA group. In an exploratory analysis restricted to the DA group, the incidence of distal embolization was 0% (0/96) in procedures performed with embolic protection devices (EPDs) and 3.9% (2/51) in those without EPD use (P = 0.119).

Analysis of risk factors associated with stent implantation

The univariate and multivariate logistic regression results for stent implantation are presented in Table S4. Multivariate analysis identified smoking history (OR 1.55, 95% CI 1.14–2.10, P = 0.005), advanced TASC classification (B–D vs. A: OR 2.29–5.37, all P < 0.05), chronic total occlusion (OR 1.64, 95% CI 1.14–2.37, P = 0.008), and moderate-to-severe arterial calcification (OR 1.74–1.95, P < 0.05) as independent predictors of stent implantation. Diabetes mellitus was inversely associated with stenting (OR 0.66, 95% CI 0.49–0.89, P = 0.006). Importantly, the use of DA remained strongly protective against stent implantation (OR 0.33, 95% CI 0.18–0.60, P < 0.001).

Subgroup analysis and interaction test

Comparisons between DA and POBA across subgroups with different lesion characteristics are illustrated in Fig. 4. Subgroup analyses showed that the benefit of DA in reducing stent implantation was significant in CTO lesions (OR 0.17, 95% CI 0.07–0.42, P < 0.001), but not in non-CTO lesions (OR 1.24, 95% CI 0.50–3.07, P = 0.643), with a significant interaction (P for interaction = 0.002). In contrast, the protective effect of DA was generally consistent across subgroups defined by P2/P3 involvement, calcification severity, and lesion length, without evidence of interaction (all P for interaction > 0.05).

Fig. 4.

Subgroup analysis of stent implantation across predefined lesion characteristics. Forest plot showing odds ratios (ORs) and 95% confidence intervals (CIs) for the association between treatment modality (DA vs. POBA) and stent implantation across prespecified subgroups. The benefit of DA in reducing stent implantation was significant in patients with chronic total occlusion (CTO) lesions (interaction P = 0.002), but not across other lesion characteristics. Abbreviations: DA, directional atherectomy; POBA, plain old balloon angioplasty; CTO, chronic total occlusion; OR, odds ratio; CI, confidence interval.

Discussion

The present study provides real-world evidence from a large Chinese cohort comparing DA versus POBA as vessel preparation before DCB therapy in FPOD. The main finding was that DA significantly reduced both the rate and the total length of stent implantation compared with POBA (8.8% vs. 20.4%; P = 0.001), facilitated the subsequent application of larger diameter DCB (5.3 mm vs. 5.0 mm; P < 0.001), and achieved comparable procedural success and mid-term patient-oriented outcomes, including CD-TLR, major amputation, and all-cause mortality. At a median follow-up of two years, Kaplan–Meier survival analyses demonstrated no significant differences between the two groups in freedom from CD-TLR, major amputation, or death, indicating that both strategies provided satisfactory clinical durability in contemporary endovascular practice.

Dissections after conventional balloon angioplasty occur in more than 50% of femoropopliteal interventions and are strongly associated with increased risk of restenosis and target lesion revascularization.¹³Flow-limiting dissections (type C–E) usually need stent implantation, which may subsequently lead to restenosis or even stent fracture.¹⁴DA improves luminal gain by directly removing atherosclerotic plaque, facilitates subsequent balloon dilation, reduces elastic recoil, enhances vessel compliance, and minimizes the occurrence of flow-limiting dissections, thereby lowering the need for stent implantation.

Our findings are consistent with previous evidence supporting the role of DA in vessel preparation. The DEFINITIVE AR trial demonstrated that DA followed by DCB significantly reduced the incidence of flow-limiting dissections (2% vs. 19%, P = 0.01) and the need for adjunctive therapy (5% vs. 37%, P < 0.001) compared with DCB alone.⁷Similarly, a Chinese randomized study reported fewer flow-limiting dissections with DA versus POBA (4% vs. 25%, P = 0.006).¹⁵ In complex lesions, the VIVA REALITY registry confirmed that even in patients with long and heavily calcified lesions (mean length 18 cm; 39% chronic total occlusion [CTO]), DA achieved a low stent implantation rate (8.8%), closely aligned with the 8.8% observed in our DA group.¹⁶Multivariable regression analysis in our study further identified DA as an independent protective factor against stent implantation (adjusted OR 0.33, 95% CI 0.18–0.60, P < 0.001) after accounting for lesion complexity (TASC classification, CTO, calcification, and lesion length).

The assessment of procedural success after vessel preparation remains suboptimal. In current practice, post-preparation evaluation primarily focuses on angiographically apparent dissection and elastic recoil, which also constitute the main indications for stent implantation. However, standardized and objective criteria for assessing these features are still lacking. Dissection severity is commonly inferred from angiographic flow characteristics and residual stenosis, both of which are subject to substantial limitations. Flow assessment is inherently subjective and may be influenced by inflow and outflow conditions, distal vascular bed resistance, and cardiac output, thereby introducing uncertainty in determining whether a dissection is truly flow-limiting. Moreover, as angiography provides two-dimensional projections, the apparent degree of residual stenosis caused by dissection may vary considerably across different viewing angles, particularly in the presence of spiral dissections.¹⁷.

Another key component of vessel preparation assessment is elastic recoil, the delayed and non-immediate nature of which further complicates accurate intra-procedural evaluation. Taken together, these limitations highlight an unmet need for more objective and reproducible assessment frameworks to characterize vessel response after preparation. The development and validation of standardized evaluation tools for post-vessel preparation outcomes warrant further investigation.^17,18.

In terms of safety, DA was associated with a higher rate of minor perioperative complications (4.1% vs. 0.8%, P = 0.017), predominantly puncture-site hematoma or distal embolization. This may be related to two factors: (1) incomplete use of distal embolic protection devices (EPDs) in real-world practice, only 65.1% of DA cases in our study employed EPDs and (2) the requirement for larger sheath sizes during atherectomy, which can increase access-site complication risk. Notably, EPD utilization during DA was determined at the operator’s discretion, primarily based on intraoperative risk assessment, including distal runoff status and perceived embolic risk. In our study, an exploratory analysis within the DA cohort demonstrated a numerically lower incidence of distal embolization in procedures performed with EPDs compared with those without EPD use, although the difference did not reach statistical significance. Together with prior reports identifying DA as a risk factor for distal embolization, these findings underscore the potential importance of more consistent EPD utilization in optimizing procedural safety during atherectomy…¹⁹.

To further contextualize our findings, we performed an indirect comparison between the POBA + DCB control group in the present study and outcomes reported in large, previously published real-world datasets. Notably, patients in our POBA cohort appeared to have more complex baseline characteristics, with an older mean age (71.7 ± 9.5 years vs. 68.6 ± 10.1 years) and substantially longer lesions (18.6 ± 10.8 cm vs. 12.1 ± 9.5 cm) compared with those reported in the IN.PACT Global registry. Despite this higher lesion complexity, key clinical outcomes in our control group were broadly comparable to those observed in IN.PACT Global. Specifically, 1-year freedom from clinically driven target lesion revascularization (92.3% vs. 92.6%), freedom from major amputation (99.1% vs. 99.8%), and provisional stent implantation rates (20.9% vs. 21.2%) were of similar magnitude.²⁰These findings suggest that the performance of the POBA + DCB control arm in the present study is consistent with established real-world evidence. Taken together, this indirect comparison supports the external validity of our results and reinforces the interpretation that the observed reduction in stent implantation associated with directional atherectomy reflects a procedural benefit rather than an unexpectedly poor performance of the control group.

Subgroup analyses revealed a significant interaction between lesion morphology and the effect of DA on stent reduction. The protective effect of DA was evident in CTO lesions (OR 0.17, 95% CI 0.07–0.42, P < 0.001) but not in non-CTO lesions (OR 1.24, P = 0.643; P for interaction = 0.002), suggesting that DA may be particularly beneficial for plaque modification in occlusive or high-burden lesions. In contrast, the benefit remained stable across subgroups stratified by popliteal artery involvement (P2/P3) and calcification severity, supporting its potential utility even in “no-stenting zones.” These findings have practical implications: selective DA use tailored to lesion complexity may optimize cost-effectiveness while minimizing unnecessary device use.

This study has several limitations. First, standardized protocols are lacking for precise post-procedural assessment following vessel preparation. Evaluation of vessel preparation efficacy remains contentious due to substantial interobserver variability in interpreting dissections via intraoperative angiography. Quantitative measurement of luminal gain requires intravascular ultrasound (IVUS) or other imaging modalities to ensure uniform assessment—a methodology not yet routinely implemented in Chinese clinical practice. Consequently, stent implantation decisions and the maximum diameter of DCBs were utilized as surrogate markers reflecting operators’ judgment of post-preparation luminal adequacy.

Second, the study did not include a systematic imaging-based follow-up for vessel patency assessment. Objective evaluations such as duplex ultrasound or repeat angiography were not routinely performed during follow-up, and patency data were therefore not consistently available for analysis. Unlike prior studies primarily focused on angiographic or duplex-derived patency endpoints, our study design emphasized patient-oriented clinical outcomes, including CD-TLR, major amputation, and all-cause death. Although this approach may have overlooked potential differences in anatomical patency between treatment strategies, it reflects real-world follow-up practice and provides a broader, clinically meaningful assessment of therapeutic effectiveness. This limitation should be considered when interpreting the mechanistic implications of plaque-modification strategies.

Third, although the data originated from a multicenter, real-world registry with comprehensive covariate adjustment and propensity score matching, residual confounding cannot be completely ruled out due to the retrospective nature of the analysis. Moreover, the study population consisted exclusively of Chinese patients. The extension and application of the conclusions of this study require further research in a wider range of regions. This report presents the results of a two-year follow-up, and a longer follow-up is currently underway.

In addition, although directional atherectomy represents one of the most widely studied plaque-modification strategies for vessel preparation prior to DCB therapy, it is important to recognize that other lesion preparation techniques, including rotational and laser atherectomy, have also been explored in this setting. Previous small retrospective studies and registry analyses have suggested that such plaque-modification approaches may improve technical success and reduce flow-limiting dissections, particularly in complex and heavily calcified femoropopliteal lesions.⁴.

However, the current evidence supporting lesion preparation strategies before DCB therapy in peripheral artery disease remains limited and largely derived from single-center retrospective cohorts or non-randomized registry studies, rather than adequately powered randomized controlled trials.²¹Therefore, the potential benefits of plaque modification should be interpreted cautiously, and our findings should be viewed as hypothesis-generating rather than definitive. Further prospective, randomized studies are warranted to clarify the optimal lesion preparation strategy and to determine whether the observed procedural advantages translate into durable clinical benefit. Furthermore, although higher upfront procedural costs were observed with directional atherectomy, reduced bailout stenting may offer potential long-term economic advantages by simplifying future reinterventions; however, this hypothesis warrants further evaluation in dedicated cost-effectiveness studies.

In conclusion, this multicenter real-world analysis demonstrated that directional atherectomy before DCB therapy significantly reduced the need for stent implantation and total stent length without compromising mid-term clinical outcomes. The benefit was particularly pronounced in complex CTO lesions, supporting a selective, lesion-specific use of DA for vessel preparation in FPOD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Study design: Zibo Feng, Meng YeData collection: Chunshui He, Xin Fang, Ziheng Wu, Qiang Li, Weihao Shi, Zhenyu Shi, Lianrui Guo, Meng YeData analysis: Wei Zhang, Xinyuan WangWriting: Meng Ye, Xinyuan Wang.

Funding

Intramural Research Program of Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology (NO. 2023LYYYKYPT0001). Hubei Provincial Health Science and Technology Project (NO. WJ2025M020).

Data availability

The datasets generated and/or analyzed during the current study are derived from an ongoing institutional registry (www.coachpvd.com) and are not publicly available due to confidentiality agreements. Data access or login credentials to the study platform can be provided by the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical statement

The protocol for the PROMISING registry was approved by the institutional review boards of the participating centers, all patients provided informed consent, and the study was approved by the local ethics committees.