Drug-coated balloon alone versus atherectomy plus drug-coated balloon for treatment of femoropopliteal atherosclerotic disease: a meta-analysis

Abstract

Background

We performed a systematic review and meta-analysis to compare the safety and efficacy of drug-coated balloon (DCB) with atherectomy (AT) plus DCB alone for the treatment of femoropopliteal artery disease (FAD).

Methods

We searched PubMed, the Cochrane Library, and EMBASE up to 31 December 2024. Observational studies and randomized controlled trials comparing DCB alone versus AT combined with DCB in the treatment of FAD were included. Outcomes were categorized as procedural outcomes, including bailout stenting, flow-limiting dissection, distal embolization, perforation, and hematoma, and clinical outcomes, including primary patency at 12-month, freedom from target lesion revascularization (fTLR) at 12-month, leg amputation, and all-cause mortality. Two reviewers independently assessed study eligibility, extracted data, and evaluated the risk of bias.

Results

Eleven studies were included: 3 randomized controlled trials and 8 observational cohort studies, comprising 1296 participants. Compared to the AT + DCB group, when treating FAD, the DCB-alone group showed a higher incidence of bailout stenting (OR = 2.93, 95% CI [1.95, 4.39]; heterogeneity: I² = 7%) and flow-limiting dissection (OR = 4.46, 95% CI [2.53, 7.86]; heterogeneity: I² = 0%), but a lower risk of distal embolism (OR = 0.37, 95% CI [0.16, 0.84]; heterogeneity: I² = 0%). For other procedural outcomes, no significant differences were observed in perforation (OR = 0.66, 95% CI [0.28, 1.58]; heterogeneity: I² = 0%) or hematoma (OR = 0.83, 95% CI [0.38, 1.81]; heterogeneity: I² = 0%). For clinical outcomes, there were no significant differences between the two groups in fTLR at 12-month (OR = 0.81, 95% CI [0.57, 1.16]; heterogeneity: I² = 24%), primary patency at 12-month (OR = 0.92, 95% CI [0.68, 1.24]; heterogeneity: I² = 0%), all-cause mortality (OR = 1.15, 95% CI [0.74, 1.78]; heterogeneity: I² = 31%), or amputation (OR = 1.79, 95% CI [0.90, 3.56]; heterogeneity: I² = 0%).

Conclusion

Our meta-analysis demonstrated that AT + DCB did not show significant clinical superiority over DCB alone in terms of 12-month primary patency or fTLR. This combined approach did, however, significantly reduce the occurrence of flow-limiting dissections and decrease the need for bailout stenting. This highlights its potential as a strategy for achieving “graft-free residual” treatment in FAD. If further technical refinements can reduce the higher risk of distal embolization seen with the addition of atherectomy, longer term clinical benefits may be realized by reducing the risk of ISR. Importantly, these findings emphasize the necessity for large-scale, multicenter randomized controlled trials with extended follow-up periods to confirm the clinical efficacy and safety of this combined therapeutic approach.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13019-026-04481-3.

Background

Femoropopliteal atherosclerotic disease (FAD) constitutes a substantial clinical burden in aging populations, frequently leading to high morbidity of arterial stenosis and occlusion [1, 2]. While endovascular advancements—ranging from percutaneous transluminal angioplasty to stent implantation—have improved outcomes, challenges such as in-stent restenosis (ISR) and limited long-term patency remain unresolved. In this landscape, Drug-coated balloon (DCB) has emerged as a compelling “leave nothing behind” strategy. By delivering antiproliferative agents (e.g., paclitaxel or sirolimus) directly to the vessel wall, DCB have demonstrated superior primary patency and freedom from target-lesion revascularization (fTLR) compared to standard balloon angioplasty [3, 4].

However, the efficacy of DCBs can be severely attenuated in complex lesions characterized by heavy calcification or significant plaque burden, which physically impede drug absorption and vascular remodeling [5, 6]. To overcome these barriers, atherectomy (AT) has been proposed as a synergistic adjunctive therapy. By debulking plaque and mitigating elastic recoil, AT aims to optimize vessel compliance and facilitate deeper drug penetration. Some studies’ results demonstrate that combining AT with DCB achieved higher primary patency rates and symptom improvement compared to DCB alone in FAD [7–10]. Nevertheless, the benefits of combined therapy remain controversial. A retrospective analysis of common femoral artery lesions showed preparation of the artery with directional AT was not associated with statistically significantly higher primary patency or fTLR compared to DCB angioplasty alone at 12-month [11]. Additionally, concerns persist regarding the lack of standardized protocols for lesion preparation.

Given this controversy and the lack of standardized lesion preparation protocols, this meta-analysis aims to systematically synthesize existing evidence regarding the safety and efficacy of DCB alone versus DCB combined with AT in FAD treatment. By scrutinizing primary patency, fTLR, and other outcomes, this study seeks to clarify the role of hybrid therapies in optimizing vascular outcomes and advancing the paradigm of “minimal metal implantation” in peripheral interventions.

Methods

Study design

The protocol for this systematic review was prospectively registered with the International Prospective Register of Systematic Reviews (PROSPERO, ID: CRD42025649950). To ensure methodological rigor, the search strategy, study selection processes, and statistical analyses were executed in strict adherence to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [12]. The overarching objective was to evaluate whether adjunctive AT prior to DCB yields superior long-term clinical outcomes—specifically regarding patency, fTLR, and the mitigation of bail-out stenting—compared to DCB monotherapy in patients with FAD, without compromising procedural safety. Study eligibility was defined using the P.I.C.O. framework: Patients—symptomatic individuals with de novo FAD; Intervention—DCB angioplasty combined with AT; Comparison—DCB angioplasty alone; and Outcomes—bail-out stenting, 12-month primary patency, 12-month fTLR, distal embolization, perforation, hematoma, flow-limiting dissection, amputation, and all-cause mortality [13]. It should be noted that the AT + DCB group pooled four mechanistically distinct atherectomy modalities—directional, rotational, orbital, and laser atherectomy, as well as a mixed AT approach in some studies. These devices differ in mechanism of action, safety profile, operator learning curve, and optimal lesion targets.

Search strategy

For this study, a comprehensive search was performed across three databases: PubMed, the Cochrane Library, and EMBASE. The literature search strategy incorporated the following keywords: “atherectomy,” “directional atherectomy,” “orbital atherectomy,” “rotational atherectomy,” “laser atherectomy,” “debulking devices,” “peripheral artery disease,” “femoral,” “popliteal,” “femoropopliteal,” “drug-coated balloon,” “drug-eluting balloon,” and “paclitaxel-coated balloon.” The search encompassed studies published up to 31 December 2024. Additionally, eligible studies were identified by reviewing the references of included articles and relevant meta-analyses addressing similar topics. Additional limits included English language and full-text availability. The full search strategy is detailed in Supplementary Appendix.

Inclusion and exclusion criteria

Studies were deemed eligible for inclusion if they satisfied the following conditions: [1] A comparative design evaluating atherectomy (rotational, directional, or orbital) combined with DCB versus DCB monotherapy for FAD management; [2] Enrollment of symptomatic patients (presenting with claudication or rest pain) with femoropopliteal lesions measuring ≥ 5 cm in length; and [3] Provision of longitudinal follow-up data for a minimum of 12-month.

We excluded studies based on: [1] Non-original research or summary publications, including case reports, conference abstracts, literature reviews, and studies lacking extractable outcome data; [2] Significant data sparsity (defined as < 25% of predefined variables being retrievable); and [3] Redundant publications involving overlapping patient cohorts. In instances of serial publications from the same dataset, priority was given to the most recent report, unless the outcomes analyzed were mutually exclusive.

Literature screening and data extraction

Following the removal of duplicate records, titles and abstracts were screened for relevance by an independent reviewer. Full-text manuscripts satisfying the inclusion criteria were subsequently retrieved for detailed assessment. To ensure rigor, any discrepancies regarding study eligibility were adjudicated by a third investigator (Fig. 1).

Fig. 1.

Flow diagram adhering to the Preferred Reporting Items for Systematic Review and Meta-Analysis guidelines to illustrate the search and selection process during the initial stages of our review

Data were systematically extracted using a standardized data collection form. This process was executed by the primary author and cross-validated by co-authors to ensure accuracy. The captured dataset included bibliographic details (first author, publication year), study design, cohort size, baseline patient characteristics, procedural interventions, length of follow-up, and specific clinical outcome measures.

Risk of bias assessment

The methodological integrity of included Randomized Controlled Trials (RCTs) was appraised using the Cochrane Risk of Bias tool [14]. This assessment scrutinized key dimensions including randomization processes, deviations from intended interventions, missing outcome data, and selective reporting. Conversely, for non-randomized observational studies, quality was gauged using the Newcastle-Ottawa Scale (NOS) [15]. This scale specifically evaluates the selection of study groups, the comparability of cohorts, and the ascertainment of outcomes.

Data synthesis and statistical analyzes

Both qualitative and quantitative syntheses were performed. Meta-analysis was considered when data from two or more studies could be pooled. Clinical homogeneity among the selected studies was evaluated by comparing population characteristics, intervention groups, control groups, and outcomes. Statistical heterogeneity was assessed using the I² statistic. For low heterogeneity (I² < 50%), the Mantel-Haenszel model was applied for meta-analysis, while the Random-Effects model was used for moderate or high heterogeneity (I² ≥ 50%).

Continuous variables were compared between groups using the t-test, and categorical variables were analyzed using the χ²-test. The random-effects model was employed to evaluate the results. Statistical significance was determined by assessing the overlap of 95% confidence intervals (CI) between groups. Data management and analysis were conducted using Review Manager software (version 5.2; the Cochrane Collaboration) and STATA software (version 12.0; STATA Corporation, College Station).

Result

Literature screening

A total of 562 references published within the predefined time span of the analysis were searched. After a full-text review of 29 articles, 11 studies comprising 1296 participants were chosen for review based on the full text (Table 1). 3 RCTs [8, 16, 17] and 8 cohort trials [7, 9–11, 18–21] comparison of different AT + DCB vs. DCB alone in the treatment of FAD were identified: 6 on directional atherectomy (n = 748) [7, 8, 11, 16, 18, 19], 5 on rotational atherectomy (n = 717) [10, 17–20], 2 on orbital atherectomy (n = 155) [9, 21], and 1 on laser atherectomy ( n = 119) [18] prior to application of DCB. Regarding lesion type, most included studies primarily enrolled de novo femoropopliteal lesions. However, Yang et al. included both de novo lesions and in-stent restenosis lesions [18]. Because outcomes stratified by lesion type were not separately available in that study, ISR-specific data could not be excluded from the pooled estimates. Pooled data from 11 studies demonstrated comparable baseline demographics between the AT + DCB and DCB alone groups (Table 2). Known FAD risk factors, including sex, smoking, dyslipidemia, and hypertension, show no significant difference between the two groups. However, diabetes mellitus was significantly higher in the DCB group (59.20% vs. 52.48%; χ²=5.46, P = 0.02). Rutherford category analysis (8 studies) revealed a higher proportion of category 4 (20.78% vs. 13.79%; χ²=5.55, P = 0.02) and 5 (18.42% vs. 7.62%; χ²=10.63, P < 0.01) lesions in the DCB group, while there was no difference in the TASC (Trans-Atlantic Inter-Society Consensus). The degree of calcification burden was significantly higher in the AT + DCB groups (62.30% vs. 77.52%; χ²=26.64, P < 0.01). Distal protection device usage was more frequent in the AT + DCB group (3.84% vs. 73.15%; χ²=532.49, P < 0.01).

Table 1.

General characteristics of the included studies

First author	Year	Design	Country	Comparison	Relevant patients (n)		Lesion location	Lesion length (mm, mean ± SD)		Follow-up (Month)
					DCB	AT + DCB		DCB	AT + DCB
Stavroulakis	2016	Retrospective cohort	Germany	DCB VS DA + DCB	31	41	Popliteal lesions	47 ± 24	42 ± 24	12
Stavroulakis	2017	Retrospective cohort	Germany	DCB VS DA + DCB	26	21	SFA, deep femoral artery	39 ± 14	34 ± 16	12
T. Raymond Foley	2017	Retrospective cohort	American	DCB VS OA + DCB	61	28	SFA or popliteal	139 ± 100	135 ± 100	12
Thomas Zeller	2017	RCT	Germany, Poland, and Switzerland	DCB VS DA + DCB	54	67	SFA or popliteal	96.6 ± 40.9	112.3 ± 40.3	12
Zhiwen Cai	2019	RCT	China	DCB VS DA + DCB	49	45	SFA or popliteal	111 ± 63	113 ± 66	24
Lin Yang	2024	Retrospective cohort	China	DCB VS DA/LA/RA + DCB	89	52	Femoropopliteal artery	170.6 ± 141.5	185.6 ± 170.1	24
Jung-Joon Cha	2022	Retrospective cohort	Korea	DCB VS DA/RA + DCB	266	82	Femoropopliteal artery	181.9 ± 100.4	176 ± 120.9	24
Nicolas W Shammas	2022	RCT	American	DCB VS RA + DCB	16	31	Femoropopliteal artery	112.1 ± 76.4	108.3 ± 42.6	24
Patricia Wischmann	2024	Prospective cohort	England	DCB VS RA + DCB	84	66	Common femoral artery, proximal SFA, and proximal PFA	N.A.	N.A.	36
Damianos G.	2020	Retrospective cohort	American	DCB VS OA + DCB	33	33	Femoropopliteal artery	169.9 ± 105.3	177.7 ± 93.8	36
Orhan Rodoplu	2021	Retrospective cohort	Turkey	DCB VS RA + DCB	63	58	Femoropopliteal artery	102 ± 51	144 ± 52	24

AT, atherectomy; DCB, drug-coated balloon; SD, standard ratio; RCT, randomized controlled study; DA, directional atherectomy; OA, orbital atherectomy; LA, laser atherectomy; RA, rotational atherectomy; SFA, superficial femoral artery; PFA, profunda femoris artery

Table 2.

Demographic characteristics of the included studies

	Number of studies included (n)	DCB alone group	AT + DCB group	X-squared	P-value
Male (n, %)	11	610/772(79.02%)	398/524(75.95%)	1.52	0.22
Comorbidities (n, %)
Hypertension	11	647/772(83.81%)	432/524(82.44%)	0.33	0.57
Diabetes mellitus	11	457/772(59.20%)	275/524(52.48%)	5.46	0.02
Coronary heart disease	11	374/772(48.45%)	260/524(49.62%)	0.13	0.72
Dyslipidemia	10	478/711(67.23%)	353/496(71.17%)	1.94	0.17
Cerebrovascular disease	8	117/592(19.76%)	60/367(16.35%)	1.54	0.22
Smoking	10	402/746(53.89%)	270/503(53.68%)	0.00	0.99
Rutherford category (n, %)	8
2		36/176(20.45%)	32/177(18.08%)	0.19	0.67
3		226/361(62.60%)	236/348(67.82%)	1.90	0.17
4		75/361(20.78%)	48/348(13.79%)	5.55	0.02
5		42/228(18.42%)	17/223(7.62%)	10.63	0.00
6		12/148(8.11%)	13/106(12.26%)	0.78	0.38
TASC (n, %)	6
A		64/332(19.28%)	48/229(20.96%)	0.15	0.70
B		104/332(31.33%)	90/281(32.03%)	0.01	0.92
C		113/332(34.04%)	88/281(31.32%)	0.39	0.53
D		88/278(31.65%)	45/183(24.59%)	2.35	0.13
Calcification (n, %)	9	395/634(62.30%)	331/427(77.52%)	26.64	0.00
Distal protection (n, %)	9	23/599(3.84%)	297/406(73.15%)	532.49	0.00

AT, atherectomy; DCB, drug-coated balloon; TASC, Trans-Atlantic Inter-Society Consensus

Literature quality assessment

The risk of bias was rated as low for most items in RCTs, except for the detection bias, which was assessed as uncertain due to the lack of mention by the authors (Fig. 2). The quality of the literature for non-randomized controlled studies was assessed using the NOS scale and was high in all cases (Table 3).

Fig. 2.

Risk of bias. Methodological quality assessment with the latest version of the cochrane risk of bias tool for randomized controlled study

Table 3.

Risk of bias. Methodological quality assessment with the latest version of the Newcastle-Ottawa scale

Author	Year	Selection	Comparability	Outcome	Total score
Stavroulakis	2016	3	2	3	8
Stavroulakis	2017	3	2	3	8
T. Raymond Foley	2017	3	1	3	7
Lin Yang	2024	4	2	2	8
Jung-Joon Cha	2022	4	2	3	9
Patricia Wischmann	2024	4	2	3	9
Damianos G.	2020	3	2	3	8
Orhan Rodoplu	2021	3	2	3	8

Meta-analysis results

10 studies reported fTLR and the results indicated that at the 12-month postoperative mark, AT + DCB did not significantly improve fTLR compared to DCB alone (OR = 0.81, 95% CI= [0.57, 1.16]; heterogeneity: I² = 24%) (Fig. 3). The primary patency was reported in 9 of the included studies. There was no difference in primary patency between the two groups (OR = 0.92, 95% CI= [0.68, 1.24]; heterogeneity: I² = 0%) (Fig. 4). However, there is a significant increase in bailout stenting in the DCB alone group (OR = 2.93, 95% CI= [1.95, 4.39]; heterogeneity: I² = 7%) (Fig. 5).

Fig. 3.

Forest plot of freedom from target lesion revascularization (fTLR) in comparison of DCB only vs. AT + DCB within a 12-month risk period. AT, atherectomy; DCB, drug-coated balloon; M-H, Mantel-Haenszel fixed-effects model (I² = 24%); CI, confidence interval

Fig. 4.

Forest plot of primary patency in comparison of DCB only vs. AT + DCB within a 12-month risk period. AT, atherectomy; DCB, drug-coated balloon; M-H, Mantel-Haenszel fixed-effects model (I² = 0%); CI, confidence interval

Fig. 5.

Forest plot of bailout stenting in comparison of DCB only vs. AT + DCB. AT, atherectomy; DCB, drug-coated balloon; M-H, Mantel-Haenszel fixed-effects model; CI, confidence interval

When we conducted a subgroup analysis based on the type of AT, we found that all types of AT performed prior to DCB decreased the use of bailout stenting postoperatively (Supplementary Figure S1). All-cause mortality did not differ significantly between the two groups: 10.7% (76/711) in the DCB-alone group versus 6.9% (34/496) in the AT + DCB group (OR = 1.15, 95% CI = [0.74, 1.78]; heterogeneity: I² = 31%) (Figure S2A). Similarly, no difference was observed in the amputation rate between the two treatment groups (OR = 1.79, 95% CI = [0.90, 3.56]; heterogeneity: I² = 0%) (Figure S2B). Perforation rates also showed no significant difference between the groups (OR = 0.66, 95% CI = [0.28, 1.58]; heterogeneity: I² = 0%) (Figure S2C). Hematoma rates were comparable between the two groups (OR = 0.83, 95% CI = [0.38, 1.81]; heterogeneity: I² = 0%) (Figure S2D). However, a statistically higher incidence of flow-limiting dissection rates was observed in patients treated with DCB alone compared to AT + DCB (OR = 4.46, 95% CI = [2.53, 7.86]; heterogeneity: I² = 0%) (Figure S2E). Subgroup analysis based on the type of AT revealed that the incidence of flow-limiting dissections was higher in the DCB-alone group compared to all AT + DCB subgroups, except for the orbital AT + DCB subgroup, which showed no significant difference (Supplementary Figure S3). In contrast, subgroup analysis indicated no significant difference in the incidence of distal embolism between different AT subgroups combined with DCB and DCB alone (Supplementary Figure S4). However, the overall effect demonstrated that DCB alone was associated with a decreased risk of distal embolism compared to AT + DCB (OR = 0.37, 95% CI= [0.16, 0.84]; heterogeneity: I² =0%) (Supplementary Figure S4).

Subgroup analyses based on study design demonstrated generally consistent findings between RCTs and retrospective cohort studies, supporting the robustness of the primary conclusions. For bailout stenting, the benefit of AT + DCB was significant in both RCTs (OR = 4.06, 95% CI= [1.62, 10.20]; heterogeneity: I² =72%) and cohort studies (OR = 2.82, 95% CI= [1.93, 4.13]; heterogeneity: I² =0%), with no evidence of subgroup disparity (P for interaction = 0.47) (Supplementary Figure S5). Similarly, for flow-limiting dissection, both RCTs (OR = 6.46, 95% CI= [2.33, 17.91]; heterogeneity: I² =31%) and cohort studies (OR = 3.76, 95% CI= [1.90, 7.45]; heterogeneity: I² =0%) confirmed a significantly higher risk with DCB alone, with consistent effects across designs (P for interaction = 0.39) (Supplementary Figure S6). Regarding distal embolism, the overall protective effect of AT + DCB (OR = 0.37, 95% CI= [0.16, 0.84]; heterogeneity: I² =0%) was not significantly modified by study design (P for interaction = 0.44) (Supplementary Figure S7). For fTLR, while no significant overall difference was detected (OR = 0.81, 95% CI= [0.57, 1.16]; heterogeneity: I² = 24%), the test for subgroup differences approached significance (P = 0.05), suggesting a potential variation in treatment effect that warrants further investigation. Collectively, these analyses confirm that the observed advantages of adjunctive atherectomy with DCB for periprocedural outcomes are robust and not disproportionately influenced by the inclusion of observational data.

Publication bias

When analyzing the publication bias of bailout stenting, the funnel plot results (Fig. 6) showed that the samples on both sides were essentially symmetrical, and the Egger’s test (p > 0.05) and Begg’s test (p > 0.05) indicated that no evidence of publication bias was found. Formal assessment of publication bias using funnel plots and Egger’s/Begg’s tests was performed only for bailout stenting, as the primary procedural outcome. For 12-month primary patency and fTLR, the number of contributing studies was fewer than 10, which is below the recommended threshold for reliable funnel plot asymmetry tests. Therefore, formal publication bias assessment for these endpoints was not conducted.

Fig. 6.

Effect size funnel on publication bias

Discussion

The therapeutic landscape for FAD has been fundamentally transformed by the widespread adoption of DCB. Championing the “leave nothing behind” philosophy, DCB function by utilizing a lipophilic matrix to facilitate the rapid transfer of antiproliferative agents—predominantly paclitaxel or sirolimus—into the arterial media during inflation, thereby effectively inhibiting neointimal hyperplasia [22]. Several large RCTs have demonstrated that DCB was effective for the treatment of FAD [23–25]. Despite these accolades, the “DCB-only” strategy is not a panacea; its efficacy is notably attenuated in complex anatomical subsets, particularly those with severe calcification [26, 27]. Consequently, the paradigm of avoiding permanent implants is frequently compromised by the necessity for bailout stenting, primarily due to flow-limiting dissection or significant residual stenosis [28–31]. Notably, a recent meta-analysis by Lin et al. [32] involving 470 patients reported a bailout rate of 27.74%, a figure that closely aligns with the 26.52% observed in our analysis. These findings underscore a persistent clinical challenge: relying solely on balloon dilation often fails to achieve adequate luminal gain without incurring vessel damage. To overcome these mechanical and pharmacological limitations, intravascular plaque debulking techniques has been proposed as a synergistic adjunctive therapy to optimize vessel preparation.

A salient finding of our meta-analysis is that adjunctive AT significantly reduced the incidence of bailout stenting (11.57% vs. 26.52%) and flow-limiting dissections (3.52% vs. 13.91%) compared to DCB monotherapy. This benefit is mechanistically attributed to aggressive “vessel preparation.” By physically excising plaque and modifying the calcified ring, AT alters vessel compliance, thereby enabling effective luminal gain at lower balloon inflation pressures. This “low-pressure strategy” minimizes severe barotrauma and elastic recoil, which are the primary drivers of dissection and subsequent need for stents [29, 33, 34]. The reduction in permanent metal implants is particularly clinically relevant for “no-stent zones,” such as the common femoral artery and the popliteal segment. These anatomical regions are subjected to significant biomechanical forces—including flexion, extension, and torsion—making them prone to stent fracture and kinking [35, 36]. Consequently, the AT + DCB hybrid approach offers a durable solution for complex, long, and calcified lesions in these mobile segments, preserving future surgical options. Furthermore, long-term data, such as the 3-year results from retrospective analyses, suggest that this combined strategy maintains satisfactory patency even in severely calcified cohorts [37]. Future research should pivot towards defining a “calcium-specific algorithm,” comparing the differential efficacy of rotational, orbital, and directional atherectomy devices to determine which modality best addresses specific plaque morphologies [38].

While the efficacy of the hybrid approach is evident, the safety profile regarding distal embolization requires nuanced interpretation. Our analysis noted a substantial disparity in the utilization of Distal Embolic Protection Devices (EPDs) between the groups (73.15% in AT + DCB vs. 3.84% in DCB alone; P < 0.01). Despite this protective coverage, a statistical signal regarding embolization risk was observed (OR = 0.37, 95% CI [0.16, 0.84]), highlighting the inherent mechanical risks of debulking. his phenomenon aligns with the mechanistic reality of atherectomy. By design, rotational or directional devices liberate plaque fragments, creating a “shower” of particulate debris. A seminal meta-analysis of 7 RCTs indicated that while clinically significant macro-embolization is mitigated by EPDs, distal micro-embolization (debris captured within the filter) is nearly ubiquitous in atherectomy arms (occurring in up to 64.7% of cases), whereas it is negligible in balloon-only angioplasty [4]. These suggest that vessel preparation may disturb unstable plaques or calcified fragments, leading to peripheral vascular embolism, especially in cases with high thrombus burden or complex calcified lesions [39]. Consequently, future guidelines should explicitly define the criteria for mandatory EPDs deployment to neutralize the embolic risk associated with “aggressive” vessel preparation [40].

Another issue relevant to clinical homogeneity is the inclusion of ISR lesions in at least one study. Although the present meta-analysis was designed to evaluate de novo femoropopliteal artery disease, Yang et al. included ISR lesions in addition to de novo lesions [18]. ISR differs from de novo atherosclerotic disease in biological behavior, mechanical substrate, restenosis mechanism, and response to endovascular therapy. Therefore, inclusion of ISR lesions may have influenced the pooled estimates for 12-month primary patency and fTLR. Because lesion-type-specific outcomes were not separately reported, we were unable to isolate the effect of ISR lesions or exclude them from the pooled event data without removing the entire study. Accordingly, the clinical outcome estimates should be interpreted cautiously, and this issue has been highlighted as a limitation.

It should be further emphasized that, in the absence of clear long-term clinical benefits, the increased procedural complexity and cost associated with AT may not be justified, particularly given the rising incidence of distal embolization. However, when focusing on PAD management, we observe that the increasing prevalence of endovascular interventions has correspondingly elevated the incidence of ISR, creating a highly challenging clinical scenario that often necessitates complex reinterventions. As a complex subtype of PAD, ISR lesions have been systematically excluded from numerous RCTs evaluating DCB [41]. A recent large randomized controlled trial involving 417 patients with complex PAD lesions demonstrated that ISR correlates with the poorest long-term outcomes, including the lowest clinically driven target lesion revascularization rate and five-year all-cause mortality-free rate, with 15.9% of patients undergoing two or more target lesion revascularizations during follow-up [42]. These data underscore that avoiding stent implantation—and thereby preventing new ISR lesions—may represent a clinically important goal beyond immediate procedural success. Our finding suggests that AT + DCB significantly reduces bailout stenting may therefore have important implications for the lifetime management of PAD patients, particularly those with complex lesion morphologies where stent avoidance is desirable. While the current follow-up duration (predominantly 12-month) may be insufficient to capture differences in long-term outcomes, the reduction in stent implantation could theoretically translate into improved outcomes over extended follow-up by mitigating the cumulative burden of ISR and repeat interventions. It is noteworthy that although AT + DCB did not demonstrate significantly superior 12-month patency or fTLR, the majority of included studies showed no statically difference compared to DCB alone, suggesting that the procedural benefits were not achieved at the expense of compromised mid-term efficacy.

Limitations

The most important limitation of this meta-analysis is the substantial baseline imbalance between the DCB-alone and AT + DCB groups. Because most included studies were observational, treatment selection was likely influenced by clinical severity, lesion morphology, calcification burden, and operator preference. The DCB-alone group included more patients with diabetes mellitus and advanced ischemia, including Rutherford categories 4 and 5, whereas the AT + DCB group included more heavily calcified lesions and substantially more frequent use of distal embolic protection devices. These imbalances may have affected the primary clinical endpoints in opposing directions: diabetes mellitus and advanced ischemia may worsen patency and increase the likelihood of reintervention in the DCB-alone group, whereas severe calcification and greater lesion complexity may adversely affect procedural and clinical outcomes in the AT + DCB group. Therefore, the observed lack of significant difference in 12-month primary patency and fTLR should be interpreted cautiously and should not be considered proof of true equivalence between the two treatment strategies. Another important limitation is the mechanistic heterogeneity among the AT modalities included in the meta-analysis. The pooled AT + DCB group comprises directional, rotational, orbital, laser, and mixed AT interventions, each with distinct procedural characteristics, safety profiles, and lesion suitability. Consequently, the summary estimates reflect an average effect across heterogeneous devices, which may not apply uniformly to any single atherectomy type. Our subgroup analyses mitigate this concern to some extent, but the heterogeneity should be considered when interpreting the overall findings. Finally, the available follow-up was predominantly limited to 12 months, and data beyond 3 years were sparse, restricting our ability to determine whether the procedural advantages of AT + DCB translate into durable long-term clinical benefit.

Conclusion

Our meta-analysis demonstrated that AT + DCB did not show significant clinical superiority over DCB alone in terms of 12-month primary patency or fTLR. This combined approach did, however, significantly reduce the occurrence of flow-limiting dissections and decrease the need for bailout stenting. This highlights its potential as a strategy for achieving “graft-free residual” treatment in FAD. If further technical refinements can reduce the higher risk of distal embolization seen with the addition of atherectomy, longer term clinical benefits may be realized by reducing the risk of ISR. Importantly, these findings emphasize the necessity for large-scale, multicenter randomized controlled trials with extended follow-up periods to confirm the clinical efficacy and safety of this combined therapeutic approach.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Abbreviations

DCB

Drug-coated balloon

AT

Atherectomy

FAD

Femoropopliteal atherosclerotic disease

fTLR

Freedom from target lesion revascularization

ISR

In-stent restenosis

PRISMA

Preferred reporting items for systematic reviews and meta-analyses

P.I.C.O.

Patient, intervention, comparison, outcome

RCT

Randomized controlled trial

NOS

Newcastle-Ottawa scale

TASC

Trans-Atlantic Inter-Society consensus

EPD

Embolic protection device

OR

Odds ratio

CI

Confidence interval

Author contributions

CSB, WJY, and HT were responsible for designing the research study. Data collection was carried out by YY, CSB, and WXL, while data analysis was conducted by WXL, WJY, and CSB. The manuscript was prepared by YY, CZG, and HT. All authors contributed to editorial revisions, reviewed and approved the final version, and ensured their active participation and accountability in all aspects of the work.

Funding

This work is supported by the Wuhan Central Hospital Research Fund (23YJ30). No relationship with industry.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.