Drug-coated balloons are the preferred treatment for in-stent restenosis (ISR), avoiding the need for a second metallic layer. While both paclitaxel-coated balloons (PCB) and limus-coated balloons (LCB, typically sirolimus or biolimus) are used, direct comparative evidence is limited, particularly regarding their long-term angiographic efficacy and clinical safety. We conducted a systematic review and study-level meta-analyzis of studies comparing LCB with PCB for coronary ISR. A comprehensive search was performed across PubMed, Embase, Cochrane, Scopus, Web of science for randomized controlled trials. Outcomes were compared using risk ratios (RR) for categorical data and mean differences for continuous data in a random-effects model. Heterogeneity was assessed using the I 2 statistic. Six randomized controlled trials involving 1,038 patients were included (552 in the LCB group and 481 in the PCB group). Compared with PCB, LCB were associated with a significantly higher risk of clinically driven target lesion revascularization (RR: 1.48; 95% CI: 1.02 to 2.14; p = 0.04; I 2 = 0%) and target lesion failure (RR: 1.19; 95% CI: 0.87 to 1.62; p = 0.27; I 2 = 0%). In contrast, both platforms demonstrated no differences in all-cause mortality (RR: 0.98; p = 0.96), myocardial infarction (RR: 0.73; p = 0.46), stent thrombosis (RR: 0.66; p = 0.57), or MACE (RR: 1.09; p = 0.58). Angiographic outcomes were comparable, including late lumen loss (mean differences: −0.00 mm; p = 0.65; I 2 = 74.7%) and minimal lumen diameter (mean differences: −0.12 mm; p = 0.07). In patients with ISR, PCB reduced repeat revascularization compared with LCB. These data suggest a modest advantage for PCB in preventing restenosis following drug-coated balloon angioplasty for ISR.
Despite the advances achieved with second- and third-generation drug-eluting stents (DES), in-stent restenosis (ISR) remains a clinically relevant complication, with an estimated incidence of 5% to 10% of all percutaneous cardiovascular interventions procedures, depending on lesion complexity and patient profile. , The management of ISR seeks to balance antiproliferative efficacy with the preservation of vascular function. The conventional “stent-in-stent” approach restores luminal flow promptly but adds another metallic layer, increasing the risk of late thrombosis, impairing vasomotion, and complicating future reinterventions. In this context, “leave-nothing-behind” strategies have emerged as physiologically conservative alternatives, avoiding metal accumulation and promoting functional vessel restoration.
Drug-coated balloons (DCBs) and coronary brachytherapy allow treatment of ISR without additional stent layers. Unlike DES, DCBs deliver antiproliferative agents locally without leaving a permanent implant, thereby reducing inflammation and improving endothelial healing. The European Society of Cardiology (2024) guidelines and the European Association of Percutaneous Cardiovascular Interventions (2023) consensus recommend DCBs as first-line therapy for ISR, with a Class I recommendation and Level of Evidence A , although DES are preferred over DCB for DES ISR. ,
There is controversy on the angiographic efficacy and clinical safety of limus- versus paclitaxel-coated DCBs in ISR. Published studies had methodological variability, differences in lesion preparation, coating technologies, and endpoint definitions, as well as small sample sizes and short follow-up durations (<12 months). Previous meta-analyzes were limited by small sample sizes with limited power for clinical outcomes and focused primarily on angiographic outcomes like late lumen loss (LLL).
Methods
We performed a systematic review and study-level meta-analyzis of randomized controlled trials (RCTs) comparing limus-coated-balloons (sirolimus or biolimus-coated) versus paclitaxel-coated balloons for the treatment of coronary in stent restenosis. The study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and registered in PROSPERO (CRD420261285762).
Search strategy and study selection
A comprehensive search was conducted across PubMed, Embase, Cochrane Central Register of Controlled Trials, Scopus, and Web of Science (WOS) databases from January 2000 to October 2025. The final search was performed on October 15, 2025. The following medical subject heading were used ([“Sirolimus” OR “Sirolimus-eluting”] AND [“Paclitaxel”OR “Paclitaxel-eluting”] AND [“In-stent restenosis” OR ISR OR “coronary restenosis”] AND [“Percutaneous Coronary Intervention” OR PCI OR “Coronary Angioplasty”]). We identified a total of 948 records, of which 706 were removed as duplicates. Two reviewers independently conducted the search, screened titles and abstracts, and assessed full-text articles, resolving disagreements through consensus. References of all included studies were manually reviewed to identify additional eligible studies, though gray literature was not searched. Data extraction was performed independently by 3 reviewers using a standardized extraction form, and inconsistencies were resolved by discussion.
Eligibility criteria and endpoints
Studies were eligible for inclusion if they RCTs that compared limus-coated balloons with paclitaxel-coated balloons for the treatment of coronary in-stent restenosis in adult patients undergoing percutaneous coronary intervention. Only studies that reported at least one angiographic or clinical outcome of interest, such as late lumen loss, percentage diameter stenosis, minimal lumen diameter, target lesion revascularization (TLR) defined as Repeat percutaneous intervention or surgical bypass of the original target lesion, target lesion failure (TLF), myocardial infarction, cardiac death, or major adverse cardiovascular events, were included. Studies with overlapping populations from the same recruitment centers, conference abstracts, case reports, editorials, reviews, animal or laboratory investigations, or those without extractable quantitative outcomes were excluded. When essential information was missing or unclear, attempts were made to contact study authors for clarification.
Regarding the definition of clinical endpoints, revascularization was defined as clinically driven in the majority of contemporary trials, such as REFORM, BIO ASCEND ISR, and SIBLINT-ISR. ,, In contrast, the Mohd Ali (2019) and Scheller (2022) trials utilized mandatory 6-month angiographic follow-up to assess their primary endpoint of late lumen loss. In these studies, any reintervention performed at the time of the scheduled angiogram, even in the absence of worsening symptoms, was recorded as a TLR. This ‘angiographic trigger’ likely contributed to higher overall revascularization rates in these specific cohorts compared to trials using purely symptom-driven criteria.
Bias assessment
To ensure the internal validity of the included studies, we performed a rigorous assessment of the risk of bias using the Cochrane Risk of Bias 2 (RoB 2) tool for randomized controlled trials. This assessment was conducted independently by 2 reviewers, with any discrepancies resolved through consensus or by consultation with a third senior investigator. The RoB 2 tool was applied across 5 essential domains to evaluate the integrity of the trial results. Each domain was classified as having a “low risk of bias,” “some concerns,” or a “high risk of bias.” Finally, an overall risk of bias was assigned to each trial based on these criteria to determine the strength of the evidence generated by our meta-analysis.
Statistical analyzis
Continuous outcomes were compared using mean differences (MD) with 95% confidence intervals, while categorical outcomes were compared using risk ratios (RR) with 95% confidence intervals. A random-effects model was chosen a priori due to anticipated heterogeneity in study populations, devices, and angiographic methodologies. Heterogeneity was assessed using the Cochran Q test and the I² statistic. To evaluate the robustness of the primary findings and ensure that the results were not driven by a single outlier, sensitivity analyzes were performed using the leave-one-out method. This involved iteratively recalculating the pooled effect size by omitting one study at a time to assess the stability of the RRs and MDs. Finally, potential publication bias and small-study effects were visually inspected using funnel plots. The symmetry of these plots was evaluated to determine the likelihood of missing or unpublished data affecting the validity of the meta-analytical results. All statistical analyzes were performed using R version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) and the meta package.
Results
Study selection and characteristics
The study selection process is shown in Figure 1 . A total of 948 records were identified through the systematic database search. After removal of duplicate records (242 studies) and ineligible studies by title and abstract, 60 remained and were fully reviewed. Ultimately, 6 studies (6 RCTs) met the inclusion criteria and were incorporated into the final analyzis. ,,,,, Tables 1 and 2 summarize 6 studies (2017 to 2022) comparing Limus versus PCB, with 50 to 280 patients per study (about 1,038 total, 552 in the LCB group and 481 in the PCB group); populations were predominantly male (about 73% to 97 %) and mean age was similar between groups (about 59 to 70 years). Previous MI (A clinical condition defined by the presence of acute myocardial injury detected by abnormal cardiac biomarkers) prevalence varied widely (about 16 % to 100 %). Most studies assessed angiographic primary endpoints, mainly late lumen loss at 6 to 12 months or percent diameter stenosis at 6 months, while one study reported cardiac death. Overall, follow-up percent diameter stenosis and late lumen loss were broadly comparable between LCB and PCB, but in-stent restenosis was heterogeneous, notably higher with LCB in Byrne (2025) (33.9 % vs 9.3 %), whereas other studies showed more similar rates when reported (about 11% to 32% vs 13% to 30%).
PRISMA flow diagram.
Table 1
Characteristics of the included studies
| Authors (year) | Country | Study design (year range) | N total (LCB/PCB) | Agen, mean (LCB/PCB) | Male, n (%) LCB/PCB | Limus coated balloon type |
|---|---|---|---|---|---|---|
| REFORM STUDY (2025) | 6 countries | RCTs (2020-2022) | 202 (135/67) | 68.9/68.3 | 110 (81.5)/52 (77.6) | Biolimus |
| BIO ASCEND ISR (2024) | China | RCTs (2020-2022) | 280 (138/137) | 63.64/64.24 | 103 (74.6)/101 (73.7) | Biolimus |
| SIBLINT-ISR (2025) | – | RCTs (2020-2021) | 260 (132/128) | 63.8/63.7 | 95 (73.1)/97 (75.8) | Sirolimus |
| Mohd (2019) | – | RCTs (2020-2021) | 50 (25/25) | 58.6/61.6 | 19 (79)/22 (88) | Sirolimus |
| TIS 2 STUDY (2025) | Czech Republic | RCTs (2019-2022) | 145 (72/73) | 69.7/68.0 | 56 (77.8)/59 (80.8) | Sirolimus |
| Scheller (2022) | Malaysia and German-Swiss | RCTs (2017-2020) | 101 (50/51) | 67/63 | 43 (86)/39 (76) | Sirolumus |
Table 2
Baseline characteristics of the patients in each study
| Authors (year) | Previous MI, n (%) LCB/PCB | Primary end point | Follow-up, %DS in segment (LCB/PCB) | Follow-up, %DS in stent (LCB/PCB) | Follow-up, LLL in segment (LCB/PCB) | Follow-up, %DS in stent (LCB/PCB) | Binary restenosis in stent (LCB/PCB) |
|---|---|---|---|---|---|---|---|
| REFORM STUDY (2025) | 69 (51.1)/37 (55.2) | % diameter stenosis at 6 months | 43.25 ± 22.87/31.4 ± 17.7 | 42.76 ± 22.90/27.9 ± 19.49 | 0.54 ± 0.62/0.20 ± 0.54 | 0.69 ± 0.62/0.31 ± 0.54 | 41 (33.9%)/5 (9.3%) |
| BIO ASCEND ISR (2024) | 22 (15.9)/33 (24.1) | Late lumen loss at 9 months | 32.59 ± 18.78/31.61 ± 16.9 | 31.25 ± 19.31/29.54 ± 17.3 | 0.25 ± 0.40/0.27 ± 0.39 | 0.26 ± 0.42/0.29 ± 0.40 | 17 (11.3%)/20 (13.1%) |
| SIBLINT-ISR (2025) | 132 (100)/128 (100) | Late lumen loss at 9 months | 31.81 ± 20.92/31.46 ± 8.44 | 27.99 ± 21.62/28.77 ± 18.72 | 0.37 ± 0.48/0.30 ± 0.38 | 0.35 ± 0.47/0.31 ± 0.36 | 20 (17.9%)/18 (15%) |
| Mohd (2019) | 9 (36)/8 (32) | Late lumen loss at 6 months | – | 16.1 ± 15.5/17.8 ± 20.9 | 0.21 ± 0.54/0.17 ± 0.55 | 0.28 ± 0.51/0.25 ± 0.58 | – |
| TIS 2 STUDY (2025) | 48 (66.7)/50 (68.5) | Late lumen loss at 12 months | – | 32 (3%)/38 (15%) | – | 0.45 ± 0.77/ 0.50 ± 0.56 | 25 (31.6%)/24 (30.4%) |
| Scheller (2022) | 50 (100)/51 (100) | Late lumen loss at 6 months | – | 20.8 ± 22.9/18.3 ± 18.0 | 0.26 ± 0.43/0.25 ± 0.42 | 0.30 ± 0.47/0.30 ± 0.47 | 7 (13.5%)/4 (7.7%) |
LCB = limus-coated balloon; PCB = paclitaxel coated balloon.
Clinical outcomes
Clinical outcomes were comparable between the 2 technologies across most safety and efficacy parameters. As compared with PCB, LCB demonstrated similar rates of all-cause mortality (RR 0.98; 95% CI: 0.35 to 2.70; p = 0.96; I 2 = 0%, Figure 3 ), TLF (RR 1.19; 95% CI: 0.87 to 1.62; p = 0.27; I 2 = 0%, Figure 4 ) and MACE (A standard composite clinical endpoint used to evaluate the cumulative safety and efficacy of cardiovascular interventions. It is defined as cardiovascular death, non-fatal myocardial infarction, and stroke, though some protocols may include target lesion revascularization [TLR]) (RR 1.09; 95% CI: 0.81 to 1.47; p = 0.58; I 2 = 0%, Figure 5 ). Regarding secondary safety endpoints, no significant differences were observed in the incidence of myocardial infarction (RR 0.73; 95% CI: 0.31 to 1.70; p = 0.46, Figure 6 ) or stent thrombosis (RR 0.66; 95% CI: 0.15 to 2.81; p = 0.57, Figure 7 ), with 0 heterogeneity (I 2 = 0%) across all safety analyzes.
In contrast, treatment with LCB was associated with significantly higher TLR compared with PCB (RR 1.48; 95% CI: 1.02 to 2.14; p = 0.04; I 2 = 0%, Figure 2 ).
