This study analyzed medical records of consecutive patients who underwent ELCA or RA at Guangdong Provincial People's Hospital (July 2020–August 2023). The choice between ELCA and RA was made at the operator's discretion, based on a comprehensive assessment of specific lesion characteristics. From 525 initially screened procedures, we identified 89 patients with CTO exhibiting device-uncrossable or balloon-undilatable lesions. Ultimately, 75 eligible patients were included in this study. The inclusion and exclusion process of the study subjects was detailed in Figure 1. The inclusion criteria for this study strictly adhered to the following points: (1) age ≥18 years; (2) CTO-PCI indications including angina symptoms and/or objective evidence of reversible myocardial ischemia (perfusion imaging/stress testing). The exclusion criteria were: (1) age >85years or high bleeding risk contraindicating surgery; (2) contraindications to dual antiplatelet therapy; (3) history of radiation skin injury; (4) severe renal impairment without dialysis commitment. This study has been strictly approved by the Ethics Committee of Guangdong Provincial People's Hospital (approval number: KY2023-716-01), and in accordance with regulations, participants are not required to sign an informed consent form. Figure 2 presents representative cases of CTO-PCI, one performed with ELCA and the other with RA.

A CTO was defined as a coronary artery occlusion with Thrombolysis in Myocardial Infarction (TIMI) flow grade 0 persisting for >3 months, as determined by symptom onset, documented myocardial infarction (MI) in the territory, or prior angiographic evidence. Lesion complexity was assessed using the Japanese CTO (J-CTO) score, with collateral circulation graded by the Werner classification (8). Moderate/severe calcification required clearly visible high-density shadows on angiography during both systolic and diastolic phases (9), while significant tortuosity was defined as ≥2 bends >70° or one bend >90° (10). The retrograde technique involved using specialized guidewires and microcatheters to approach the occlusion distally through collateral channels. Procedure time was measured from catheterization laboratory entry to exit. Technical success required successful stent/balloon delivery with <30% residual stenosis and TIMI 3 flow; procedural success additionally mandated absence of in-hospital MACCE, including death, periprocedural MI [Fourth Universal Definition (11)], urgent revascularization, pericardial drainage, surgical tamponade, or stroke. Follow-up MACCE comprised cardiovascular death, non-fatal MI, ischemia-driven revascularization, and stroke, with stent thrombosis defined by Academic Research Consortium criteria (12).

ELCA was performed using the CVX-300 excimer laser system (Spectranetics Corporation, Colorado Springs, CO, USA), which operates at 308 nm wavelength with 135 ns pulse width and 165 mJ/pulse energy output (13). The Vitesse CTM laser atherectomy catheter (diameter range: 0.9–1.7 mm) was selected based on angiography or intracoronary imaging, maintaining a catheter-to-vessel diameter ratio of 0.67 ± 0.21. All procedures employed saline-flush technique: (1) pre-irrigation with 0.9% saline to clear contrast media; (2) bolus injection of 5–10 ml saline before each laser activation; (3) continuous saline infusion (1 ml/s) during lasing. The catheter was advanced at 0.5–1 mm/s with pulsed activation (≤10 s/pulse, ≥5 s intervals) at energy densities of 35–80 mJ/mm² and frequencies of 35–80 Hz.

RA was performed using the Rotablator system (Boston Scientific, Marlborough, MA, USA). Burr selection was based on vessel diameter and plaque burden, with diameters ranging from 1.25 to 1.75 mm, and the maximum burr size was confirmed by angiography or intracoronary imaging to ensure adequate debulking. The rotational speed was maintained at 170,000–180,000 rpm, increased to 200,000 rpm for severely calcified lesions when needed. Following RA, the rotational guidewire was exchanged for a working guidewire, followed by balloon angioplasty or stent implantation to optimize angiographic results and minimize residual stenosis.

Follow-up data were collected through outpatient visits, hospital admissions, or telephone interviews, with clinical documentation during each encounter. The primary endpoint was MACCE, defined as a composite of cardiovascular mortality, non-fatal MI, ischemia-driven target vessel revascularization (TVR), and stroke. All procedures followed the Declaration of Helsinki principles.

Continuous variables with normal distribution were presented as mean ± standard deviation and compared using Student's t-test, while non-normally distributed variables were expressed as median (interquartile range) and analyzed with Mann–Whitney U-test. Categorical variables were reported as frequencies (%) and compared using χ² or Fisher's exact tests. Time-to-event outcomes were assessed using Kaplan–Meier methodology (log-rank test) and Cox proportional hazards models with adjustment for key clinical confounders. A two-tailed p < 0.05 was considered statistically significant. All analyses were conducted using MedCalc (v14.12.0, Ostend, Belgium).

Seventy-five consecutive patients undergoing PCI for CTO were treated with either ELCA (n = 25) or RA (n = 50). Baseline characteristics were well-balanced between groups (Table 1), with no significant differences in mean age (ELCA: 64.76 ± 9.42 vs. RA: 67.22 ± 8.58 years), comorbidities (hypertension: 68% vs. 72%; dyslipidemia: 52% vs. 58%; diabetes mellitus: 36% vs. 2%), smoking status (28% vs. 34%), or prior revascularization [PCI: 24% vs. 30%; coronary artery bypass grafting (CABG): 8% vs. 6%] (all p > 0.05). Serum creatinine (1.12 ± 0.31 vs. 1.08 ± 0.28 mg/dl) and left ventricular ejection fraction (LVEF) (52.4 ± 8.7% vs. 53.1 ± 9.2%) were also comparable.

Angiographic characteristics of coronary artery lesions are compared in Table 2. Significant differences were observed between the groups in several parameters. The proportion of lesions longer than 20 mm was significantly higher in the ELCA group (56% vs. 30%, p = 0.044). CTO caused by in-stent restenosis (ISR-CTO) were more prevalent in the ELCA group (20% vs. 2%, p = 0.024). Most notably, moderate/severe calcification was significantly more frequent in the RA group (76% vs. 48%, p = 0.020). Lesion complexity assessed by J-CTO score was comparable between groups (1.88 ± 1.2 vs. 1.86 ± 0.9, p = 0.942), as was moderate/severe tortuosity (28% vs. 12%, p = 0.161). Although the distribution of CTO target vessels did not differ significantly between groups, the RA group showed a numerically higher rate of prior failed CTO-PCI (20% vs. 4%, p = 0.134). Both groups had high and comparable rates of collateral circulation and multivessel disease. No significant differences were observed in the distribution of target vessels or lesion locations between the two groups.

Table 3 summarizes the procedural characteristics of both groups. During guidewire crossing of occluded lesions, severe plaque burden frequently prevented device advancement (microcatheters or balloons), with significantly higher device-crossing failure rates observed in the ELCA group compared to the RA group prior to atherectomy (72% vs. 48%, p = 0.048). The two groups demonstrated comparable numbers of balloons used for pre-dilation (5.76 ± 3.11 vs. 5.90 ± 2.48, p = 0.846) and achieved similar maximum pre-dilation pressures (15.52 ± 3.4 atm vs. 15.73 ± 4.3 atm, p = 0.845). Post-stent dilation pressures were also equivalent (19.10 ± 4.56 atm vs. 21.24 ± 3.69 atm, p = 0.053). Intravascular ultrasound (IVUS) or optical coherence tomography (OCT) guidance was frequently used in both groups, with comparable utilization rates between the ELCA and RA groups (44% vs. 40%, p = 0.740). Both groups required implantation of comparable numbers of stents (2.5 ± 0.9 vs. 2.3 ± 0.8) with similar total stent lengths (72.8 ± 35.6 vs. 68.4 ± 26.4 mm; all p > 0.05). No significant differences were observed in technical success (88% vs. 96%, p = 0.413) or procedural success rates (84% vs. 90%, p = 0.706) between the two groups. Fluoroscopy time, total procedure duration, and radiation dose were marginally higher in the ELCA group compared to the RA group, although without statistical significance. Total procedural costs were significantly higher in the ELCA group (US$11,147.48 ± 2,334.89 vs. US$9,266.65 ± 2,711.36; p = 0.007), while device-excluded costs showed no significant difference (US$6,433.19 ± 2,334.89 vs. US$7,291.65 ± 2,711.36; p = 0.210).

Table 4 summarizes procedural complications and in-hospital outcomes. No significant difference in major procedural complications was observed between groups (ELCA 24% vs. RA 16%, p = 0.600). However, the ELCA group showed a higher incidence of acute slow/no-reflow phenomena (12% vs. 0%, p = 0.061), while coronary perforation with cardiac tamponade occurred exclusively in the RA group (4% vs. 0%, p = 0.550), including one case requiring emergent pericardiocentesis followed by coil embolization. Two in-hospital deaths (multi-organ failure) occurred in the ELCA group, both adjudicated as unrelated to the laser procedure. No in-hospital mortality was observed in the RA group, though one patient required implantable cardioverter-defibrillator (ICD) implantation for recurrent ventricular tachycardia before discharge.

The clinical outcomes during follow-up are presented in Table 5. Patients with procedural failure were excluded from analysis, with the remaining cohort (follow-up rate: 88%) followed for a median of 16 months (IQR 12–26) in the ELCA group vs. 18 months (IQR 13.5–25.5) in the RA group. No significant between-group differences were observed in MACCE, cardiac rehospitalization, or all-cause mortality (all p > 0.05). No cardiac deaths occurred in either group. Non-cardiac deaths in the RA group (n = 3) included acute alcohol poisoning, complications of end-stage diabetes mellitus, and one undetermined cause. In the ELCA group, two stroke cases (lacunar infarction and brainstem infarction) were managed medically without residual disability. The RA group had one case of gastrointestinal bleeding (1-month post-PCI), potentially attributable to antiplatelet therapy, which resolved with hospitalization.

Cox proportional hazards regression models were used to assess the association between atherectomy modality (ELCA vs. RA) and clinical outcomes, with adjustment for baseline covariates (moderate/severe calcification, lesion length >20 mm, ISR-CTO, and diabetes mellitus). As detailed in Table 6, the adjusted analysis demonstrated no increased risk of MACCE with ELCA vs. RA (adjusted HR 1.53, 95% CI 0.35–6.65, p = 0.569). Similarly, adjusted models showed no significant differences in cardiac rehospitalization (HR 1.39, 95% CI 0.33–5.84, p = 0.653), recurrent angina (HR 0.78, 95% CI 0.09–6.87, p = 0.824), recurrent MI (p = 0.742), or TVR (p = 0.754).

Figure 3 displays unadjusted Kaplan–Meier curves comparing event-free survival between ELCA and RA after CTO-PCI. Over a median follow-up of 17.5 months (IQR 13–25.25), MACCE-free survival did not differ significantly between groups (χ² = 0.41, p = 0.523). Stroke-free survival showed a nominal difference (χ² = 4.12, p = 0.042), with two stroke events in the ELCA group (lacunar infarction at 13 months and brainstem infarction at 25 months post-PCI), both managed medically without residual deficits, vs. no strokes in the RA group. One RA patient developed target vessel-related acute inferior wall MI at 12 months (successfully treated with PCI), though MI-free survival demonstrated no intergroup difference (χ² = 0.46, p = 0.496). TVR-free survival rates declined comparably from 100% to 85.1% (ELCA) and 81.4% (RA) (χ² = 0.02, p = 0.901).