Article ID: CJ-20-1222
Background: The aim of the study was to assess anatomical and procedural predictors of clinical and procedural failure of rotational atherectomy (RA) in an all-comers population.
Methods and Results: A total of 534 consecutive patients who underwent RA were included in a double-center observational study. The primary composite endpoint consisted of: rota-wire introduction failure, burr-passage failure, periprocedural complications and procedure-related major adverse events. The second primary endpoint included rota-wire introduction failure and burr-passage failure. The primary endpoint occurred in 76 (14.2%) patients and the second primary endpoint occurred in 64 (12%) Periprocedural complications occurred in 23 (4.3%) and procedure-related adverse events in 23 (4.3%) patients. Multivariable analysis revealed angulation on lesion ≤90° (HR=2.18, 95% CI: 1.21–3.94, P=0.0096) and sequential lesion (HR=1.89, 95% CI: 1.01–3.54, P=0.046) as independent predictors of no clinical success of RA. Multivariable analysis revealed again that angulation on lesion ≤90° (HR=2.26, 95% CI: 1.16–4.40, P=0.02) and sequential lesion (HR=3.77, 95% CI: 1.64–8.69, P<0.01) as independent predictors of no procedural success of RA.
Conclusions: The presence of an acute angulation on lesion and sequential lesion are independent determinants of clinical and procedural failure of RA. Further research is necessary to establish a score predicting RA failure, which can help in preprocedural risk stratification of patients undergoing complex percutaneous coronary intervention with RA.
Rotational atherectomy (RA), recently accompanied by orbital atherectomy and intravascular lithotripsy, is the last resort for successful percutaneous coronary intervention (PCI) of highly calcified or fibrotic coronary lesions. Many studies revealed that the extent of coronary calcifications strongly correlated with the rate of future cardiac events, and the high prevalence of calcifications in patients with coronary artery disease makes PCI difficult to perform.1,2 Coronary calcifications are related to unsuccessful lesion preparation, stent delivery and expansion with subsequent procedure failure, target vessel failure, risk of restenosis, and stent thrombosis.3,4 The community of interventional cardiologists is growing around the world and the interest in RA has increased over the past decade as a consequence of more calcified and complex coronary lesions being treated with PCI. There is a constant increase in the number of RA procedures in more demanding patients subgroups, which constitutes 1–5% of all PCI implantations.5–7
However, RA is a complex procedure with a relatively long learning curve. Unfortunately, there is a lack of a dedicated score, which could predict the probability of RA failure and help the operator with preprocedural risk assessment. Most established predictors of outcome after RA of heavily fibro-calcified lesions are related to patients general risk and clinical conditions, such as age, decreased left ventricle ejection fraction (LVEF), diabetes mellitus, chronic kidney disease, and peripheral artery disease.8–11 Notwithstanding, there is a paucity of data concerning specific predictors of unsuccessful RA, related to detailed lesions anatomy like tortuosity, calcification degree and procedural factors, which could be helpful in risk stratification before the procedure.
The aim of the study was to determine predictors of clinical and procedural failure of RA in an all-comers population, focusing on a visual analysis of coronary lesions anatomy and on RA procedure-related factors.
Between April 2008 and November 2018, all consecutive patients who underwent PCI with accompanying RA were enrolled in a double-center observational study. There were no exclusion criteria. Baseline demographic, clinical characteristics, detailed anatomical and procedural data were collected, including indication for procedure, urgency, access site, all PCI equipment and lesion anatomy with basic quantitative coronary angiography (QCA) parameters, and characteristics such as length, tortuosity, length and severity of calcifications, sequential character of the lesion and Syntax Score. Anatomical parameters were obtained by a visual evaluation of a coronary angiography by 2 independent interventional cardiologists; in case of inconsistency, a third assessment was conducted by the supervising cardiologist.
Data concerning the reason of unsuccessful RA, periprocedural complications after each intervention and in-hospital major adverse cardiovascular events were collected. All patients gave informed consent for the procedure. The study protocol was approved by each participating center, accepted by the local ethics committee and was in accordance with the Declaration of Helsinki.
Study DefinitionsSevere calcifications were identified when radiopaque densities were visible without heart motion and affecting both sides of the vessel, and moderate calcifications were identified when densities were visible during heart motion and affecting one side of the vessel.
Calcium length was a length of visible radiopaque densities associated with the lesion.
A sequential lesion was identified when at least 2 significant stenoses were separated by a non-stenosed part of the vessel.
Vessel tortuosity was defined angiographically as a bending angle before, on and after the lesion. Angulation of the vessel was measured as a proximal angle between straight lines drawn along the proximal and distal part of the culprit lesion, then along the vessel before and after the lesion. An example of angle measurements on the lesion, pre lesion and post lesion is presented in Figure 1.
Examples of angulation measurements: (A,C) baseline coronary angiography, (B) angle measurements of an acute angle on a lesion (85°), pre lesion (112°) and post lesion (78°), (D) angle measurements of an obtuse angle on a lesion (114°), pre lesion (118°) and post lesion (133°).
We defined an undilatable lesion as the lesion that could not be adequately dilated by a balloon, whereas an uncrossable lesion was defined as a lesion that could be crossed by a wire, but could not be crossed with even the smallest balloons.
Syntax Score was calculated according to the online scoring algorithm (http:/www.syntaxscore.com).12 Lesion Syntax Score was a calculation of Syntax Score only in the vessel through which the RA was planned to be perform.
Rota-wire introduction failure was determined as an unsuccessful insertion of the device despite using a microcatheter or other techniques.
The definition of burr-passage failure included the inability to introduce a burr into the vessel or lesion, burr entrapment, lack of full lesion dilatation, inability to deliver or full expansion of a stent with residual stenosis >30% or final thrombolysis in myocardial infarction (TIMI) flow grade <3.
PCI-related myocardial infarction was defined according to the universal definition of myocardial infarction.13
Endpoint DefinitionsThe primary composite endpoint consisted of rota-wire introduction failure, burr-passage failure, periprocedural complications and procedure-related major adverse events, which occurred within 24 h post procedure (RA clinical failure). The second primary endpoint included rota-wire introduction failure and burr-passage failure (RA procedural failure). The secondary endpoint consisted of periprocedural complications with adverse events within 24 h post procedure. Periprocedural complications comprise of no/slow flow phenomenon, side branch occlusion, perforation, emergency coronary artery bypass grafting (CABG) or atrio-ventrocular block requiring pacemaker implantation. Major adverse events consisted of PCI-related myocardial infarction, target vessel revascularization, stroke and death.
ProcedureThe RA procedure was performed using a standard Boston Scientific Rotablator system (Boston Scientific, Marlborough, MA, USA). The radial or femoral route was used according to operator discretion. Burr speeds were between 140,000 and 180,000 rpm, with a run duration of approximately 20–30 s. In all procedures, an intracoronary continuous infusion of heparin and isosorbide dinitrate via the burr sheath was used. Heparin was given to maintain an activated clotting time of >250 s. In-hospital treatment before and after RA was conducted according to current standards, including an adequate dual antiplatelet therapy.
Statistical AnalysisContinuous variables with a normal distribution are presented as a mean±standard deviation, continuous variables with skewed distribution as median with an interquartile range and categorical variables as numbers and percentages. For continuous variables, intergroup differences were compared by using a Student’s t-test or the Mann-Whitney U-test, depending on the type of distribution. The χ² test was used to compare categorical variables. Both univariate and multivariate logistic regression models were used to determine the predicting factors of a composite endpoint. The multivariate model included all variables with P<0.05 in the univariate model. A P value <0.05 was considered statistically significant. All statistical analyses were performed using the Statistica 10.0 (StatSoft, USA) software.
A total of 534 consecutive patients who required PCI with RA were included in the study. In 458 patients (85.8%), we observed a procedural and clinical success (successful RA group) and in 76 (14.2%) patients, the procedure was unsuccessful (RA clinical failure group). Complete demographics, comorbidities, and laboratory results of both groups are presented in Table 1.
| Patients | Successful RA (n=458) |
RA clinical failure (n=76) |
P value |
|---|---|---|---|
| Age, years | 71.9 (±9.4) | 71.9 (±8.3) | 0.80 |
| Male | 313 (68) | 45 (59) | 0.12 |
| Hypertension | 396 (87) | 64 (84) | 0.60 |
| Diabetes mellitus | 235 (51) | 41 (54) | 0.85 |
| Prior stroke/TIA | 52 (11) | 8 (11) | 0.83 |
| Hyperlipidemia | 244 (53) | 41 (54) | 0.91 |
| Thyroid disease | 52 (11) | 8 (11) | 0.83 |
| Cancer disease | 157 (34) | 9 (12) | 0.87 |
| COPD | 27 (6) | 5 (7) | 0.81 |
| Atrial fibrillation | 119 (26) | 14 (19) | 0.15 |
| Peripheral artery disease | 106 (23) | 21 (28) | 0.39 |
| Severe valve disease | 58 (12) | 11 (15) | 0.51 |
| LVEF <50% | 168 (37) | 34 (45) | 0.16 |
| LVEF ≤35% | 76 (17) | 14 (19) | 0.67 |
| Impaired renal function | 94 (21) | 19 (25) | 0.37 |
| Hemodialysis | 13 (3) | 2 (3) | 0.92 |
| Prior acute coronary syndrome | 252 (55) | 49 (65) | 0.12 |
| Prior PCI | 307 (67) | 56 (74) | 0.25 |
| Prior CABG | 96 (21) | 17 (22) | 0.78 |
| Laboratory parameters | |||
| White blood cell count (103/uL) | 7.3 (5.9–9.1) | 7.5 (6.2–8.9) | 0.62 |
| Red blood cell count (106/uL) | 4.5 (4.1–4.8) | 4.5 (4.0–4.8) | 0.93 |
| Hemoglobin (g/dL) | 13.4 (12.1–14.4) | 13.2 (12.4–14.6) | 0.65 |
| Platelet count (103/uL) | 204 (169–244) | 216 (173–262) | 0.23 |
| Creatinine (mg/dL) | 0.94 (0.80–1.16) | 0.94 (0.80–1.15) | 0.72 |
| Glucose (mg/dL) | 108 (94–137) | 114 (96–143) | 0.13 |
| eGFR (mL/min/1.73 m2) | 72.5 (±26.0) | 71.7 (±21.0) | 0.87 |
| Risk scores | |||
| Logistic Euroscore II | 2.4 (1.4–4.5) | 2.7 (1.3–5.8) | 0.31 |
| Syntax Score | 15 (9–23) | 19.5 (11.5–29.0) | <0.01 |
| Syntax Score low (≤22) | 340 (74) | 43 (57) | <0.01 |
| Syntax Score intermediate (23–32) | 80 (18) | 21 (28) | 0.04 |
| Syntax Score high (≥33) | 38 (8) | 12 (16) | 0.04 |
| Lesion Syntax Score | 9 (7–12) | 10 (7–13.5) | 0.37 |
| Syntax Score II PCI | 36.4 (28.9–46.3) | 36.7 (29.4–52.7) | 0.22 |
| Syntax Score II PCI mortality | 19% (±19) | 24% (±24) | 0.22 |
| Syntax Score II CABG | 35.4 (±10.9) | 36.5 (±11.3) | 0.38 |
| Syntax Score II CABG mortality | 14% (±13) | 16% (±13) | 0.30 |
| Medication at discharge | |||
| Aspirin | 446 (97) | 69 (97) | 0.32 |
| P2Y12-inhibitor | 450 (98) | 65 (92) | <0.01 |
| β-blocker | 415 (90) | 66 (93) | 0.24 |
| ACE-inhibitor/ARB | 410 (90) | 63 (83) | 0.65 |
| Statin | 429 (94) | 68 (96) | 0.16 |
| Diuretic | 238 (52) | 38 (54) | 0.61 |
| Nitrates | 62 (14) | 13 (18) | 0.16 |
| Oral anticoagulation | 89 (19) | 10 (14) | 0.44 |
| Proton pump inhibitor | 295 (65) | 45 (63) | 0.94 |
Data are presented as numbers and percentages for categorical variables, mean±standard deviation for continuous variables with normal distribution and median with interquartile range for continuous variables with skewed distribution. ACE, angiotensin-converter enzyme; ARB, angiotensin receptor blocker; CABG, coronary artery bypass grafting; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; LVEF, left ventricle ejection fraction; PCI, percutaneous coronary intervention; TIA, transient ischemic attack.
There were no differences in demographics, comorbidities and laboratory parameters between the 2 study groups. There was also no difference in logistic Euroscore II on admission (P=0.31). However, the median Syntax Score was higher in the RA failure group than in the successful RA group (19.5 vs. 15.0, P<0.001), as was a high Syntax Score ≥33 (16% vs. 8%, P=0.038). Lesion Syntax Score and Syntax Score II were similar in both groups. The lower use of a second antiplatelet drug in the RA failure group was related to aborted procedures, when no final PCI was performed.
Procedure CharacteristicsDetailed procedure characteristics are summarized in Table 2 and a precise lesions angulation assessment is presented in Table 3. When RA was unsuccessful, the lesions were more complex: longer (30 mm vs. 25 mm, P<0.01), more severely calcified (91% vs. 79%, P=0.02 with calcium length >20 mm in 74% vs. 58%, P<0.01), more often sequential (72% vs. 47%, P<0.001) and totally occluded (16% vs. 7%, P=0.01). Importantly, these lesions were also more tortuous with more often acute angles than lesions from the successful RA group (angulation on lesion ≤90° in 36% vs. 15%, P<0.0001).
| Successful RA (n=458) |
RA clinical failure (n=76) |
P value | |
|---|---|---|---|
| Acute coronary syndrome | 151 (33) | 20 (27) | 0.28 |
| Radial access | 286 (63) | 58 (76) | 0.02 |
| Disqualification from CABG | 149 (33) | 30 (40) | 0.24 |
| Temporary pacing | 32 (7) | 8 (11) | 0.28 |
| MCS use | 6 (1) | 5 (7) | <0.01 |
| Reason for RA | |||
| Uncrossable lesion | 110 (34) | 27 (44) | 0.10 |
| Undilatable lesion | 213 (66) | 35 (57) | 0.19 |
| Direct RA | 109 (25) | 12 (16) | 0.10 |
| Target vessel | |||
| LM | 32 (7) | 5 (7) | 0.90 |
| LAD | 204 (45) | 28 (37) | 0.27 |
| Cx | 73 (16) | 13 (17) | 0.89 |
| RCA | 149 (33) | 30 (40) | 0.23 |
| Lesion characteristics | |||
| Lesion type B2/C | 361 (79) | 71 (93) | <0.01 |
| Ostial lesion | 67 (15) | 15 (20) | 0.25 |
| Bifurcation lesion | 184 (40) | 34 (45) | 0.45 |
| Chronic total occlusion | 32 (7) | 12 (16) | 0.01 |
| Moderate calcifications | 52 (11) | 5 (7) | 0.21 |
| Severe calcifications | 360 (79) | 69 (91) | 0.02 |
| Sequential lesion | 216 (47) | 55 (72) | <0.001 |
| Calcium length >20 mm | 266 (58) | 56 (74) | <0.01 |
| Lesion length >20 mm | 307 (68) | 59 (83) | <0.01 |
| Lesion length, mm | 25 (15–40) | 30 (22–51) | <0.01 |
| Diameter stenosis, % | 90 (85–98) | 90 (90–99) | 0.06 |
| Minimum lumen diameter, mm | 0.28 (0.04–0.45) | 0.20 (0.03–0.36) | 0.24 |
| Reference diameter, mm | 3.4 (3.0–3.6) | 3.1 (3.0–3.5) | <0.01 |
| Procedural data | |||
| Number of burrs | 1.2 (±0.4) | 1.0 (±0.7) | <0.001 |
| Burr-to-artery ratio | 0.44 (0.40–0.50) | 0.43 (0.42–0.50) | 0.26 |
| Maximum burr diameter | 1.5 (±0.2) | 1.4 (±0.2) | 0.05 |
| Number of stents | 1.5 (±0.7) | 1.0 (±1.3) | <0.001 |
| Total contrast volume, mL | 200 (180–280) | 250 (200–300) | 0.10 |
| Fluoroscopy time, min | 22 (15–30) | 24 (18–31) | 0.06 |
| Procedure time, min | 80 (60–105) | 80 (65–110) | 0.43 |
| Radiation exposure, μGy | 2,410 (1,436–3,999) | 2,588 (1,346–3,917) | 0.80 |
| Discharge after RA, days | 2.8 (1–3) | 3 (1–5) | 0.01 |
Data are presented as numbers and percentages for categorical variables, mean±standard deviation for continuous variables with normal distribution and median with interquartile range for continuous variables with skewed distribution. Cx, circumflex artery; LAD, left anterior descending; LM, left main; MCS, mechanical circulatory support; RA, rotational atherectomy; RCA, right coronary artery. Other abbreviations as in Table 1.
| Angle | Successful RA (n=458) |
RA clinical failure (n=76) |
P value |
|---|---|---|---|
| Pre lesion | |||
| ≤90° | 14 (3) | 6 (8) | 0.04 |
| 91°–135° | 57 (13) | 12 (16) | 0.42 |
| 136°–180° | 387 (85) | 58 (76) | 0.08 |
| On lesion | |||
| ≤90° | 69 (15) | 27 (36) | <0.0001 |
| 91°–135° | 129 (28) | 25 (33) | 0.40 |
| 136°–180° | 260 (57) | 24 (32) | <0.0001 |
| Post lesion | |||
| ≤90° | 11 (2) | 2 (3) | 0.90 |
| 91°–135° | 51 (11) | 13 (17) | 0.14 |
| 136°–180° | 396 (87) | 61 (80) | 0.15 |
Data are presented as numbers and percentages for categorical variables. Abbreviations as in Table 2.
When RA was successful, more burrs (1.2 vs. 1.0, P<0.001) as well as more stents (1.5 vs. 1.0, P<0.001) were used and hospital stay after the procedure was shorter (2.8 days vs. 3 days, P=0.01). There were no differences in procedure and fluoroscopy time, radiation exposure, total contrast volume and other procedural characteristics, except for more frequent use of radial access (76% vs. 63%, P=0.02) and mechanical circulatory support (MCS) (7% vs. 1%, P<0.01) in the RA failure group. Indication for MCS was elective in 35% of patients with high-risk and very low LVEF and rescue in remaining patients in case of hemodynamic deterioration during PCI. Intra-aortic balloon pump, Impella system or extracorporeal membranous oxygenation was used according to clinical presentation.
Periprocedural OutcomesRA failure reasons and periprocedural outcomes are presented in Table 4. Primary composite endpoint occurred in 76 (14.2%) patients. Second primary endpoint occurred in 64 (12%) patients including rota-wire introduction failure (2.4%) and burr-passage failure (9.6%). Periprocedural complications occurred in 23 (4.3%) and procedure-related adverse events in 23 (4.3%) patients.
| RA clinical failure (n=76) |
|
|---|---|
| Rota-wire failure | 13 (2.4) |
| Burr-passage failure | 51 (9.6) |
| Periprocedural complications | |
| Slow/no-flow phenomenon | 10 (1.9) |
| Side branch occlusion | 6 (1.1) |
| Perforation | 6 (1.1) |
| Emergency CABG | 0 (0) |
| Permanent pacing | 1 (0.2) |
| Post procedure 24-h outcomes | |
| Death | 6 (1.1) |
| PCI-related MI | 15 (2.8) |
| Stroke | 0 (0) |
| Target vessel revascularization | 2 (0.4) |
Data are presented as numbers and percentages for categorical variables. MI, myocardial infarction. Other abbreviations as in Tables 1,2.
Univariable analysis identified the following determinants of RA clinical failure: high Syntax Score ≥33 (P=0.04), angulation on lesion ≤90° (P<0.0001), severe calcifications (P=0.02), calcium length >20 mm (P=0.01), lesion length >20 mm (P=0.01), presence of chronic total occlusion (CTO) (P=0.01) and sequential lesion (P<0.0001). Multivariable analysis revealed angulation on lesion ≤90° (HR=2.18, 95% CI: 1.21–3.94, P<0.01) and sequential lesion (HR=1.89, 95% CI: 1.01–3.54, P<0.05) as independent predictors of no clinical success of RA (Table 5, Figure 2).
| Univariate model | Multivariate model | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P value | OR | 95% CI | P value | |
| Composite endpoint | ||||||
| High Syntax Score (≥33) | 2.07 | 1.03–4.18 | 0.04 | 1.31 | 0.59–2.93 | 0.51 |
| Angulation on lesion ≤90° | 3.11 | 1.82–5.31 | <0.0001 | 2.18 | 1.21–3.94 | <0.01 |
| Severe calcifications | 2.69 | 1.19–6.04 | 0.02 | 1.87 | 0.76–4.61 | 0.17 |
| Calcium length >20 mm | 2.02 | 1.17–3.48 | 0.01 | 1.40 | 0.74–2.27 | 0.30 |
| Lesion length >20 mm | 2.33 | 1.22–4.49 | 0.01 | 1.36 | 0.66–2.81 | 0.40 |
| Chronic total occlusion | 2.47 | 1.21–5.06 | 0.01 | 1.77 | 0.80–3.89 | 0.15 |
| Sequential lesion | 2.93 | 1.71–5.02 | <0.0001 | 1.89 | 1.01–3.54 | <0.05 |
| Endpoint consisted of rota-wire introduction and burr-passage failure | ||||||
| High Syntax Score (≥33) | 2.00 | 0.88–4.56 | 0.10 | – | – | – |
| Angulation on lesion ≤90° | 3.94 | 2.14–7.3 | <0.0001 | 2.26 | 1.16–4.40 | 0.02 |
| Severe calcifications | 2.99 | 1.05–8.54 | 0.04 | 2.13 | 0.72–6.27 | 0.17 |
| Calcium length >20 mm | 2.87 | 1.40–5.87 | <0.01 | 1.74 | 0.80–3.73 | 0.17 |
| Lesion length >20 mm | 1.37 | 0.67–2.78 | 0.39 | – | – | – |
| Chronic total occlusion | 2.79 | 1.25–6.21 | 0.01 | 1.81 | 0.76–4.32 | 0.17 |
| Sequential lesion | 5.84 | 2.68–12.73 | <0.0001 | 3.77 | 1.64–8.69 | <0.01 |
| Endpoint consisted of periprocedural complications and adverse events | ||||||
| High Syntax Score (≥33) | 1.08 | 0.37–3.17 | 0.88 | – | – | – |
| Angulation on lesion ≤90° | 2.09 | 1.02–4.29 | 0.04 | 1.58 | 0.75–3.34 | 0.23 |
| Severe calcifications | 1.75 | 0.67–4.6 | 0.25 | – | – | – |
| Calcium length >20 mm | 1.24 | 0.63–2.44 | 0.53 | – | – | – |
| Lesion length >20 mm | 5.8 | 1.76–19.19 | <0.01 | 4.90 | 1.43–16.74 | 0.01 |
| Chronic total occlusion | 1.25 | 0.42–3.70 | 0.69 | – | – | – |
| Sequential lesion | 2.13 | 1.07–4.23 | 0.03 | 1.31 | 0.63–2.74 | 0.47 |
CI, confidence interval; OR, odds ratio.
Predictors of rotational atherectomy clinical failure in multivariable analysis.
Determinants of rota-wire introduction and burr-passage failure in univariable analysis were: angulation on lesion ≤90° (P<0.0001), severe calcifications (P=0.04), calcium length >20 mm (P<0.01), presence of chronic total occlusion (CTO) (P=0.01) and sequential lesion (P<0.0001). Multivariable analysis revealed again angulation on lesion ≤90° (HR=2.26, 95% CI: 1.16–4.40, P=0.02) and sequential lesion (HR=3.77, 95% CI: 1.64–8.69, P<0.01) as independent predictors of no procedural success of RA (Table 5, Figure 3). The independent determinant of periprocedural complications and adverse events in multivariable analysis was lesion length (HR 4.90, 95% CI: 1.43–16.74, P=0.01).
Predictors of rotational atherectomy procedural failure in multivariable analysis.
In our study, we analyzed a high-risk, all-comers population of patients who required PCI with RA, exploring the reasons and predictors of procedural and clinical failure. Compared to other studies, our analysis uniquely focused on anatomical and procedural determinants of not only adverse events, but the lack of RA success in general. The novel findings of our study are: (1) the prevalence of RA clinical failure is noticeable, reaching 14.2% and procedural failure in 12%; and (2) severe acute angulation on lesion ≤90° and sequential character of lesion are independent predictors of no clinical and procedural success of RA.
The studied groups were clinically similar and had high cardiovascular risk associated with advanced age (mean 72 years) and numerous comorbidities, irrespective of final RA success. Baseline risk scores were relatively high: median logistic Euroscore II was 2.6 and Syntax Score was 17.5 with higher Syntax Score (19.5) in the RA failure group. Earlier publications investigated a moderate-risk population undergoing RA, but recent studies analyzed high-risk patients as well, with an estimated Euroscore II of 2.1 and a Syntax Score of 19.5.8,9,14 The frequency of acute coronary syndromes in recent studies reached 20–40%, which is similar to our study population (32%) and it was not associated with worse outcome. The distribution of target lesions in a coronary tree was equal in both study groups, including a left main coronary artery. It is noteworthy that a radial approach was used in 65% of our patients, compared to 30–50% in aforementioned studies, albeit with a more frequent use when RA failed. The median burr-to-artery ratio was relatively low (0.44), which was proven to be effective enough for plaque modification according to the expert consensuses on RA.15 One-third of our patients were disqualified from CABG treatment by a local Heart Team, meaning that we analyzed a real-world, high-risk group of patients, with frequent RA use as the last revascularization option. Nonetheless, despite an additional inclusion to our study group also patients in whom a successful introduction of rota-wire or burr across the lesion failed, the procedural and clinical success was 85.8%. The periprocedural complication rate was 4.3% and mortality 1.1%, which is comparable to other research results.16–19
Most of the previous studies focused mainly on clinical, not procedural and lesion anatomy-related predictors of RA outcome. Decreased LVEF, age, chronic kidney disease, dialysis, diabetes mellitus and peripheral vascular disease are well-documented clinical predictors not only of RA failure but also of PCI failure in general.9,10,14,20 Bouisset et al, in a recent large multicenter registry, demonstrated also that the presence of a significant left main coronary artery lesion, female gender and acute coronary syndrome at admission are independent predictors of complications, but in a 1-year follow up.21 Notwithstanding, there are only a few studies concerning some anatomical or procedure-related factors that are associated with procedural success or adverse events during or immediately post RA. Therefore, we decided to analyze, in our database, a more detailed anatomy of the artery and the lesion, including precise angulation assessment, as presented in Table 3.
Still, at the end of the 20th century, even before the stent era, it was pointed out that RA failure was correlated independently with an outflow obstruction, lesion irregularity, stenosis bend and female sex.22 Another study revealed that complex lesions, such as those that are eccentric, long, and calcified are associated with a vessel dissection during RA and procedural failure.23 And until recently, in one study, in-hospital adverse outcomes were more frequent in patients with tortuous target vessels and lesions >20 mm, as indicated by a high target vessel Syntax Score, along with bailout RA and reduced LVEF.24 In contrast, treating long lesions >25 mm with RA did not impact in-hospital outcome, and long-term outcomes in a multicenter registry.25 In another publication, patients with high Syntax scores were more likely to suffer peri-procedural complications.8 Procedure-related complication rates were also determined by emergency procedures and triple-vessel disease (vs. single-vessel).18 Abovementioned scores, however, did not correlate with clinical and procedural failure of RA in our multivariable analysis. Lesion length correlated only with periprocedural complications alone, but not with RA success in general.
We can find some similarities of tight, uncrossable highly fibro-calcified lesions with typical CTOs, although in the majority of RA cases, an antegrade flow is initially maintained. In CTO procedures, lesion tortuosity is a predictor of PCI failure.26,27 Severe vessel angulation makes the whole CTO procedure much more complex, irrespective of RA usage. One of the principles of RA is differential displacement of friction, allowing the burr to be advanced easily and atraumatically through tortuous segments. Conceivably, bends of the rota-wire can reduce the ability of the burr to track “naturally” toward the less compliant atherosclerotic lesion, forfeiting the ability to abrade diseased rather than normal vessel differentially and thereby inducing dissection or perforation. Tortuosity is one of the variables of the widely used Syntax Score, which determines the complexity of coronary artery disease, in general. One significant bend or a sum of 3 mild bends are defined as severe tortuosity in this score. We have proven that a severe angulation on lesion ≤90° is independently associated also with RA failure. Severe bends make the passage of different interventional tools, especially the rotablator system, much more difficult. Interestingly, acute angulation proximal or distal to a target lesion did not influence the outcome. Proximal tortuosity of the vessel could be a problem for wire and burr passage towards a culprit lesion, and distal bends could make correct wire placement difficult; however, it seems that the most important characteristic is tortuosity of tight calcified culprit stenosis.
To the best of our knowledge, this is the first study showing the presence of sequential coronary stenoses as an independent predictor of the lack of procedural and clinical success of RA. In a majority of our rota-wire and burr-passage failure cases, the sequential lesion was present, indicating that the passage through such stenoses is far more challenging. This may also be related to lesion length; as shown in the abovementioned CTO studies, a length >20 mm was an independent predictor of procedural failure. Such anatomy could be a reason of pressure variation of the burr on the vessel, resulting in greater endothelium damage and thus a greater risk of distal embolization. It is also unclear how proximal stenosis reacts to many polishing runs during distal stenosis atherectomy. Additionally, blood located between stenoses may change its features, leading to increased inflammation reaction and interference in coagulation balance during the procedure. In our analysis, lesion length was associated only with periprocedural complications, whereas the anatomy corresponding to a sequential lesion was independently associated with clinical and procedural failure of RA.
Recently, the role of intravascular imaging-guided PCI is significantly growing. Intravascular ultrasonography (IVUS) and optical coherence tomography (OCT) guidance enhances not only the acute procedural result, but also improves clinical outcomes. One OCT-based study suggested that lesions with calcium pools with a maximum angle >180°, maximum thickness >0.5 mm, and length >5 mm are at increased risk for stent under-expansion.28 An additional issue is the presence of a calcified nodule (CN), best visible during OCT, which can pose significant challenges for stent deployment and optimization. The use of RA could not improve the outcome of a CN after PCI and it can result in unfavorable clinical outcomes.29,30 Although coronary angiography shows the amount of calcifications and CNs only semi-quantitative, our findings could be helpful in the initial assessment before PCI, especially when intracoronary imaging is unavailable or unable to be performed, such as is the case for tortuous tight lesions.
Therefore, we conclude that apart from general and clinical-related factors, coronary artery anatomy and RA procedure-related factors can constitute a better risk stratification in this specific high-risk population undergoing RA. The presence of anatomical factors like acute angulation on a lesion and sequential stenosis, which can be indicated before the RA procedure, can help in identifying a higher risk sub-group, with the potential necessity of more advanced treatment. Further studies are required to confirm the role of the aforementioned risk factors and to create a more standardized risk stratification protocol for patients qualifying for RA, like it was done for CTO lesions.26,27 A score identifying a risk of RA failure would be helpful also for low-volume centers, facilitating the cooperation with tertiary centers with greater expertise in the treatment of more complex cases.
Study LimitationsThis was an observational study, without the use of intravascular imaging like IVUS or OCT in a meaningful group of patients. The relationship between angiographical heavy calcium, intravascular imaging and procedural outcome is of great practical importance and we hope to explore this topic in future research. A core lab analysis of coronary angiographies was not performed. The inclusion period was relatively long (∼10 years), so the increasing experience of the operators could influence the complications rate.
The presence of acute angulation on lesions and sequential lesions are independent determinants of clinical and procedural failure of RA. Further research is necessary to establish a score predicting the RA failure, which can help in a preprocedural risk stratification of patients undergoing complex PCI with RA.
The publication costs were financially supported from funding from the Department of Heart Diseases, Wroclaw Medical University, Poland (No. SUB.E190.19.052).
The authors declare that there are no conflicts of interest.
This study was approved by the Bioethics Committee of Wroclaw Medical University (approval number KB 143/2016).
