2023 Volume 30 Issue 9 Pages 1229-1241
Aims: Complex and high-risk intervention in indicated patients (CHIP) is an emerging concept in the contemporary percutaneous coronary intervention (PCI). CHIP is known to consist three factors, namely, (1) patient factors, (2) complicated heart disease, and (3) complex PCI. However, it remains unclear whether additional CHIP factors further increase the incidence of complications in complex PCI. Thus, in this study, we aim to compare the incidence of complications among definite CHIP, possible CHIP, and non-CHIP in terms of complex PCI and to further investigate the association between CHIP and complications.
Methods: The primary aim of this study was to determine the major complications in PCI. We included 989 PCI lesions and divided those into definite CHIP (n=140), possible CHIP (n=397), and the non-CHIP groups (n=452).
Results: The incidence of major complications was noted to be the highest in the definite CHIP, followed by the possible CHIP, and lowest in the non-CHIP (p=0.001). The multivariate logistic regression analysis using a generalized estimating equation revealed definite CHIP (versus non-CHIP: odds ratio (OR) 2.099, 95% confidence interval (CI) 1.062–4.150, p=0.033) was significantly associated with major complications after controlling for confounding factors. Another multivariate logistic regression analysis revealed immunosuppressive drugs (OR 3.040, 95% CI 1.251–7.386, p=0.014), unstable hemodynamics (OR 5.753, 95% CI 1.217–27.201, p=0.027), and frailty (OR 2.039, 95% CI 1.108–3.751, p=0.022) were significantly associated with major complications among CHIP factors.
Conclusions: The incidence of major complications in complex PCI was determined to be the highest in the definite CHIP, followed by the possible CHIP and lowest in the non-CHIP. Thus, more attention should be given to the three components of CHIP to prevent major complications in complex PCI.
Coronary artery disease (CAD) has been identified as the leading cause of morbidity and mortality in developed countries1-3). Despite the availability of guideline-directed medical therapy (GDMT), a significant number of patients still require coronary revascularization including percutaneous coronary intervention (PCI)4-6). In the beginning of PCI, non-complex coronary lesions were often the main target, and complex coronary lesions were often referred to coronary artery bypass surgery (CABG)7-8). However, the emergence of drug-eluting stent and the development of PCI devices were able to expand the indications of PCI to complex coronary lesions such as left main disease, triple vessel disease, severely calcified lesions, and chronic total occlusion (CTO)9-15). Complex and high-risk intervention in indicated patients (CHIP) is an emerging concept in contemporary PCI16). CHIP is known to be composed of three factors, namely, (1) patient factors such as frailty or active malignancy, (2) complicated heart disease such as left ventricular dysfunction or severe valvular disease, and (3) complex PCI such as left main disease or CTO16-17). Therefore, CHIP-percutaneous coronary intervention (CHIP-PCI) is seen to have more complications as compared to non-CHIP-PCI18). However, as complex PCI itself has been reported to have more complications than non-complex PCI, it remains unclear whether the addition of CHIP factors (patient factors and/or complicated heart disease) further increase the incidence of complications in complex PCI19-20). Thus, in this study, we aim to compare the incidence of complications among definite CHIP, possible CHIP, and non-CHIP groups in complex PCI and to further investigate the association between CHIP and complications in complex PCI.
We reviewed all PCI lesions treated at our institution (Saitama Medical Center, Jichi Medical University) between January 2017 and December 2021. The inclusion criteria were PCI lesions during the study period. Meanwhile, the exclusion criteria were as follows: (1) non-complex PCI lesions, (2) emergent or urgent PCI, and (3) PCI for patients who did not undergo echocardiogram at our institution. Complex PCI was defined as (1) PCI to chronic total occlusion (CTO), (2) PCI to left main trunk, (3) PCI using rotational atherectomy, (4) PCI using orbital atherectomy, and (5) PCI for three-vessel disease with ≥ 90% stenosis or proven ischemia. When three-vessel disease was included as complex PCI, only the PCI for the first vessel was included.
Thereafter, the final study population was further divided into three groups, namely, definite CHIP group, possible CHIP group, and non-CHIP group according to the following definitions: patient factors were defined as (1) frailty, (2) active malignancy, (3) use of immunosuppressive drugs including corticosteroids, (4) pulmonary disease requiring inhalants, (5) liver cirrhosis, (6) chronic renal failure on hemodialysis, and (7) history of cerebral infarction. Complicated heart disease was defined as (1) unstable hemodynamics, (2) left ventricular ejection fraction (LVEF) ＜40%, and (3) moderate or severe valvular disease. Based on the above definition, definite CHIP was defined as complex PCI with both patient factors and complicated heart disease, whereas possible CHIP was defined as complex PCI with either patient factors or complicated heart disease. Lastly, non-CHIP was defined as complex PCI with neither patient factors nor complicated heart disease. The primary aim of this study was to determine the major complications in PCI, which were defined as the composite of (1) in-hospital death, (2) cardiac tamponade requiring pericardial drainage, (3) emergent surgery, (4) stroke within 48 hours from PCI, (5) emergent use of mechanical circulatory support during PCI procedure, (6) periprocedural myocardial infarction (PMI), (7) puncture site complications requiring blood transfusion, (8) acute heart failure during PCI procedure, and (9) unsuccessful PCI to non-CTO lesions. Information regarding the above clinical outcomes were acquired from the hospital records. This study was approved by the Institutional Review Board of the Saitama Medical Center, Jichi Medical University (S22-008), and written informed consent was waived because of the retrospective design of this study.
DefinitionsIn our institution, the patients and/or their representatives were routinely interviewed as regards pre-admission activities of daily living (ADL): having meals, using toilets, maintaining cleanliness, washing face, changing clothes, and moving around. These six variables, which reflect real ADL in Japan21), were evaluated as either independent or dependent at admission. If at least one variable was judged as dependent, the patient was defined as frail. Active malignancy was defined according to the definition proposed by the National Institute for Health and Care Excellence (NICE), i.e., receiving active antimitotic treatment, diagnosed within the past 6 months, recurrent or metastatic, or inoperable22). Patients with a history of cirrhosis were identified from hospital records. Unstable hemodynamics was defined as systolic blood pressure ＜90 mmHg at admission, requiring catecholamines to maintain blood pressure, using mechanical support to maintain circulation, or using mechanical ventilation irrespective of respiratory failure. Stroke was defined as an acute episode of focal or global neurological dysfunction caused by brain, spinal cord, or retinal vascular injury due to hemorrhage or infarction23). Emergent use of mechanical circulatory support was defined as the initiation of intra-aortic balloon pumping, veno-arterial extracorporeal membrane oxygenation (VA-ECMO), and Impella during PCI. In patients whose baseline CK levels were normal, PMI was defined as an elevation of CK levels twice of the upper limit of normal (ULN) with an elevation of CK-MB levels above the ULN at the next day of PCI24-25). If baseline CK levels were already elevated, PMI was defined as the further increase of CK levels at the next day of PCI than CK level at baseline26). LVEF and valvular disease were measured via transthoracic echocardiography. LVEF was calculated by either using a modified Simpson’s rule, Teichholz method, or eyeball estimation27). Teichholz method was only adopted when a modified Simpson method is not available27); meanwhile, an eyeball estimation was only adopted when both modified Simpson method and Teichholz method are not available. Acute heart failure during PCI was defined as conditions that required the initiation of dobutamine or mechanical ventilator including noninvasive positive pressure ventilation during PCI. Hypertension was defined as systolic blood pressure ＞140 mmHg, diastolic blood pressure ＞90 mmHg, or medical treatment for hypertension28). Dyslipidemia was defined as a total cholesterol level ≥ 220 mg/dL, low-density lipoprotein cholesterol level ≥ 140 mg/dL, medical treatment for dyslipidemia, or a history of dyslipidemia28). Diabetes mellitus was defined as hemoglobin A1c ≥ 6.5% or treatment for diabetes mellitus28). AMI was defined according to the universal definition29). Diagnostic ST elevation was defined as the new ST elevation at the J point in at least two contiguous leads of 2 mm (0.2 mV), and the AMI patients with ST elevation were diagnosed with ST-segment elevation myocardial infarction (STEMI)30). We have also calculated estimated glomerular filtration rate (eGFR) using serum creatinine (Cr), age, weight, and gender according to the following formula: eGFR=194×Cr−1.094×age−0.287 (male), or eGFR=194×Cr−1.094×age−0.287×0.739 (female)31).
Statistical AnalysisData were presented as a percentage for categorical variables, a mean±standard deviation (SD) for normally distributed continuous variables, and median (quartile 1-quartile 3) for nonparametric variables. Categorical variables were presented as numbers (percentage), and these were compared using chi-square test. The Shapiro–Wilk test was performed to determine whether the continuous variables were normally distributed or not. Normally distributed continuous variables were compared using one-way ANOVA. Otherwise, continuous variables were compared using a Kruskal–Wallis test. This study population was not a patient-level basis, but a lesion-level basis, which included 816 patients and 989 PCI lesions. We were able to perform two models of multivariate logistic regression analysis using a generalized estimating equation to adjust within-patient clustering. In the first multivariate logistic regression analysis using a generalized estimating equation, the selection of parameters was derived from the results of comparing the clinical characteristics between patients with and without major complications (p-value ＜0.05). We did not include variables with substantial missing values. In the second multivariate logistic regression analysis using a generalized estimating equation, the parameters were CHIP components. The odds ratio (OR) and the 95% confidence interval (CI) were calculated. P-value of ＜0.05 was considered statistically significant. All analyses were performed using the Statistical Package for the Social Sciences (SPSS 25/Windows) (SPSS, Chicago Illinois).
In total, 4024 PCI lesions were performed in our medical center from January 2017 to December 2021. After excluding PCI lesions as per the exclusion criteria, the final study population consisted of 989 complex PCI lesions, and these were divided into definite CHIP group (n=140), the possible CHIP group (n =397), and the non-CHIP group (n =452) (Fig.1).
Abbreviations: CHIP=complex and high-risk intervention in indicated patients, CTO=chronic total occlusion, EF=ejection fraction, LMT=left main trunk, PCI=percutaneous coronary intervention, TTE=transthoracic echocardiography.
The comparison of patient characteristics between the three groups is shown in Table 1. The prevalence of peripheral vascular disease, heart failure, and previous myocardial infarction was found to be the highest in the definite CHIP group, followed by possible CHIP group and lowest in the non-CHIP group. The prevalence of PCI for culprit lesion of NSTEMI, cardiopulmonary arrest out of hospital, the use of diuretics, and the use of insulin were highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group. The level of serum creatinine, C-reactive protein, and brain natriuretic peptide was noted to be the highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group.
| Definite CHIP (n=140) | Possible CHIP (n=397) | Non-CHIP (n=452) | P value | |
|---|---|---|---|---|
| Age, years | 76 (70-81) | 75 (67-80) | 72 (66-78) | ＜0.001 |
| Male, n (%) | 109 (77.9) | 311 (78.3) | 356 (78.8) | 0.971 |
| Body mass index (kg/m2) | 22.4 (20.9-25.0) (n=139) | 23.6 (21.6-25.8) | 24.2 (22.2-26.7) | ＜0.001 |
| Comorbidities | ||||
| Hypertension, n (%) | 108 (77.1) | 324 (81.6) | 371 (82.1) | 0.410 |
| Hyperlipidemia, n (%) | 107 (76.4) | 331 (83.4) | 411 (90.9) | ＜0.001 |
| Diabetes mellitus, n (%) | 77 (55.0) | 211 (53.1) | 255 (56.4) | 0.634 |
| Current smoker, n (%) | 30 (21.7) (n=138) | 99 (25.2) (n=393) | 117 (26.0) (n=450) | 0.599 |
| Frailty, n (%) | 51 (36.4) | 61 (15.4) | 0 (0.0) | ＜0.001 |
| Chronic renal failure on hemodialysis, n (%) | 61 (43.6) | 102 (25.7) | 0 (0.0) | ＜0.001 |
| Active malignancy, n (%) | 12 (8.6) | 25 (6.3) | 0 (0.0) | ＜0.001 |
| Taking immunosuppressive drugs, n (%) | 13 (9.3) | 25 (6.3) | 0 (0.0) | ＜0.001 |
| Pulmonary disease requiring inhalants, n (%) | 15 (10.7) | 17 (4.3) | 0 (0.0) | ＜0.001 |
| Cirrhosis of liver, n (%) | 1 (0.7) | 10 (2.5) | 0 (0.0) | 0.002 |
| History of cerebral infarction, n (%) | 43 (30.7) | 77 (19.4) | 0 (0.0) | ＜0.001 |
| History of peripheral vascular disease, n (%) | 64 (45.7) | 120 (30.2) | 88 (19.5) | ＜0.001 |
| History of heart failure, n (%) | 89 (63.6) | 130 (32.7) | 36 (8.0) | ＜0.001 |
| History of previous PCI, n (%) | 72 (51.4) | 218 (54.9) | 261 (57.7) | 0.387 |
| History of previous CABG, n (%) | 7 (5.0) | 28 (7.1) | 20 (4.4) | 0.237 |
| History of previous myocardial infarction, n (%) | 55 (39.3) | 155 (39.0) | 120 (26.5) | ＜0.001 |
| Reason for PCI | ＜0.001 | |||
| Culprit lesion of STEMI, n (%) | 4 (2.9) | 9 (2.3) | 4 (0.9) | |
| Culprit lesion of NSTEMI, n (%) | 54 (38.6) | 78 (19.6) | 35 (7.7) | |
| Culprit lesion of non-AMI stenosis, n (%) | 82 (58.6) | 310 (78.1) | 413 (91.4) | |
| Cardiopulmonary arrest out of hospital, n (%) | 4 (2.9) | 4 (1.0) | 0 (0.0) | 0.004 |
| Vital signs | ||||
| Systolic blood pressure at admission, mmHg | 126 (108.5-145) | 132 (116-150) | 128 (117-142) | 0.003 |
| Diastolic blood pressure at admission, mmHg | 69 (60-81) | 72 (62-83) | 72 (64-80) | 0.380 |
| Heart rate at admission, bpm | 81.5 (70.5-96) | 70 (62-82) | 68 (61-77) | ＜0.001 |
| Laboratory data | ||||
| Hemoglobin levels, g/dL | 11.4 (10.5-12.7) | 12.3 (11.0-13.7) | 13.3 (12.3-14.4) | ＜0.001 |
| Platelets, ×103/uL | 18.4 (14.5-24.5) | 19.8 (16.1-24.4) | 20.3 (16.9-24.2) | 0.031 |
| Serum creatinine, mg/dL | 1.86 (0.93-6.49) | 1.06 (0.81-4.23) | 0.82 (0.70-1.03) | ＜0.001 |
| eGFR, mL/min/1.73 m2 | 26.4 (7.25-58.7) | 51.0 (12.0-68.8) | 67.1 (53.4-81.5) | ＜0.001 |
| Hemoglobin A1c, % | 6.3 (5.8-7.4) | 6.3 (5.8-7.1) | 6.5 (6.0-7.1) | 0.014 |
| C-reactive protein, mg/dL | 0.43 (0.14-1.72) | 0.18 (0.08-0.54) | 0.10 (0.05-0.21) | ＜0.001 |
| Brain natriuretic peptide, pg/ml | 785.8 (234.3-2011.9) | 198.7 (79.0-443.4) (n=393) | 49.2 (25.8-112.6) | ＜0.001 |
| Creatine kinase, U/L | 73.0 (50.0-136.5) | 91.0 (62.0-136.0) | 93.0 (72.0-136.0) | 0.004 |
| Left ventricular ejection fraction, % | 35.5 (26.8-44.1) | 53.9 (38.6-63.0) | 62.5 (55.1-67.3) | ＜0.001 |
| Valvular disease, n (%) | 71 (50.7) | 56 (14.1) | 0 (0.0) | ＜0.001 |
| moderate to severe AS, n (%) | 31 (22.1) | 23 (5.8) | 0 (0.0) | ＜0.001 |
| moderate to severe AR, n (%) | 18 (12.9) | 18 (4.5) | 0 (0.0) | ＜0.001 |
| moderate to severe MS, n (%) | 0 (0.0) | 0 (0.0) | 0 (0.0) | - |
| moderate to severe MR, n (%) | 34 (24.3) | 24 (6.0) | 0 (0.0) | ＜0.001 |
| Medication at admission | ||||
| Aspirin, n (%) | 140 (100) | 397 (100) | 451 (99.8) | 0.552 |
| Thienopyridine, n (%) | 140 (100) | 397 (100) | 452 (100) | - |
| Statin, n (%) | 129 (92.1) | 368 (92.7) | 441 (97.6) | 0.002 |
| ACE inhibitors or ARBs, n (%) | 87 (62.1) | 275 (69.3) | 304 (67.3) | 0.302 |
| Beta-blockers, n (%) | 97 (69.3) | 300 (75.6) | 347 (76.8) | 0.197 |
| Calcium channel blocker, n (%) | 57 (40.7) | 164 (41.3) | 194 (42.9) | 0.848 |
| Diuretics, n (%) | 68 (48.6) | 147 (37.0) | 69 (15.3) | ＜0.001 |
| Oral antidiabetic, n (%) | 59 (42.1) | 152 (38.3) | 186 (41.2) | 0.609 |
| Insulin, n (%) | 26 (18.6) | 62 (15.6) | 44 (9.7) | 0.006 |
| Direct oral anticoagulants., n (%) | 7 (5.0) | 24 (6.0) | 30 (6.6) | 0.774 |
| Warfarin, n (%) | 3 (2.1) | 11 (2.8) | 11 (2.4) | 0.907 |
Data were presented as a percentage for categorical variables, a mean±standard deviation (SD) for normally distributed continuous variables, and median (quartile 1–quartile 3) for nonparametric variables. One way ANOVA was used for normally distributed continuous variables and a Kruskal Wallis test was used for abnormally distributed continuous variables. A Chi-square test was used for categorical variables. Abbreviations: ACE inhibitors = angiotensin- converting enzyme inhibitor, AMI = Acute Myocardial Infarction, AR = Aortic Regurgitation, ARBs = angiotensin receptor blockers, AS = Aortic Stenosis, CABG = coronary artery-bypass grafting, eGFR = estimated glomerular filtration rate, MR = Mitral Regurgitation, MS = Mitral stenosis, NSTEMI = non- ST-segment elevation myocardial infarction, PCI = percutaneous coronary intervention, STEMI = ST-segment elevation myocardial infarction.
Table 2 shows the comparison of lesion and procedural findings between the three groups. The prevalence of PCI to CTO and trans-radial approach was determined to be the highest in the non-CHIP group, followed by the possible CHIP group and lowest in the definite CHIP group. The prevalence of using rotational atherectomy and scheduled mechanical support was highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group.
| Definite CHIP (n = 140) | Possible CHIP (n = 397) | Non-CHIP (n = 452) | P value | |
|---|---|---|---|---|
| PCI to chronic total occlusion, n (%) | 26 (18.6) | 118 (29.7) | 157 (34.7) | 0.001 |
| PCI to left main trunk, n (%) | 17 (12.1) | 53 (13.4) | 71 (15.7) | 0.459 |
| Use of rotational atherectomy, n (%) | 86 (61.4) | 191 (48.1) | 171 (37.8) | ＜0.001 |
| Use of orbital atherectomy, n (%) | 2 (1.4) | 6 (1.5) | 12 (2.7) | 0.431 |
| PCI for three vessel disease with ≥ 90% stenosis or proven ischemia, n (%) | 30 (21.4) | 76 (19.1) | 71 (15.7) | 0.214 |
| Number of narrowed coronary arteries (≥ 75% stenosis) | 0.197 | |||
| Single, n (%) | 80 (57.1) | 254 (64.0) | 306 (67.7) | |
| Double, n (%) | 26 (18.6) | 57 (14.4) | 65 (14.4) | |
| Triple, n (%) | 34 (24.3) | 86 (21.7) | 81 (17.9) | |
| ≥ 50% stenosis at left main, n (%) | 19 (13.6) | 56 (14.1) | 77 (17.0) | 0.407 |
| First TIMI flow (0,1,2,3) | 0.004 | |||
| 0, n (%) | 26 (18.6) | 115 (11.6) | 156 (34.5) | |
| 1, n (%) | 3 (2.1) | 10 (2.5) | 6 (1.3) | |
| 2, n (%) | 14 (10.0) | 18 (4.5) | 22 (4.9) | |
| 3, n (%) | 97 (69.3) | 254 (64.0) | 268 (59.3) | |
| Final TIMI flow (0,1,2,3) | 0.672 | |||
| 0, n (%) | 3 (2.1) | 8 (2.0) | 9 (2.0) | |
| 1, n (%) | 0 (0.0) | 0 (0.0) | 3 (0.7) | |
| 2, n (%) | 1 (0.7) | 2 (0.5) | 4 (0.9) | |
| 3, n (%) | 136 (97.1) | 387 (97.5) | 436 (96.5) | |
| Use of aspiration catheter, n (%) | 0 (0.0) | 2 (0.5) | 1 (0.2) | 0.590 |
| Final PCI Procedure | 0.180 | |||
| POBA only, n (%) | 0 (0.0) | 2 (0.5) | 0 (0.0) | |
| Drug coated balloon, n (%) | 21 (15.0) | 48 (12.1) | 45 (10.0) | |
| Bare metal stent, n (%) | 2 (1.4) | 1 (0.3) | 0 (0.0) | |
| Drug eluting stent, n (%) | 114 (81.4) | 337 (84.9) | 398 (88.1) | |
| POBA and aspiration, n (%) | 0 (0.0) | 1 (0.3) | 1 (0.2) | |
| No wire cross, n (%) | 3 (1.5) | 8 (2.0) | 8 (1.8) | |
| Approach site | ＜0.001 | |||
| Radial, n (%) | 40 (28.6) | 136 (34.3) | 223 (49.3) | |
| Brachial, n (%) | 8 (5.7) | 12 (3.0) | 1 (0.2) | |
| Femoral, n (%) | 92 (65.7) | 249 (62.7) | 228 (50.4) | |
| Guide-Catheter size (Fr) | 0.243 | |||
| 6Fr, n (%) | 19 (13.6) | 40 (10.1) | 56 (12.4) | |
| 7Fr, n (%) | 105 (75.0) | 285 (71.8) | 312 (69.0) | |
| 8Fr, n (%) | 16 (11.4) | 72 (18.1) | 84 (18.6) | |
| Dual artery access, n (%) | 53 (37.9) | 163 (41.1) | 202 (44.7) | 0.295 |
| Unstable hemodynamics, n (%) | 6 (4.3) | 6 (1.5) | 0 (0.0) | ＜0.001 |
| IABP, n (%) | 5 (3.6) | 4 (1.0) | 0 (0.0) | 0.001 |
| VA-ECMO, n (%) | 1 (0.7) | 1 (0.3) | 0 (0.0) | 0.249 |
| Impella, n (%) | 0 (0.0) | 0 (0.0) | 0 (0.0) | - |
| Catecholamine, n (%) | 4 (2.9) | 1 (0.3) | 0 (0.0) | ＜0.001 |
| Ventilator, n (%) | 4 (2.9) | 2 (0.5) | 0 (0.0) | 0.001 |
| Scheduled IABP or Impella, n (%) | 8 (5.7) | 15 (3.8) | 5 (0.5) | 0.005 |
| Scheduled Impella, n (%) | 1 (0.7) | 0 (0.0) | 0 (0.0) |
Data were expressed as numbers (percentages). A Chi-square test was used for categorical variables. Abbreviations: CTO = chronic total occlusion, IABP = Intra-aortic balloon pumping, PCI = percutaneous coronary intervention, POBA = plain old balloon angioplasty, TIMI = thrombolysis in myocardial infarction, VA-ECMO = veno-arterial extracorporeal membrane oxygenation.
Table 3 shows the comparison of clinical outcomes between the three groups. The incidence of major complications, in-hospital death, and stroke was determined to be the highest in the definite CHIP group, followed by the possible CHIP group, and lowest in the non-CHIP group. Based on the comparison of clinical characteristics between patients with and without major complications (Supplementary Table 1), the first multivariate logistic regression analysis using a generalized estimating equation was used to find the factors associated with major complications. Table 4 shows that definite CHIP (versus non-CHIP: OR 2.099, 95% CI 1.062–4.150, p=0.033) was significantly associated with major complications.
| Definite CHIP (n=140) | Possible CHIP (n=397) | Non-CHIP (n=452) | P value | |
|---|---|---|---|---|
| Major complications, n (%) | 22 (15.7) | 38 (9.6) | 26 (5.8) | 0.001 |
| In hospital death, n (%) | 5 (3.6) | 1 (0.3) | 0 (0.0) | ＜0.001 |
| Cardiac tamponade requiring pericardial drainage, n (%) | 0 (0.0) | 2 (0.5) | 0 (0.0) | 0.224 |
| Emergency surgery, n (%) | 0 (0.0) | 0 (0.0) | 0 (0.0) | - |
| Stroke within 48 hours since PCI, n (%) | 3 (2.1) | 1 (0.3) | 0 (0.0) | 0.002 |
| Emergent use of mechanical circulatory support during PCI procedure, n (%) | 2 (1.4) | 2 (0.5) | 0 (0.0) | 0.061 |
| Periprocedural myocardial infarction, n (%) | 14 (10.0) | 33 (8.3) | 25 (5.5) | 0.122 |
| Puncture site complications requiring blood transfusion, n (%) | 2 (1.4) | 0 (0.0) | 1 (0.2) | 0.028 |
| Acute heart failure during PCI procedure, n (%) | 2 (1.4) | 3 (0.8) | 0 (0.0) | 0.076 |
| Unsuccessful PCI, n (%) | 3 (2.1) | 8 (2.0) | 10 (2.2) | 0.980 |
| Unsuccessful PCI to CTO lesions, n (%) | 3 (2.1) | 8 (2.0) | 10 (2.2) | 0.980 |
| Unsuccessful PCI to non-CTO lesions, n (%) | 0 (0.0) | 0 (0.0) | 0 (0.0) | - |
Data were expressed as numbers (percentages). A Chi-square test was used for categorical variables. Abbreviations: CTO = chronic total occlusion, PCI = percutaneous coronary intervention.
| All (n=989) | Had major complications (n=86) | Had no complication (n=903) | P value | |
|---|---|---|---|---|
| Age, years | 74 (67-79) | 75.5 (70-79) | 73 (67-79) | 0.128 |
| Male, n (%) | 776 (78.5) | 63 (73.3) | 713 (79.0) | 0.219 |
| Body mass index (kg/m2) | 23.8 (21.7-26.2) (n=988) | 22.7 (20.8-25.7) (n=85) | 23.9 (21.8-26.2) | 0.033 |
| Hypertension, n (%) | 803 (81.2) | 72 (83.7) | 731 (81.0) | 0.530 |
| Hyperlipidemia, n (%) | 849 (85.8) | 76 (88.4) | 773 (85.6) | 0.482 |
| Diabetes mellitus, n (%) | 543 (54.9) | 43 (50.0) | 500 (55.4) | 0.339 |
| Current smoker, n (%) | 246 (25.1) (n=981) | 17 (20.5) (n=83) | 229 (25.5) (n=898) | 0.313 |
| Frailty, n (%) | 112 (11.3) | 18 (20.9) | 94 (10.4) | 0.003 |
| Chronic renal failure on hemodialysis, n (%) | 163 (16.5) | 18 (20.9) | 146 (16.2) | 0.257 |
| Active malignancy, n (%) | 37 (3.7) | 4 (4.7) | 33 (3.7) | 0.642 |
| Taking immunosuppressive drugs, n (%) | 38 (3.8) | 8 (9.3) | 30 (3.3) | 0.006 |
| Pulmonary disease requiring inhalants, n (%) | 32 (3.2) | 1 (1.2) | 31 (3.4) | 0.256 |
| Cirrhosis of liver, n (%) | 11 (1.1) | 1 (1.2) | 10 (1.1) | 0.963 |
| History of cerebral infarction, n (%) | 120 (12.1) | 16 (18.6) | 104 (11.5) | 0.054 |
| History of peripheral vascular disease, n (%) | 272 (27.5) | 34 (39.5) | 238 (26.4) | 0.009 |
| History of heart failure, n (%) | 255 (25.8) | 34 (39.5) | 221 (24.5) | 0.002 |
| History of previous PCI, n (%) | 551 (55.7) | 55 (64.0) | 496 (54.9) | 0.107 |
| History of previous CABG, n (%) | 55 (5.6) | 6 (7.0) | 49 (5.4) | 0.549 |
| History of previous myocardial infarction, n (%) | 330 (33.4) | 38 (44.2) | 292 (32.3) | 0.026 |
| Culprit lesion of STEMI, n (%) | 17 (1.7) | 1 (1.2) | 16 (1.8) | |
| Culprit lesion of NSTEMI, n (%) | 167 (16.9) | 22 (25.6) | 145 (16.1) | |
| Culprit lesion of non-AMI stenosis, n (%) | 805 (81.4) | 63 (73.3) | 742 (82.2) | |
| Cardiopulmonary arrest out of hospital, n (%) | 8 (0.8) | 1 (1.2) | 7 (0.8) | 0.701 |
| Unstable hemodynamics, n (%) | 5 (5.8) | 7 (0.8) | ＜0.001 | |
| Systolic blood pressure at admission, mmHg | 129 (115-146) | 128.5 (118-145) | 129 (115-146) | 0.762 |
| Diastolic blood pressure at admission, mmHg | 72 (63-81) | 70 (63-83) | 72 (63-81) | 0.964 |
| Heart rate at admission, bpm | 70 (62-81) | 76 (64-89) | 70 (62-81) | 0.006 |
| Hemoglobin levels, g/dL | 12.8 (11.4-14.0) | 12.1 (11.0-13.0) | 12.9 (11.4-14.1) | ＜0.001 |
| Platelets, ×103/uL | 19.9 (16.2-24.4) | 20.2 (16.4-24.6) | 19.9 (16.2-24.4) | 0.781 |
| Serum creatinine, mg/dL | 0.94 (0.74-1.35) | 1.06 (0.76-2.27) | 0.94 (0.74-1.32) | 0.056 |
| Hemoglobin A1c, % | 6.4 (5.9-7.1) | 6.4 (5.9-7.2) | 6.4 (5.9-7.1) | 0.768 |
| C-reactive protein, mg/dL | 0.14 (0.06-0.43) | 0.20 (0.09-0.62) | 0.14 (0.06-0.41) | 0.016 |
| Brain natriuretic peptide, pg/ml | 110.5 (39.2-358.2) (n=985) | 166.3 (58.5-734.2) (n=85) | 107.8 (38.4-343.9) (n=900) | 0.015 |
| Creatine kinase, U/L | 91.0 (65.0-136.0) | 97 (68-179) | 90 (64-134) | 0.069 |
| Left ventricular ejection fraction, % | 57.2 (41.7-65.1) | 55 (39-66) | 57.3 (42.3-65.1) | 0.382 |
| Valvular disease, n (%) | 127 (12.8) | 11 (12.8) | 116 (12.8) | 0.988 |
| Medication at admission | ||||
| Aspirin, n (%) | 988 (99.9) | 85 (98.8) | 903 (100.0) | 0.001 |
| Thienopyridine, n (%) | 989 (100) | 86 (100.0) | 903 (100.0) | - |
| Statin, n (%) | 938 (94.8) | 81 (94.2) | 857 (94.9) | 0.773 |
| ACE inhibitors or ARBs, n (%) | 666 (67.3) | 61 (70.9) | 605 (67.0) | 0.458 |
| Beta-blockers, n (%) | 744 (75.2) | 60 (8.1) | 684 (75.7) | 0.220 |
| Calcium channel blocker, n (%) | 415 (42.0) | 44 (51.2) | 371 (41.1) | 0.070 |
| Diuretics, n (%) | 284 (28.7) | 23 (26.7) | 261 (28.9) | 0.672 |
| Oral antidiabetic, n (%) | 397 (40.1) | 32 (37.2) | 365 (40.4) | 0.562 |
| Insulin, n (%) | 132 (13.3) | 17 (19.8) | 115 (12.7) | 0.067 |
| Direct oral anticoagulants., n (%) | 61 (6.2) | 5 (5.8) | 56 (6.2) | 0.886 |
| Warfarin, n (%) | 25 (2.5) | 2 (2.3) | 23 (2.5) | 0.900 |
| PCI to Chronic Total Occlusion, n (%) | 301 (30.4) | 18 (20.9) | 283 (31.3) | 0.045 |
| PCI to Left main artery, n (%) | 141 (14.3) | 10 (7.1) | 131 (14.5) | 0.466 |
| Use of rotational atherectomy, n (%) | 448 (45.3) | 50 (58.1) | 398 (44.1) | 0.012 |
| Use of orbital atherectomy, n (%) | 20 (2.0) | 2 (2.3) | 18 (2.0) | 0.834 |
| PCI for triple vessel disease with 90% stenosis or proven ischemia, n (%) | 177 (17.9) | 15 (17.4) | 162 (17.9) | 0.908 |
| Final PCI Procedure | 0.085 | |||
| POBA only, n (%) | 2 (0.2) | 0 (0.0) | 2 (0.2) | |
| Drug coated balloon, n (%) | 114 (11.5) | 15 (17.4) | 99 (11.0) | |
| Bare metal stent, n (%) | 3 (0.3) | 0 (0.0) | 3 (0.3) | |
| Drug eluting stent, n (%) | 849 (85.8) | 70 (81.4) | 779 (91.8) | |
| POBA and aspiration, n (%) | 2 (0.2) | 1 (1.2) | 1 (0.1) | |
| No wire cross, n (%) | 19 (1.9) | 0 (0.0) | 19 (2.1) | |
| Approach site | 0.407 | |||
| Radial, n (%) | 399 (40.3) | 30 (34.9) | 369 (40.9) | |
| Brachial, n (%) | 21 (2.1) | 3 (3.5) | 18 (2.0) | |
| Femoral, n (%) | 569 (57.5) | 53 (61.6) | 516 (57.1) | |
| Guide-Catheter size (Fr) | 0.664 | |||
| 6Fr, n (%) | 115 (11.6) | 11 (12.8) | 104 (11.5) | |
| 7Fr, n (%) | 702 (71.0) | 63 (73.3) | 639 (70.8) | |
| 8Fr, n (%) | 172 (17.4) | 12 (14.0) | 160 (17.7) | |
| First TIMI flow (0,1,2,3) | 0.179 | |||
| 0, n (%) | 297 (30.0) | 19 (22.1) | 278 (30.8) | |
| 1, n (%) | 19 (1.9) | 3 (3.5) | 16 (1.8) | |
| 2, n (%) | 54 (5.5) | 3 (3.5) | 51 (5.6) | |
| 3, n (%) | 619 (62.6) | 61 (70.9) | 558 (61.8) | |
| Final TIMI flow (0,1,2,3) | 0.400 | |||
| 0, n (%) | 20 (2.0) | 0 (0.0) | 20 (2.2) | |
| 1, n (%) | 3 (0.3) | 0 (0.0) | 3 (0.3) | |
| 2, n (%) | 7 (0.7) | 0 (0.0) | 7 (0.8) | |
| 3, n (%) | 959 (97.0) | 86 (100) | 873 (96.7) | |
| Dual artery access, n (%) | 418 (42.3) | 29 (33.7) | 389 (43.1) | 0.093 |
| Scheduled IABP or Impella, n (%) | 28 (2.8) | 3 (3.5) | 25 (2.8) | 0.701 |
Data were presented as a percentage for categorical variables, a mean±standard deviation (SD) for normally distributed continuous variables, and median (quartile 1–quartile 3) for nonparametric variables. A Student’s t test was used for normally distributed continuous variables and Mann–Whitney U test was used for abnormally distributed continuous variables. A Chi-square test was used for categorical variables. Abbreviations: ACE inhibitors = angiotensin-converting enzyme inhibitor, AMI = Acute Myocardial Infarction, AR = Aortic Regurgitation, ARBs = angiotensin receptor blockers, AS = Aortic Stenosis, CABG = coronary artery-bypass grafting, eGFR = estimated glomerular filtration rate, IABP = Intra-Aortic Balloon Pumping, MR = Mitral Regurgitation, MS = Mitral stenosis, NSTEMI = non-ST-segment elevation myocardial infarction, PCI = percutaneous coronary intervention, POBA = plain old balloon angioplasty, STEMI = ST-segment elevation myocardial infarction, TIMI = thrombolysis in myocardial infarction.
| parameters | Dependent variable: Major Complications | p value | |
|---|---|---|---|
| Odds ratio | 95% confidence interval | ||
| CHIP | |||
| Possible CHIP vs Non-CHIP | 1.396 | 0.806-2.418 | 0.234 |
| Definite CHIP vs Non-CHIP | 2.099 | 1.062-4.150 | 0.033 |
| Heart rate at admission (per 1bpm) | 1.009 | 0.999-1.020 | 0.081 |
| Hemoglobin levels (per 1g/dL) | 1.006 | 0.953-1.062 | 0.833 |
| C-reactive protein (per 1mg/dL) | 0.960 | 0.900-1.024 | 0.212 |
| History of heart failure | 1.188 | 0.674-2.097 | 0.551 |
| History of previous myocardial infarction | 1.559 | 0.958-2.537 | 0.074 |
| History of peripheral vascular disease | 1.427 | 0.829-2.457 | 0.200 |
| PCI to Chronic Total Occlusion | 0.781 | 0.420-1.450 | 0.433 |
| Use of rotational atherectomy | 1.373 | 0.795-2.371 | 0.255 |
Multivariate logistic regression model using a generalized estimating equation was performed for factors with no missing values that met p＜0.05 in univariate analysis. Abbreviations: CHIP = complex and high-risk intervention in indicated patients.
The multivariate logistic regression analysis using a generalized estimating equation was used to identify the CHIP factors associated with major complications. As shown in Table 5, the use of immunosuppressive drugs (OR 3.040, 95% CI 1.251–7.386, p=0.014), unstable hemodynamics (OR 5.753, 95% CI 1.217–27.201, p=0.027), and frailty (OR 2.039, 95% CI 1.108–3.751, p=0.022) were significantly associated with major complications among CHIP factors.
| parameters | Dependent variable: Major Complications | p value | |
|---|---|---|---|
| Odds ratio | 95% confidence interval | ||
| Frailty | 2.039 | 1.108-3.751 | 0.022 |
| Active malignancy | 1.185 | 0.409-3.433 | 0.754 |
| Use of immunosuppressive drugs | 3.040 | 1.251-7.386 | 0.014 |
| Pulmonary disease requiring inhalants | 0.367 | 0.046-2.902 | 0.342 |
| Cirrhosis of liver | 0.990 | 0.133-7.360 | 0.992 |
| Chronic renal failure on hemodialysis | 1.329 | 0.712-2.480 | 0.371 |
| History of cerebral infarction | 1.679 | 0.890-3.167 | 0.110 |
| Unstable hemodynamics | 5.753 | 1.217-27.201 | 0.027 |
| Left ventricular ejection fraction (per 1%) | 1.002 | 0.986-1.017 | 0.833 |
| Valvular disease | 0.736 | 0.349-1.552 | 0.420 |
Multivariate logistic regression was performed among CHIP factors using a generalized estimating equation. Abbreviations: CHIP = complex and high-risk intervention in indicated patients.
In this study, we included 989 complex PCI lesions, and we divided them into the definite CHIP group (n=140), the possible CHIP group (n=397), and the non-CHIP group (n =452). The incidence of major complications of PCI was highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group. The multivariate logistic regression analysis revealed that definite CHIP (versus non-CHIP: OR 2.099, 95% CI 1.062–4.150, p=0.033) was significantly associated with major complications. Among the CHIP factors, immunosuppressive drugs (OR 3.040, 95% CI 1.251–7.386, p=0.014) unstable hemodynamics (OR 5.753, 95% CI 1.217–27.201, p=0.027), and frailty (OR 2.039, 95% CI 1.108–3.751, p=0.022) were significantly associated with major complications. The summary findings are illustrated in Fig.2.
Abbreviations: CHIP=complex and high-risk intervention in indicated patients, CTO=chronic total occlusion, EF=ejection fraction, LMT=left main trunk, PCI=percutaneous coronary intervention.
First, we should determine the difference between this present study and earlier studies. When the concept of CHIP was initially suggested, three components (patient comorbidities, complicated heart disease, and coronary lesion complexity) were identified as the framework of CHIP16). However, the definitions of these three components have not been established. Burzotta et al. reported the long-term outcome of Impella-supported PCI in patients with CHIP using the Roma-Verona Registry32). However, the definition of CHIP was not available in their report32). Thus, Brener et al. defined CHIP using CHIP criteria such as age, ejection fraction, dialysis, prior bypass surgery, and PCI for left main trunk or CTO or two-vessel disease, and found that CHIP-PCI was associated with increased risk of 1-year mortality17). An analysis of the United Kingdom BCIS Database including 313,054 elective PCI found the CHIP factors associated with in-hospital MACCE. Although this study used a large registry data, patient comorbidities were not fully assessed33). Thus, the definition of CHIP was found to vary widely among earlier studies17, 32-35), and there were only few studies clearly defining the three components (patient factor, complicated heart disease, and lesion complexity) for CHIP. The novelty of our study lies in the fact that the three components for CHIP were clearly defined, and the possible CHIP and CHIP were compared with the non-CHIP.
Second, we should discuss the association between CHIP and major complications. Our multivariate logistic regression analysis revealed that definite CHIP was significantly associated with major complications and that possible CHIP tended to be associated with major complications without reaching statistical significance. In comparing complex PCI without patient factor or complicated heart disease, the risk of major complications might be more significant in complex PCI complicated with either patient factors or complicated heart disease and would be greater in complex PCI complicated with both patient factor and complicated heart disease. In other words, it is thus important to recognize both patient factors and complicated heart disease to estimate the risk of major complications in complex PCI.
Among the CHIP factors (patient factor and complicated heart disease), our findings revealed that immunosuppressive drugs, unstable hemodynamics, and frailty were significantly associated with major complications. Although there were several literatures indicating the relationship of immunosuppressive drugs with cardiovascular risk factors including hypertension and dyslipidemia36-38), there were only few studies demonstrating the direct association between immunosuppressive drugs and major complications in PCI. Although the mechanism of the above association remains unknown, patients who had immunosuppressive drugs would have systemic inflammation. Such systemic inflammation might facilitate coronary plaque calcification, which, in turn, could cause complications in complex PCI39-40). Besides immunosuppressive drugs, frailty was also found to be associated with major complications in complex PCI. National all-payer inpatient health care database in the United States including 7,306,007 admissions showed that high frailty risk was independently associated with in-hospital death41), which was consistent with our results. On the other hand, it is not surprising that unstable hemodynamics was the determinant of complications in complex PCI. National cardiovascular data registry in the United States including 588,398 PCI procedures showed that the combination of cardiogenic shock and salvage PCI was associated with the highest mortality in PCI42). Moreover, an analysis of the United Kingdom BCIS Database revealed that upfront LV support was most associated with in-hospital MACCE in elective PCI33). These findings in large-scale registries were also found to be consistent with our results.
Clinical implications of this present study should be noted. The incidence of major complications in complex PCI was highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group. Thus, it would be important to evaluate complicated heart disease and patient factors including immunosuppressive drug and frailty before planning complex PCI. Accurate risk estimation in complex PCI would be useful for the decision-making in the heart team conference. More importantly, the accurate risk estimation should be the cornerstone for the patient’s decision whether to receive complex PCI. CHIP candidates were often referred to PCI due to the ineligibility of CABG. In such situations, the patient’s decision is not “PCI or CABG,” but “PCI or medical therapy alone.” Therefore, the accurate risk estimation considering CHIP would help each patient to give informed consent in the treatment of CAD.
This present study has several limitations. Since this study was a single-center, retrospective study, there was a potential selection bias. As the patient’s medical history such as having cirrhosis or active malignancy was identified from hospital records, the past medical diseases that were treated in other hospitals might have been overlooked. In the definition of frailty, we could not adopt the Clinical Frailty Scale43), which is the global standard for evaluating frailty because we did not check the Clinical Frailty Scale in our routine daily practice. In fact, our definition of frailty would be probably compatible to the Clinical Frailty Scale ≥ 5 (moderately or severely frail), which is the most common cut-off used to describe a person as frail44). However, we could not allocate each study patient into the exact scale (1, 2, 3, 4, 5, 6, 7, 8, and 9) retrospectively. Since all unsuccessful PCIs were derived from PCI to CTO, we excluded unsuccessful PCI for CTO from the major complications. However, our success rate for CTO-PCI in this present study was 93.0% (280 successful PCI/ 301 CTO-PCI), which was higher than the success rate (81.2%) of the recent meta-analysis including 6211 CTO-PCI45). Moreover, there were several parameters with missing values such as BNP levels. The definition of periprocedural MI was not simple, when baseline CK levels were already elevated．We have adopted the definition of periprocedural MI as further increase of CK levels at the next day of PCI than CK level at baseline if baseline CK levels were already elevated, which was defined in our previous work25). Finally, our study population was not a patient-level database, but a PCI-level database. If a patient has separately received ≥ 2 PCIs during the study period, each PCI was separately included. Although we have used a generalized estimating equation for multivariate logistic regression analysis to adjust within-patient clustering, we could not exclude the effect of the clustered nature from one patient in comparing the three groups.
The incidence of major complications in complex PCI was determined to be the highest in the definite CHIP group, followed by the possible CHIP group and lowest in the non-CHIP group. Thus, more attention should be given to the three components of CHIP to prevent major complications in complex PCI.
The authors acknowledge all staff in the catheter laboratory, cardiology units, and emergency and critical care units in Saitama Medical Center, Jichi Medical University for their technical support in this study.
Dr. Sakakura has received speaking honoraria from Abbott Vascular, Boston Scientific, Medtronic Cardiovascular, Terumo, OrbusNeich, Japan Lifeline, Kaneka, and NIPRO. Dr. Jinnouchi has received speaking honoraria from Abbott Vascular. Prof. Fujita has served as a consultant for Mehergen Group Holdings, Inc.
