Original Contribution

Incidence and Predictors of Side-Branch Compromise in Primary Percutaneous Coronary Intervention for Acute Myocardial Infarction

Valerie Khoo;  Liang Shen, PhD;  Vanessa Khoo;  Germaine Loo;  Mark Richards, MD;  Tiong-Cheng Yeo, MD;  Chi-Hang Lee, MD

Valerie Khoo;  Liang Shen, PhD;  Vanessa Khoo;  Germaine Loo;  Mark Richards, MD;  Tiong-Cheng Yeo, MD;  Chi-Hang Lee, MD

Abstract: Objective. We aimed to determine the incidence and predictors of side-branch compromise (SBC) in patients who underwent primary percutaneous coronary intervention (PCI) for acute myocardial infarction (AMI). Background. Little data exist on SBC in AMI patients, especially in the drug-eluting stent era. Methods. We recruited 174 patients who underwent primary PCI over a 12-month period. After reviewing their coronary angiograms, we included for analysis 102 patients with a side branch >2 mm arising from the culprit lesion and that was spanned by a coronary stent. SBC was defined as post-stent implantation TIMI flow of <3 in the side branch. Results. Among the 102 patients analyzed, drug-eluting stents (n = 77), bare-metal stents (n = 17), and bioresorbable vascular scaffolds (n = 8) were used to treat the culprit lesions. Final TIMI flow of the main vessel was 2 or 3 in 101 patients (99%). SBC occurred in 23 patients (final side branch TIMI flow 0, n = 6; TIMI 1, n = 4; TIMI 2, n = 13), giving an incidence of 22.5%. Multivariate analysis showed non-left anterior descending (LAD) culprit vessel (odds ratio [OR], 3.66; 95% confidence interval [CI], 1.22-10.95; P=.02), higher peak creatine kinase level (OR, 1.03 for every 100-unit increase; 95% CI, 1.01-1.05; P=.01), and Rentrop score of 2/3 (OR, 3.57; 95% CI, 0.98-13.04; P=.055) to be independent predictors of SBC. Conclusions. The incidence of SBC was 22.5%. The independent predictors of SBC were non-LAD culprit vessel, larger infarct size, and good collateral vessel formation.

J INVASIVE CARDIOL 2014;26(7):297-302

Key words: side branch, stent, acute myocardial infarction, Rentrop score, collateral


A familiar complication of primary percutaneous coronary intervention (PCI) procedures involving bifurcation lesions is the compromise of a side branch originating directly from (or within the vicinity of) the lesion, and which is within the target segment spanned by the implanted stent.1,2 Whereas the occlusion of a small side branch is of limited importance, the coverage of large side branches (≥2 mm in diameter) supplying a considerable mass of myocardium may induce subsequent myocardial ischemia and infarction.3,4 Additional clinical concerns have been expressed over the compromise of larger side branches encompassed by the stent, as these will no longer be accessible to catheterization techniques should an ulterior lesion develop more distally in this side branch. As such, interventional strategies require careful consideration of the risk of side-branch compromise (SBC) when pursuing coronary stent implantation.

Previous studies investigating SBC have aimed to identify promising independent predictors of SBC. Strong predictors of SBC, such as high-grade ostial side-branch stenosis and small reference side-branch diameter at baseline angiography, have been widely supported.5-8 However, most studies performed on SBC have focused on establishing predictors in stable patients undergoing elective stent implantation,5-7 whereas few data exist that characterize predictors of SBC in patients presenting with acute myocardial infarction (AMI). Specifically, AMI is associated with varying degrees of vasoconstriction and residual thrombus at the culprit lesion even after thrombus aspiration, rendering angiographic evaluation of the vessel size and severity of ostial side-branch stenosis difficult.

Coronary stent implantation is the preferred treatment for AMI and has affirmed benefits in reducing cardiac mortality and the need for target vessel revascularization.9 It is therefore surprising that few studies have examined the predictors and risks of SBC following stenting for AMI. Accordingly, the objective of this study was to describe the incidence and predictors of SBC following intracoronary stent implantation in AMI patients.


We performed a single-center review of consecutive patients who underwent successful primary PCI for AMI between February 2012 and February 2013 at a tertiary institution in Singapore. Prospective information from the PCI registry at the institution was independently gathered by a cardiology research nurse coordinator. Per protocol, the clinical diagnosis of AMI was based on the following parameters: (1) chest pain lasting ≥30 minutes; (2) ST-segment elevation ≥2 mm in at least 2 contiguous electrocardiographic (ECG) leads; and (3) a greater than 3-fold increase in serum creatine kinase levels. PCI was performed with the intention of restoring perfusion with Thrombolysis in Myocardial Infarction (TIMI) grade 3 flow to the infarct-related artery and/or achieving optimal stent implantation outcomes of the primary lesion in these patients presenting with AMI.

We enrolled 174 patients in this study. Of these patients, 72 were excluded from the analysis for the following reasons: balloon dilation without stent implantation during PCI (n = 6); previous coronary artery bypass graft (CABG) procedure involving the presenting infarct-related artery (n = 4); and absence of side branches, or side branches with a luminal diameter of <2 mm, emerging from the target segment spanned by the implanted stent (n = 62). Our study group comprised the remaining 102 patients with intracoronary bifurcation lesions involving side branches with a luminal diameter of ≥2 mm and whose ostia were encompassed by the stent. The National Healthcare Group Domain Specific Review Board approved the research protocol of the study (reference C/2010/00341), and written informed consent was obtained from all subjects.

The PCI procedure was performed in accordance with accepted standards. The interventional strategy was left to the discretion of the individual operator, with mandatory balloon predilation followed by stent implantation at pressures below the critical balloon burst pressure. Standard anticoagulant premedication comprised a continuous intravenous infusion of weight-adjusted heparin prior to PCI that was intended to achieve an activated clotting time of >300 seconds, or 200-250 seconds when platelet glycoprotein IIb/IIIa inhibitor was administered. All patients received life-long aspirin in combination with a loading dose of clopidogrel (600 mg) or prasugrel (60 mg) before the intervention. In addition, all patients who underwent intracoronary stenting maintained a dosage of clopidogrel (75 mg/day) or prasugrel (10 mg/day) for >1 month.

Coronary angiograms were reviewed by an investigator who was blinded to the demographic and clinical characteristics. All frames were calibrated before injection of contrast agent using the tip of the guiding catheter as the reference. The region of interest (ROI) was defined as the stent implantation site, and all side branches bridged by the longitudinal confines of the ROI and that had an estimated diameter ≥2 mm were evaluated in the angiographic assessment. Additionally, in attempting the dilation of a secondary vessel (eg, diagonal branch) proximal to its origin, the distal portion of the main branch (eg, left anterior descending [LAD] coronary artery) sufficed as a side branch in jeopardy of occlusion. Patency of side branches was assessed at baseline before PCI, after balloon predilation, subsequent to stent deployment, and in the final cineangiogram after the interventional procedure. The coronary perfusion pattern in the side branches was graded according to the TIMI classification system. Side-branch compromise was defined as a TIMI flow ≤3 at the final cineangiogram after PCI. The patients were divided into SBC (side-branch TIMI flow 0-2) and non-SBC (side-branch TIMI flow 3) groups.

Statistical analysis. Patient demographic and clinical characteristics were summarized descriptively. Numerical predictors were evaluated with the Mann-Whitney U-test, whereas categorical predictors were evaluated with the Chi-square test. Those variables showing a P-value of less than .10 were included in a multiple logistic regression with backward model selection.


Among the 102 patients (96% men; average age, 54 years old), the infarction location was anterior in 55 patients (54%). Apart from 5 patients who received 2 overlapping stents to the main vessels, the remaining 97 patients were treated with a single-stent implantation to the main vessels. Drug-eluting stents (n = 77, of which 67 were everolimus-eluting stents), bare-metal stents (n = 17), and bioresorbable vascular scaffolds (n = 8) were used to treat the culprit lesions. At the end of the procedure, final TIMI flow of the main vessel was 2-3 in 101 patients (99%) and 3 in 88 patients (86.3%).

Side-branch compromise occurred in 23 patients (final side-branch TIMI flow 0, n = 6; TIMI 1, n = 4; TIMI 2, n = 13), with an overall incidence of 22.5%. The detailed angiographic and procedural characteristics of these 23 patients are shown in Table 1. Of these 23 patients, a total of 16 received drug-eluting stents, 5 received bare-metal stents, and 2 received bioresorbable vascular scaffolds, giving incidences of SBC for the corresponding stent types of 20.8% (16/77), 31.3% (5/16), and 25% (2/8) (P=.65), respectively. Prophylactic guidewire insertion to the side branch was performed in 14 patients. Balloon angioplasty to the side branch was performed in 2 patients. No additional interventional procedure was performed in the remaining patients. An example of SBC is shown in Figure 1.

The baseline clinical and demographic characteristics of the patients in the SBC and non-SBC groups are shown in Table 2. Compared to the non-SBC group, the SBC group was associated with a higher peak creatine kinase level. There was also a trend toward higher prevalence of diabetes mellitus in the SBC group.

The angiographic characteristics of the patients in the SBC and non-SBC groups are shown in Table 3. Compared to the non-SBC group, the culprit vessel in the SBC group was more likely to be a non-LAD location, and the Rentrop score was higher. Multivariate analysis showed that culprit vessel not being in the LAD artery (OR, 3.66; 95% CI, 1.22-10.95; P=.02), peak creatine kinase level (OR, 1.03 for every 100-unit increase; 95% CI, 1.01–1.05, P=.01), and Rentrop score of 2 or 3 (OR, 3.57; 95% CI, 0.98-13.04; P=.04) were independent predictors of SBC.


SBC during PCI has been implicated as a contributing factor in the development of periprocedural myocardial infarction, which has been associated with unfavorable late clinical outcomes, including an increased risk of cardiac mortality.10-12 We evaluated the incidence and predictors of SBC in a cohort of patients who underwent primary PCI for AMI. Among the 174 patients recruited, a side branch (≥2 mm) arose from the target lesion and was spanned by a stent in 102 patients. Most (75%) of these 102 patients were treated with drug-eluting stents. At the end of the procedure, SBC occurred in 23 patients (22.5%). Culprit vessel not being the LAD artery, Rentrop score of 2 or 3, and higher peak creatine kinase level were independent predictors of SBC by multivariate analysis. To our knowledge, this is the first study evaluating SBC in AMI patients treated in the drug-eluting stent era.

Various mechanisms have been proposed to rationalize the development of SBC following stenting of the main branch. “Snow-plowing,” involving the axial redistribution of plaque from the main vessel into the orifice of a lesion-related branch, and carina shift have been extensively justified by studies reporting SBC in branches with diffuse ostial stenosis in continuity with (and developing after high-pressure stent dilations of) the parent lesion.6,13,14 Nevertheless, Meier et al proposed that SBC followed the dissection of a dilated artery involving the origin of the side branch, having visualized a dissection line extending from the target lesion to the take-off of the side branch in all cases of SBC in their study.15 However, a high incidence of spontaneous reperfusion in initially obstructed side branches lends support to the influence of transient coronary spasm on SBC. These mechanisms apply, however, primarily to elective PCI patients, and when considering AMI patients, a different mechanism of SBC may be operating. Thrombus embolization from a ruptured plaque in the main vessel into the side branch seems the most plausible pathogenesis accounting for SBC in our AMI patients. AMI is most often attributable to blockage of a coronary artery by a thrombus; subsequent wiring of the vessel during PCI may dismantle the clot and cause remnants of the plaque from the parent lesion to drift into and plug the side-branch ostium, jeopardizing its flow.

Most studies investigating SBC have focused on its occurrence in elective patients receiving PCI,5-7 whereas few have delineated the incidence of SBC among AMI patients undergoing the procedure. The reported incidences of side-branch occlusion/compromise ranged from 10%-21%,5-7 depending on the definitions of “side branch” and “occlusion/compromise.” Apart from the present research, only 1 other study has concentrated on SBC during interventional stenting for AMI.8 Interestingly, the incidence of SBC in AMI patients reported by both Kralev et al (12.5%) and the present study (22.5%) are higher than the majority of SBO rates reported by others examining elective patients. This discrepancy may be due to the lower rates of side-branch wiring carried out during the interventional procedure for AMI patients because of the acute nature of the condition and the urgent need to promptly restore perfusion of myocardium supplied by the main vessel. Thus, because ensuring the patency of side branches is secondary to restoring flow in the main vessel in AMI, the absence of side-branch protection would result in a comparatively elevated rate of SBC. Furthermore, the higher incidence of SBC found in this study compared with that reported by Kralev et al may be attributable to the different definitions of SBC used in the two studies.8 We defined SBC as a persistent reduction in side-branch flow to TIMI 0-2, whereas the parallel study considered only a final TIMI flow of 0 or 1 to be indicative of SBC.

We may have indeed underestimated the true rates of SBC in this study. Wiring of the main vessel during PCI for AMI may disperse components of the parent lesion and cause fragments to plug associated side branches. This may, in turn, cause the complete disappearance of the side branch on the cineangiogram and, as the side branch is not visualized, this instance of SBC may not have been taken into account. Although the incidence of SBC during stenting for AMI is generally higher than for elective patients, the true incidence may certainly be even higher than previously conceived.

Our study describes SBC in the context of the predominant use of drug-eluting stents to treat AMI, in comparison to a previous study in which bare-metal stents were used.8 Concern has been expressed over the tendency of first-generation drug-eluting stents to compromise the ostium of side branches because of their polymer coating, smaller cell-surface area, and the increased metal-to-artery ratio.13,16 In comparison to the patients who were treated with multiple bare-metal stents, those treated with multiple paclitaxel-eluting stents had more frequent occurrence of significant side-branch narrowing or progression to total occlusion (42.6% vs 30.6%; P=.03) and a greater likelihood of reduced side-branch TIMI flow (41.9% vs 28.6%; P=.02). Fortunately, the current generation of drug-eluting stents has been shown to be substantially safer in this regard.3,17 In the current study, the incidence of SBC was non-significantly lower in the drug-eluting stent cohort than in the bare-metal stent cohort. Although the number the patients treated with bioresorbable vascular scaffolds was small, our data suggest that the incidence of SBC was similar to those treated with drug-eluting stents.

The present study extends previous findings by exploring the relationship between coronary collateral Rentrop scoring and the risk of SBC. The Rentrop score is utilized to provide an index of collateral vessel development, and in particular, our results indicate a Rentrop score of 2 or 3, reflecting good collateral formation from the non-infarct related artery, to be a significant independent predictor of SBC in AMI patients.18-20 These results support the notion that a critical plaque burden may be the responsible pathogenesis underlying the progression of SBC in AMI patients. As plaque build-up impedes perfusion through the main vessel, the consequential myocardial ischemia induces the formation of coronary collateral vessels to provide supplementary blood flow to the ischemic area normally supplied by the vessel. Thus, a Rentrop score of 2 or 3 indicates increased collateral formation to compensate for an increased plaque burden in the main vessel, hence increasing the probability of plaque embolization from the main vessel into the side branch to cause SBC. Thus, it is conceivable that a Rentrop score of 2 or 3 may be a promising independent predictor of acute SBC in AMI patients undergoing stenting.

The possible clinical implications of these findings relate to the protection of side branches during intracoronary stenting for AMI. Although side-branch protection is often disregarded in AMI, safeguarding their patency may be necessary in cases where coronary collateral formation with Rentrop grade 2 or 3 is observed, as these side branches are at greatest risk of occlusion following the stenting procedure. Where side branches supplying a substantial amount of critical myocardium are compromised, SBC may prompt the development of a moderate-sized infarct. As such, to prevent the occurrence of any consequential adverse cardiac events due to SBC, specialized techniques such as kissing-balloon angioplasty, T-stenting, cutting-balloon angioplasty, and debulking of atherosclerotic plaque may be used to protect side branches in AMI patients showing significant collateral development.

Study limitations. There are some limitations to this study. This was a single-center, retrospectively conducted study and was subject to the restrictions associated with such a study type. Coronary angiograms extracted from the database were not originally intended for research analysis, and image quality may have been substandard for visualizing side-branch perfusion. In addition, the relatively small sample size may have introduced a selective bias, impeding the generalization of our findings to the public. The lack of serial follow-up evaluation of side-branch perfusion reflects an inability to detect the subsequent restoration of flow in initially compromised side branches. Spontaneous reperfusion following SBC is a widely documented phenomenon, occurring in as frequently as 44% of SBC cases. A high incidence of spontaneous reperfusion would reinforce the notion of temporary coronary spasm underlying the pathogenesis of SBC and hence weaken the argument for “snow plowing” as the likely mechanism. It is therefore meaningful to quantify the extent to which spontaneous reperfusion occurs in follow-up angiographic analysis to strengthen or diminish the validity of our proposed mechanism of SBC. The prevailing practice at our institution was to restore coronary perfusion to the main vessel in patients undergoing primary PCI for AMI. This explains the small number (n = 2) of patients who underwent balloon angioplasty to the side branch. We do not routinely prolong the procedure to intervene all the occluded side branches, as the additional prognostic benefit is controversial. 


In conclusion, we found that among the patients who underwent primary PCI, a side branch of ≥2 mm was identified in 58.6% of the patients. The incidence of SBC was 22.5%, and the independent predictors of SBC were non-LAD culprit lesion, larger infarct size, and good collateral vessel formation. 


  1. Vetrovec GW, Cowley MJ, Wolfgang TC, Ducey KC. Effects of percutaneous transluminal coronary angioplasty on lesion-associated branches. Am Heart J. 1985;109(5 Pt 1):921-925.
  2. Fischman DL, Savage MP, Leon MB, et al. Fate of lesion-related side branches after coronary artery stenting. J Am Coll Cardiol. 1993;22(6):1641-1646.
  3. Lansky AJ, Yaqub M, Hermiller JB, et al. Side branch occlusion with everolimus-eluting and paclitaxel-eluting stents: three-year results from the SPIRIT III randomised trial. EuroIntervention. 2010;(6Suppl J):J44-J52.
  4. Park DW, Kim YH, Yun SC, et al. Frequency, causes, predictors, and clinical significance of periprocedural myocardial infarction following percutaneous coronary intervention. Eur Heart J. 2013;34:1662-1669.
  5. Aliabadi D, Tilli FV, Bowers TR. Incidence and angiographic predictors of side branch occlusion following high-pressure intracoronary stenting. Am J Cardiol. 1997;80(8):994-997.
  6. Alfonso F, Hernández C, Pérez-Vizcayno MJ, et al. Fate of stent-related side branches after coronary intervention in patients with in-stent restenosis. J Am Coll Cardiol. 2000;36(5):1549-1556.
  7. Poerner TC, Kralev S, Voelker W, et al. Natural history of small and medium-sized side branches after coronary stent implantation. Am Heart J. 2002;143(4):627-635.
  8. Kralev S, Poerner TC, Basorth D, et al Sidebranch occlusion after coronary stent implantation in patients presenting with ST-elevation myocardial infarction: clinical impact and angiographic predictors. Am Heart J. 2006;151(1):153-157.
  9. De Luca G, Suryapranata H, Stone GW, et al. Coronary stenting versus balloon angioplasty for acute myocardial infarction: a meta-regression analysis of randomized trials. Int J Cardiol. 2008;126(1):37-44. Epub 2007 Jun 4.
  10. Califf RM, Abdelmeguid AE, Kuntz RE, et al. Myonecrosis after revascularization procedures. J Am Coll Cardiol. 1998;31(2):241-251.
  11. Prasad A, Singh M, Lerman A, Lennon RJ, Holmes DR Jr, Rihal CS. Isolated elevation in troponin T after percutaneous coronary intervention is associated with higher long-term mortality. J Am Coll Cardiol. 2006;48(9):1765-1770.
  12. Jensen LO, Maeng M, Kaltoft A, et al. Stent thrombosis, myocardial infarction, and death after drug-eluting and bare-metal stent coronary interventions. J Am Coll Cardiol. 2007;50(5):463-470.
  13. Furukawa E, Hibi K, Kosuge M, et al. Intravascular ultrasound predictors of side branch occlusion in bifurcation lesions after percutaneous coronary intervention. Circ J. 2005;69(3):325-330.
  14. Muramatsu T, Onuma Y, García-García HM, et al; ABSORB-EXTEND investigators. Incidence and short-term clinical outcomes of small side branch occlusion after implantation of an everolimus-eluting bioresorbable vascular scaffold: an interim report of 435 patients in the ABSORB-EXTEND single-arm trial in comparison with an everolimus eluting metallic stent in the SPIRIT first and II trials. JACC Cardiovasc Interv. 2013;6(3):247-257.
  15. Meier B, Gruentzig AR, King SB 3rd, et al. Risk of side branch occlusion during coronary angioplasty. Am J Cardiol. 1984;53(1):10-14.
  16. Stone GW, Ellis SG, Cannon L, et al; TAXUS V investigators. Comparison of a polymer-based paclitaxel-eluting stent with a bare metal stent in patients with complex coronary artery disease: a randomized controlled trial. JAMA. 2005;294(10):1215-1223.
  17. Popma JJ, Mauri L, O’Shaughnessy C, et al. Frequency and clinical consequences associated with side branch occlusion during stent implantation using zotarolimus-eluting and paclitaxel-eluting coronary stents. Circ Cardiovasc Interv. 2009;2(2):133-139.
  18. Cohen M, Rentrop KP. Limitation of myocardial ischemia by collateral circulation during sudden controlled coronary artery occlusion in human subjects: a prospective study. Circulation. 1986;74(3):469-476.
  19. Mizuno K, Horiuchi K, Matui H, et al. Role of coronary collateral vessels during transient coronary occlusion during angioplasty assessed by hemodynamic, electrocardiographic and metabolic changes. J Am Coll Cardiol. 1988;12(3):624-628.
  20. Pérez-Castellano N, García EJ, Abeytua M, et al. Influence of collateral circulation on in-hospital death from anterior acute myocardial infarction. J Am Coll Cardiol. 1998;31(3):512-518.


From the National University Heart Centre Singapore, Singapore.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted September 18 2013, provisional acceptance given January 13, 2014, final version accepted April 21, 2014.

Address for correspondence: Dr Chi-Hang Lee, MBBS, National University Hospital Cardiac Department, Level 3, Main Building, National University Hospital, 5, Lower Kent Ridge Road, Singapore, Singapore 119074, Singapore. Email: leerch@hotmail.com