Differential Longitudinal Outcomes Following Percutaneous Coronary Intervention to the Left Internal Mammary Artery and Other Bypass Grafts of the LAD: Findings From the NCDR

Abstract: Objectives. Limited studies of percutaneous coronary intervention (PCI) of the left internal mammary artery (LIMA) graft exist. We compared outcomes of different bypass grafts to the left anterior descending (LAD) coronary artery. Methods. Participants ≥65 years old in the CathPCI Registry who underwent PCI of a bypass graft to the LAD between 2009 and 2014 were included. Individuals were divided by graft type: LIMA; saphenous vein graft (SVG); or other. Clinical characteristics and outcomes using Medicare claims data for mortality, rehospitalization for myocardial infarction (MI), stroke, or unplanned repeat revascularization at 1 year were examined. Results. There were 10,051 PCIs performed on grafts to the LAD: 6797 SVGs (67.6%), 3011 LIMA grafts (30.0%); and 243 other (2.4%). Procedural success rates (SVG 92.9%, LIMA 91.1%, other 93.4%; P=.65) and in-patient mortality rates (SVG 3.0%, LIMA 2.7%, other 2.1%; P=.61) were similar. However, dissection rates were higher in LIMA interventions (SVG 0.7%, LIMA 2.8%, other 2.5%; P<.001). At 1 year, mortality, MI, and repeat revascularization were lower in arterial grafts (mortality: SVG 16.6%, LIMA 14.8%, other 11.8% [P<.001]; MI: SVG 9.9%, LIMA, 6.6%, other 8.1% [P<.001]; revascularization: SVG 14.4%, LIMA 9.5%, other 10.4% [P<.001]). After multivariable adjustment, LIMA patients had lower rates of MI (hazard ratio, 0.71; 95% confidence interval, 0.60-0.84) and repeat revascularization (hazard ratio, 0.68; 95% confidence interval, 0.59-0.79) compared with the SVG group. Mortality was not significantly different. Conclusions. Despite similar procedural success rates compared with SVG and other graft types, LIMA interventions were independently associated with lower rates of recurrent MI and repeat revascularization at 1 year.  

J INVASIVE CARDIOL 2020;32(6):E143-E150. 

Key words: arterial, coronary artery bypass graft, coronary intervention, saphenous vein graft

Coronary artery bypass graft (CABG) surgery remains the dominant therapy for multivessel coronary artery disease, despite the advent of drug-eluting stents.^1,2 The use of the left internal mammary artery (LIMA) graft to the left anterior descending (LAD) artery is felt to be a key determinant of the efficacy and durability advantage of CABG.^3,4 Surgical case series show the durability of the LIMA to be as high as 85%-95% at 10 years.^5-9 However, graft failure can occur, especially at the anastomotic site of the LIMA to the LAD. Progression of distal LAD disease can also occur. 

LIMA graft interventions are uncommon and little has been published in the literature. They are generally perceived to be risky and technically challenging. Current literature is limited to isolated case reports and small case series, with only two case series in the recent literature.^10,11 The aim of this study is to leverage a large and contemporary real-world percutaneous coronary intervention (PCI) database — the American College of Cardiology (ACC) National Cardiovascular Data Registry (NCDR) CathPCI Registry — to accrue sufficient observations of LIMA interventions to enable a more systematic description of the clinical, anatomical, and procedural characteristics, as well as in-hospital outcomes associated with PCI of the LIMA graft when compared with other graft types to the LAD. 

Methods

Participants ≥65 years old in the ACC-NCDR CathPCI Registry who underwent PCI of a bypass graft to the LAD between July 2009 and December 2014 were included. PCI to grafts to any other vessel (including diagonals) were excluded. Skip (ie, jump) grafts were excluded. Individuals were divided into three groups based on the graft type (to LAD): LIMA graft; saphenous vein graft (SVG); or “other” arterial graft, including the right internal mammary artery and radial arteries. Any interventions in the graft, including the aortic anastomosis, the body, and the distal anastomosis, were attributed to the assigned category. Graft location was defined as follows: aortic indicates that the most severe stenosis is at the aortic anastomosis of the graft (≤3 mm from insertion point); body indicates that the most severe stenosis is in the body of the graft; and distal indicates that the most severe stenosis is at the distal anastomosis of the graft (≤3 mm from insertion point). PCI to the LAD distal to the graft anastomosis was excluded. 

Clinical and procedural characteristics, longitudinal outcomes using Medicare claims data, including mortality, rehospitalization for myocardial infarction (MI), stroke, or unplanned repeat revascularization at 1 year, were compared across groups. CathPCI data were linked to longitudinal Centers for Medicare & Medicaid (CMS) claims data through 2014. Analyzed data were limited to patients ≥65 years old. 

Primary outcomes were procedural success and 1-year mortality. Procedural success was defined as the number of lesions successfully dilated equal to the number of lesions attempted. Secondary outcomes were rehospitalization for MI, stroke, or unplanned repeat revascularization at 1 year. One-year follow-up data began at discharge and patients were censored at the end of fee-for-service eligibility and end of follow-up. Unplanned repeat revascularization was defined as readmission with PCI or CABG procedure, and a discharge diagnosis that was inconsistent with elective readmission (ie, heart failure, acute MI, unstable angina, arrhythmia, or cardiac arrest). According to the methodology described by Krumholz et al,¹² PCI or CABG procedures with other principal discharge diagnoses performed within 60 days of discharge were identified as staged procedures and were excluded. 

Statistical analysis. Patient characteristics, including demographics, clinical history and risk factors, catheterization laboratory visit details, diagnostic catheterization procedure, coronary anatomy details, PCI procedure details, lesions and devices, laboratory values, intra- and postprocedure events, and discharge details were compared across groups. Categorical variables were presented as frequencies (percentages) and differences by graft were assessed using the Chi-square test when the sample size was sufficient; otherwise, an exact test was used. Continuous variables were presented as median with interquartile range (IQR) and differences by graft were compared using the Kruskal-Wallis test.

To determine 1-year mortality, revascularization rates, rates of rehospitalization for MI, and rates of rehospitalization for stroke, cumulative incidence of first event by categories of PCI graft was calculated. Kaplan-Meier methods were used to calculate the cumulative incidence and the log-rank test was used to assess differences by PCI graft group. Cox models were used for mortality. The Fine-Gray model was used to calculate the cumulative incidence for non-fatal outcomes by accounting for death as a competing risk. Both the unadjusted and adjusted hazard ratios (HRs) comparing PCI graft to SVG graft to the LAD (reference) are presented. The adjustment covariates included were those from the validated CathPCI mortality risk model: gender; race; ST-segment elevation MI vs non-ST segment elevation MI; age ≤70 years; age >70 years; body mass index (BMI) ≤30 kg/m²; BMI >30 kg/m²; cerebrovascular disease; peripheral arterial disease; chronic lung disease; prior PCI; prior congestive heart failure; insulin-dependent diabetes vs no diabetes; non-insulin dependent diabetes vs no diabetes; glomerular filtration rate (GFR); renal failure or GFR<30 mL/min/1.73 m² or dialysis; left ventricular ejection fraction (LVEF); sustained shock and salvage; sustained shock or salvage; transient shock but not salvage; emergency PCI without shock/salvage; urgent PCI without shock/salvage; New York Heart Association (NYHA) class IV heart failure within 2 weeks; NYHA class I/II/III heart failure within 2 weeks; cardiac arrest within 24 hours; and at least one lesion previously treated in the past month now presenting with in-stent thrombosis.¹³ Note that spline functions were used for the modeling of LVEF, BMI, age, and GFR. Follow-up was defined to begin at date of discharge. Patients were censored at the end of fee-for-service eligibility for Medicare and end of follow-up. A P-value threshold of .05 was used to define statistical significance for all tests. All statistical analyses were performed by the Duke Clinical Research Institute using SAS, version 9.3. 

Results

Between July 2009 and December 2014, there were 1,812,321 CathPCI CMS link-eligible patients. Linkage was successful in 1,295,185 patients. Of these, 10,051 PCIs were performed on bypass grafts to the LAD: 6797 SVGs (67.6%); 3011 LIMA grafts (30.0%), and 243 other grafts (2.4%) (Supplemental Figure S1). Table 1 shows the clinical and procedural characteristics of the groups. Table 2 shows the presentation and indications for PCI. Of note, patients with SVG-PCI were more likely to present with acute coronary syndromes and more likely to be urgent or emergent procedures. Table 3 shows the periprocedural and procedural characteristics. LIMA patients were more likely to have distal anastomotic lesions, while SVGs were more likely to have lesions in the body of the graft (distal anastomosis: SVG 19.4%, LIMA 54.3%, other 28.0%; body of graft: SVG 55.2%, LIMA 27.2%, other 51.0%; P<.001 for both). Thrombus was more likely to be present for SVG vs LIMA or LAD (SVG 16.7%, LIMA 5.6%, other 6.2%; P<.001). SVGs were less likely to be treated with drug-eluting stent (DES) implantation compared with LIMA and other grafts (SVG 67.4%, LIMA 71.4%, other 73.7%; P<.001).

PCI outcomes and complications. Table 4 shows PCI outcomes and complications. Procedural success was similar between the groups (SVG 92.9%, LIMA 91.1%, other 93.4%; P=.65). However, the LIMA and other groups were more likely to be complicated by dissection compared with SVGs (SVG 0.7%, LIMA 2.8%, other 2.5%; P<.001). Redo CABG was also more common in the LIMA group (SVG 0.8%, LIMA 2.4%, other 0%; P<.001). Same-day and in-hospital mortality were similar in all three groups.  

At 1 year, mortality, MI, and repeat revascularization were all lower in arterial grafts (mortality: SVG 16.6%, LIMA 14.8%, other 11.8%; MI: SVG 9.9%, LIMA 6.6%, other 8.1%; revascularization: SVG 14.4%, LIMA 9.5%, other 10.4%; P<.001 for all). There was no significant difference in stroke rates. Cumulative incidence of events is shown in Figure 1. After multivariable adjustment, LIMA interventions had significantly lower MI rates (HR, 0.71; 95% CI, 0.60-0.84) and repeat revascularization at 1 year vs the SVG group (HR, 0.68; 95% CI, 0.59-0.79). There was no statistically significant difference in 1-year mortality between the three groups. 

Discussion

This is the largest study of procedural outcomes related to PCI of the LIMA graft to the LAD. Despite similar procedural success and in-hospital mortality rates compared with SVGs and other interventions, LIMA interventions were independently associated with lower rates of recurrent MI and repeat revascularization at 1 year.   

Graft interventions are generally more complex than interventions in native coronaries. Challenges include engagement of the graft. In SVGs, there is the additional risk of distal embolization (usually requiring the use of an embolic filter), while LIMA grafts have a propensity to dissect. Indeed, our results show that 16.7% of the SVGs had thrombus. In the LIMA group, dissection rates were significantly higher than in SVG interventions. Interestingly, there appeared to be a higher incidence of periprocedural MI in the other group. However, given the small number in the other group, interpretation of this remains speculative. Despite the technical differences in SVG and LIMA interventions, procedural success, in-hospital mortality, and major adverse outcomes, such as stroke, MI, bleeding, cardiogenic shock, and renal failure, were similar. The overall success rate reported is consistent with published literature.¹⁴ 

Our results also show differences in the location of the target lesion within the graft. For SVG patients, this was most commonly at the body of the graft (55.2%); while for the LIMA, this was most commonly at the distal anastomotic site (54.3%). These observations hint at the underlying pathophysiology of the disease. For SVGs, the probable mechanisms include thrombosis, neointimal hyperplasia, and accelerated atherosclerosis, consistent with the greater propensity of finding thrombus in these patients.^15-17 This is also consistent with the nature of the presentation of these patients, with SVG patients more likely to present with acute coronary syndromes (SVG 79.2% vs LIMA 70.3%) and as urgent, emergency or salvage cases, compared with LIMA patients (SVG 61.0% vs LIMA 52.2%). On the other hand, for the LIMA group, the preponderance of distal anastomotic lesions suggests that there may be a technical contribution from the bypass surgery.¹⁷ Alternatively, the distal anastomotic site may represent an area of greater shear stress or turbulence, predisposing to atherosclerosis.^18,19 

Intuitively, LIMA patients should have superior graft patency and better survival than SVG patients, even though the target vessel for both is the LAD.^20,21 It is therefore reassuring to find that intervention to the LIMA graft is associated with superior medium-term outcomes with lower rates of recurrent MI and repeat revascularization at 1 year. The different rates of recurrent MI and repeat revascularization are likely to be multifactorial. However, we note that SVG interventions were less likely to be treated with DES compared with the LIMA and other groups. DES implantation is well established with lower rates of restenosis and repeat revascularization and thus may contribute to this finding. On the other hand, there are mixed data on the benefits of DES in SVGs. As such, the use of DES in SVG may or may not have made a difference in restenosis rates.^22-24 Recurrent MI may be explained by a combination of different factors; we hypothesize that this includes degeneration and thrombus formation within SVGs, less use of DES with higher rates of restenosis, and perhaps more sluggish flow given the larger caliber of SVGs compared with arterial grafts.

While unadjusted mortality at 1 year was lower in patients who had PCI to arterial grafts compared with SVG grafts, this was no longer significant after multivariable adjustment. 

Study limitations. There are several limitations to this study. First, this is a retrospective study with limited data on the anatomical and procedural considerations, and no data on patients who did not undergo a procedure. We were also unable to specifically analyze patients who had intervention in the distal LAD beyond the anastomotic site. Nevertheless, the CathPCI database provides some of the most detailed procedural data available for analysis. Second, these data only examine patients who require a PCI to LAD grafts and not the entire cohort of patients who had bypass of the LAD. Therefore, inferences about the durability of LIMA-LAD and other grafts cannot be made. Third, the study could not examine the effect of operator experience in treating these lesions. Graft interventions are not common and limited experience can affect procedural outcomes.^25,26 In addition, the available data do not provide sufficient information as to whether subsequent events, such as death or MI, were related to the PCI to the graft. Next, this study examined only patients ≥65 years old and may not be applicable to younger patients. Finally, given the retrospective nature of this study, these results should only be considered hypothesis generating.

Conclusion

In patients who needed PCI to a graft to the LAD, despite similar procedural success rates and in-hospital mortality rates compared with SVG and other interventions, LIMA interventions were independently associated with lower rates of recurrent MI and repeat revascularization at 1 year. Given the retrospective nature of this study, further studies are required to validate these findings.

From the ¹UC Davis Medical Center, Sacramento, California; ²National Heart Center Singapore, Singapore; ³UC San Francisco Medical Center, San Francisco, California; ⁴Duke Clinical Research Institute, Durham, North Carolina; and ⁵San Francisco Veterans Affairs Medical Center, San Francisco, California. 

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Yeo reports research funding from Amgen, AstraZeneca, and Medtronic; personal fees from Medtronic, AstraZeneca, Amgen, Bayer, Boston Scientific, Abbott Vascular, Philips, Alvimedica, and Menarini. Dr Low reports advisory board income from Abbott Vascular and Boston Scientific. Dr Roe reports research funding from Eli Lilly, Sanofi-Aventis, Daiichi-Sanko, Janssen Pharmaceuticals, Ferring Pharmaceuticals, Myokardia, AstraZeneca, the American College of Cardiology, the American Heart Association, and the Familial Hypercholesterolemia Foundation; consulting income or honoraria from PriMed, AstraZeneca, Boehringer-Ingelheim, Merck, Actelion, Amgen, Myokardia, Eli Lilly, Novartis, Daiichi-Sanyko, Quest Diagnostics, and Elsevier Publishers. Dr Shunk reports grant support from Siemens Medical Systems, Svelte Medical, and Cardiovascular Systems, Inc; consulting income from TransAortic Medical and MedeonBio; non-financial support from the American College of Cardiology NCDR, outside the submitted work. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted December 13, 2019, provisional acceptance given December 19, 2019, final version accepted December 26, 2019.

Address for correspondence: Kendrick Shunk, MD, PhD, San Francisco Veterans Affairs Medical Center, 4150 Clement Street, San Francisco, CA 94121. Email: kendrick.shunk@va.gov

1. Farkouh ME, Domanski M, Sleeper LA, et al. Strategies for multivessel revascularization in patients with diabetes. N Engl J Med. 2012;367:2375-2384.
2. Serruys PW, Morice MC, Kappetein AP, et al. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med. 2009;360:961-972.
3. Hillis LD, Smith PK, Anderson JL, et al. 2011 ACCF/AHA guidelines for coronary artery bypass graft surgery: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2011;124:e652-e735.
4. Hlatky MA, Boothroyd DB, Bravata DM, et al. Coronary artery bypass surgery compared with percutaneous coronary interventions for multivessel disease: a collaborative analysis of individual patient data from ten randomised trials. Lancet. 2009;373:1190-1197.
5. Barner HB, Standeven JW, Reese J. Twelve-year experience with internal mammary artery for coronary artery bypass. J Thorac Cardiovasc Surg. 1985;90:668-675.
6. Bourassa MG, Enjalbert M, Campeau L, Lesperance J. Progression of atherosclerosis in coronary arteries and bypass grafts: ten years later. Am J Cardiol. 1984;53:102C-107C.
7. Cameron A, Davis KB, Green G, Schaff HV. Coronary bypass surgery with internal-thoracic-artery grafts — effects on survival over a 15-year period. N Engl J Med. 1996;334:216-219.
8. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med. 1986;314:1-6.
9. Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PC. Long-term (5 to 12 years) serial studies of internal mammary artery and saphenous vein coronary bypass grafts. J Thorac Cardiovasc Surg. 1985;89:248-258.
10. Sharma AK, McGlynn S, Apple S, et al. Clinical outcomes following stent implantation in internal mammary artery grafts. Catheter Cardiovasc Interv. 2003;59:436-441.
11. Gruberg L, Dangas G, Mehran R, et al. Percutaneous revascularization of the internal mammary artery graft: short- and long-term outcomes. J Am Coll Cardiol. 2000;35:944-948.
12. Krumholz HM, Normand SL. Public reporting of 30-day mortality for patients hospitalized with acute myocardial infarction and heart failure. Circulation. 2008;118:1394-1397.
13. Brennan JM, Curtis JP, Dai D, et al. Enhanced mortality risk prediction with a focus on high-risk percutaneous coronary intervention: results from 1,208,137 procedures in the NCDR (National Cardiovascular Data Registry). JACC Cardiovasc interv. 2013;6:790-799.
14. Marx R, Klein RM, Horlitz M, et al. Angioplasty of the internal thoracic artery bypass-graft an alternative to reoperation. Int J Cardiol. 2004;94:143-149.
15. Harskamp RE, Lopes RD, Baisden CE, de Winter RJ, Alexander JH. Saphenous vein graft failure after coronary artery bypass surgery: pathophysiology, management, and future directions. Ann Surg. 2013;257:824-833.
16. Kim FY, Marhefka G, Ruggiero NJ, Adams S, Whellan DJ. Saphenous vein graft disease: review of pathophysiology, prevention, and treatment. Cardiol Rev. 2013;21:101-109.
17. Gaudino M, Antoniades C, Benedetto U, et al. Mechanisms, consequences, and prevention of coronary graft failure. Circulation. 2017;136:1749-1764.
18. Ghista DN, Kabinejadian F. Coronary artery bypass grafting hemodynamics and anastomosis design: a biomedical engineering review. Biomed Eng Online. 2013;12:129.
19. Nordgaard H, Swillens A, Nordhaug D, et al. Impact of competitive flow on wall shear stress in coronary surgery: computational fluid dynamics of a LIMA-LAD model. Cardiovasc Res. 2010;88:512-519.
20. Boylan MJ, Lytle BW, Loop FD, et al. Surgical treatment of isolated left anterior descending coronary stenosis. Comparison of left internal mammary artery and venous autograft at 18 to 20 years of follow-up. J Thorac Cardiovasc Surg. 1994;107:657-662.
21. Goldman S, Zadina K, Moritz T, et al. Long-term patency of saphenous vein and left internal mammary artery grafts after coronary artery bypass surgery: results from a Department of Veterans Affairs cooperative study. J Am Coll Cardiol. 2004;44:2149-2156.
22. Brilakis ES, Edson R, Bhatt DL, et al. Drug-eluting stents versus bare-metal stents in saphenous vein grafts: a double-blind, randomised trial. Lancet. 2018;391:1997-2007.
23. Colleran R, Kufner S, Mehilli J, et al. Efficacy over time with drug-eluting stents in saphenous vein graft lesions. J Am Coll Cardiol. 2018;71:1973-1982.
24. Patel NJ, Bavishi C, Atti V, et al. Drug-eluting stents versus bare-metal stents in saphenous vein graft intervention. Circ Cardiovasc Interv. 2018;11:e007045.
25. Fanaroff AC, Zakroysky P, Dai D, et al. Outcomes of PCI in relation to procedural characteristics and operator volumes in the United States. J Am Coll Cardiol. 2017;69:2913-2924.
26. Fanaroff AC, Zakroysky P, Wojdyla D, et al. Relationship between operator volume and long-term outcomes after percutaneous coronary intervention. Circulation. 2019;139:458-472.