Coronary Artery Bypass Graft Patency and Survival in Patients on Dialysis.

BACKGROUND: Little is known about graft patency after coronary artery bypass grafting (CABG) performed in patients on dialysis. Our aim was to assess patency of internal thoracic artery (ITA) grafts and saphenous vein grafts (SVGs) in these patients. METHODS: From 1/1997 to 1/2018, 500 patients on dialysis underwent primary CABG with or without concomitant procedures at Cleveland Clinic, 40 of whom had 48 postoperative angiograms for recurrent ischemic symptoms. Complete follow-up was obtained on all but 1 patient lost to follow-up 1 y after CABG. Thirty-six ITA grafts and 65 SVGs were evaluable for stenosis and occlusion. RESULTS: Two of 40 patients (5%) had emergency CABG; 3 (7.5%) with calcified aortas had a change in operative strategy to avoid ascending aortic manipulation, 2 (5%) had poor conduit quality, and 12 (30%) had severe diffuse atherosclerotic disease with calcification of the coronary targets causing technical difficulties. Thirty-three patients (82%) were bypassed with an in situ ITA and 3 (7.5%) had a free ITA graft. Three of 36 ITA grafts were occluded at 0.78, 1.8, and 9.4 y (too few to model). SVG patency was 52% and 37% at 1 and 2 y, respectively. CONCLUSIONS: Among patients on dialysis who underwent CABG, coronary angiography for ischemic symptoms in a select subset revealed that SVG patency was lower than expected from published reports in the general CABG population and may contribute to the poor prognosis of this cohort. Further work is needed to guide graft selection and improve graft patency in dialysis patients.

Risks and Outcomes of Reoperative Cardiac Surgery in Patients With Patent Bilateral Internal Thoracic Artery Grafts.

BACKGROUND: Reoperative cardiac surgery in patients with patent bilateral internal thoracic artery (ITA) grafts is technically challenging. METHODS: From 2008 to 2017, of 7640 patients undergoing reoperative cardiac surgery, 116 (1.5%) had patent bilateral ITA grafts, including 28 with a right ITA crossing the midline. Mean age was 70 ± 9.6 years, and 111 patients (96%) were men. Reoperations included isolated coronary artery bypass grafting (n = 11), isolated valve (n = 55), valve + coronary artery bypass grafting (n = 26), and other procedures (n = 24). Clinical details, intraoperative management, and perioperative outcomes were analyzed. RESULTS: Aortic cannulation was central in 64 patients (56%) and through the femoral or axillary artery in 50 (44%). Four patients (3.4%) had planned transection and reattachment of ITAs crossing the midline, and 4 (3.4%) had ITA injuries, all right ITAs, 3 crossing the midline; 3 were repaired with an interposition vein graft, and 1 was managed by translocating the right ITA as a Y-graft off another graft. Patent ITAs were managed by atraumatic occlusion during aortic clamping in 90 patients (78%) and by systemic cooling without ITA occlusion in 19. There were 6 operative deaths, all due to low cardiac output syndrome (5.2%); 4 strokes (3.4%); and 5 cases of new postoperative dialysis (4.3%). CONCLUSIONS: Risk of injury to bilateral ITA grafts during reoperation is high, and right ITAs crossing the midline present a particular risk of injury and should inform planning for primary coronary artery bypass grafting. Risk of low cardiac output syndrome underscores the challenge of ensuring adequate myocardial protection.

Right Internal Thoracic Artery Patency Is Affected More by Target Choice Than Conduit Configuration.

BACKGROUND: Although coronary artery bypass grafting using bilateral internal thoracic arteries (ITA) maximizes long-term survival, knowledge of the effect of different right ITA (RITA) inflow configurations on graft patency is limited. We have compared RITA occlusion among these configurations and identified its risk factors while adjusting for outflow coronary target location. METHODS: From January 1972 to January 2016, of 7092 patients undergoing bilateral ITA grafting at a single center, 1331 received one ITA to the left anterior descending coronary artery and had one or more evaluable postoperative coronary angiograms: 835 (63%) in situ, 496 free RITA grafts (311 [63%] originating from aorta; 98 [20%] left ITA [LITA], 76 [15%] saphenous vein graft, 11 [2%] radial graft). RITA occlusion reported on 1983 angiograms performed a median of 5.8 years later was estimated using nonlinear mixed-effects longitudinal modeling. RESULTS: RITA patency was 90% at 1 year, 87% at 5 years, and 86% at 10 and 15 years. At 15 years, in situ RITA patency was 91% and free RITA patency from aorta was 91%, LITA 89%, and saphenous vein graft 77%. After adjusting for coronary target location and degree of stenosis, occlusion was similar in free RITAs from aorta (P = .15), LITA (P = .4), saphenous vein grafts (P = .13), and in situ RITAs. However, RITAs grafted to the left anterior descending coronary artery had fewer occlusions (P < .001), with patency similar to LITAs. CONCLUSIONS: Among patients with bilateral ITA grafting requiring interval coronary angiography, long-term RITA patency was high and independent of its inflow configuration. Therefore, priority should be a RITA configuration optimizing its reach to important coronary targets, including the left anterior descending coronary artery.

Coronary Artery Target Selection and Survival After Bilateral Internal Thoracic Artery Grafting.

BACKGROUND: The importance of a coronary artery, based on the myocardial mass it perfuses, is well documented, but little is known about the importance of a vessel that has been bypassed and its effect on survival in the context of bilateral internal thoracic artery (BITA) grafting. OBJECTIVES: This study determined the effect of a dominant left anterior descending (LAD) artery and important non-LAD targets on outcomes after BITA grafting. METHODS: From January 1972 to January 2011, of 6,127 patients who underwent BITA grafting, 2,551 received 1 ITA grafted to the LAD and had an evaluable coronary angiogram. A dominant LAD was defined as one that was wrapped around the left ventricular apex. Non-LAD targets were graded based on their terminal reach toward the apex: important: >75% (n = 1,698); and less important: ≤75% (n = 853). Mean follow-up was 14 ± 8.7 years. Multivariable analysis was performed to identify risk factors for time-related mortality. RESULTS: A dominant LAD was present more frequently in patients with less important additional targets (51% vs. 35%; p < 0.0001). A total of 179 patients (7.0%) received a second ITA to multiple targets, 77 (43%) of which were to multiple important target vessels. Unadjusted late survival was similar regardless of degree of importance of the second ITA target-77% at 15 years (p = 0.70) for the important and less important targets, respectively. In the multivariable model, grafting the second ITA to multiple important targets was associated with better long-term survival (p = 0.005). In patients with a nondominant LAD, a second ITA grafted to a less important artery was associated with higher risk of operative mortality (2.4% vs. 0.51%; p = 0.007). A saphenous vein graft to an important or less important target did not influence long-term survival. CONCLUSIONS: In BITA grafting, bypassing multiple important targets to maximize myocardium supplied by ITAs improved long-term survival. In patients with a nondominant LAD, selecting an important target for the second ITA lowered operative mortality.

Long-Term Patency of Individual Segments of Different Internal Thoracic Artery Graft Configurations.

BACKGROUND: Internal thoracic artery (ITA) grafts are the most durable conduits available for coronary artery bypass grafting (CABG). However, little is known about long-term angiographic outcomes of ITA grafts used in different configurations and whether sequential or Y grafting compromises patency of the inflow ITA graft. METHODS: From January 1972 to August 2016, 60,500 patients underwent primary isolated CABG, of whom 326 received ITA grafts placed in sequential or Y configuration and were studied angiographically (median 4.8 years to first follow-up angiogram). Each sequential or Y segment was studied individually using a mixed-effects longitudinal model with the patient as the random effect. RESULTS: At 15 years, patency of the proximal ITA segment (n = 331) was 99%; of a sequential segment (n = 222), 97%; and of the segment beyond anastomosis of a Y graft (n = 109), 99%. Patency of the Y grafts (n = 109) was 92% at 5 years, 91% at 10 years, and 90% at 15 years. After adjusting for proximal stenosis and graft location, Y grafts were associated with greater occlusion than the inflow segment of ITA grafts (odds ratio; 51, 95% confidence interval, 6.1 to 422; p = 0.003) and of sequential grafts (odds ratio, 12; 95% confidence interval, 1.14 to 120; p = 0.04). CONCLUSIONS: Long-term patency of ITA grafts in sequential or Y configuration is similar qualitatively, but not quantitatively, to the known patency of single ITA-to-left anterior descending grafts. Sequential or Y grafting does not compromise patency of the inflow portion of an ITA graft. Y-graft patency is lower than sequential graft patency but is still better than known patency of saphenous vein grafts.

Natural History of Moderate Coronary Artery Stenosis After Surgical Revascularization.

BACKGROUND: It remains controversial whether grafting moderately stenosed coronary arteries (MSCAs) influences native-vessel disease progression and whether grafting may protect against late myocardial ischemia. METHODS: From 1972 to 2011, 55,567 patients underwent primary isolated coronary artery bypass grafting (CABG); 1,902 had a single coronary artery with angiographically moderate (50% to 69%) stenosis and ≥1 postoperative angiogram. Disease progression was studied in 489 nongrafted, 371 internal thoracic artery (ITA)-grafted, and 957 saphenous vein (SV)-grafted MSCAs, as well as patency of 376 ITA and 1,016 SV grafts to these MSCAs. RESULTS: At 1, 5, 10, and 15 years, native-vessel disease progressed from moderate to severe stenosis/occlusion in 32%, 52%, 66%, and 72% of nongrafted MSCAs; 55%, 73%, 84%, and 87% of ITA-grafted MSCAs; and 67%, 82%, 90%, and 92% of SV-grafted MSCAs. After adjusting for patient characteristics, MSCA disease progressed 3.6 times faster with ITA and 10 times faster with SV grafting compared with nongrafting. At these same time points, occlusion of ITA grafts to MSCAs was 8%, 9%, 11%, and 15% and for SV grafts, 13%, 32%, 46%, and 56%; protection from myocardial ischemia by ITA-grafted versus nongrafted MSCAs was 29%, 47%, 59%, and 61%. CONCLUSIONS: Most MSCAs progress to severe stenosis or occlusion in the long term. Progression is faster in grafted than nongrafted MSCAs, more so with SV than ITA grafts. However, ITA grafts to such arteries have excellent patency, providing long-term protection from myocardial ischemia. Therefore, ITA grafting of MSCAs should be considered.

Similar Outcomes in Diabetes Patients After Coronary Artery Bypass Grafting With Single Internal Thoracic Artery Plus Radial Artery Grafting and Bilateral Internal Thoracic Artery Grafting.

BACKGROUND: The purpose of this study was to determine in patients with diabetes mellitus whether single internal thoracic artery (SITA) plus radial artery (RA) grafting yields outcomes similar to those of bilateral internal thoracic artery (BITA) grafting. METHODS: From January 1994 to January 2011, 1,325 diabetic patients underwent primary isolated coronary artery bypass graft surgery with either (1) SITA plus RA with or without saphenous vein (SV) grafts (n = 965) or (2) BITA with or without SV grafts (n = 360); an internal thoracic artery was used in all patients to graft the left anterior descending coronary artery. Endpoints were in-hospital outcomes and time-related mortality. Median follow-up was 7.4 years, with a total follow-up of 9,162 patient-years. Propensity score matching was performed to identify 282 well-matched pairs for adjusted comparisons. RESULTS: Unadjusted in-hospital mortality was 0.52% for SITA plus RA with or without SV grafts and 0.28% for BITA with or without SV grafts, and prevalence of deep sternal wound infection was 3.2% and 1.7%, respectively. Unadjusted survival at 1, 5, 10, and 14 years was 97%, 88%, 68%, and 51% for SITA plus RA with or without SV grafts, and 97%, 95%, 80%, and 66% for BITA with or without SV grafts, respectively. Among propensity-matched patients, in-hospital mortality (0.35% versus 0.35%) and prevalence of deep sternal wound infection (1.4% versus 1.4%) were similar (p > 0.9) in the two groups, as was 1-, 5-, 10-, and 14-year survival: 97%, 90%, 70%, and 58% for SITA plus RA with or without SV grafting versus 97%, 93%, 79%, and 64% for BITA with or without SV grafting, respectively (early p = 0.8, late p = 0.2). CONCLUSIONS: For diabetic patients, SITA plus RA with or without SV grafting and BITA with or without SV grafting yield similar in-hospital outcomes and long-term survival after coronary artery bypass graft surgery. Therefore, both SITA plus RA and BITA plus SV grafting should be considered for these patients.

Value of surgery for infective endocarditis in dialysis patients.

OBJECTIVES: To determine the value of surgery for infective endocarditis (IE) in patients on hemodialysis by comparing the nature and invasiveness of endocarditis in hemodialysis and nonhemodialysis patients and their hospital and long-term outcomes, and identifying risk factors for time-related mortality after surgery. METHODS: From January 1997 to January 2013, 144 patients on chronic hemodialysis and 1233 nonhemodialysis patients underwent valve surgery for IE at our institution. Propensity matching identified 99 well-matched hemodialysis and nonhemodialysis patient pairs for comparison of outcomes. RESULTS: Staphylococcus aureus infection was more common in hemodialysis patients than in nonhemodialysis patients (42% vs 21%; P < .0001), but invasive disease was similar in the 2 groups (47%; P = .3). Hospital mortality was 13% and 5-year survival was 20% for hemodialysis patients, 20% below that expected in a general hemodialysis population but 15% above that of hemodialysis patients treated nonsurgically for IE. For matched patients, hospital mortality was 13% for hemodialysis patients versus 5.1% for nonhemodialysis patients (P = .05), and survival at 1 and 5 years was 56% versus 83% and 24% versus 59%, respectively (P < .004). Use of an arteriovenous graft for dialysis access (P = .01) and preoperative placement of a pacemaker (P < .0001) were risk factors for late mortality in hemodialysis patients. For matched patients, freedom from reoperation was similar in the hemodialysis and nonhemodialysis groups (P > .9). CONCLUSIONS: Intermediate-term survival after surgery for IE in hemodialysis patients is substantially worse than that in nonhemodialysis patients, but only slightly worse than that in the general hemodialysis population and substantially better than that in hemodialysis patients with IE treated nonsurgically, supporting continued surgical intervention for IE.