Response to Letter Regarding Article, “The Impact of Diabetic Status on Coronary Artery Bypass Graft Patency: Insights From the Radial Artery Patency Study”
We thank Dr Nezic and colleagues for their interest in our work. They suggest that our conclusions on the relative patency of radial artery (RA) and saphenous vein (SV) grafts in diabetic patients1 may differ if string sign grafts were considered occluded rather than patent. The predefined primary end point of the Radial Artery Patency Study2,3 was the proportion of RA and SV grafts that were completely occluded at follow-up angiography, where complete occlusion was defined as the absence of visible opacification of the target coronary vessel (ie, Thrombolysis in Myocardial Infarction [TIMI] 0 flow).4 Perfect graft patency (TIMI 3 flow) was a secondary angiographic end point, as was the presence of diffuse narrowing of the graft to <1 mm in diameter but with TIMI flow of at least 1 (ie, angiographic “string sign”). The study was planned in this manner because we expected the frequency of string sign to be <7.0%, the value observed. The inclusion criteria mandated that the proximal target vessel stenosis was >70%, and postoperative angiography was performed 1 year postoperatively rather than early after surgery. In addition, we considered complete occlusion a more definitive outcome.
Using our original prespecified end point definitions, RA grafting was protective in the diabetic cohort (RA: 11 of 115 occluded [9.5%] versus SV: 22 of 115 occluded [19.1%], McNemar corrected P=0.05; relative risk, 0.42; 95% confidence interval, 0.16 to 1.01) and the nondiabetic cohort (RA: 25 of 325 occluded [7.7%] versus SV: 38 of 325 occluded [11.7%], McNemar corrected P=0.11; relative risk, 0.63; 95% confidence interval, 0.35 to 1.10).
We repeated our analysis twice to incorporate broader definitions of imperfect grafts, as suggested by Nezic and colleagues. We were mindful that our study was not adequately powered to analyze these end points, which we assume would diminish differences between conduits but only that of absolute 1-year patency. In the first analysis, string sign grafts with TIMI 1 were considered occluded, whereas in the second analysis, every string sign graft was counted as occluded.
Using a broader imperfect graft definition of TIMI 0 or TIMI 1 + string sign, overall imperfect grafts were present in 51 of 440 RAs and 62 of 440 SVs (McNemar P=0.26). Among diabetic patients, imperfect grafts (TIMI 0 or TIMI 1 + string sign) were present in 13 of 115 RAs and 22 of 115 SVs (McNemar P=0.09). Among nondiabetic patients, imperfect grafts (TIMI 0 or TIMI 1 + string sign) were present in 38 of 325 RAs and 40 of 325 SVs (McNemar P=0.8).
Using an even broader imperfect graft definition of TIMI 0 or any string sign, overall imperfect grafts were present in 67 of 440 RAs and 64 of 440 SVs (McNemar P=0.8). Among diabetic patients, imperfect grafts (TIMI 0 or any string sign) were present in 17 of 115 RAs and 23 of 115 SVs (McNemar P=0.39). Among nondiabetic patients, imperfect grafts (TIMI 0 or any string sign) were present in 49 of 325 RAs and 40 of 325 SVs (McNemar P=0.37).
Although diabetes was associated with poorer graft patency in our analysis, it was not an important risk variable of radial string sign: 6 of 115 in diabetic patients versus 25 of 325 in nondiabetic patients. We previously showed this finding in multivariate analysis as well.5 In fully adjusted generalized estimating equation models of graft failure, diabetes was not a significant risk factor for either definition of imperfect grafts (TIMI 0 or TIMI 1 + string sign: relative risk =1.2, 95% confidence interval 0.9 to 1.7, P=0.27; TIMI 0 or any string sign: relative risk =1.2, 95% confidence interval 0.8 to 1.7, P=0.31). This result was expected given the increasing relative influence of string signs on the patency end point and the lack of association between diabetes and string sign.
In summary, we conclude that at 1 year RA grafts provide improved overall graft patency in diabetic and nondiabetic patients; diabetes does not influence the incidence of RA string signs; and broader definitions of imperfect grafts that include string signs show similar results between conduits regardless of diabetic status.
Singh S, Desai N, Petroff S, Deb S, Cohen E, Radhakrishnan S, Schwartz L, Dubbin J, Fremes SE, for the Radial Artery Patency Study Investigators. The impact of diabetic status on coronary artery bypass graft patency: insights from the radial artery patency study. Circulation. 2008; 118: S-222–S-225.
Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J, Dodge HT, Francis CK, Hillis D, Ludbrook P. Thrombolysis in myocardial infarction (TIMI) trial, phase I: a comparison between intravenous tissue plasminogen activator and intravenous streptokinase: clinical findings through hospital discharge. Circulation. 1987; 76: 142–154.