long term patency

LVGRSGNST

Tdvuoptsro51poaw1e

APPW

a

Journal of the American College of Cardiology Vol. 44, No. 11, 2004© 2004 by the American College of Cardiology Foundation ISSN 0735-1097/04/$30.00P

Downloa

ong-Term Patency of Saphenousein and Left Internal Mammary Arteryrafts After Coronary Artery Bypass Surgery

esults From a Department of Veterans Affairs Cooperative Studyteven Goldman, MD, FACC,* Karen Zadina, RN, MA,* Thomas Moritz, MS,† Theron Ovitt, MD,*ulshan Sethi, MD,* Jack G. Copeland, MD,* Lizy Thottapurathu, MS,† Barbara Krasnicka, PHD,†ancy Ellis, MS,† Robert J. Anderson, PHD,† William Henderson, PHD,† for the VA Cooperativetudy Group #207/297/364ucson, Arizona; and Hines, Illinois

OBJECTIVES This study defined long-term patency of saphenous vein grafts (SVG) and internal mammaryartery (IMA) grafts.

BACKGROUND This VA Cooperative Studies Trial defined 10-year SVG patency in 1,074 patients and leftIMA patency in 457 patients undergoing coronary artery bypass grafting (CABG).

METHODS Patients underwent cardiac catheterizations at 1 week and 1, 3, 6, and 10 years after CABG.RESULTS Patency at 10 years was 61% for SVGs compared with 85% for IMA grafts (p � 0.001). If

a SVG or IMA graft was patent at 1 week, that graft had a 68% and 88% chance, respectively,of being patent at 10 years. The SVG patency to the left anterior descending artery (LAD)(69%) was better (p � 0.001) than to the right coronary artery (56%), or circumflex (58%).Recipient vessel size was a significant predictor of graft patency, in vessels �2.0 mm indiameter SVG patency was 88% versus 55% in vessels �2.0 mm (p � 0.001). Other positivesignificant predictors of graft patency were use of aspirin after bypass, older age, lower serumcholesterol, and lowest Canadian Functional Class (p � 0.001 to 0.058).

CONCLUSIONS The 10-year patency of IMA grafts is better than SVGs. The 10-year patency for SVGs isbetter and the 10-year patency for IMA grafts is worse than expected. The 10-year patencyof SVGs to the LAD is better than that to the right or circumflex. The best long-termpredictors of SVG graft patency are grafting into the LAD and grafting into a vessel that is�2.0 mm in diameter. (J Am Coll Cardiol 2004;44:2149–56) © 2004 by the American

ublished by Elsevier Inc. doi:10.1016/j.jacc.2004.08.064

College of Cardiology Foundation

wy

M

SCA1hRiwSssailAiaida

he success of coronary artery bypass grafting (CABG) isependent on the long-term patency of the arterial andenous grafts. Worldwide, more than 800,000 patientsndergo CABG annually, with more than 350,000 patientsperated each year in the U.S. The majority of theseatients receive left internal mammary artery (IMA) graftso the left anterior descending (LAD) coronary artery andaphenous vein grafts (SVGs) or other conduits to theemaining vessels. Based on small studies of selected groupsf patients, it is generally believed that SVGs have a 40% to0% 10-year patency and that the IMA has a 90% to 95%0-year patency (1–7). To address the question of long-termatency of saphenous vein and IMA grafts, the Departmentf Veterans Affairs Cooperative Studies Program supported10-year angiographic follow-up study of patients who

ere participants in two VA Cooperative Trials in the980s. These trials were originally designed to examine theffect of antiplatelet agents on graft patency after CABG

From the *Southern Arizona VA Health Care System and the University ofrizona Sarver Heart Center, Tucson, Arizona; and the †VA Cooperative Studiesrogram Coordinating Center, Hines, Illinois. Supported by the Cooperative Studiesrogram of the Department of Veterans Affairs Research and Development Service,ashington, DC.

nManuscript received May 6, 2004; revised manuscript received August 16, 2004,

ccepted August 25, 2004.

ded From: http://content.onlinejacc.org/pdfAccess.ashx?url=/data/Journals/

ith postoperative coronary angiography one week, oneear, and in some cases three years after surgery (8–12).

ETHODS

tudy population. This trial, organized and funded by theooperative Studies Program of the Department of Veteransffairs Research and Development Service, obtained data from,254 male patients entered into two studies at 13 participatingospitals from July 1983 to September 1988. The Institutionaleview Board approved each study, the subjects gave written

nformed consent, and procedures were followed in accordanceith institutional guidelines (8–13).urgery. Coronary artery bypass grafting was done in atandard fashion as previously described (8–12). The deci-ion to use arterial or venous conduits was made by thettending surgeon. The patient was eligible for the trial onlyf the determination was made preoperatively to utilize ateast one SVG.

ngiographic analysis. The angiographic analysis wasdentical to our earlier trials (8–12). Briefly, the left IMAnd each aortic anastomosis were selectively engaged andnjected. When the status of a vein graft could not beetermined by graft or stump injection, an aortic rootngiogram was performed. Selective angiography of the
ative coronary arteries was performed during the one-week
JAC/23029/ on 10/06/2013

acwtctresvfabipSgfsri

poi

(Srottoaatog

Fo(fig

Fo((*

Fo(

2150 Goldman et al. JACC Vol. 44, No. 11, 2004Long-Term SVG and IMA Graft Patency December 7, 2004:2149–56

Downloa

nd one-year catheterization only when a graft was oc-luded. Selective angiography of the native coronary arteriesas routinely performed at the 3-year and 10-year cathe-

erization. Graft patency and stenosis were defined in aentral angiographic laboratory with a computer-based sys-em (8–12). The data are presented as time-based occlusionates where occlusion is defined in two ways. The primaryvent of interest is 100% stenosis (Figs. 1 to 4). Theecondary event is 50% to 99% stenosis, not includingessels totally occluded (Fig. 5). Because of long-termollow-up and the study group being composed of middle-ged men with coronary artery disease, some patients diedefore the completion of the study. We obtained autopsiesn as many patients as possible and included the autopsyatency data in our analysis.tatistical analysis. Traditionally, reports of long-termraft patency rates have used coronary angiography resultsrom a single time point after CABG. Patency rates from aingle time period distant from the original operation mayesult in biased estimates if there is no accounting fornterceding interventions.

The data from this study posed two major analyticroblems. The first problem was that the exact time of graftcclusion could not be known. We addressed this by usingnterval-censored observations in the survival analysis

Abbreviations and AcronymsCABG � coronary artery bypass graftCX � circumflexIMA � internal mammary arteryLAD � left anterior descendingRCA � right coronary arterySVG � saphenous vein graft

igure 1. Plot of time-related graft patency (or freedom from graftcclusion) for saphenous vein grafts (SVG) and internal mammary arteryIMA) grafts. The number of patients at each time point is listed in the
gure. *p � 0.001 (IMA vs. SVG). CABG � coronary artery bypassrafting.
oC


PROC LIFETEST in SAS [SAS Institute Inc. SAS/TAT Version 8, Cary, North Carolina]). This analysisequires identification of the time interval in which thecclusion occurred. If a patient had only one angiogram andhis showed an occluded graft, the time between the date ofhe CABG and date of angiography was used as thecclusion occurrence interval. If a patient had multiplengiograms, the time between the date of the most recentngiogram that showed a graft was patent and the date ofhe angiogram that showed occlusion was used as thecclusion occurrence interval. The date of the latest angio-ram was used as the right censored time for grafts, which

igure 2. Plot of time-related graft patency (or freedom from graftcclusion) for saphenous vein grafts (SVG) and internal mammary arteryIMA) grafts if the graft was patent at one week after coronary bypassCABG). The number of patients at each time point is listed in the figure.p � 0.001 (IMA vs. SVG).

igure 3. Plot of time-related graft patency (or freedom from graftcclusion) for saphenous vein grafts (SVG) to the left anterior descendingLAD), circumflex (CX), and right coronary (RCA) arteries. The number
f patients at each time point is listed in the figure. *p � 0.001 (LAD vs.X and/or RCA). CABG � coronary artery bypass grafting.
JAC/23029/ on 10/06/2013

raipcaCw

tna

cobiote

gsnpbbcrtrpdvttt

p1

R

Pohipp3caetc6ty6fc1iahdtt

Fodp�

FmaeL

2151JACC Vol. 44, No. 11, 2004 Goldman et al.December 7, 2004:2149–56 Long-Term SVG and IMA Graft Patency

Downloa

emained patent. For patients who had patency data from anutopsy, the date of death was used. If there was anntervention on a stenotic but still patent graft by eitherercutaneous coronary intervention or repeat surgical revas-ularization, the graft was right censored on the date of thengiogram and no further data were collected for that graft.omparisons of Kaplan-Meier product-limit survival curvesere made with the log-rank test.Although time-related analyses of graft patency data used

he exact date of each postoperative angiogram, for conve-ience of presentation, some information is presented inrbitrarily defined time frames. The 1-week data included

igure 4. Plot of time-related graft patency (or freedom from graftcclusion) for saphenous vein grafts (SVG) to recipient vessel withiameters �2.0 mm versus �2.0 mm. The number of patients at each timeoint is listed in the figure. *p � 0.001 (�2.0 mm vs. �2.0 mm). CABG

coronary artery bypass grafting.

igure 5. Plot of time to development of 50% to 99% stenosis in internalammary artery (IMA) and single saphenous vein graft (SVG) to the left

nterior descending coronary artery (LAD). The number of patients at

dach time point is listed in the figure. *p � 0.001 (IMA vs. single SVG toAD). CABG � coronary artery bypass grafting.


atheterizations performed between 1 and 60 days post-peration, the 1-year included catheterizations performedetween 61 days and 1.5 years post-operation, the 3-yearncluded catheterizations after 1.5 to 4.5 years post-peration, the 6-year included catheterizations after 4.5 upo 8 years post-operation, and the 10-year included cath-terizations after 8 to 12 years post-operation.

The second analytic problem was that there were multiplerafts (i.e., clustered observations, within a patient). Wehowed previously (14) that graft patency within a patient isot independent. This does not affect the estimates foratency rates, but it does cause the standard error terms toe underestimated. Recently, the SAS macro IWM (16) hasecome available for the analysis of clustered, interval-ensored survival data (15,16). This approach producesobust estimates of the standard error terms by adjusting forhe correlated nature of the clustered observations. Patient-elated risk variables, graft-related risk variables, and CABGrocesses of care variables were used as candidate indepen-ent variables in the IWM macro to identify the set ofariables that jointly predict 10-year graft patency. Variableshat were significant at the p � 0.10 level were included inhe final models. The Appendix gives a description on howo interpret the importance of the variables.

The t test and chi-square test were used to comparere-CABG patient risk factors for patients who received a0-year catheterization versus those who did not.

ESULTS

atient data. Of the 1,254 patients undergoing CABGperations during the study period, 1,079 patients (86%)ad at least one postoperative cardiac catheterization. This

ncluded 1,074 patients with one or more SVG and 457atients with an IMA graft. Two hundred twenty-fiveatients had a single postoperative catheterization (18%),28 patients had two catheterizations (26%), 384 had threeatheterizations (31%), 119 had four catheterizations (9%),nd 23 had five catheterizations (2%). Patient status duringach of the five angiographic periods is given in Table 1. Aotal of 919, 799, 367, 170, and 369 patients had cardiacatheterizations nominally at the 1-week, 1-year, 3-year,-year, and 10-year periods, respectively. The median timeso the catheterizations were 8 days and 1.0, 3.0, 6.2, and 9.7ears, respectively. Ninety-six and sixty-seven percent of the- and 10-year catheterizations, respectively, were per-ormed for clinical, not study-related, reasons. The 10-yearatheterizations occurred between May 1991 and January999. There were 523 deaths before the 10-year catheter-zation. Of these 523 patients, serial catheterizations orutopsy data were obtained on 396 patients (76%). Oneundred thirty patients had a single catheterization beforeeath (25%), 168 had two catheterizations (32%), 87 hadhree catheterizations (17%), and 11 had four catheteriza-ions (2%). There were autopsy data with graft patency
efined on 71 patients. The details of the patient data,
JAC/23029/ on 10/06/2013

ioi(tb�0O(6wpCwnggpSawc

(dv0piRL(Pd1v�vtmIvcppl

T

PPD

T 1,


Downloa

ncluding clinical outcomes and complications of the previ-us catheterizations, have been reported (8–12). The clin-cal characteristics of patients with SVGs and IMA graftsTable 2) are presented. The patients who did not undergohe 10-year catheterization had a higher incidence of dia-etes (p � 0.022) and a trend toward more hypertension (p

0.057) and a higher incidence of left main disease (p �.084) (Table 3).verall patency. The 10-year patency for SVGs was 61%

Fig. 1). If a vein graft was patent at the 1-week study, the-year patency rate was 76% and the 10-year patency rateas 68% (Fig. 2). We note that the sudden drop inercentage patency in the graphs around six months post-ABG occurs because a large amount of new informationas obtained at the one-year angiogram and hence shouldot be viewed as having physiologic relevance. For IMArafts, the 10-year patency was 85% (Fig. 1). If an IMAraft was patent at 1 week, the 6-year and the 10-yearatency rates were 90% and 88%, respectively (Fig. 2).elected distal sites. The data for a specific vessel graftedt 10 years show that vein graft patency to the LAD (69%)as better (p � 0.001) than vein graft patency to the right

oronary artery (RCA) (56%) or circumflex (CX) (58%)

able 1. Status of Study Patients at Each Angiographic Period

7–10-DayAngiogram

atients having coronary angiography 919atients alive but not having angiography 302eaths 33

otals 1,254

Table 2. Characteristics of Patients Undergoinor IMA Grafts

Variable n S

Age (yrs) 368Cholesterol (mg/dl) 338 2Triglycerides (mg/dl) 316 2HDL (mg/dl) 220Left main disease 366Diseased vessels

1 162 823 270

CFCI 8II 55III 132IV 169

Angina 368Hypertension 367Prior MI 368Diabetes 368Smoking history

Never 79Previous 186Current 101

CABG � coronary artery bypass graft; CFC � Canadian Functionmammary artery; MI � myocardial infarction; SVG � saphenous


Fig. 3). There was no difference in the recipient vesseliameters among SVGs to the LAD (1.67 � 0.43 mm)ersus the CX (1.68 � 0.44 mm) versus the RCA (1.71 �.50 mm), p � 0.179. If the vein graft to the LAD wasatent at 1 week, the 10-year patency was 74%. Similarmprovement was seen at 10 years with early patency for theCA (64%) and for the CX (66%). Left IMA grafts to theAD had a higher patency than single SVGs to the LAD

85% vs. 69%, p � 0.001).redictors of graft patency. Along with the location of theistal site, the best predictor of graft patency over the0-year post-bypass period was the diameter of the recipientessel by angiographic measurement (Fig. 4). In vessels2.0 mm the 10-year patency was 88%, versus 55% in

essels with diameters �2.0 mm (p � 0.001). For SVGs tohe LAD, the 10-year patency was 90% for vessels �2.0m versus 52% for vessels �2.0 mm (p � 0.001). For the

MA, the 10-year patency was 100% for vessels �2.0 mmersus 82% for vessels �2.0 mm (p � 0.008). Afterontrolling for hospital effects, other positive significantredictors of graft patency over the 10-year post-bypasseriod were older age, use of aspirin after CABG surgery,

ower serum cholesterol, and lowest Canadian Functional

earogram

3-YearAngiogram

6-YearAngiogram

10-YearAngiogram

799 367 170 369389 738 797 36266 149 287 523

254 1,254 1,254 1,254

Year Post-CABG Catheterization by SVG

atients n IMA Patients

6.8 137 59.0 � 6.845.3 124 226.8 � 47.4118.5 114 230.7 � 137.611.0 76 35.3 � 9.0

.6 137 15.3

.4 5 3.7

.3 32 23.4

.4 100 73.0

.2 2 1.5

.1 22 16.4

.3 40 29.9

.4 70 52.2

.1 137 98.5

.2 137 46.0

.0 137 56.2

.7 137 6.6

.6 33 24.1

.8 66 48.2

.6 38 27.7

1-YAngi

g 10-

VG P

59.2 �24.4 �24.8 �35.7 �

12

42273

21536469845565

215027

al Class; HDL � high-density lipoprotein; IMA � internalvein graft.

JAC/23029/ on 10/06/2013

Cppdcppdvtpvh

idgabcPoc9gas


Downloa

lass (Table 4). If a graft was patent at one week, significantositive predictors of graft patency rates over the 10-yearost-bypass period included IMA grafts to the LAD, largeriameters of the recipient vessel, older age, and lower serumholesterol (Table 5). Lower platelet count was mildlyredictive (p � 0.089) of graft patency. Interestingly, theresence of diabetes requiring insulin and cigarette smokingid not predict graft patency independent of whether theein graft was patent early or not. Other variables that wereested in the model that were not predictive of vein graftatency included the following: race, number of diseasedessels, presence of peripheral edema, systemic emboli,ypertension, white blood cell count, prior myocardial

Table 3. Characteristics of Patients UndergoinThose Who Did Not Undergo a 10-Year Post

Variable nNo 10-Ye

Catheteriza

Age (yrs) 883 59.3 � 7.Cholesterol (mg/dl) 812 221.7 � 50Triglycerides (mg/dl) 753 217.1 � 14HDL (mg/dl) 571 34.4 � 10Grafted sites 882 3.5 � 1.Left main disease 878 16.4Diseased vessels

1 23 2.62 178 20.13 684 77.3

CFCI 39 4.5II 130 14.9III 280 32.2IV 421 48.4

Angina 882 96.8Hypertension 882 51.0Prior MI 882 60.5Diabetes 882 9.6Smoking history

Never 173 19.7Previous 430 48.9Current 277 31.5

Abbreviations as in Table 1.

Table 4. Predictors of Long-Term Graft Occlu

VariableResponse

Level

Graft location SVGIMA

Distal anastomosis diameter �2.0 mm�2.0 mm

Age Per 10 yrsRandomized treatment Placebo

AspirinCholesterol Per 50 mg/dCanadian Functional Class I

II–IV

For each variable, the acceleration factor represents the changefactor �1.0 means that an increase in the value of the assoocclusion; an acceleration factor �1.0 means that an increaseto graft occlusion. For example, adjusting for all other variabthe time to graft occlusion. Thus, if a 50-year old man will o
graft at 1.84 � 5 or 9.2 years, all other variables being equal.
CI � confidence interval; SVG � saphenous vein graft.


nfarction, proximal anastomotic site, temperature of car-ioplegia solution, whether preservation solution given viaraft or not, heart size, use of positive inotropic, vasodilator,ntiarrhythmic agent perioperatively, use of an intraaorticalloon, lowest body temperature, and total volume ofardioplegia.resence of disease within grafts. The primary end pointf this study was the graft patency at 10 years, but diseaselearly develops in grafts. In SVGs, which developed 50% to9% stenosis by 10 years, there was no difference amongrafts to the LAD, CX, and RCA (17% to 18%). There issignificant difference in IMA grafts with 50% to 99%

tenosis compared with SVGs to the LAD (5% vs. 22%,

Year Post-CABG Catheterization andBG Catheterization

n10-Year

Catheterization p Value

369 59.2 � 6.8 0.835339 224.4 � 45.2 0.406317 224.5 � 118.4 0.387221 35.7 � 11.0 0.144369 3.5 � 1.1 0.823367 12.5 0.084

0.10917 4.682 22.2

270 73.20.176

8 2.255 15.1

132 36.2170 46.6369 98.1 0.211368 45.1 0.057369 55.8 0.122369 5.7 0.022

0.36679 21.5

187 51.0101 27.5

Using All Grafts

Acceleration Factor(95% CI) p Value

Reference �0.00122.77 (9.83–52.72)Reference �0.00147.66 (18.91–120.15)1.84 (1.35–2.51) �0.001

Reference 0.0152.01 (1.15–3.53)0.76 (0.60–0.96) 0.022

Reference 0.0580.64 (0.40–1.02)

e to graft occlusion for the increment noted. An accelerationvariable leads to an increase in the length of time to graft

value of the associated variable leads to a decrease in the timeevery increase in 10 years of age, there is a 1.84 increase inhis graft at 5 years, then a 60-year old man will occlude his

g 10--CA

artion

9.96.5.80

sion

l

in timciatedin theles, forcclude

JAC/23029/ on 10/06/2013

p(s

D

TppsyIorSys8lSbg

rSgspipcrrotafucl

f(yaifssatsbatdT6rtebbh

wnotlispsi

uch


Downloa

� 0.001) and with IMA grafts compared with all SVGs5% vs. 18%, p � 0.001). The difference appears to startomewhere after three years postoperatively (Fig. 5).

ISCUSSION

hese data report 10-year serial angiographic follow-up ofatients operated on in the 1980s. For SVGs, the 10-yearatency is 61%, and if a SVG is patent at 1 week afterurgery, that graft has a 68% chance of being patent at 10ears. For IMA grafts, the 10-year patency is 85% and if anMA graft is patent at 1 week, that graft has a 88% chancef being patent at 10 years. Combining the data in thiseport with our previous work (8–12), we know that serialVG patency is 95% at 1 week, 84% at 1 year, 80% at 3ears, 69% at 6 years, and 61% at 10 years (Fig. 1). Theerial data for IMA grafts are 99%, 95%, 93%, 87%, and5% at each of the same time points. Although theong-term patency of IMA grafts is better than that forVGs, the 10-year patency of SVGs is better than what haseen previously reported, and the 10-year patency of IMArafts is somewhat worse than previously thought.

The most often quoted results in published reportsegarding graft patency after CABG are that at 10 years,VGs have about a 40% to 50% patency and that IMArafts have a 90% to 95% 10-year patency. The initialtudies that are the basis of this information were notrospective and reported on selected patients operated uponn the 1970s (1–6). In these retrospective studies, graftatency was not determined on everyone in the initialohort, but rather on those patients who received angiog-aphy most often because of symptoms. In 1996 there was aeport from a larger database, which included a combinationf first surgery and re-operation surgery (7). Even thoughhat trial was designed to follow patients prospectively, theuthors state that their plans “were thwarted by lack ofunds” and their graft patency data, in the subset thatnderwent catheterization, were similar to the numbersited above. When the present study was planned, the

Table 5. Predictors of Long-Term Graft Occlu

VariableResponse

Level

Graft location SVGIMA

Distal anastomosis diameter �2.0 mm�2.0 mm

Age Per 10 yearCholesterol Per 50 mg/Platelet count Per 100,000

For each variable, the acceleration factor represents the changefactor �1.0 means that an increase in the value of the assoocclusion; an acceleration factor �1.0 means that an increaseto graft occlusion. For example, adjusting for all other variabin the time to graft occlusion by a factor of 0.82. Thus, if agraft at 5 years, then a man with a serum cholesterol level ofvariables being equal.

IMA � internal mammary artery; other abbreviations as

ongest angiographic follow-up studies of IMA grafts were b


rom the Cleveland Clinic (6,17,18), Karolinska Hospital19), and the Montreal Heart Institute (3,20). At or past 10ears, each of these centers has reported studying 37, 39,nd 20 patients, respectively. Although these reports aremportant and are obviously responsible for the enthusiasmor using IMA grafts today, they are relatively small retro-pective studies. Perhaps the most telling criticism of thesetudies is that the type of patients operated on 10 to 15 yearsgo is not the same as patients undergoing CABG now. Inhe report from the Karolinska Hospital, the 37 patientstudied at 11 years had an 88% IMA patency rate (19). Theaseline characteristics of this patient population includedn average ejection fraction of 64% and 42 of 99 (42%) ofhe original patients had only IMA grafts. This obviouslyoes not represent the current cardiac surgical population.he patients in our study had average ejection fractions of1%, but they received 2.9 vein grafts per patient. Anothereason why the earlier data may not be applicable now ishat the patient population operated on today has morextensive disease than those operated on in the 1970secause today surgeons must operate on patients who haveeen denied percutaneous coronary intervention or areaving second and third revascularization procedures.All of the studies from the 1970s were done prior to the

idespread use of antiplatelet therapy after CABG, and it isow well established that aspirin helps to prevent SVGcclusion (8–12). In the 1970s, very little attention was paido risk factor control and aggressive lowering of low-densityipoprotein cholesterol, which may alter SVG patency andmprove clinical outcomes after CABG (21–23). Lastly,urgical techniques have been modified with regard toreparation of the vein graft before implantation; for in-tance, we know that cold vein graft preservation solutionmproves long-term vein graft patency (13).

Comparisons are often made between veins and arteriessed as conduits, with the general belief that arterialonduits have a better long-term patency. In fact, the IMAas been proclaimed the graft of choice to bypass the LAD

Given Grafts Were Open Early

Acceleration Factor(95% CI) p Value

Reference �0.0015.11 (2.85–9.18)Reference �0.0016.78 (3.86–11.89)1.28 (1.04–1.58) 0.0210.82 (0.69–0.98) 0.0310.82 (0.66–1.01) 0.069

e to graft occlusion for the increment noted. An accelerationvariable leads to an increase in the length of time to graft

value of the associated variable leads to a decrease in the timer every 50-mg/dl increase in cholesterol, there is a decreaseith a serum cholesterol level of 200 mg/dl will occlude his

g/dl will occlude his graft at 0.82 � 5 or 4.1 years, all other

le 4.

sion

sdl

in timciatedin theles, foman w250 m

in Tab

ecause of its presumed excellent long-term patency rate. A

JAC/23029/ on 10/06/2013

1sisiaIgsTohIgdrgscLAtsgttfapcg

Wsplrttsttpytnrucigeptua

hl

ppps(ft3tcrsnfiaaopraLga

ednapatac1dwign1aola

Ip1lip


Downloa

986 editorial in the New England Journal of Medicine (23)tated, “at 10 years internal mammary artery grafts remainedn excellent condition; nearly 95% were patent, with noigns of deterioration.” Although this suggests that the IMAs an ideal conduit, it is important to emphasize that therere no randomized prospectively controlled trials comparingMA to SVGs. In the past surgeons tended to use IMArafts in better candidates (i.e., the patient was clinicallytable, without severe lung disease, diabetes, and so on).his clinical approach is borne out by our data and that ofthers, which show that patients receiving SVGs tend toave more severe disease initially than the patients receivingMA grafts (9,10,13). This argument is probably notermane today, when the characteristics of patients haveramatically increased toward more complex disease as aesult of percutaneous coronary intervention, and yet IMArafts are still used in the majority of patients. In the presenttudy, where patients were not randomized to differentonduits, the 10-year patency of left IMA grafts to theAD is 85%, as opposed to 69% for SVGs to the LAD.nother argument that is often used to support the IMA as

he conduit of choice is that retrospective studies havehown that survival is improved in patients receiving IMArafts as opposed to vein grafts (4,6,25,26). Even thoughhese studies adjust for demographic and clinical differences,hey are still retrospective nonrandomized analyses. Datarom our study group have shown that use of the IMA isssociated with a longer operative time as well as increasedostoperative bleeding (27). Thus, the sicker, more compli-ated patients 10 to 15 years ago would have received veinrafts and not IMA grafts.

The data on predictors of graft patency are interesting.e and other investigators have previously shown that the

ize of the recipient vessel is an important predictor of graftatency (13). The simple explanation has been that the

arger the recipient vessel, the better the flow and distalunoff. The size of the recipient vessel is also thought to behe explanation for why SVGs to the LAD do better thanhose to the right or CX (Tables 4 and 5, Fig. 3). The otherignificant patient specific predictors of graft patency are, forhe most part, a reflection of the disease in the patient (i.e.,here was worse SVG patency in younger patients, whichrobably is a reflection of worse native coronary disease inounger patients, and patients with elevated serum choles-erols). Interestingly, diabetes and cigarette smoking wereot predictors of poor graft patency. We have previouslyeported on the beneficial effects of aspirin on SVG patencyp to one year after surgery (11–13). That observation isonsistent with our present data that show the use of aspirins predictive when looking at all grafts (Table 4), but not inrafts that were patent originally at one week (Table 5). Ourarlier work also showed that the temperature used toreserve the vein graft was important, with improvedhree-year SVG patency when colder temperatures weresed (13). We hypothesized that the colder temperatures
nd the type of preservation fluid used in the vein might r

ave preserved endothelial function and thus improvedong-term graft patency.

We believe that the estimates of 10-year graft patencyresented here are less biased than any results previouslyublished. First, although it is generally believed thatatients who die may have much higher occlusion rates thanurvivors, the patency rate from 71 autopsies in our study13.6% of all deaths) was 76% (156 of 205 grafts). There-ore, the effect of lost observations due to death on long-erm patency rates may be minimal. Second, although only3% of the 10-year catheterizations were done strictly forhe purposes of this study, this is a substantial proportionompared with previous reported 10-year data, which usedetrospective data collection and based their results solely onymptomatic patients. The use of survival analysis tech-iques to estimate long-term patency is a third technique inne-tuning the estimates of graft patency. Angiography atny time point can only tell that a graft has become occludedt that point in time, but it cannot determine when thecclusion occurred. Thus, to use the time of the angiogra-hy to estimate graft patency will overestimate the patencyate. The use of interval censoring is an attempt to moreccurately estimate the time that the occlusion occurred.astly, as in all our previous reports, we acknowledge thatraft patency within an individual is correlated and we havedjusted for this association.

It may be reasonable to assume that graft patency influ-nced survival. The data are only from the survivors who, byefinition, may have had better graft patency than theonsurvivors. We acknowledge that not being able toddress whether graft patency influenced survival is aotential weakness in our study, but this would be true forny long-term study of graft patency after CABG. Owing tohe long-term nature of the study and the fact that coronaryrtery disease is a progressive disease, we used data fromlinically indicated angiograms, in part, for the 6- and0-year follow-up. In this older patient population it isifficult to get patients to agree to repeat invasive procedureshen they already have had a clinically indicated catheter-

zation, are asymptomatic, or when the attending cardiolo-ist would not consider any interventional procedure. Twootes of caution: the patients who did not undergo the0-year catheterization had more diabetes and hypertensionnd, therefore, the current estimates of patency may beverestimated. Lastly, all patients were men and, therefore,ong-term angiographic follow-up of CABG grafts is notvailable for a female cohort.

In conclusion, we report long-term analysis of SVG andMA patency for 10 years after operation, showing thatatency for IMA grafts is better than for SVGs, but the0-year patency for SVGs is better and 10-year patency foreft IMA grafts is worse than previously thought. The mostmportant predictors of long-term graft patency are initialatency at one week after CABG and the diameter of the
ecipient vessel into which the graft is placed.
JAC/23029/ on 10/06/2013

RCSs

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

A

ms


Downloa

eprint requests and correspondence: Dr. Steven Goldman,ardiology Section (1-111C), Southern Arizona VA Health Careystem, 3601 South 6th Avenue, Tucson, Arizona 85723. E-mail:[email protected].

EFERENCES

1. Bourassa MG, Campeau L, Lesperance J, Grondin CM. Changes ingrafts and coronary arteries after saphenous vein aortocoronary bypasssurgery: results at repeat angiography. Circulation 1982;65 SupplII:II90–7.

2. Campeau L, Enjalbert M, Lesperance J, Vaislic C, Grondin CM,Bourassa MG. Atherosclerosis and late closure of aortocoronarysaphenous vein grafts: sequential angiographic studies at 2 weeks, 1year, 5 to 7 years, and 10 to 12 years after surgery. Circulation 1983;68Suppl II:II1–7.

3. Grondin CM, Campeau L, Lesperance J, Enjalbert M, Bourassa MG.Comparison of late changes in internal mammary artery and saphenousvein grafts in two consecutive series of patients 10 years after operation.Circulation 1984;70 Suppl I:I208–12.

4. Cameron A, Kemp HG, Green GE. Bypass surgery with the internalmammary artery graft: 15-year follow-up. Circulation 1986;74 SupplIII:III30–6.

5. Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PC.Long-term (5–12 years) serial studies of internal mammary artery andsaphenous vein coronary artery bypass grafts. J Thorac Cardiovasc Surg1990;15:15–20.

6. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormstic M,Williams GW, et al. Influence of the internal-mammary graft on10-year survival and other cardiac events. N Engl J Med 1986;314:1–6.

7. Fitzgibbon GM, Kafka HP, Leach AJ, Keon WJ, Hooper GD, BurtonJR. Coronary bypass graft fate and patient outcome: angiographicfollow-up of 5,056 grafts related to survival and reoperation in 1,388patients during 25 years. J Am Coll Cardiol 1996;28:616–26.

8. Goldman S, Copeland J, Moritz T, Henderson W, Zadina K, Ovitt T,et al. Improvement in early saphenous vein graft patency after coronaryartery bypass surgery with antiplatelet therapy. Circulation 1988;77:1324–32.

9. Goldman S, Copeland J, Moritz T, Henderson W, Zadina K, Ovitt T,et al. Saphenous vein graft patency one year after coronary arterybypass surgery and effects of antiplatelet therapy. Circulation 1989;80:1190–7.

0. Goldman S, Copeland J, Moritz T, Henderson W, Zadina K, et al.Internal mammary artery and saphenous vein graft patency: effects ofaspirin. Circulation 1990;82 Suppl IV:IV237–42.

1. Goldman S, Copeland J, Moritz T, et al. Starting aspirin therapy afteroperation: effects on early graft patency. Circulation 1991;84:520–6.

2. Goldman S, Copeland J, Moritz T, et al. Long term graft patency (3
years) after coronary artery surgery: effects of aspirin. Results of a VAcooperative study. Circulation 1994;89:1138–43. w

3. Goldman S, Zadina K, Krasnicka B, et al., for VA Cooperative StudyGroup #297. Predictors of graft patency 3 years after coronary arterybypass surgery. J Am Coll Cardiol 1997;29:1563–8.

4. Henderson W, Moritz T, Goldman S, et al. The statistical analysis ofgraft patency data in a clinical trial of anti-platelet agents followingcoronary artery bypass grafting. Control Clin Trials 1988;9:189–205.

5. Bogaerts K, Leroy R, Lesaffre E, Declerck D. Modelling toothemergence data based on multivariate interval-censored data. Stat Med2002;21:3775–87.

6. Huster WJ, Brookmeyer R, Self, SG. Modelling paired survival datawith covariates. Biometrics 1998;45:145–56.

7. Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PC.Long-term (5–12 years) serial studies of internal mammary artery andsaphenous vein coronary bypass grafts. J Thorac Cardiovasc Surg1985;89:248–58.

8. Ivert T, Huttunen K, Landou C, Bjork VO. Angiographic studies ofinternal mammary artery grafts 11 years after coronary artery bypassgrafting. J Thorac Cardiovasc Surg 1988;96:1–12.

9. Grondin CM, Campeau L, Thornton JC, Engle JC, Cross FS,Schreiber H. Coronary artery bypass grafting with saphenous vein.Circulation 1989;79 Suppl I:I24–9.

0. Blankenhorn DH, Johnson RL, Mack WJ, El-Zein HA, Vailas LI.The influence of diet on the appearance of new lesions in humancoronary arteries. JAMA 1990;263:1646–52.

1. Knatterud GL, Rosenberg Y, Campeau L, et al., and Post CABGInvestigators. Long-term effects on clinical outcomes of aggressivelowering of low-density lipoprotein cholesterol levels and low-doseanticoagulation in the Post Coronary Artery Bypass Graft Trial.Circulation 2000;102:157–65.

2. Post Coronary Artery Graft Trial Investigators. The effect of aggres-sive lowering of low-density lipoprotein cholesterol levels and low-dose anticoagulation on obstructive changes in saphenous-veincoronary-artery bypass grafts. N Engl J Med 1997;336:153–62.

3. Spencer FC. The internal mammary artery: the ideal coronary bypassgraft? N Engl J Med 1986;314:50–1.

4. Acinapura AJ, Rose DM, Jacobowitz IJ, et al. Internal mammaryartery bypass grafting: influence on recurrent angina and survival in2,100 patients. Ann Thorac Surg 1989;48:186–91.

5. Cameron C, Davis KB, Green GE, Myers WO, Pettinger M. Clinicalimplications of internal mammary artery bypass grafts: the CoronaryArtery Surgery Study experience. Circulation 1988;77:815–9.

6. Loop FD, Lytle BW, Cosgrove DM. New arteries for old. Circulation1989;79:I40–5.

7. Sethi GK, Copeland JG, Moritz T, Henderson W, Zadina K,Goldman S. Comparison of postoperative complications betweensaphenous vein and IMA grafts to left anterior descending coronaryartery. Ann Thorac Surg 1991;51:733–8.

PPENDIX

For the Statistical Appendix and a list of the Depart-ent of Veterans Affairs Study Group members, please

ee the December 7, 2004, issue of JACC at http://
ww.onlinejacc.org.
JAC/23029/ on 10/06/2013

http://www.onlinejacc.org

http://www.onlinejacc.org

long term patency

Documents