Continuing evolution of therapy for coronary artery disease. Initial results... coronary angioplasty.

Continuing evolution of therapy for coronary artery disease. Initial results from the era of
coronary angioplasty.
D B Mark, C L Nelson, R M Califf, F E Harrell, Jr, K L Lee, R H Jones, D F Fortin, R S Stack, D D
Glower and L R Smith
Circulation. 1994;89:2015-2025
doi: 10.1161/01.CIR.89.5.2015
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1994 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/89/5/2015
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click Request
Permissions in the middle column of the Web page under Services. Further information about this process is
available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
2015
Continuing Evolution of Therapy for
Coronary Artery Disease
Initial Results From the Era of Coronary Angioplasty
Daniel B. Mark, MD, MPH; Charlotte L. Nelson, MS; Robert M. CalifT, MD;
Frank E. Harrell, Jr, PhD; Kerry L. Lee, PhD; Robert H. Jones, MD; Donald F. Fortin, MD;
Richard S. Stack, MD; Donald D. Glower, MD; L. Richard Smith, PhD; Elizabeth R. DeLong, PhD;
Peter K. Smith, MD; J.G. Reves, MD; James G. Jollis, MD; James E. Tcheng, MD;
Lawrence H. Muhlbaier, PhD; James E. Lowe, MD; Harry R. Phillips, MD; David B. Pryor, MD
Background Survival after coronary artery bypass graft
surgery (CABG) and medical therapy in patients with coronary artery disease (CAD) has been studied in both randomized trials and observational treatment comparisons. Over the
past decade, the use of coronary angioplasty (PTCA) has
increased dramatically, without guidance from either randomized trials or prospective observational comparisons. The
purpose of this study was to describe the survival experience of
a large prospective cohort of CAD patients treated with
medicine, PTCA, or CABG.
Methods and Results The study was designed as a prospective nonrandomized treatment comparison in the setting of an
academic medical center (tertiary care). Subjects were 9263
patients with symptomatic CAD referred for cardiac catheterization (1984 through 1990). Patients with prior PTCA or
CABG, valvular or congenital disease, nonischemic cardiomyopathy, or significant (.75%) left main disease were excluded.
Baseline clinical, laboratory, and catheterization data were
collected prospectively in the Duke Cardiovascular Disease
Databank. All patients were contacted at 6 months, 1 year, and
annually thereafter (follow-up 97% complete). Cardiovascular
death was the primary end point. Of this cohort, 2788 patients
were treated with PTCA (2626 within 60 days) and 3422 with
CABG (3080 within 60 days). Repeat or crossover revascularization procedures were counted as part of the initial treatment strategy. Kaplan-Meier survival curves (both unadjusted
and adjusted for all known imbalances in baseline prognostic
factors) were used to examine absolute survival differences,
and treatment pair hazard ratios from the Cox model were
used to summarize average relative survival benefits. For the
latter, a 13-level CAD prognostic index was used to examine
the relation between survival and revascularization as a function of CAD severity. The effects of revascularization on
survival depended on the extent of CAD. For the least severe
forms of CAD (ie, one-vessel disease), there were no survival
advantages out to 5 years for revascularization over medical
therapy. For intermediate levels of CAD (ie, two-vessel disease), revascularization was associated with higher survival
rates than medical therapy. For less severe forms of two-vessel
disease, PTCA had a small advantage over CABG, whereas for
the most severe form of two-vessel disease (with a critical
lesion of the proximal left anterior descending artery), CABG
was superior. For the most severe forms of CAD (ie, threevessel disease), CABG provided a consistent survival advantage over medicine. PTCA appeared prognostically equivalent
to medicine in these patients, but the number of PTCA
patients in this subgroup was low.
Conclusions In this first large-scale, prospective observational treatment comparison of PTCA, CABG, and medicine,
we confirmed the previously reported survival advantages for
CABG over medical therapy for three-vessel disease and severe
two-vessel disease. For less severe CAD, the primary treatment
choices are between medicine and PTCA. In these patients,
there is a trend for a relative survival advantage with PTCA,
although absolute survival differences were modest. In this
setting, treatment decisions should be based not only on survival
differences but also on symptom relief, quality of life outcomes,
and patient preferences. (Circulation. 1994;89:2015-2025.)
Key Words * angioplasty * bypass * coronary disease
T he effects of both coronary artery bypass graft
surgery (CABG) and medical therapy on survival in patients with coronary artery disease
(CAD) have been well studied.'-5 During the 1980s, the
use of percutaneous transluminal coronary angioplasty
(PTCA) as a third alternative grew to the point that
more than 250 000 PTCAs are now performed annually
in the United States.67 Although PTCA was initially
viewed as a treatment that would allow patients to defer
or eliminate the need for CABG, there is now strong
Received July 14, 1993; revision accepted January 14, 1994.
From the Heart Center, the Departments of Medicine, Community and Family Medicine, Anesthesiology, and Surgery, Duke
University Medical Center, Durham, NC.
Correspondence to Daniel B. Mark, MD, MPH, PO Box 3485,
Duke University Medical Center, Durham, NC 27710.
evidence that the procedure has been used predominantly in patients who would formerly have been treated
medically.8 Thus, the rising incidence of PTCA in the
United States has actually been accompanied by an
increase in the use of CABG.7
Many have pointed out that the current patterns of
PTCA use have evolved without guidance from randomized trial data.9 The first major randomized trial involving PTCA was published in 1992.10 This trial, Angioplasty Compared to Medicine (ACME), found that
PTCA was associated with better functional status and
more complete relief of symptoms in patients with
angina who had single-vessel disease. However, ACME,
which randomized 212 patients, was not large enough to
address the issue of possible survival differences. Re-
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
2016
Circulation Vol 89, No 5 May 1994
cently, an interim analysis of the Randomized Intervention Treatment of Angina (RITA) trial was published."1
This trial randomized 1011 patients with significant
CAD to either PTCA or CABG. At a mean follow-up of
2.5 years, cardiac mortality in the PTCA group was
1.6% and in the CABG group, 2% (P=NS). Of the four
additional randomized trials comparing PTCA and
CABG in multivessel-disease patients, only BARI was
designed with survival as the primary end point.12,13
In 1989, we described our survival experience with
therapy for CAD starting with the initiation of CABG at
Duke Medical Center in 1969 and continuing to 1984.4
The purpose of the present report is to extend those
results into the era of PTCA by describing the 5-year
survival experience of patients with ischemic heart
disease treated with medicine, PTCA, or CABG at
Duke between 1984 and 1990.
Methods
Patient Population
Between March 1984 and August 1990, 17 309 consecutive
patients were referred to the Duke Heart Center for initial
cardiac catheterization for evaluation of suspected ischemic
heart disease. Of these, 7710 patients were excluded for the
following reasons: absence of significant (275% diameter)
stenosis in at least one major epicardial coronary segment
(n=5817), 275% left main stenosis (n=691), significant (ie,
3+ to 4+) ischemic mitral regurgitation (n=422), prior PTCA
(n=253), and prior CABG (n=527). Patients with primary
valvular, congenital, or cardiomyopathic disease were also
excluded. Thus, the final study population consisted of 9263
patients with significant CAD comprising all eligible patients
with stable angina, progressive angina, unstable angina, postinfarction angina, an acute evolving myocardial infarction, or a
recent myocardial infarction.
Data Collection and Information System
Pertinent baseline variables from the history, physical examination, laboratory tests, chest x-ray, and 12-lead ECG
were collected prospectively on standardized forms as part of
the patient care process and stored in the Duke Cardiovascular Disease Databank.14'15 The results of cardiac catheterization and the procedural details of each PTCA and CABG have
also been collected prospectively. Definitions of key prognostic
variables (Table 1) have been published previously.4'15-18
Cardiac Catheterization
Significant coronary artery stenosis was defined as 75% or
greater obstruction of a major epicardial coronary segment.4
Arterial lesions were graded by subjective visual consensus of
at least two experienced observers on the following ordinal
scale: 0%, 10% (minor irregularities), 25%, 50%, 75%, 95%,
or 100% luminal diameter stenosis. The amount of coronary
disease present was summarized in two ways for this study.
First, we used the traditional one-, two-, or three-vessel
disease classification.19 Recognizing that this widely used
classification groups together categories of patients with differing prognoses and (presumably) differing responses to
revascularization, we also used a more detailed characterization of coronary disease severity. In this latter system, the
Coronary Artery Disease Prognostic Index (Table 2), possible
combinations of coronary lesions were ranked and weighted
using a Cox regression analysis of medical prognosis in patients
treated at Duke between 1969 and 1984.18 This CAD index
considers not only the number of diseased vessels but also any
significant involvement of the left anterior descending coronary artery (LAD), particularly when there was involvement
of the proximal segment and/or the proximal segment stenosis
was severe (ie, .95% stenosis).
TABLE 1. BaselIne Prognostic Characerlstics
Left ventricular ejection fraction
Coronary disease severity (No. of diseased vessels, proximal
left anterior descending stenosis, severe [95%] proximal left
anterior descending stenosis, coronary artery disease index)
Noninvasive markers of myocardial damage (Q wave on ECG,
history of myocardial infarction, cardiomegaly on chest x-ray,
S3 gallop, severity of congestive heart failure)
Severity of angina (stable, progressive, unstable,
postinfarction, frequency, nocturnal)
Acute myocardial infarction (<6 hours, 6-24 hours, 24 hours to
6 weeks)
Demographics (age)
Mitral regurgitation (none, 1+, 2+)
Conduction disturbances on ECG (right bundle branch block,
left bundle branch block, hemiblocks)
Extracardiac vascular disease (cerebrovascular, peripheral
vascular)
Comorbidity (renal insufficiency, chronic obstructive lung
disease, cancer [excluding skin], liver disease)
Year of cardiac catheterization
Coronary Revascularization and Medical Therapy
Standard PTCA techniques used at Duke during the period
of this study have been described in detail elsewhere.20 During
this study, the procedural success rate (c50% residual stenosis) for angioplasty was 91% for acute myocardial infarction
patients and unstable angina patients undergoing an emergency procedure and 95% for nonacute procedures (overall
procedural success rate, 94%). The emergency bypass surgery
rate for this cohort was 2.8%. The 30-day unadjusted mortality
rate for PTCA was strongly influenced by mode of clinical
TABLE 2. Coronary Artery Diseae Pronsic Index*
Extent of CAD
No CAD .50%
1 VD 50%-74%
> 1 VD 50%-74%
1 VD (75%)
1 VD (.95%)
2 VD
2 VD (both .95%)
1 VD, .95% proximal LAD
2 VD, .95% LAD
2 VD, >95% proximal LAD
3 VD
3 VD, .95% in at least
Prognostic Weight (0-100)
0
19
23
23
32
37
42
48
48
56
56
63
67
74
82
100
CAD indicates coronary artery disease; VD, vessel disease;
and LAD, left anterior descending coronary artery.
*Note that the CAD index is hierarchical and that patents are
assigned to the most severe category that applies to them.18
one
3 VD, 75% proximal LAD
3 VD, .95% proximal LAD
Left main (75%)
Left main (.95%)
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Mark et al Evolution of Therapy: Angioplasty
TABLE 3. Baseline Clinical Characteristics of 9263 Study Patients
Medicine (n=3557)
PTCA (n=2626)
Male patients, %
Age, y (range)
Anginal symptoms, % or No.
Typical angina
Stable angina
71
61 (53-68)
65
27
32
28
5
8
<1
2 (0-5)
11
11 (1-58)
70
58 (50-66)
71
11
28
22
6
2017
CABG (n=3080)
76
62 (56-69)
78
19
42
24
6
9
Progressive/unstable
Recent Ml, stable
Recent Ml, unstable
Acute Ml (s24 hours)
33
Variant angina
1
Pain frequency/week
4 (1-7)
3 (0-5)
Nocturnal pain
14
8
Duration of CAD symptoms, mo
1 (0.1-12)
12 (0.8-68)
Risk factors, %
Family history of CAD
47
47
51
Smoker
67
72
65
Hypertension
55
46
53
Diabetes
24
15
22
Hyperlipidemia
29
27
31
Clinical measures of myocardial damage, %
63
History of Ml
71
59
17
History of CHF
7
12
NYHA CHF class IV
3
1
6
S3 gallop
4
4
Major comorbidity*
11
7
9
PTCA indicates percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft surgery; Ml,
myocardial infarction; CAD, coronary artery disease; CHF, congestive heart failure; and NYHA, New York Heart
Association.
*Specified in Table 1.
presentation: acute myocardial infarction (s24 hours), 10.4%;
recent myocardial infarction (>24 hours but c6 weeks), 2.1%;
unstable angina, 0.8%; and stable angina, 0.3%. The standard
cardiac surgical and anesthesia techniques used at Duke
during the period of this study have also been described in
detail previously.21 The unadjusted 30-day mortality rates in
the surgical patients included in this study were acute myocardial infarction, 6.2%; recent myocardial infarction, 3.7%;
unstable angina, 2.8%; and stable angina, 1.6%. The medical
therapy given to each study patient was determined by the
patient's attending Duke cardiologist and primary care
physician.
Follow-up Procedures
Patients were contacted at 6 months and 1 year after
presentation and then annually.4"15 The follow-up anniversary
was timed from the date of catheterization for medical patients
and from the time of the initial revascularization procedure for
PTCA and CABG patients. Follow-up was performed primarily with self-administered questionnaires; telephone contact
was reserved for patients who did not return the questionnaires. Survival data were 97% complete for all time intervals.
Cause of death was classified by an independent mortality
committee into cardiovascular and noncardiovascular causes.
The primary end point for this study was cardiovascular death.
Importantly, our conclusions were unchanged when all-cause
mortality was substituted as the end point for analysis.
Data Analysis
Descriptive Stasics
Baseline characteristics were summarized with medians and
interquartile ranges for continuous variables and percentages
for discrete variables. For the purposes of describing baseline
characteristics (Tables 3 and 4), patients were assigned to a
treatment group defined by the initial revascularization procedure performed within 60 days of cardiac catheterization.
Patients not receiving revascularization within 60 days of catheterization composed the medical group. Patients were assigned
to a therapeutic group according to the first revascularization
procedure performed, regardless of success of the procedure or
subsequent crossover.
Treatment Assignment for Survival Analysis
For survival analysis, Kaplan-Meier survival estimates were
used to describe patterns of mortality by treatment group.22
Because the >60-day treatment group definition unfairly
penalizes the medical group by ascribing medical therapy to all
patients who die before they are able to undergo revascularization, we used a method of treatment assignment that
reduces this waiting-time bias. With this method, all 9263
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
2018
Circulation Vol 89, No S May 1994
TABLE 4. Baelane CMwlzaion of 93 Study Psents
MedLcine (n=3557)
PTCA (n=2626)
CABG (n=3080)
Coronary artery disease, %
1 VD
48
61
10
2 VD
29
30
34
22
3 VD
10
56
LAD disease (>75%), %
62
56
91
Prox LAD disease (.75%), %
22
23
46
CAD index*, %
1 VD (275%)
11
8
1
32
1 VD (.95%)
43
5
11
12
2 VD
7
4
2 VD, 2 severe
5
4
1 VD, prox LAD 295% or 2 VD, .95% LAD
18
15
15
4
6
2 VD, .95% prox LAD or 3 VD
11
14
6
3 VD, 1 severe
30
1
3
3 VD, >75% prox LAD
9
2
5
3 VD, 295% prox LAD
18
53 (45-60)
50 (39-59)
52 (42-60)
Ejection fraction, %
71
76
Abnormal LV contraction, %
76
Mitral insufficiency, %
85
80
Grade 0
81
10
13
Grade 1
13
7
5
Grade 2
6
PTCA indicates percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft surgery; VD,
vessel disease; LAD, left anterior descending coronary artery; prox, proximal; and LV, left ventricular.
*The CAD index is hierarchical -patients are placed into the most severe category appropriate to their anatomy.
patients were initially considered to be medically treated.
When patients underwent their initial revascularization procedure, their medical follow-up was censored and their follow-up was restarted at time zero in the appropriate revascularization group (Fig 1). This method of assigning treatment is
identical to that used in our previous observational treatment
comparison of medicine and CABG4 and is also the same as
"method A" used by the CASS group in their nonrandomized
initil Rx A "eigm
M edicie
n
lot ~
~
*
CASO
[m
| lcA--->J)
|imi Rx
_g
|
9263
| PTCA Mod~
4---PTCA
(4%)
in
CABG
FiG 1. Diagram showing method of assignment of therapeutic
for treatment comparison that reduces the waiting time
bias against medical survival. All patients are initially counted as
medically treated. When patients cross over for the first time to
either percutaneous transluminal coronary angioplasty (PTCA)
or coronary artery bypass graft surgery (CABG), they are censored (ie, removed alive) from the medical group and their
follow-up is restarted in the appropriate revascularizatlon group.
If an intervening repeat cardiac catheterization has been performed, their covariates are updated to the newer values. Once
assigned to a revascularization group, patients are counted in
that group regardless of subsequent crossovers.
group
treatment comparisons.23 Once a patient was assigned to a
revascularization group, either PTCA or CABG, he or she was
maintained in that group until death or the end of follow-up
regardless of subsequent therapeutic crossovers. Thus, a
PTCA patient with a failed procedure requiring emergency
CABG would be analyzed as a PTCA patient. Similarly, a
CABG patient requiring PTCA of a graft stenosis would be
counted as a CABG patient. If patients who crossed over from
medicine to a revascularization procedure underwent a repeat
catheterization before that procedure, the baseline variables
used in our prognostic calculations for that revascularization
group were those of the most recent preprocedural catheterization study. Of the 2788 patients undergoing initial PTCA
during the study, 162 (6%) had their procedure more than 60
days after the index cardiac catheterization. Similarly, of the
3422 CABG patients in the study, 342 (10%) had their
operation more than 60 days after cardiac catheterization.
Covarite Adjusens and Ststical Modeling
Because treatment in our study was not randomly assigned,
observed survival differences among groups may be due to
imbalances in baseline prognostic factors rather than to a
specific effect of treatment on outcome. To control for or
"subtract out" the effects of such baseline imbalances before
comparing therapies, we performed a standard covariate adjustment that included all identifiable prognostic factors (Table 1) in a multivariable survival model. The objective of this
adjustment was to ensure that any prognostic differences
between treatments were additional to the effects of these
nontreatment variables. For these analyses, we used a Cox
proportional-hazards regression model stratified on treatment
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Mark et al Evolution of Therapy: Angioplasty
1 Vessel Disease
Unadjusted
1.0
1.0 L_
t~ ~~~~eetee-"~
@
'
0.91-
E 0.9
.0
.0
2 0.8
A.
0
Adjusted
0.7 1
(0.
- PTCA
-CABG
-Medicine
0.81-
0.71-
2919
96
425
1693
944
3C9p
339
219
{X-fr
|7
l
0
1
2
3
4
5
C
06 2919
1693
339
(ls
0
1
2
94
M
944
219
0P
3
c
4
5
2019
FIG 2. Survival curves for one-vessel
disease showing unadjusted (left) and
adjusted (right) comparisons of the three
treatment groups to demonstrate absolute survival differences. X axis shows
follow-up time out to 5 years. Y axis
shows cardiovascular survival probability from 1.0 to 0.5. Numbers at the bottom of the plots show number of patients
in each treatment group remaining to be
followed at 0, 3, and 5 years of follow-up.
PTCA indicates percutaneous transluminal coronary angioplasty (P); CABG, coronary artery bypass graft surgery (C);
and M, medically treated.
Years
Years
and number of diseased vessels,24 using an approach described
in detail previously.416 24 Basically, this model allows separate
underlying hazard functions for the three treatments with
common coefficients for each of the other covariates in the
model across the three treatments, unless covariate-treatment
interactions were present. By stratification on treatment, the
need for assuming proportional hazards among treatments is
eliminated.
In addition to standard covariate adjustments, we adjusted
for factors favoring selection of one treatment over another
using propensity scores.25-27 These scores were developed
from binary logistic regression models and were represented in
the final Cox regression models by three variables consisting of
the linear score or logit from each of the three logistic models
(ie, CABG versus PTCA, CABG versus medical therapy,
PTCA versus medical therapy).
Statistical Testing and Treatment Comparisons
Treatment comparisons are presented in this paper in two
main formats: Kaplan-Meier survival curves (unadjusted and
adjusted) to contrast absolute survival differences and hazard
ratios from the Cox model to compare relative survival differences of the three treatments (taken two at a time) according
to levels of the Duke prognostic CAD index (Table 2).
Because the hazard ratios involve a direct treatment comparison, the calculations required refitting Cox models in which
the treatment group was used as a covariate rather than as a
stratification factor. Since early procedural mortality creates a
pattern of crossing survival curves, the proportional-hazards
assumptions of the Cox model are violated. The typical effect
of this violation is to make statistical treatment comparisons
more conservative and confidence limits on the hazard ratios
wider.
Hazard ratios are presented with 99% confidence limits
rather than 95% limits because a more conservative position is
warranted by the multiple comparisons involved in calculating
separate hazard ratios for each level of the CAD index.
Confidence limits that do not intersect the 1.0 line of treatment equivalence are indicative of a treatment difference that
is significant at the .01 level.
Results
Baseline clinical characteristics are shown in Table 3
for the 2626 patients receiving PTCA as their initial
revascularization procedure within 60 days of catheterization, the 3080 patients initially treated with CABG
within 60 days, and the 3557 patients who were treated
medically for at least 60 days. Although there was much
overlap among groups in these clinical characteristics,
some trends were evident (Table 3). Median age was
lowest in the PTCA group and highest in the CABG
group. CABG patients had more progressive or unstable anginal symptoms, whereas the PTCA group had the
highest prevalence of acute myocardial infarction (catheterization within 24 hours of presentation). CABG patients had been symptomatic the greatest period of time,
whereas PTCA patients had a median of 1 month between
the date of symptom onset and catheterization. The
medical patients had the highest prevalence of any congestive heart failure and of class IV heart failure. Major
comorbidity, including renal insufficiency, chronic obstructive lung disease, and cancer, was also slightly more
prevalent in the medical group.
The biggest differences among groups were found in
the distribution of baseline catheterization characteristics (Table 4). In particular, 61% of the PTCA group
had one-vessel disease and 10% three-vessel disease,
whereas for CABG the proportions were essentially
reversed (10% and 56%, respectively). Examination of
the CAD index provided more details regarding differences in the distribution of coronary disease among the
three groups (Table 4). PTCA patients had the best and
medical patients the worst left ventricular function,
judged in terms of the ejection fraction. The prevalence
of mild mitral regurgitation was similar in the three
groups.
Fig 2 displays survival curves for the patients with
one-vessel disease according to treatment group with
(right) and without (left) adjustment for imbalances in
baseline characteristics. Although there was a very
slight trend toward higher survival in the PTCA group
compared with the other two treatments, this difference
was small at the 5-year follow-up point (adjusted 5-year
survivals: PTCA, 95%; CABG, 93%; medicine, 94%). In
the patients with two-vessel disease (Fig 3), revascularization with either PTCA or CABG showed a small
advantage over medicine emerging at 1 year of follow-up (adjusted 5-year survivals: PTCA, 91%; CABG,
91%; medicine, 86%). In contrast, in the group with
three-vessel disease (Fig 4), which had the smallest
number of PTCA patients (n=260), CABG was superior to both medicine and PTCA (adjusted 5-year
survivals: CABG, 89%; PTCA, 81%; medicine, 72%).
To ensure that these results were not due to inappropriate combination of survival experience from acute or
recent myocardial infarction patients with other symptomatic CAD patients, we generated the same survival
curves after excluding patients with a myocardial infarc-
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Circulation Vol 89, No S May 1994
2020
2 Vessel Disease
Unadjusted
1.0
U
.0
2 0.8
W
0.91
m0.9
I.
Adjusted
1.0
PTCA
CABG
- dicine
F
0.7 rL
53 .6
1216
0.7 I
22m 06
174P
29 C
708
1160
a
0
1
2
FIG 3. Survival curves for two-vessel
disease. See Fig 2 for orientation. PTCA
indicates percutaneous transluminal coronary angioplasty (P); CABG, coronary
artery bypass graft surgery (C); and M,
medically treated.
0.8 F
537
425
anJ-
A_~~~~~~~~~~~~~-
3
4
5
2568
Sm
1169
u.5.
537
226 M
174 P
289 C
425
708
l
0
1
2
Years
tion within 2 weeks of the diagnostic catheterization
(Fig 5). The patterns evident in Figs 2 through 4 were
also clearly evident in this subset of the population.
Additional subsets examined, including patients with
unstable angina, older patients, and those with low
ejection fraction, also displayed these same relations.
Because the traditional classification of one-, two-,
and three-vessel disease is not sufficiently detailed to
reflect all the important prognostic aspects of coronary
anatomy, we used a recently derived CAD index (Table
2) to calculate Cox model estimated hazard or mortality
ratios for pairs of treatments against nine levels of
coronary disease severity. Hazard ratios summarize the
relative effect of one treatment versus another on
mortality averaged over the 5-year follow-up period. Fig
6 shows the hazard ratios for CABG and medical
therapy along with 99% confidence limits. In the least
severe categories of CAD (left side of Fig 6), there is a
trend favoring medicine, although the confidence limits
include the possibility of no effect (ie, hazard ratio of 1)
or even benefit from CABG. In the more severe forms
of one-vessel disease and the less severe forms of
two-vessel disease, the two therapies appear equivalent,
whereas in severe two-vessel disease (with a 95% proximal LAD lesion) and all forms of three-vessel disease,
CABG clearly improves survival relative to medicine by
36% to 63%.
A similar comparison is shown in Fig 7 for medical
therapy and PTCA. In one-vessel disease categories and
A
3
Years
4
5
the less severe forms of two-vessel disease, there is a
trend for PTCA to reduce mortality relative to medicine
by approximately 20%, although in every case the
confidence limits cross the 1.0 hazard ratio line of
identity. In severe two-vessel disease (with 95% proximal LAD involvement) and all forms of three-vessel
disease, the estimated benefit of PTCA is 10% to 15%
and the confidence limits suggest an equivalent effect on
survival.
The final set of hazard ratios involving CABG and
PTCA is shown in Fig 8. In the less severe forms of oneand two-vessel disease, PTCA reduced mortality by
almost 50% relative to CABG. In severe one-vessel
disease (with 95% proximal LAD involvement) and
moderate two-vessel disease, the two therapies had an
equivalent effect on survival, whereas in severe twovessel disease (with 95% proximal LAD involvement)
and all forms of three-vessel disease, there was a clear
survival benefit for CABG, with the magnitude of the
benefit (25% to 60% mortality reduction) proportional
to the severity of the coronary disease.
Discussion
This study is the first large-scale prospective observational treatment comparison of 5-year survival outcomes with PTCA, CABG, and medical therapy. The
major findings are as follows. First, we confirmed the
survival benefits of CABG relative to medicine that had
been reported for CAD patients treated during the
3 Vessel Disease
Unadjusted
I
Adjusted
1.0
FIG 4. Survival curves for three-vessel
disease. See Fig 2 for orientation. PTCA
indicates percutaneous transluminal coronary angloplasty (P); CABG, coronary
artery bypass graft surgery (C); and M,
medically treated.
0.9
0.8
IL
0.7
60
0.6 267
n}-s
-- rW
v._
F
2
Years
-
- PTCA
- CABG
- Mkdlicna
E~U~55w
A-
260
1814
w
=I
1
2
302
118m
117
46 P
510 c
1151
3
4
5
Years
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Mark et al Evolution of Therapy: Angioplasty
1 Vessel Disease
> 1.0
=
.0
-0
MU.. MMUM
0.9
-
b> 0.6
CO n=
U.*
PTCA
CABG
Medicine
-
2 0.8 I11 0.7
1606
672
719
223
447
0
o
1
1
154
2
2
3
3
4
5
4
5
Years
2 Vessel Disease
> 1.0
n
0.9
2 0.8
~j0.7cn 0.6 1589
295
394
109
0.5-753
1
A
0
1
2
14901
3
Years
3 Vessel Disease
1.0
0
.0
0
0.9
0.8
CD 0.7
CL)
0.
1746
215
,74
37
504
12U"
0.5L0
95
1
1A
1
2
3
md
20 PTCA
373 CABO
a~~~~~~~~~~~~
4
5
Years
FIG 5. Adjusted survival curves for one-, two-, and three-vessel
disease excluding patients with acute myocardial infarction
within 2 weeks. Patterns seen in Figs 2 through 4 are clearly
preserved in the non-acute myocardial infarction portion of the
study population. PTCA indicates percutaneous transluminal
coronary angioplasty; CABG, coronary artery bypass graft surgery; and Med, medically treated.
1970s and early 1980s.1-5'28 Specifically, in this more
recent (1984 through 1990) cohort of patients, CABG
was found to improve survival for the more severe forms
of CAD, starting with severe two-vessel disease and
extending through all forms of three-vessel disease (Fig
6). Furthermore, the magnitude of survival benefits was
clearly a function of the severity of the CAD present,
with more severe CAD associated with greater benefit.4,U Second, in less severe forms of CAD, PTCA
showed a modest survival advantage over CABG (Fig
8). In the most severe category of one-vessel disease and
moderate two-vessel disease (both involving LAD lesions), however, PTCA and CABG appeared prognostically equivalent, whereas in the most severe forms of
CAD, PTCA had prognostic effects equivalent to medicine (Fig 4). Finally, comparing PTCA and medicine
(Fig 7), there was a definite trend favoring PTCA in the
less severe forms of CAD, although this did not reach
statistical significance. Starting with severe two-vessel
2021
disease and continuing with all forms of three-vessel
disease, PTCA and medicine appeared prognostically
equivalent.
Typically, when survival differences between treatment alternatives are present, they are largest in absolute terms in high-risk patients and smallest in low-risk
patients.4 In the case of low-risk patients, such as those
with one-vessel disease, the small absolute differences
present in the survival outcomes of alternative therapies
may be of less consequence in decision making than
effects on symptom status or patient preferences,
whereas in high-risk patients, survival considerations
may dominate. Although our results by themselves are
not sufficient for individual patient treatment selection
(since they do not include either symptom status or
patient preferences), they do provide a conceptual
framework for such decisions. Generally, medicine or
PTCA should be the preferred initial strategies for
low-risk CAD (eg, one-vessel disease), whereas CABG
should be preferred in many high-risk CAD patients
(eg, three-vessel disease). In the intermediate levels of
CAD severity, revascularization offers a modest survival
advantage by 5 years relative to medicine, with CABG
offering an advantage for the highest-risk patients of the
group (ie, those with a critical proximal LAD lesion)
and PTCA preferable for many of the remainder.
To date, only one randomized trial has compared
outcomes between PTCA and medical therapy for patients with angina.10 The ACME trial, involving 212
patients with single-vessel disease, reported better
6-month symptom and functional status after PTCA.
However, the study was not intended to be large enough
to evaluate mortality outcomes. No other randomized
comparison of PTCA and medicine is currently under
way, although the ACME group is planning a trial in
lower-risk multivessel-disease patients. Five randomized trials are being conducted to compare PTCA and
CABG, four of which are restricted to multivessel CAD.
The largest of these trials, the National Heart, Lung,
and Blood Institute (NHLBI) BARI trial, enrolled 1829
patients; the other trials are substantially smaller. Two
of these trials, the NHLBI BARI trial12 and the UK
RITA trial,29 have "hard event" primary end points of
death and myocardial infarction, whereas the other
three (EAST, CABRI, and GABI) are smaller and have
composite end points that include both cardiac events
and other measures such as thallium defects or anginal
symptoms.13 Recently published interim results from
the RITA trial reveal no survival differences between
PTCA and CABG at a mean follow-up of 2.5 years.11
To date, no prospective nonrandomized treatment
comparisons of long-term survival with medicine versus
PTCA have been reported. A nonrandomized treatment comparison from Emory of PTCA and CABG in
patients with two-vessel disease reported identical
adjusted 5-year survival results for PTCA and CABG at
93%.3° Our study had very similar results, with adjusted
5-year survival rates of 91% for both PTCA and CABG
(Fig 3, right). In our study, there was heterogeneity in
the relative effects of PTCA and CABG on different
subtypes of two-vessel disease (Fig 8). This possibility
was not examined in the Emory study. Recently, investigators from the Cleveland Clinic conducted a 588patient matched-pair comparison of PTCA and CABG
for multivessel CAD using the patient selection criteria
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Circulation Vol 89, No 5 May 1994
2022
2.5
I 2.0
.2
1U
cc
N
Th0
U
I
1.5
a
1.0
FIG 6. Hazard (mortality) ratios for coronary
artery bypass graft surgery (CABG) vs medicine
calculated from the Cox regression model to
evaluate relative survival differences. Points indicate hazard ratios for each level of the coronary artery disease index (see Table 2); bars
indicate 99% confidence intervals. Horizontal
line at ratio of 1.0 indicates point of prognostic
equivalence between treatments. Hazard ratios
below the line favor CABG, those above the line
favor medicine. VD indicates vessel disease;
Prox LAD, proximal left anterior descending
coronary artery.
.
-
I 0.5 h
0.0
WvD
1VD 2VD 2VD
5%
b
IYD
y%
3vD 3vD
2VD
>05% >95%
>95*
Prox
LAD
Prox
LAD
or
Prox
WvD
>95%
LAD
PrOxLAD
or
2VD
SYD
>95% LAD
of the BARI trial.31 Overall, they found no mortality
difference between the two treatment strategies out to 3
years. BARI eligible patients were not specifically examined in the present study, but our results suggest that
the likelihood of a true difference in survival between
PTCA and CABG in multivessel disease would depend
on the mix of CAD subtypes enrolled in the study.
It should be strongly emphasized that our analysis, as
well as every randomized trial involving revascularization therapies, compares treatment strategies rather
than the "pure effects" of the treatments themselves.32
A clinician selects a particular treatment for a given
patient with the explicit understanding that, if necessary, that patient may require an alternative treatment
in the future. For example, PTCA may be selected
initially for a patient with two-vessel disease, but if the
procedure is complicated by abrupt closure, the patient
may undergo emergency CABG. The same patient
might have an uncomplicated PTCA but experience
several subsequent episodes of symptomatic restenosis
g to the need for CABG in a more elective fashion
2 months after the initial procedure. In fact, if the
ian's goal is to defer CABG as long as possible, an
nplicated crossover could readily be viewed as the
ssful culmination of that strategy.
the present study, PTCA patients had a fivefold
r incidence of crossovers than CABG patients
[) in addition to the frequent need for repeat
i to treat restenosis. Because such events cannot
)ided or eliminated with currently available therthere are two remaining options for dealing with
in a treatment comparison analysis: incorporate
into the treatment plan and view the whole
nent course as a "strategy," the approach we
td in this study, or treat them as outcome events
failure of therapy). Although repeat PTCA is now
only accepted as a part of PTCA therapy, views on
vers to CABG are mixed. In the absence of a
licating death or (otherwise avoidable) major morit is unclear that crossover to CABG should be
2.5
2.0
1.5
1.0
_r
40
1
00
1.
0.
40
00
cates vessel disease; Prox LAD, proximal left anterior
descending coronary artery.
1
1
0.5
0.0
a
a
A
1WD
IVO
2VD
A
2V
1
1
IVD
2VD
PMo
or
PMo
LAD
or
2VD
SYD
LAD
1
A
3VM 3YD
k^
FIG 7. Hazard ratios for percutaneous transluminal
coronary angioplasty (PTCA) vs medicine. See Fig 6
for orientation. Points beiow 1.0 favor PTCA. VD indi-
L-i
SVD
Prox
LAD
PoxLAD
LAD
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Mark et al Evolution of Therapy: Angioplasty
2023
2.5
I 2.0
0
.2
1.5
90
N
U
.
4
cc
FIG 8. Hazard ratios for coronary artery bypass graft surgery (CABG) vs percutaneous
transluminal coronary angioplasty (PTCA). See
Fig 6 for orientation. Points below 1.0 favor
CABG. VD indicates vessel disease; Prox LAD,
proximal left anterior descending coronary
artery.
1.0
0.5
0.0
ND
IVD
>i%
2VD
a
2VD VD
% >%
Prox
LAD
2VD
VD 3VD
M
MVD
>% >05% Prox
>05%
Prox
LAD
ProxLAD
LAD
or
or
2VD
05% LAD
SYD
regarded more adversely than initial selection of CABG
therapy.
Our analysis strategy of starting all CABG and PTCA
patients in the medical therapy group and then restarting them at time zero in the appropriate revascularization group with updated baseline covariates is one of
several ways that have been used to control waiting-time
bias in observational treatment comparisons. We have
also performed our treatment comparisons using arbitrary time windows (eg, 60 days, 90 days). To correct for
waiting-time bias using this approach, which was used
by the CASS group as "method B." we started medical
follow-up at the average value of the time from catheterization to revascularization (7 days for PTCA and 14
days for CABG in our population).23 This approach
yielded results that are virtually the same as those
presented. Thus, we do not feel that the choice of
analytical methods used in our study affects our final
results.
In this study, we have presented only data on cardiovascular mortality end points. Additional analyses (not
presented) have shown similar results for all-cause
mortality. As noted above, repeat procedures and crossovers between PTCA and CABG and their consequences have been analyzed as part of the initial
treatment strategy (Fig 1). We have not included nonfatal myocardial infarction events in the present analysis. There is now strong evidence from the CABG
literature that revascularization does not prevent myocardial infarctions, although it does make them smaller
and more survivable.33-35 Whether PTCA has equivalent
effects relative to CABG in this regard is unsettled.
A related issue is that of incomplete revascularization. The present analysis does not consider the effects
of incomplete revascularization on PTCA outcomes, but
it is tempting to speculate that the diminished survival
benefits of PTCA relative to CABG in the more complex, severe CAD patients are at least partially attributable to this factor. Recent data, including results from
the NHLBI 1985-1986 PTCA Registry, have indicated
that complete revascularization is intended in only
about one third of multivessel-disease patients under-
going PTCA and is actually achieved in only two thirds
of the patients in whom it is attempted (ie, 25% to 50%
of two-vessel disease and 9% to 25% of three-vessel
disease patients).36-38 One of the main reasons for these
low figures appears to be chronic total occlusions, which
are associated with lower PTCA success rates and
relatively high complication rates.36,38 Another potential
reason is the substantial increase in the number of
angiographically identifiable individual lesions that are
present in more severe CAD patients that must be
attempted so as to provide complete revascularization.
Each site of dilation represents another opportunity for
abrupt closure or other acute complications and for
eventual restenosis. Consequently, the vast majority of
PTCA patients currently being treated receive three or
fewer individual lesion dilations.39A40
There are several important limitations that must be
considered in the interpretation of this study. First,
despite extensive efforts to control for treatment selection bias by use of both standard covariate adjustments
and covariate adjustment with treatment propensity
scores, we cannot rule out the possibility that residual
selection bias accounts for some of our observed treatment differences. Covariate adjustment can only correct
for observed imbalances. Unobserved or unmeasured
factors affecting both treatment selection and outcome
could have influenced our results. The consistency of
our CABG/medicine results with those reported previously by US4 and others provides at least a measure of
reassurance against this concern. However, this success
of a previous observational treatment comparison does
not guarantee the accuracy of the present analysis, and
direct validation of our results against the larger PTCA/
CABG randomized trials will not be possible for several
more years.
Second, the effects of lesion morphology on outcome
are not considered in our analysis. There is no evidence
that lesion-specific details, such as are contained in the
American Heart Association/American College of Cardiology classification, affect outcome with medicine or
CABG, but the long-term prognostic effects in the
PTCA population are uncertain. Certainly, the relation
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
2024
Circulation Vol 89, No S May 1994
between specific morphological characteristics and
acute success and complication rates have been well
described.41,42 However, unless such factors are strongly
associated with subsequent mortality, no long-term
prognostic effects may be evident. In a preliminary
analysis of this issue, we found no relation between
lesion morphology and long-term survival in our PTCA
cohort,43 so we do not feel that variations in lesionspecific characteristics in the PTCA cohort affected our
results.
Finally, treatment selection in CAD is a complex
decision involving consideration of multiple dimensions
of outcome, including functional status, symptom relief,
and perceived quality of life. Data on survival effects of
therapy, whether from observational studies or randomized trials, can only be one part of this decision and must
be interpreted in light of those other dimensions. In
particular, among low-risk patients with coronary disease, consideration of likely functional status and quality
of life outcomes should carry particular weight and are
currently being analyzed in our patient population. Ultimately, such multidimensional outcome data will need to
be combined with subjective patient preference data to
define not only the benefits of different treatment options
but also whether they are worthwhile.44
Acknowledgments
This study was supported by research grants HL-36587,
HL-45702, and HL-17670 from the National Heart, Lung, and
Blood Institute, Bethesda, Md; research grants HS-05636 and
HS-06503 from the Agency for Health Care Policy and Research, Rockville, Md; and a grant from the Robert Wood
Johnson Foundation, Princeton, NJ. The authors wish to
acknowledge the leadership of Drs Joseph C. Greenfield, Jr,
and David C. Sabiston, whose support and vision have made it
possible for us to continually evaluate and test the benefits of
the treatments we provide to our patients. We are also
indebted to our patients, without whose cooperation this study
would not have been possible. We appreciate critical reviews
of the manuscript provided by Dr Mark A. Hlatky and Dr Eric
J. Topol. Finally, we appreciate the editorial and technical
assistance of Lisa Breslau and Lori Baysden in the preparation
of the manuscript.
Contributing Cardiologists: Thomas M. Bashore, MD; Victor
S. Behar, MD; Robert M. Califf, MD; Stephen Culp, MD;
Julie K. Fetters, MD; Walter L. Floyd, MD; Donald F. Fortin,
MD; David J. Frid, MD; Christopher B. Granger, MD; J.
Kevin Harrison, MD; Sheila Kim Heinle, MD; Michael Higginbotham, MD; Mark A. Hlatky, MD; James G. Jollis, MD;
Yi-Hong Kong, MD; Mitchell W. Krucoff, MD; Daniel B.
Mark, MD, MPH; Kenneth G. Morris, MD; James J. Morris,
MD; Christopher M. O'Connor, MD; E. Magnus Ohman, MD;
Jose A. Perez, MD; Robert H. Peter, MD; Edward L.C.
Pritchett, MD; Harry R. Phillips, MD; David B. Pryor, MD;
Robert A. Rosati, MD; Michael H. Sketch, Jr, MD; Richard S.
Stack, MD; James E. Tcheng, MD; Thomas C. Wall, MD;
Robert A. Waugh, MD; Robert E. Whalen, MD; James P.
Zidar, MD.
Contributing Surgeons: James M. Douglas, Jr, MD; Donald
D. Glower, MD; Robert H. Jones, MD; James E. Lowe, MD;
H. Newland Oldham, Jr, MD; J. Scott Rankin, MD; David C.
Sabiston, Jr, MD; Peter K. Smith, MD; Peter Van Trigt, MD;
Andrew S. Wechsler, MD; Walter G. Wolfe, MD.
Contributing CardiacAnesthesiologists: Fiona Clements, MD;
Norbert P. Debruijn, MD; Joannes H. Kavis, MD; J.G. Reves,
MD; Thomas E. Stanley III, MD.
Follow-up: Charles B. McCants, Mary Rainey.
ComputerProgramming: Patricia Blunden, Donald F. Fortin,
MD, William E. Hammond, Jr, PhD.
Biostatistics: Elizabeth R. DeLong, PhD; Frank E. Harrell,
PhD; Kerry L. Lee, PhD; Lawrence H. Muhlbaier, PhD;
Charlotte L. Nelson, MS; Karen S. Pieper, MS; Linda K. Shaw;
Richard Smith, PhD.
Clinical Epidemiology: Robert M. Califf, MD; Donald F.
Fortin, MD; David Frid, MD; Mark A. Hlatky, MD; James G.
Jollis, MD; Daniel B. Mark, MD; E. Magnus Ohman, MD;
David B. Pryor, MD; Robert A. Rosati, MD.
References
1. Alderman EL, Bourassa MG, Cohen LS, Davis KB, Kaiser GG,
Killip T, Mock MB, Pettinger M, Robertson TL. Ten-year follow-up
of survival and myocardial infarction in the randomized Coronary
Artery Surgery Study. Circulation. 1990;82:1629-1646.
2. Varnauskas E, European Coronary Surgery Study Group.
Twelve-year follow-up of survival in the randomized European
Coronary Surgery Study. N Engl J Med. 1988;319:332-337.
3. The Veterans Administration Coronary Artery Bypass Surgery
Cooperative Study Group. Eleven-year survival in the Veterans
Administration randomized trial of coronary bypass surgery for
stable angina. N Engi J Med. 1984;311:1333-1339.
4. Califf RM, Harrell FE Jr, Lee KL, Rankin JS, Hlatky MA, Mark
DB, Jones RH, Muhlbaier LH, Oldham HN Jr, Pryor DB. The
evolution of medical and surgical therapy for coronary artery
disease: a 15-year perspective. JAMA. 1989;261:2077-2086.
5. Gersh BJ, Califf RM, Loop FD, Akins CW, Pryor DB, Takaro TC.
Coronary bypass surgery in chronic stable angina. Circulation.
1989;79(suppl I):I-464I-59.
6. Gruentzig AR, Senning A, Siegenthaler WE. Nonoperative dilatation of coronary-artery stenosis: percutaneous transluminal
coronary angioplasty. N Engi J Med. 1979;301:61-68.
7. 1991 Heart and Stroke Facts. Dallas, Tex: American Heart Association; 1992:2-19.
8. Hartz AJ, Kuhn EM, Pryor DB, Krakauer H, Young M, Heudebert G, Rimm AA. Mortality after coronary angioplasty and
coronary artery bypass surgery (the national medicare experience). Am J Cardiol. 1992;70:179-185.
9. Mock MB, Reeder GS, Schaff HV, Holmes DR Jr, Vlietstra RE,
Smith HC, Gersh BI. Percutaneous transluminal coronary angioplasty versus coronary artery bypass: isn't it time for a randomized
trial? N Engi J Med. 1985;312:916-919.
10. Parisi AF, Folland ED, Hartigan P. A comparison of angioplasty
with medical therapy in the treatment of single-vessel coronary
artery disease. N Engl J Med. 1992;326:10-16.
11. RITA Trial Participants. Coroilary angioplasty versus coronary
artery bypass surgery: the Randomised Intervention Treatment of
Angina (RITA) trial. Lancet. 1993;341:573-580.
12. The BARI Investigators. Protocol for the Bypass Angioplasty
Revascularization Investigation. Circulation. 1991;84(suppl VI):
VI-1-VI-27.
13. Editorial. BARI, CABRI, EAST, GABI, and RITA: coronary
angioplasty on trial. Lancet. 1990;335:1315-1316.
14. Rosati RA, McNeer JF, Starmer CF, Mittler BS, Morris JJ Jr,
Wallace AG. A new information system for medical practice. Arch
Intern Med. 1975;135:1017-1024.
15. Harris PJ, Harrell FE Jr, Lee KL, Behar VS, Rosati RA. Survival
in medically treated coronary artery disease. Circulation. 1979;60:
1259-1269.
16. Harrell FE Jr, Lee KL, Califf RM, Pryor DB, Rosati RA.
Regression modeling strategies for improved prognostic prediction. Stat Med 1984;3:143-152.
17. Pryor DB, Harrell FE Jr, Rankin JS, Lee KL, Muhlbaier LH,
Oldham HN Jr, Hlatky MA, Mark DB, Reves JG, Califf RM. The
changing survival benefits of coronary revascularization over time.
Circulation. 1987;76(suppl V):V-13-V-21.
18. Smith LR, Harrell FE Jr, Rankin JS, Califf RM, Pryor DB,
Muhlbaier LH, Lee KL, Mark DB, Jones RH, Oldham HN, Glower
DD, Reves JG, Sabiston DC Jr. Determinants of early versus late
cardiac death in patients undergoing coronary artery bypass graft
surgery. Circulation. 1991;84(suppl III):III-245-III-253.
19. Mark DB, Califf RM, Stack RS, Phillips HR. Cardiac catheterization. In: Sabiston DC, ed. The Davis - Christopher Textbook of
Surgery. 13th ed. Philadelphia, Pa: W.B. Saunders Co; 1986:
2135-2165.
20. Buller C, Mark DB, Phillips HR, Stack RS. Cardiac catheterization
and percutaneous coronary angioplasty. In: Sabiston DC, ed. The
Davis - Christopher Textbook of Surgery. 14th ed. Philadelphia, Pa:
W.B. Saunders Co; 1991:1826-1843.
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014
Mark et al Evolution of Therapy: Angioplasty
21. Rankin JS, Sabiston DC Jr. The coronary circulation. In: Sabiston
DC Jr, ed. The Davis - Christopher Textbook of Surgery. 14th ed.
Philadelphia, Pa: WB Saunders Co; 1991:1957-1972.
22. Kaplan EL, Meier P. Nonparametric estimation from incomplete
observations. JAm Stat Assoc. 1958;53:457-481.
23. Alderman EL, Fisher LD, Litwin P, Kaiser GC, Myers WO,
Maynard C, Levine F, Schloss M. Results of coronary artery
surgery in patients with poor left ventricular function (CASS).
Circulation. 1983;68:785-795.
24. Cox DR. Regression models and life-tables (with discussion). J R
Stat Soc. 1972;34:187-220.
25. Rosenbaum PR, Rubin DB. Reducing bias in observational studies
using subclassification on the propensity score. J Am Stat Assoc.
1984;79:516-524.
26. Cook EF, Goldman L. Asymmetric stratification: an outline for an
efficient method for controlling confounding in cohort studies. Am
JEpidemioL 1988;127:626-639.
27. Meyers WO, Gersh BJ, Fisher LD, Mock MB, Holmes DR, Schaff
HV, Gillispie S, Ryan TJ, Kaiser GC, Other CASS Investigators.
Time to first new myocardial infarction in patients with mild angina
and three-vessel disease comparing medicine and early surgery: a
CASS registry study of survival. Ann Thorac Surg. 1987;43:
599-612.
28. Kirklin JW, Akins CW, Blackstone EH, Booth DC, Califf RM,
Cohen LS, Hall RJ, Harrell FE, Kouchoukos NT, McCallister BD,
Naftel DC, Parker JO, Sheldon WC, Smith HC, Wechsler AS,
Williams JF, Fisch C, Beller GA, DeSanctis RW, Dodge HT,
Kennedy JW, Reeves TJ, Weinberg SL. Guidelines and indications
for coronary artery bypass graft surgery. JAm Coll Cardiol. 1991;
17:543-589.
29. Henderson RA. The Randomised Intervention Treatment of
Angina (RITA) trial protocol: a long term study of coronary angioplasty and coronary artery bypass surgery in patients with angina.
Br Heart J. 1989;62:411-414.
30. Weintraub WS, King SB III, Jones EL, Douglas JS Jr, Craver JM,
Liberman HA, Morris DC, Guyton RA. Coronary surgery and
coronary angioplasty in patients with two-vessel coronary artery
disease. Am J Cardiol. 1993;71:511-517.
31. Gersh BR. Efficacy of percutaneous transluminal coronary angioplasty (PTCA) in coronary artery disease: why we need randomized trials. In: Topol EJ, ed. Textbook of Interventional Cardiology. 2nd ed. Philadelphia, Pa: W.B. Saunders Co; 1994:251-273.
32. Fisher LD, Kaiser GC, Davis KB, Mock MB. Crossovers in
coronary artery bypass grafting trials: desirable, undesirable, or
both? Ann Thorac Surg. 1989;48:465-466.
33. Muhlbaier LH, Pryor DB, Rankin JS, Smith LR, Mark DB, Jones
RH, Glower DD, Harrell FE Jr, Lee KL, Califf RM, Sabiston DC
Jr. Observational comparison of event-free survival with medical
34.
35.
36.
37.
38.
39.
40.
41.
2025
and surgical therapy in patients with coronary artery disease: 20
years of follow-up. Circulation. 1992;86(suppl II):II-198-II-204.
Davis KB, Alderman EL, Kosinski AS, Passamani E, Kennedy JW.
Early mortality of acute myocardial infarction in patients with and
without prior coronary revascularization surgery. Circulation. 1992;
85:2100-2109.
Wiseman A, Waters DD, Walling A, Pelletier GB, Roy D, Theroux
P. Long-term prognosis after myocardial infarction in patients with
previous coronary artery bypass surgery. JAm Coil Cardiol. 1988;
12:873-880.
Bourassa MG, Holubkov R, Yeh W, Detre KM, The Co-investigators of the National Heart, Lung, and Blood Institute PTCA
Registry. Strategy of complete revascularization in patients with
multivessel coronary artery disease (a report from the 1985-1986
NHLBI PTCA registry). Am J Cardiol. 1992;70:174-178.
Vandormael M, Deligonul U, Taussig S, Kern MJ. Predictors of
long-term cardiac survival in patients with multivessel coronary
artery disease undergoing percutaneous transluminal coronary
angioplasty. Am J Cardiol 1991;67:1-6.
Bell MR, Bailey KR, Reeder GS, Lapeyre AC III, Holmes DR.
Percutaneous transluminal angioplasty in patients with multivessel
coronary disease: how important is complete revascularization for
cardiac event-free survival? JAm Coil Cardiol. 1990;16:553-562.
Detre K, Holubkov R, Kelsey S, Bourassa M, Williams D, Holmes
D Jr. One-year follow-up results of the 1985-1986 National Heart,
Lung, and Blood Institute's Percutaneous Transluminal Coronary
Angioplasty Registry. Circulation. 1989;80:421-428.
Vacek JL, Rosamond TL, Stites HW, Rowe SK, Robuck W,
Dittmeier G, Beauchamp GD. Comparison of percutaneous transluminal coronary angioplasty versus coronary artery bypass
grafting for multivessel coronary artery disease. Am J Cardiol.
1992;69:592-597.
Ellis SG, Vandormael MG, Cowley MJ, DiSciascio G, Deligonul
U, Topol EJ, Bulle TM, Multivessel Angioplasty Prognosis Study
Group. Coronary morphologic and clinical determinants of procedural outcome with angioplasty of multivessel coronary disease:
implications for patient selection. Circulation. 1990;82:1193-1202.
42. Myler RK, Shaw RE, Stertzer SH, Hecht HS, Ryan C, Rosenblum
J, Cumberland DC, Murphy MC, Hansell HN, Hidalgo B. Lesion
morphology and coronary angioplasty: current experience and
analysis. JAm Coll Cardiol. 1992;19:1641-1652.
43. Fortin DF, Nelson CL, Gardner LH, Harrell F Jr, Frid DJ, Tcheng
JE, Phillips HR III, Califf RM, Mark DB, Stack RS. Preangioplasty lesion morphology: a case of diminishing effect on
survival? JAm Coil Cardiol 1993;21:35A. Abstract.
44. Wong JB, Sonnenberg FA, Salem DN, Pauker SG. Myocardial
revascularization for chronic stable angina: analysis of the role of
percutaneous transluminal coronary angioplasty based on data
available in 1989. Ann Intern Med. 1990;113:852-871.
Downloaded from http://circ.ahajournals.org/ by guest on September 9, 2014