DISCUSSION
This is the first study in patients with class III obesity to directly
compare the feasibility and interpretability of DSE with and without the
use of intravenous contrast enhancement, both cumulatively and on a
segment-by-segment basis.13-19 The results indicate
that in patients with class III obesity, DSE performed using intravenous
contrast enhancement is superior to DSE performed without intravenous
enhancement, particularly with regard to visualization of the lateral
and anterior segments. Moreover, DSE without intravenous contrast was
associated with a significantly higher frequency of poor- or
non-visualization of ≥ 2 segments (rendering them uninterpretable) than
DSE performed with intravenous contrast. Our study consisted of patients
with class III obesity without chest pain and without previously
documented CAD. Most of our
Alpert
patients had at least 1 traditional CV risk factor and many had
>1 risk factor. Of the 128 patients studied, 55 were
diabetic. Thus, our study population simulated a group of patients with
class III obesity who may be referred for stress testing prior to
bariatric surgery to screen for myocardial ischemia in the absence of
chest pain or documented pre-existent CAD. This comprises the majority
of such referrals. Our results further suggest that myocardial ischemia
is extremely uncommon in such patients.
Nine prior studies have evaluated patients with various degrees of
obesity with transthoracic exercise stress echocardiography (ESE) or DSE
to determine the feasibility of these studies in this population (4
studies), to assess prognosis based on the presence or absence of
myocardial ischemia and other CV outcomes (8
studies).12-18 or both (3 studies). Lerakis et al
studied 611 patients whose BMI was ≥ 40 kg/m2 to
determine the feasibility of DSE prior to and the short- and long-term
outcomes following bariatric surgery based on the results of the stress
test.12 The study population consisted of 86.6% women
and 13.4% men (mean age: 42 ± 10 years). The mean BMI was 48.0 ± 6.1
kg/m2 and mean body weight was 136 ± 18 kg. The
end-points were all-cause mortality and multiple adverse cardiac events
(cardiac death, acute coronary syndrome, urgent coronary
revascularization) at 30 days and at 6 months following surgery.
Adequate baseline echocardiographic images were achieved in 590 patients
(96.6%). Intravenous contrast was used in 72.2% of cases. Those
without adequate baseline images were referred for other procedures to
assess for myocardial ischemia. DSE was negative in 545 patients
(92.4%), was inconclusive in 6.4%, and was positive in 1 patient
(1.2%). Laparoscopic surgery was performed in 77% of cases. There were
3 early deaths (all from sepsis). There were no major cardiac events
within 30 days of surgery. One patient developed acute coronary
Alpert
syndrome between 30 days and 6 months. Shah and colleagues reported the
results of the multicenter Study of Ultrasound in Morbid Obesity
(SUMO).13 Morbid obesity was defined as a BMI ≥ 35
kg/m2. The study population consisted of 209 subjects
who were referred for evaluation of suspected CAD. The mean BMI was 39.3
± 4.1 kg/m2. ESE was performed in 60% and DSE in 40%
of subjects. Intravenous contrast was used in 96% of patients. Mean
follow-up was 13 months. Outcomes studied were all-cause mortality,
myocardial infarction, and late coronary revascularization. Inducible
ischemia was detected in 32 patients (15%). Coronary angiography was
performed in 25 patients. Twenty-two had obstructive coronary artery
stenotic lesions in distributions that corresponded with the location of
myocardial ischemia during ESE or DSE and 77% of whom received coronary
revascularization. Predictors of the outcomes were an LV ejection
fraction <50% and inducible ischemia. Supariwala and
colleagues retrospectively studied 652 consecutive obese patients (BMI ≥
30 kg/m2) with multiple CV risk factors who underwent
stress echocardiography.14 DSE was performed on 423
patients (65%) and 229 (35%) underwent ESE. The study population
consisted of 84% women (mean age: 47 ± 10 years). Mean BMI was 47 ± 9
kg/m2 and mean body weight was 280 ± 69 pounds. The
study population was 12% Caucasian, 56% Hispanic, and 30%
African-American. Mean follow-up was 3.2 ± 2.7 years. The studies were
classified as technically difficult in 293 patients (45%) and
intravenous contrast was used in 229 subjects (35%). Target heart rate
was achieved in 553 patients (85%). The test was considered suboptimal
in 64 patients (10%). During the follow-up period, there were 8 deaths.
Deaths occurred in 1 patient with obesity (BMI: 30.0-39.9
kg/m2), in 3 patients with extreme obesity (BMI:
40.0-49.9 kg/m2), and in 4 patients with super-obesity
(BMI ≥ 50 kg/m2). Hu and co-workers studied 62
patients who were overweight or obese (BMI 26-33
Alpert
kg/m2).15 DSE was performed with and
without intravenous contrast injection. Visualization of LV segments was
scored as 0 (not-visualized), 1 (poorly-visualized), and 2
(well-visualized). A total of 992 segments were assessed. Each patient
served as their own control with regard to acquisition of images with
and without intravenous contrast enhancement (this process was not
clearly-described). Compared to non-utilization of intravenous contrast,
use of intravenous contrast was associated with higher sensitivity (82%
vs. 70%), specificity (78% vs. 67%), and accuracy (81% vs. 69%).
Visualization scores without contrast were 0 in 179 segments (18%), 1
in 328 segments (38%), and 2 in 433 segments (44%). In the group that
received intravenous contrast visualization scores were 0 were in 0
segments, 1 in 50 segments (5%) and 2 in 942 segments (95%). Silviera
and colleagues performed physical stress echocardiography on 945 obese
and 3105 non-obese patients.16 No significant
differences were noted in the incidence of myocardial ischemia in obese
patients (19.0%) compared to non-obese patients (17.9%). Logistic
regression analysis identified age, female gender, diabetes mellitus,
and hypertension, but not obesity as predictors of myocardial ischemia.
Murphy et al performed a retrospective chart review on 704 patients to
determine the prognostic value of a normal stress echocardiogram in
overweight and obese subjects.17 The study population
consisted of 366 obese patients (BMI ≥ 30 kg/m2), 196
overweight subjects, (BMI 25.0-29.9 kg/m2), and 142
normal weight patients (BMI 18.5-24.9 kg/m2) The
follow-up period was 1 year. The end-point was multiple adverse cardiac
events (myocardial infarction, cardiac death, cardiac hospitalization,
or emergency department visit). No adverse cardiac events were observed
during the follow-up period. Khan and co-workers performed DSE on 555
African American patients.18 Intravenous contrast was
used when non-contrast images of LV segments were poorly-visualized.
There were 409 obese
Alpert
and 146 non-obese subjects. Multivariate analysis showed that myocardial
scar independently predicted all-cause mortality and multiple adverse
cardiac events. There were no significant differences in these outcomes
between obese and non-obese groups. In a recently-published
case-control study these outcomes
between obese and non-obese groups. In a recently-published case-control
study of 1018 patients with a BMI > 35
kg/m2 who had suffered prior myocardial infarction and
who were followed for a median of 4.6 years, Naislund et al compared
long-term CV outcomes in 509 patients undergoing metabolic (bariatric)
surgery (predominately Roux-en-Y gastric bypass) and 509 patients with
severe obesity (BMI ≥ 40 kg/m2) who did not undergo
metabolic surgery.19 Primary outcomes were all-cause
deaths and re-admission for myocardial infarction. Secondary outcomes
were ischemic or hemorrhagic stroke. Kaplan-Meier analysis showed that
metabolic surgery was associated with significantly higher myocardial
infarction- and stroke-free survival at 8 years compared to non-surgical
controls. Those undergoing metabolic surgery also experienced
significantly fewer re-admissions for new-onset heart failure compared
to controls. In another recently-published propensity score matched
cohort study of 2638 patients with prior CV disease followed for a
median of 4.6 years, Doumouras and colleagues reported CV outcomes in
1319 patients with class II and class III obesity undergoing bariatric
surgery and 1319 patients who did not undergo bariatric
surgery.20 BMI thresholds were > 35
kg/m2 with a co-morbidity or ≥ 40
kg/m2 without a co-morbidity. Patients receiving
bariatric surgery had significantly lower incidence values for multiple
adverse CV events (all-cause mortality, myocardial infarction, coronary
revascularization, cerebrovascular events, and heart failure
hospitalizations compared to those of patients who did not undergo
bariatric
Alpert
surgery. Similar results were noted for a 3-component multiple adverse
CV event model (all-cause mortality, myocardial infarction and ischemic
stroke.
Our study differed from previous studies evaluating transthoracic stress
echocardiography in severely-obese patients in multiple ways. Six of the
previous 9 studies were retrospective and relied heavily on chart review
or registry information.14,16-19 Our study was
prospective and provided a head-to-head comparison of patients with
class III obesity receiving intravenous contrast with those who did not
receive intravenous contrast. In prior studies, intravenous contrast was
used only when baseline non-intravenous contrast images were
unsatisfactory except for the study by Hu et al which focused on
overweight and mildly obese patients.12-14,16-18 The
presence of poor- or non-visualization of ≥2 LV segments, a marker of
reduced diagnostic accuracy and interpretability21,
occurred with significantly greater frequency in the group that did not
receive intravenous contrast group than in the group that received
intravenous contrast. In previous studies, ESE or DSE were performed
without regard as to whether patients had a history of chest pain or
pre-existent CAD.13-18 In our study, we excluded
patients with chest pain or pre-existent CAD. We did so to answer the
question whether screening lower risk patients with class III obesity
for myocardial ischemia was necessary prior to elective low-risk
surgery. Our results suggest that they do not require routine screening
for ischemia. Regarding feasibility, several prior studies provided the
number of patients who received intravenous contrast, but did not cite
the precise criteria for these decisions.13,14,17 We
not only provided the numbers and percentages of segments
well-visualized and interpretable in the intravenous contrast and
non-intravenous contrast groups, but also provided a segment-by-segment
analysis to determine which LV segments were most likely to be
poorly-visualized and
Alpert
not interpretable in both groups. Previous studies performing ESE and
DSE and did not comment on whether outcomes differed based on the
technique that was chosen.13,14,17 Finally, in 7 of
the 9 previous studies, the BMI threshold for entry into the study was
<40 kg/m2 which indicates that not all
patients studied were class III obese
patients.13,14,16-18 The BMI in all of our patients
was ≥ 40 kg/m2.