Thirty-two patients with chronic ischemic cardiomyopathy and LV ejection fraction (EF) <40% on gated resting Tc-99m-SPECT, were included. All patients were in sinus rhythm. Patients with a recent infarction (<3 months), unstable angina, valvular disease pacemakers and/or intracranial clips were excluded.

Patients were included consecutively. Patients that did not qualify for revascularization were assigned to the β-blocker therapy group. Patients did not qualify for revascularization because: 1. Patients had poor target vessels (small vessels, not amendable for revascularization) 2. Patients had prior existing co-morbidity (e.g. renal failure); 3. Patients refused to undergo revascularization. β-blocker therapy was started at an initial dose of 3.125 mg carvedilol twice daily. Subsequently, carvedilol was titrated at 1-week intervals as tolerated, up to target dose of 25 mg twice daily [¹⁹].

β-Blocker treated patients were compared to patients who underwent revascularization. In the revascularization patients, coronary artery bypass surgery was performed in 75% and percutaneous coronary intervention in 25%. Each patient gave informed consent to the study protocol that was approved by the local ethics committee.

Patients were studied by CMR before therapy and at 8 ± 2 months after therapy. At baseline, the CMR protocol consisted of a resting cine CMR, Late Gadolinium Enhancement (LGE) and low-dose dobutamine (10 μg/kg/min) cine CMR. Follow-up CMR included rest function evaluation only. The congestive heart failure classification by the New York Heart Association (NYHA) was determined at baseline and follow-up by the patient's cardiologist, who was blinded to the CMR data.

A 1.5-Tesla Gyroscan ACS-NT CMR scanner (Philips Medical Systems, The Netherlands) and a 5 elements cardiac synergy coil were used. Patients were studied in supine position. All images were acquired during breath-holds of approximately 15 seconds and were gated to the vector ECG; blood pressure was monitored continuously during the examination (Millennia, Invivo Research, Orlando Fla, USA). The heart was imaged from apex to base with 10 to 12 imaging levels (dependent on heart size) in short-axis view using a steady-state free-precession sequence. Typical acquisition parameters were: field of view 400 × 400 mm², matrix size 256 × 256, slice thickness 10.00 mm, flip angle 50°, time to echo 1.82 ms and time to repeat 3.65 ms. Temporal resolution was 25-39 ms, depending on heart rate. Geometry settings of rest cine CMR scans were stored and repeated for LGE and low-dose dobutamine stress CMR.

LGE images were acquired approximately 15 minutes after bolus injection of gadopentate dimeglumine (Gadolinium-DTPA, Magnevist, Schering/Berlin, Germany, 0.15 mmol/kg) with an inversion-recovery gradient echo sequence; inversion time was determined using real time planscan. Typical parameters were the following: field of view 400 × 400 mm², matrix size 256 × 256, slice thickness 5.00 mm, slice gap - 5 mm, flip angle 15°, time to echo 1.36 ms and time to repeat 4.53 ms.

For evaluation of myocardial function under pharmacological stress, intravenous dobutamine infusion was started at a rate of 5 μg/kg/min and increased after 5 min to 10 μg/kg/min. Then, after 5 min (at steady state), low-dose dobutamine images were acquired in 2- and 4-chamber and short-axis views. The same parameters were applied as described for rest imaging.

Data were analyzed on a remote workstation using MASS software (MASS, Medis, The Netherlands). The endo- and epicardial borders of the end-diastolic and end-systolic frames were manually traced with exclusion of papillary muscles, trabeculae, and epicardial fat [²⁰]. LV end-diastolic and end-systolic volumes were calculated and LVEF was derived.

The amount of infarcted tissue was determined by drawing (manually) regions of interest around the scar tissue. In addition, the percentage of myocardium was calculated that was affected by infarction, relative to the total LV mass. LGE images were scored visually by two experienced observers (blinded to other CMR and clinical data) using a 17-segment model. Each segment was graded on a 5-point scale (segmental scar score) with 0: absence of LGE, 1: LGE of 1-25% of LV wall thickness, 2: LGE extending to 26-50%, 3: LGE extending to 51-75%, and 4: LGE extending to 76-100% of the LV wall thickness [²¹].

To determine regional wall motion at rest, cine CMR images were visually interpreted by two experienced observers (blinded to other CMR and clinical data) using a 17-segment model. Each segment was assigned a wall motion score using a 5-point scale with 0: normal wall motion, 1: mild hypokinesia, 2: severe hypokinesia, 3: akinesia, and 4: dyskinesia [²²]. In the dysfunctional segments at rest (score 1-4), the presence or absence of contractile reserve was based on visual analysis of the difference in myocardial wall motion between CMR acquisitions at rest and during infusion of low-dose dobutamine. An improvement in segmental wall motion score by 1 grade or more was considered indicative of contractile reserve.

Myocardial segments with a visual wall motion score at rest from 1 to 4 were considered as dysfunctional segments. The myocardial segments next to these dysfunctional segments in 3-dimensions, were considered as adjacent segments. The remaining myocardial segments were considered as remote tissue (Figure 1). Wall thickness was then quantified in the above defined three regions: (1) dysfunctional, (2) adjacent, and (3) remote segments, using the centerline method as described before [²³]. Wall thickening was calculated based on the difference in wall thickness between end-diastole and end-systole. The change in wall thickening was calculated between baseline and dobutamine stress acquisitions, or between baseline and follow-up acquisitions.

Continuous data were expressed as mean ± SD and compared using the Student's t test for (un-) paired data when appropriate. Stepwise multivariate analysis was performed to determine the relation between regional improvement in myocardial function and improvement in global LVEF at follow up. All tests were two-tailed and a P-value < 0.05 was considered statistically significant.

The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.

Baseline characteristics are summarized in Table 1. The variables listed in Table 1 were not statistically significantly different between patients in the β-blocker and revascularization groups (Table 1). Table 2 summarizes LV dimensions and global systolic function. At baseline, there were no statistically significant differences between both patient groups (Table 2). Table 3 shows scar burden and scar morphology. At baseline, there were no statistically significant differences between both patient groups (Table 3).

The response to low-dose dobutamine at baseline in the dysfunctional, adjacent, and remote myocardial segments is summarized in Figure 2 (left panel). There were no statistically significant differences in wall thickening in any myocardial segment, when comparing the β-blocker and revascularization groups at baseline (Figure 2).

Clinical assessment revealed that the NYHA classification changed from 2.3 ± 0.5 at baseline to 1.8 ± 0.6 at follow-up (P < 0.05) in the β-blocker treated patients, and from 2.2 ± 0.5 to 1.5 ± 0.5 in the revascularization patient group (P < 0.05). After β-blocker or revascularization therapy, LV end-diastolic volume did not change significantly in both patient groups while LV end-systolic volume and LVEF improved significantly in both the β-blocker and revascularization groups to the same extent (Table 2).

Comparison of myocardial wall thickening between follow-up and baseline, revealed that in the β-blocker group the remote segments showed the largest improvement (Figure 2, right upper panel). Stepwise multivariate analysis in β-blocker patients revealed that improvement in LVEF after therapy was mainly related to improvement in function of the remote region (y = 2 × Rmt + 1.8, where Rmt = difference in wall thickening at rest in remote tissue between follow-up and baseline, P= 0.002, acceptance value 0.05). In the revascularization patient group, most improvement in myocardial wall thickening between follow-up and baseline was achieved in the dysfunctional segments (Figure 2, right lower panel). Stepwise multivariate analysis in revascularization patients revealed that improvement in LVEF was mainly related to improvement in function of the dysfunctional and adjacent segments (y = 2.6 × Dsf + 1.8 × Adj + 0.6, where Dsf = difference in wall thickening at rest in dysfunctional segments between follow-up and baseline, Adj = difference in wall thickening at rest in adjacent segments between follow-up and baseline, P= 0.001, acceptance value 0.05).

Direct comparison between the β-blocker and revascularization groups concerning the myocardial segments shows a statistically significant difference (P < 0.05) in wall thickening between the dysfunctional and remote segments (Figure 2 right panel, top and bottom). As a consequence, a reversed pattern is observed in the contribution of the dysfunctional and remote myocardial segments to improvement in wall thickening after β-blocker or revascularization therapy. This differential pattern is also illustrated in Figure 3, showing the relative contributions of the myocardial segments to improvement in LVEF after β-blocker or revascularization therapy. In the β-blocker treated patients (Figure 3, top), the remote tissue contributes for 60% to the improvement in LVEF, whereas in the revascularized patients (Figure 3, bottom), the dysfunctional segments contributed for 56% to the improvement in LVEF after therapy.

At baseline, NYHA classification was similar for β-blocker treated and revascularization patients. In addition, global systolic function, the % scar tissue or the number of segments with contractile reserve were similar in both patient groups at baseline. Furthermore, both patient groups showed the same response to low-dose dobutamine at baseline in the dysfunctional, adjacent, and remote myocardial segments. Accordingly, the clinically defined patient groups were comparable at baseline concerning myocardial function.

Following β-blocker or revascularization therapy, LV systolic function improved significantly, and to a similar extent. Accordingly, the global functional response to therapy was comparable between both patient groups. This finding is consistent with results from previous studies [⁵,⁶,⁹,¹²,²⁴,²⁵]. Furthermore, improvement in LVEF was accompanied by an improvement in NYHA classification from baseline to follow up, to a similar extent in both patient groups.

Stepwise multivariate analysis in β-blocker patients revealed that improvement in LVEF after therapy was mainly related to improvement in function of the remote myocardium. The present results concerning β-blocker therapy effect on remote tissue are in agreement with previous studies by Reiken et al. Both in a canine model [²⁶] and in a patient study (performed in explanted hearts) [²⁷], the authors showed that β-blockers normalized Ca⁺⁺-channel function in failing myocardium. Remote myocardium in the current study may be regarded as myocardium with relatively preserved myocyte function as compared to adjacent and dysfunctional regions.

However, in remote tissue, myocardial hypertrophy may develop when excessive pressure or volume overload is imposed to sustain the burden of a dysfunctional segment [²⁸]. The remote hypertrophied tissue may appear to contract normally, but could in fact represent "pseudonormalized" myocardium. Therefore, the relatively mildly failing remote myocardium can still show a positive response to restoration of the Ca⁺⁺-channel function by administration of β-blockers, whereas the adjacent and dysfunctional tissue cannot. Furthermore, recent data demonstrated that the effect of β-blocker therapy could be predicted by the increase in LVEF during low-dose dobutamine infusion [²⁹]. Since dobutamine is a β-receptor agonist, it may be able to temporarily mimic sympatic nervous system stimulation, and predict the effect of β-blockers on LVEF. Interestingly, the regional response pattern after β-blocker therapy in the current study (Figure 2, right upper panel) is similar to the regional response to dobutamine in β-blocker treated patients (left upper panel). The latter observation supports the hypothesis that low-dose dobutamine infusion in patients with chronic ischemic LV dysfunction may predict clinical outcome after β-blocker therapy [²⁹].

Stepwise multivariate analysis in revascularization patients revealed that improvement in LVEF was mainly related to improvement in function of the dysfunctional and adjacent myocardial segments. This observation is in agreement with data obtained in previous studies [³⁰,³¹]. Recovery of hibernating myocardium occurs after successful revascularization [¹¹,³²,³³]. In the present study, no improvement in wall thickening was noted in the remote region in the revascularization patients. The latter finding may be explained by the fact that only blood flow to dysfunctional and surrounding adjacent tissue is restored. As a result, no short-term improvement in function can be expected in the remote area. An interesting finding in the present study was that adjacent myocardium also contributed significantly to the improvement of global systolic function after revascularization therapy.

The number of patients was relatively small and therefore the observations need confirmation in larger cohorts. Myocardial tagging was not performed; this could have provided a more detailed strain assessment across the myocardial segments of interest.

All comparisons between both groups were statistically non-significant.

LVEDV: left ventricular end-diastolic volume; LVESV: left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction.

*: P < 0.05 for baseline versus follow-up values. Other comparisons were non-significant.

LGE Score: 0: absence of LGE, 1: LGE of 1-25% of LV wall thickness, 2: LGE extending to 26-50%, 3: LGE extending to 51-75%, and 4: LGE extending to 76-100% of the LV wall thickness. There were no statistically significant differences between both patient groups.