Author + information
- Received November 5, 2002
- Revision received January 30, 2003
- Accepted March 12, 2003
- Published online August 20, 2003.
- Ronald G Schwartz, MD, MS, FACC*,* (, )
- Thomas A Pearson, MD, PhD, FACC*,
- Vijay G Kalaria, MD, FACC†,
- Maria L Mackin, CNMT*,
- Daniel J Williford, MD, PhD, FACC*,
- Ashish Awasthi, MD*,
- Abrar Shah, MD*,
- Adam Rains, MSc* and
- Joseph J Guido, MS*
- ↵*Reprint requests and correspondence:
Dr. Ronald G. Schwartz, University of Rochester Medical Center, Cardiology Unit, Box 679, 601 Elmwood Avenue, Rochester, New York 14642-8679, USA.
Objectives This study was designed to assess prospectively changes in serum lipid profile and myocardial perfusion with serial radionuclide single photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) during the first six months of pravastatin therapy.
Background Morbid coronary events occur despite statin therapy and lipid-lowering in patients with coronary artery disease (CAD). A reliable strategy to identify responders with effective treatment from nonresponders on statin therapy before clinical events is needed.
Methods Rest and stress SPECT MPI and lipids were assessed serially in 25 patients (36% women) with CAD and dyslipidemia during the first six months of pravastatin therapy.
Results Total cholesterol, low-density lipoprotein cholesterol, and triglycerides declined (26%, 32%, and 30%, respectively) by six weeks and remained reduced at six months. Mean stress perfusion defect (summed stress score [SSS]) was severe (13.3 ± 6.0) at baseline, showed no change at six weeks, and improved significantly at six months (10.3 ± 7.3, p < 0.01). The six-month study SSS improved in 11 (48%) patients, was unchanged in 10 (43%) patients, and worsened in 2 (9%) patients. Changes in lipid levels did not reliably predict changes in myocardial perfusion at six weeks or six months in this small pilot study.
Conclusions Serial SPECT MPI demonstrated improved stress myocardial perfusion in 48% of patients treated for six months with pravastatin. Time course of improved myocardial perfusion during pravastatin therapy is delayed compared to lipids. Direction and magnitude of changes in the myocardial perfusion vary and do not correlate closely with improvements in lipids.
Statins modify serum lipid profiles favorably and reduce coronary events substantially. Despite the documented effectiveness of statins, annual cardiovascular event rates in statin-treated patients are nontrivial: 1.7% to 2.5% in primary prevention trials (1,2)and 6.4% to 7.8% in secondary prevention trials (3–6). Identifying treated patients with effectively lowered cholesterol levels who remain at risk of coronary events remains problematic.
The complexity and pleiotropism (7)of therapeutic responses to statin therapy suggest the potential value of monitoring strategies more predictive than the cholesterol profile to determine effective therapeutic response before clinical events. An ideally predictive modality would facilitate timely adjustment of therapeutic strategy to yield improved clinical outcomes in nonresponsive patients. Radionuclide single photon emission computed tomography (SPECT) and positron emission computed tomography (PET) myocardial perfusion imaging (MPI) provides risk assessments in patients with known or suspected coronary artery disease (CAD) incremental to clinical, exercise, and/or coronary arteriographic variables (8–14). Abnormalities of stress-induced coronary vasomotion identify site-specific location of stenosis development and predict morbid coronary events (15). Effective cholesterol-lowering improves coronary endothelium-dependent vasodilation and myocardial perfusion by invasive measurement with Doppler flow wire (16)and by PET and SPECT radionuclide MPI (13,17–26), as well as increasing coronary artery diameter by contrast arteriography (13,27).
Improvements in perfusion accompanying cholesterol-lowering with serial radionuclide MPI have been reported from 30 days (20)to 5 years (13). However, the incidence and time course of serial changes in myocardial perfusion, relative to changes in lipid profile and clinical course, remain unclear. One prior PET MPI study observed that statin-induced augmentation of coronary flow reserve at six months by PET MPI is delayed compared to its lipid-lowering effects at two months (25). Given the established lipid-lowering effect of pravastatin by six weeks and its therapeutic effectiveness in reducing coronary events as early as six months in the West of Scotland trial in men (1), and in women in the Cholesterol And Recurrent Events trial (3), the present study was designed to compare the time course of changes in lipid levels and stress perfusion by SPECT MPI following the initiation of pravastatin therapy.
Twenty-five dyslipidemic (total cholesterol [TC] ≥200 mg/dl, low-density lipoprotein [LDL] ≥130 mg/dl, or TC:high-density lipoprotein [HDL] ≥4) patients with clinical CAD and stress-induced myocardial perfusion defects were enrolled. Patients were excluded if they took statin therapy during the previous two months. With approval of the Research Subjects Review Board, the informed consent process was performed for all patients.
Therapeutic intervention, compliance, and follow-up
Pravastatin 40 mg qhs for six months was initiated and compliance was assessed by monthly pill counts. Follow-up clinical evaluation at six weeks and six months was performed with lipid levels, liver function tests, and SPECT MPI. Consideration of fibrate and niacin therapy was given to patients with TC:HDL >4 at six weeks. All patients were counseled closely with regard to lifestyle modification including smoking cessation, calorie-constrained diet to achieve ideal body weight with <30% fat calories, low cholesterol, and at least 30 min aerobic exercise most days of the week.
Stress testing and pre-testing medication management
Medication usage was reviewed and beta-blockers, calcium antagonists, nitrates, and angiotensin-converting enzyme inhibitors were routinely held to standard therapeutic trough levels before stress testing. Patients were instructed to avoid caffeine consumption for 24 h before the stress test. Exercise and adenosine stress testing was performed in 11 and 12 patients, respectively. In all but two patients the stress testing modality was maintained for all serial imaging (baseline, six weeks, and six months). Rate-pressure product and Duke treadmill exercise scores were calculated in the 11 patients who underwent exercise testing.
Spect perfusion imaging and interpretation
Routine dual tracer (rest-thallium-201, exercise, or adenosine stress Tc-99m-sestamibi) SPECT MPI was performed using a dual-head SPECT DST camera (GE/SMV Medical Systems, Milwaukee, Wisconsin) with 60-s acquisitions per stop for 32 total stops (16 stops per head) with low energy high resolution collimators. Rest and stress perfusion images were processed on MIRAGE PC-based software (Segami Corp., Ellicott City, Maryland). Vertical and horizontal spatial filtering of the back-projected data was performed with a Butterworth low pass filter (frequency cut-off = 0.4, order = 6, size = 15). Consensus interpretation using quantitative analysis of SPECT MPI data with visual over-reading by three readers was employed. Semiquantitative regional scoring was performed for stress and rest images using the Cedars method (9,12), which employs a five-point scale from normal (0) to absent tracer (4). Semiquantitative regional scores were recorded in the Nuclear Cardiology Database System using PC-based Visual Foxpro (Microsoft Corp., Seattle, Washington) with automated calculation of summed perfusion scores (summed rest score [SRS], summed stress score [SSS], summed difference score [SDS]). Differences in summed perfusion scores (SSS, SRS, and SDS) from baseline to each follow-up period were rated as previously described (12)(≥3 = “improved,” >−3 to <3 = “no change,” <−3 = “worse”). Left ventricular (LV) volume and ejection fractions were obtained from the Vision PowerStation using MultiDim (GE/SMV Medical Systems) with the use of no-acquisition zoom, 2.5:1 intraprocessing zoom, and a Metz Filter (3.5 full width half maximum, order 8 resulting in a 2.7-mm pixel size) as validated in our laboratory.
To verify scintigraphic results, blinded visual analysis of perfusion images and automated quantitative analysis of the percent of LV hypoperfusion by defect score and global extent on three-dimenstional topograms was performed using the Yale Wackers-Liu CQ method (28)(Eclipse Systems, Branford, Connecticut). The direction and magnitude of changes in the myocardial perfusion and lipid profile components were noted for each patient and for the entire study cohort.
The existence of changes in lipids (TC, HDL, TC:HDL, LDL, triglycerides), scan readings (SSS, SRS, SDS), and the percent of LV hypoperfusion of defect size and global defect extent (Wackers-Liu CQ method) across the three time periods (baseline, six weeks, and six months) of the study was evaluated by analysis of variance. Values found by analysis of variance to have changed significantly within the study period were subjected to pairwise comparison of means between each time period using the paired ttest. The power of the analysis of variance and paired ttests to correctly identify deviations from the null hypothesis was determined post-hoc using the methods described by Zar (29). For all analyses, the significance level was set at 5% and dropouts were excluded from consideration. All analyses were conducted using SAS/STAT for Windows (version 8.2, SAS Institute Inc., Cary, North Carolina).
Preliminary evaluation of lipid, scan, and perfusion values by analysis of variance identified statistically significant (p < 0.05) changes in the following measures during the study period: SSS, TC, HDL, LDL, triglycerides, quantitative percent of the LV hypoperfusion of stress defect score, global extent, and TC:HDL ratio. On the basis of follow-up analysis by paired ttest, significant changes in each measure were noted as follows: TC, LDL, and triglycerides changed during the first time period (baseline to six weeks); SSS, quantitative defect extent, and HDL changed during the second time period (six weeks to six months); and TC:HDL changed during both time periods.
Clinical characteristics of the patient population
Of the 25 patients (9 women; 36%) who enrolled in the trial, 23 (92%) completed the entire study. Two study dropouts excluded from analysis included a man referred for coronary artery bypass graft surgery because of decreased perfusion at six weeks and a woman with improved perfusion at six weeks who failed subsequent follow-up. Baseline clinical characteristics of the 23 patients completing the study are summarized in Table 1. Eleven of the 23 recruited patients had a new diagnosis of CAD. Medications taken at baseline, six weeks, and six months are listed in Table 2. With the exception of three patients who received angiotensin-converting enzyme inhibitors and five patients who received niacin between six weeks and six months, most medication changes were completed before the six-week visit.
Compliance and tolerance of pravastatin
On the basis of pill counts, mean compliance with the pravastatin regimen was ≥95% (minimum 91.1%). No patient experienced significant elevations (>2 × upper normal limit) in liver function tests, or reported myalgia, fatigue, or other adverse effects attributable to pravastatin.
Eleven of the 23 study patients underwent provocative stress testing with an exercise protocol. Angina during exercise tolerance testing (ETT) was reported by three patients at baseline, one patient at six weeks, and none at six months. Group mean rate-pressure product at baseline (28,222 ± 4,752) and Duke treadmill exercise score (0.14 ± 10.91) were unchanged at six weeks and six months.
Lipid profile and SPECT MPI during pravastatin treatment
Lipid profile, the number of patients actively smoking and with hypertension, and SPECT MPI values at each time period are presented in Table 3. Statistically significant reductions were noted at six weeks for LDL (32%), TC (26%), and triglycerides (30%), which persisted at six months. Statistically significant changes were also noted at six months for HDL (15% increase), SSS (23% decrease), SDS (40% decrease), and quantitative defect size and topographic extent percent of LV hypoperfusion (35% decrease) (Figs. 1D to 2B), ⇓with no change at six weeks. The TC:HDL declined 25% from baseline ⇓to six weeks and declined 32% from six weeks to six months. No significant changes in SRS, quantitative resting perfusion defect size or extent, left ventricular ejection fraction, or left ventricular end-systolic volume index were noted at six weeks or six months (Table 3).
Average baseline SSS was severe (13.3 ± 6; mean ± SD) and declined to a moderate-average score (10.3 ± 7.3) at six months. Of the 23 patients completing the study, SSS declined in 11 (48%) patients, which included complete normalization (<4) in 5 (22%) patients; SSS remained unchanged in 10 (43%) patients and increased in 2 (9%) patients (Table 4). Four of 23 (17%) patients had worse SSS at six weeks than at six months. One of these four patients had a history of prior Q-wave myocardial infarction and developed recurrent infarction with deepening of Q waves in two of the inferior electrocardiographic leads before six weeks of therapy with pravastatin. Asymptomatic progression of the scan defects was noted in the other three patients at six weeks. Six-month scans of these four patients revealed improvement in one patient, deterioration in one patient, and no change in two patients as compared with six-week scans. The resting study was unchanged in 12 patients (52%), improved in 6 (26%) patients, and worse in 4 (17%) patients. The SDS was better in 14 (61%) patients, unchanged in 5 (22%) patients, and worse in 4 (17%) patients (Table 4).
The direction and magnitude of changes in the myocardial perfusion were variable and did not correlate with improvement of the lipid profile. Twenty-one of 23 patients showed improved LDL levels at six months; however, stress perfusion improved (reduced SSS ≥3) in only 9 of these patients. Of two patients in the study who failed to show a significant reduction of LDL cholesterol by six months, stress perfusion improved significantly (reduced SSS ≥3) in each. Responses of SSS to pravastatin therapy with either exercise or adenosine stress were similar.
Other medical therapy and changes of myocardial perfusion
Medication changes between six weeks and six months did not correlate with perfusion changes observed. Of the three patients started on angiotensin-converting enzyme inhibitor therapy, one patient showed improved SSS, one patient worsened, and one patient remained unchanged. The HDL increased at six months, but not at six weeks, potentially because of the addition of niacin after six weeks in five patients. However, SSS improved in two of these patients and remained unchanged in the other three. Thus, niacin-associated improvement in HDL did not account for all the changes in SSS between six weeks and six months.
Our laboratory first reported statin-induced improvements in stress myocardial perfusion by SPECT MPI in 1997 (17)following seminal reports of similar improvements with cholesterol-lowering therapy by PET MPI by Gould et al. (13,18)and Czernin et al. (19). Improvements in radionuclide MPI with lipid-lowering, including lifestyle changes and statins, have subsequently been confirmed independently by both PET (20,24–26)and SPECT MPI (21,23). Assessment of treatment response to statins by cholesterol levels appears inadequate, as up to 7.8% of patients with improved lipids on statin therapy experience coronary events in randomized controlled trials (1–6). Beyond cholesterol-lowering, radionuclide tomographic MPI may identify effective statin treatment response associated with improvement of abnormalities of endothelial function associated with dyslipidemia or CAD (15–26,30,31)as recently reviewed (22).
The current study is the first prospective serial monitoring trial using SPECT MPI to assess both early and late changes in myocardial perfusion and the lipid profile during the first six months of statin therapy. The principal finding of this study is pravastatin reduced abnormalities of stress perfusion (SSS, SDS, quantitative defect extent) by radionuclide SPECT MPI at six months but not six weeks in compliant dyslipidemic patients with CAD. In contrast, pravastatin reduced serum levels of TC, LDL cholesterol, and triglycerides by six weeks, and these reductions persisted at similar levels at six months. Despite a 32% reduction in LDL for the study group, the direction and magnitude of changes in stress myocardial perfusion varied widely. No change of group mean resting perfusion (SRS) was observed during the study, similar to prior SPECT (21)and PET (25)monitoring studies during statin therapy. Differences in MPI technique, study population, time course, degree of resting endothelial dysfunction, definition, or degree of dyslipidemia may account for the lack of alteration of resting perfusion observed in our study. As expected, no changes of left ventricular ejection fraction, end-diastolic volume index, or end-systolic volume index were observed during this six-month study.
The results contribute to our current understanding that: 1) scan findings parallel the time course of earliest documented coronary event reduction in the randomized clinical trials by pravastatin (1,3), rather than the earlier reduction of lipid changes that occurred at six weeks; 2) normalization of stress scans, which is known to be associated with improved prognosis during anti-ischemic therapy (32,33), and an extremely low rate of subsequent coronary events reported for patients with normal scans (8,9,11,12,14,34), was observed in 22% of patients on statin therapy; and 3) changes in the direction and magnitude of serum lipids levels did not closely parallel the improvement of stress perfusion in individual patients.
Our findings confirm a prior PET MPI study showing statin-induced improvement of stress perfusion at six months follows earlier lipid-lowering at two months and appeared unrelated to the amount of lipid-lowering (25). Statin-induced upregulation of endothelial nitric oxide synthase by Rho GTPase is an important mechanism of improved endothelial function and myocardial perfusion regulated downstream from cholesterol-lowering HMG-CoA reductase pathway (35). Additionally, numerous pleiotropic effects of statins appear to contribute to their therapeutic effectiveness (7,36).
Potential study limitations
Medications in addition to pravastatin after the baseline SPECT MPI study might have contributed to the scan changes; however, we did not observe any response trend in patients with altered medication regimen. Only modest absolute changes in myocardial perfusion were noted during this six-month study. However, these findings are consistent with prior reports of disproportionately greater effects of statin therapy on myocardial perfusion and cardiac events than on modest regression of coronary stenoses in randomized arteriographic trials (13). Because of limited sample size, this study may have insufficient power to detect minor or short-term changes in SPECT MPI variables and their correlation with serum lipid levels. Post-hoc analyses indicate a power in excess of 80% for all changes in lipid and scan values (baseline compared to six months) with the exception of HDL (71%). Lack of placebo arm is a potential weakness, but such trial design is considered unethical in the current era of documented effectiveness of statins in patients with CAD (2–4,6). Open-label study design might lead to scan interpretation bias; however, re-analyses of scans without knowledge of their chronologic order (data not shown) and automated quantitative analyses (Figs. 1B and 2B) have confirmed the results. Serial changes of high sensitivity C-reactive protein and other inflammatory markers potentially modulated by statins (36)were not assessed in the current study.
In compliant dyslipidemic patients with abnormal baseline myocardial perfusion by SPECT MPI, pravastatin therapy improved stress perfusion in 48% of the study patients by six months but not by six weeks. Divergent responses to pravastatin of lipid and scintigraphic parameters were observed and variability of responses was substantial. Scan improvement did not correlate with the magnitude of LDL reduction in this small pilot study.
Serial monitoring of stress radionuclide PET and SPECT perfusion abnormalities during pravastatin therapy may identify substantial variability in the presence and magnitude of therapeutic response not reflected by the serum lipid response. Whether patients who fail to improve stress myocardial perfusion despite appropriate lipid-lowering on statin therapy are at increased risk of coronary events is not yet known. The results of the current study support longer term prospective investigation of serial changes in lipids and radionuclide MPI during statin therapy and correlation with clinical outcomes to assess the time course and characteristics of effective therapeutic response. This combined lipid and radionuclide tomographic perfusion-guided approach may improve assessment of the residual risk on statin therapy and facilitate optimal utilization of medical and revascularization therapies.
We gratefully acknowledge the dedicated and skillful care and coordination efforts of Pat Grego, ANP, and Kris Kremer, RN, who made this study possible.
☆ The study was funded by a research grant from Bristol-Myers Squibb Company.
This work was presented in part at the 50th Annual Scientific Session of the American College of Cardiology, March 2001, Orlando, Florida.
- coronary artery disease
- high-density lipoprotein
- low-density lipoprotein
- left ventricular
- myocardial perfusion imaging
- positron emission tomography
- summed difference score
- single photon emission computed tomography
- summed rest score
- summed stress score
- total cholesterol
- Received November 5, 2002.
- Revision received January 30, 2003.
- Accepted March 12, 2003.
- American College of Cardiology Foundation
- Brown K.A.
- Hachamovitch R.,
- Berman D.S.,
- Kiat H.,
- et al.
- Sharir T.,
- Germano G.,
- Kavanagh P.B.,
- et al.
- Shaw L.J.,
- Hachamovitch R.,
- Berman D.S.,
- et al.
- Hachamovitch R.,
- Berman D.S.,
- Shaw L.J.,
- et al.
- Schachinger V.,
- Britten M.B.,
- Zeiher A.M.
- Egashira K.,
- Hirooka Y.,
- Kai H.,
- et al.
- Gould K.L.,
- Martucci J.P.,
- Goldberg D.I.,
- et al.
- Czernin J.,
- Barnard R.J.,
- Sun K.T.,
- et al.
- Huggins G.S.,
- Pasternak R.C.,
- Alpert N.M.,
- et al.
- Mostaza J.M.,
- Gomez M.V.,
- Gallardo F.,
- et al.
- O'Rourke R.A.,
- Chaudhuri T.,
- Shaw L.,
- et al.
- Guethlin M.,
- Kasel A.M.,
- Coppenrath K.,
- Ziegler S.,
- Delius W.,
- Schwaiger M.
- Sdringola S.,
- Nakagawa K.,
- Nakagawa Y.,
- et al.
- Zar J.H.
- Vita J.A.,
- Treasure C.B.,
- Nabel E.G.,
- et al.
- Zeiher A.M.,
- Drexler H.,
- Wollsclaeger H.,
- et al.
- Dakik H.A.,
- Kleiman N.S.,
- Farmer J.A.,
- et al.
- Laufs U.,
- La Fata V.,
- Plutzky J.,
- Liao J.K.
- Ridker P.M.,
- Rifai N.,
- Pfeffer M.A.,
- et al.