10.40 Comparison of gated spect imaging and 2-D echocardiography in ischemic dilated cardiomyopathy. M. Mariana Vasconcelos 1 , E. Martins 1 , T. Faria 2 , A. Oliveira 2 , N. Pardal 1 , F. Macedo 1 , J. Pereira 2 , F. Rocha-Gonc ¸alves. 11 Sa ˜o Joa ˜o Hospital, Cardiology, Porto, Portugal, 2 Sa ˜o Joa ˜o Hospital, Nuclear Medicine, Porto, Portugal Background: In coronary artery disease (CAD), gated SPECT (GSPECT) is widely used for the assessment of left ventricular ejection fraction (LVEF) and regional wall function (motion and thickening). However a number of factors may affect the reproducibility of these measurements, namely ventricular enlargement. We analyse the information obtained from GSPECT in patients with ischemic dilated cardiomyopathy (ICD) and correlate it to 2-D echocardiography, a technique used in daily practice to evaluate left ventricular function. Methods: 19 patients (16 male), mean age 69.3 years, admitted in our hospital department for acute coronary syndromes, were studied. Patients had extensive CAD on catheterization (47% with three-vessel disease). 2-D echocardiography was performed 5.5  3.3 days from GSPECT. All patients underwent one-day stress-adenosine Tc-99 tetrofosmin GSPECT. Summed stress (SSS) and summed difference scores (SDS) were determined using the 17 segment (0-4 perfusion scale) model. LVEF, wall motion (WM) and wall thickening (WT) scores were calculated from post-exercise gated images. We examined the correlation between LVEF, WM and WT scores derived from GSPECT and LVEF and WM score obtained from 2-D echocardiography. Results: GSPECT revealed reduced ejection fraction (32.8  8.2%) and ventricular enlargement (EDV 170  50 ml; ESV 116  40 ml) in all patients. An abnormal perfusion study (SSS  4) was observed in all cases, with reversible defects (SDS  4) in 10 (53%) of patients. 2-D ecocardiography estimated ejection fraction was higher (44.1  11.4%), even in patients with no ischemia (fixed defects) in GSPECT. We found no significant correlation between LVEF estimations from the two techniques (r=0.47). In WM score analysis significant correlation was present only in the lateral segments (r=o.73; p0.01). Similar results were observed comparing WT score obtained from GSPECT to WM score in 2-D echocardiography (r=0.87; p0.01). Conclusion: In our sample of patients with IDC, estimations of LVEF and WM score analysis by GSPECT and 2-D echocardiography were not significantly correlated. These data are important in clinical practice and must be confirmed in future studies. 10.41 Gated SPECT-methods variability for ejection fraction determination. G. Guilhermina Cantinho 1 , H. Pena 1 , L. Freire 2 , A. Veiga 1 , P. Gonc ¸alves 2 , F. Godinho. 1 1 Atomedical & Inst. Med. Nuc.-FML, Lisboa, Portugal, 2 Atomedical, Lisboa, Portugal Aim: Gated SPECT is recognized as a very important additional value in the information of myocardial perfusion studies. In the last years, several packages to calculate global LVEF have been available. Our aim was to compare 3 different methods: QGS(Germano), S(Smith) and 4D-MSPECT(4D)(Hamilton). Methods: The images were acquired in two different GE cameras (MG and VG). 46 patients were studied: MG (24): 18 normal; VG (22): 19 normal. Processing was done in a GE Xeleris computer, using the 3 methods. EDV and ESV were also determined by QGS and 4D. According to QGS, two groups of patients were extracted: Normal(Np)(LVEF =50%) and dilated(Dp)(LVEF  50%, EDV 130ml). Considering QGS as a standard, we calculated the differences’ average and the correlation coefficient (CC) for LVEF, EDV and ESV, for the two cameras. Results: - About LVEF: the table shows that there are statistically significant(SS) differences(dif) between LVEF by S and QGS for Np, but not for Dp. Between 4D and QGS, SS dif may be observed for Np(VG) and for Dp(MG). The CC between the dif of LVEF by S-QGS and 4D-QGS were, for Np, 0.40 and 0.53 (MG and VG) and for Dp, 0.20(MG). About volumes: (a) For Np, one has also observed a high CC between EDV by 4D and QGS: 0.97(MG) and 0.98(VG). For ESV, CC’s were still high: 0.92 and 0.95 (MG and VG). However, the dif between volume values, obtained by the two methods, were SS(pMG0.01 and pVG0.05 for EDV and pMG0.1 and pVG0.01 for ESV). (b) For Dp, the dif between EDV by 4D and QGS were not SS. The same happened for ESV. For MG, the CC between EDV by 4D and QGS was 0.68 and 0.98 between ESV. Conclusions: Results indicate several significant dif between the proposed methods. We think that further studies are necessary in order to fully test and validate the methods proposed in the literature, in order to guarantee that software variabilities will not compromise quantitative data that is important for clinical patient management. table Camera Normal Pathological Dilated GE MG (n = 18 ) 12  17 | 0  9---------- (p  0.01 | p = ns ) (n = 6 ) 21  25 | -4  2---------- (p = ns | p = 0.05 ) GE VG (n = 19 ) -5  4 | 5  5----------- (p  0.01 | p  0.01 ) (n = 2 ) -1  10 | 4  1----------- (p = ns | p = ns ) Averages and SD (%) of LVEF differences for (S-QGS) | (4D-QGS) and corresponding p-values. 10.42 Quantitative low dose dobutamine gated perfusion SPECT. T. Terrance Chua, C Y. Lee, F. Keng, Z P. Ding, T H. Koh. National Heart Centre, Cardiology, Singapore, Singapore Background: Quantitative assessment of regional left ventricular function (RF) by gated perfusion tomography (GSPECT) is now feasible using automated software. We hypothesized that the combination of low dose dobutamine with GSPECT would be feasible and potentially improve objectivity of assessment of RF over visual methods. Methods: A regional specific normal range for RF (wall motion-WM and wall thickening-WT) using quantitative GSPECT was developed from 20 patients with normal RF by echocardiography. In 23 patients, GSPECT was performed at rest (2-detector camera, 20s/stop, 30 mCi rest injection of Tc-99m tetrofosmin or sestamibi) and during low dose intravenous infusion of dobutamine (5 and 10 mcg/kg/min). For acquisiton during dobutamine, acquisition was modified to 10s/stop. Automated software (QGS) was applied to the processed data sets after filtered backprojection and re-orientation. Results: In 2 patients, the protocol was terminated after completion of acquisition at 5 mcg/kg/min, due to breathlessness. In the remaining 21 patients, the protocol was successful with no complications. Of the 21 patients, 8 had normal stress-rest perfusion and normal rest WM, while 13 had abnormal rest WM and perfusion studies due to prior infarct. GSPECT WM values were higher at rest in the 8 normal patients compared to the 13 abnormal patients (rest WM 7.22.3mm vs 3.82.7mm, p0.001) and increased with dobutamine infusion to a higher value at 10 mcg/kg/min (WM 10.2+3.2mm in normals vs 4.02.9mm in abnormals, p0.001, mean WM increase with dobutamine 2.9+2.1mm vs 0.2+1.8mm, p0.001). GSPECT WT values were higher at rest in the 8 normal patients compared to the 13 abnormal patients (rest WT 41.920.1% vs 18.413.1%, p0.001) and increased with dobutamine infusion to a higher value at 10 mcg/kg/min (WT 66.2+31.6% vs 20.3+15.0%, p0.001, mean WM increase with dobutamine 24.3+16.2% vs 1.9+8.5%, p0.001). These increases in GSPECT values from rest to peak dobutamine dose were statistically significant using a paired t-test in normal patients for both WM and WT (p0.001), and also in abnormal patients for WT (p0.001) but was of borderline significance for WM in abnormal patients (p = 0.055). Conclusion. Automated quantitative GSPECT in combination with low dose dobut- amine appears feasible and may have the potential to improve objectivity of RF assessment. 10.43 Causes of discrepancy between ejection fraction estimated by echocardiography and the obtained by gated-SPECT. P. Paula Awamleh 1 , P. Talavera 1 , V. Torres 2 , M.A. Balsa 2 , O. Gonza ´lez 1 , M.T. Alberca 1 , R. Miguel 1 , F.G. Cosı ´o. 11 Hospital Universitario de Getafe, Cardiologı ´a, Getafe, Spain, 2 Hospital Universitario de Getafe, Medicina Nuclear, Madrid, Spain Introduction. Left ventricle ejection fraction (EF) is a very important parameter in our patients. There are different techniques to calculate it. Sometimes there is disagreement between EF estimated by TTE and the obtained by Gated SPECT (GS). We sought to analise which factors could be associated with this disagreement. Patients and Methods. We compared restrospectively the EF obtained by GS and by TTE in 268 patiens along one year. We excluded the following cases: 1) More than six months between both tests, 2) inadequate echocardiographic window and 3) significant clinical changes, such us an acute myocardial infarct, between both tests. One hundred and seventy four patients were men, mean age 63,910,8 years; there were 110 treadmill exercise tests and 158 pharmacological stress tests. EF by TTE was classified, subjetively, as normal or depressed (mild, moderate or severe dysfunction). Four similar groups were established depending on the Gated-SPECT EF (50% normal, 40-49% mild dysfunction, 30-39% moderate dysfunction,30% severe dysfunction). We studied, such us theorical causes for disagreement, the following parameters: ancient myocardial infarct, left ventricle enlargement (more than 5.8 cm telediastolic diameter), bundle branch block, sex, type of test (treadmill vs pharmacological stress) and cardiac rhythm (sinus rhythm or atrial fibrillation/flutter). Results. In 211 cases (79%), there was agreement between both explorations. On the other hand, there was disagreement in 57 cases (21%). One hundred and two patients had an ancient myocardial infarct (38%). There were 46 cases with left ventricle enlargement (17.1%), 46 cases with bundle branch block. There was 34.3% of cases with discrepancy between both tests in the myocardial infarct group, whereas patients without infarct had only 13.2% of cases with discrepancy. (p0.0001). Left ventricle enlargement was associ- ated with 34.7% of cases with discrepancy. Patients without left ventricle enlargement had 18.1% of cases with discrepancy. (p0.01). We did not found significant association for the bundle branch block, sex, type of test and presence of atrial fibrillation/flutter. Conclusions. 1)The presence of myocardial infarct or left ventricle enlargement is associated with more discrepancy between both explorations. 2) The bundle branch block was not associated with more discrepancy, at least in this group of patients. 3) It would be necessary more studies, to know the accuracy of this explorations in the established situations. T U E S D A Y M A Y 10 T U E S D A Y M A Y 10 S70 Abstracts Journal of Nuclear Cardiology Tuesday, May 10, 2005 March/April 2005