- Email: [email protected]
Prognostic value of rest-redistribution tomographic thallium-201 imaging in ischemic cardiomyopathy
Prognostic Value of Rest-Redistribution Tomographic Thallium-20 1 Imaging lschemic Cardiomyopathy
in
Giuseppe Gioia, MD, Joseph Powers, MD, Jaekyeong Heo, MD, and Abdulmassih S. Iskandrian, MD, with the technical assistance of Joseph Russell, CNMT, and David Cassel, BS The relation between the resence of viable myocarditit allium imaging and prognourn by rest-redistribution sis is not well defined. This study examined the prognostic value of rest-redistribution single-photon emission computed tomographic imaging with thallium-201 in patients with coronary artery disease (CAD) and left ventricular (LV) dysfunction. Patients were divided into 2 groups: group 1 patienk (n = 47) were treated medically and group 2 patients (n = 38) underwent corona revascularization. The 2 groups were comparable in it e extent of CAD and in LV ejection fraction. Thallium images showed normal tracer uptake in 1 group 1 and 3 group 2 patients, fixed defects in 26 group 1 and 18 group 2 patients, and both reversible and fixed defeck in 20 group 1 and 17 group 2 patients (p = NS). Based on analysis of 20 segmenk/patient, reversible
defects were seen in 4 +- 4 segmenk/patient in group 1 and 5 2 5 segmenk/patient in group 2 (p = NS). Viable myocardium (defined as normal tracer uptake, reversible defeck, or mild fixed defeck) was seen in 14 f 4 segments/patient in group 1 and 15 + 5 segmenk/ patient in group 2 (p = NS). During a mean follow-up of 31 months, there were 16 roup 1 (34%) and 6 group 2 (16%) deaths. The’annua B mortality rate was 13% in group 1 and 6% in group 2. Actuarial survival analysis showed better ‘survival in group 2 than in group 1 (p = 0.056). Thus, viable myocardium in patients with CAD and LV dysfunction is associated with poor prognosis with medical therapy. Coronary revascularization improves prognosis. (Am J Cardiol 1995;75:759-762)
he determination of presence or absence of viable myocardium in patients with left ventricular (LV) dysfunction and coronary artery disease (CAD) is important. Recent data derived mostly from studies using positron emission tomography suggest that coronary revascularization is associated with improvement in LV function, functional class, and survival in patients with viable myocardium, but not in patients with nonviable myocardium.‘-” Rest-redistribution thallium imaging has been used to detect the presence of viable myocardium.4,s In 1983, WCfirst reported the use of this technique to predict improvement in LV ejection fraction alicr coronary revascularization4 This study examines the prognosis of patients with CAD and LV dysfunction in relation to the results of rest-redistribution thallium imaging and the mode of therapy (medical vs coronary rcvascularization).
of extensive nonviable myocardium. Selection bias in the study patients may explain why most patients had evidence of viable myocardium (see later). Patients were not randomized for coronary revascularization or medical therapy; such a decision was made by each treating physician based on severity of symptoms, response to therapy, suitability of coronary anatomy for coronary revascularization, and patients’ preference of 1 treatment over the other. Also, results of thallium imaging may have altered the choice of therapy. Rest-redistribution thallium-201 imaging: After an overnight fast, a dose of 3 mCi (111 MBq) of thallium-201 was injected intravenously at rest. Initial images were obtained 20 to 30 minutes later and delayed images were repeated at 4 hours. Single-photon emission computed tomo@aphic imaging was obtained over a 180” anterior arc, with acquisition of 32 images using a 64 X 64 matrix at 40 seconds/image. Standard filtered back projection tcchnicluc was used for tomographic reconstruction using methods previously described from our laboratory.” Quantitative analysis of thallium images was performed on the basis of 20 segments/patient. Apical, mid, and basal short-axis tomograms were divided into 6 segments each, and a midventricular vertical long-axis tomogram was used to assess the antero- and inferoapical segments; thus, 20 segments/patient were evaluated. A segment was defined as normal if its regional activity was 280% of the apparent normal region in that image. Segments with 50% to 79% regional activity were clelined as mild to moderate defects. Segments with ~50% activity were defined as severe. Abnormal segments that demonstrated improvement on redistribution scans (partial or complete) were considered reversible defects. A normal segment was given a score of 0, mild defect a
T
METHODS Patient selection: The patients included in this study had CAD as seen with coronary angiography (250% diameter stenosis of 21 artery) and by LV dysfunction (ejection fraction ~50%). These patients presented with stable angina pcctoris or congestive heart failure (or both), but none had recent acute myocardial infarction or unstable angina pectoris. These are not consecutive patients because patients with severe LV dysfunction did not necessarily undergo coronary angiography if the restredistribution thallium imaging suggested the presence Frorr the Ph’lacelgh’a ieart IxtitLtc, P,es.iyterian MKIICUI Cmter. Phlladel hla, Jcmsylvoria. Ad s (es5 far repnts. Ahdulrvawh S skondxn, MD, P~lladel$ia Heart Instl:Jtc, Pwqtervx Medical Center, 5 I North 396 Street, Philadelphia, Pernsylvanla 19 100
CORONARY
ARTFRY CISEAS’/PYOGNOSlIC
‘JAII!t
OF ‘XSI-REDISTKIBUTICN
[I !AlllUM
WAGING
759
TABLE I Demographic
TABLE Ill Thallium Data in Study Patients
Data in Study Group Coronary Revoscularizotion (n = 38)
Men (%) &.vave myocardiol infarction Angina pectoris (%) Congestive heort failure (%) Systemic hypertension (%) Diabetes mellitus (%)
(%)
30 23 26 13 15 13
(79) (61) (68) (34) (39) (34)
Medical Therapy (n = 47) 34 30 26 16 18 16
Coronary Revosculorizotion (n = 38)
(72) (64) (55) (34) (38) (34)
Normol imoges (“A) Abnormal images (%) Fixed defects (%) Reversible defects (* fixed defects) (%) Left ventricular dilation (%) Increased lung thallium uptake (“Y)
None of the differences were statistically significant between the 2 groups.
I TABLE II Catheterization
Pulmonary
artery wedge
pressure
(mm Hgl Meon pulmonary artery pressure (mm Hg) ’ ’ left ventricular ejection fraction (%) Number of coronary arteries narrowed 250% in diameter 1 2 3
Medical Therapy (n = 47)
14 * 8
17*
21 +9
25 * 10
29*
10
0 10 28
(‘3) (92) (47) (45)
25 (66) 20 (53)
1 121 46 (98) 26 (55) 20 (43) 33 (70) 20 (43)
Coronary revascukwization: Surgical coronary revascularization was performed using standard cardiopulmonary bypass techniques with hypothermic arrest, aortic cross-clamping, cold potassium cardioplegia, and topical cold saline solution. The left internal mammary artery was used in 25 patients, and there were 3.1 f 1.6 grafts/patient. Follow-up information was obtained by telephone interview with the patient, family member, or patient’s primary physician.” Follow-up was complete in all patients. Only cardiac deaths were considered as events in this study. Sfatisticai analysis: Results are expressed as mean f SD. Statistical analysis was performed using the Student’s t test and chi-square test. A probability value ~0.05 was considered statistically significant. Differences in survival between the medical and revascularization groups were assessed using life-table survival analysis as previously described.6
Doto in the Study Group Coronory Revascularizotion (n = 38)
3 35 18 17
Medico1 Theropy (n = 47)
8
31 * 10
7 16 24
Values ore expressed us mean * SD
score of 1, and severe defect a score of 2. The total score was the sum of total segmental scores; the higher the score, the worse the thallium images. Lung thallium uptake was assessed visually. Gated blood pool imoging: In 26 patients, LV cjection fraction was measured before and after coronary revascularization using gated blood pool imaging. All studies were performed by 1 experienced observer, with a variability of <5’% on repeated measurements.
RESULTS The demographic data in the study group are listed in Table I. The medical and surgical groups were comparable in age (66 f 10 vs 63 + 14 years), other demographic data, and history of systemic hypertension and diabetes mellitus.
P:NS I l15
I
1 T
-iI
FIGURE 1. The extent of reversible and fixed defects in the medical (M) and surgical (S) 9”PS.
M S Reversible
760
M Mild
S fixed
THt AMEKICAN JOURNAI OF CARDIOLXY
M Severe
VOL. 75
S fixed
APRIL I .5. i Wl,
The cardiac catheterization data were also comparable in the 2 groups (Table II). Results of rest-redistribution thallium imaging are listed in Table III. The 2 groups were comparable in the perfusion pattern. The thallium score was 40 f 9 in the mcdical group and 40 f 11 in the surgical group (p = NS). The extent of the perfusion abnormality is shown in Figure 1. The number of reversible, mild fixed, and severe fixed defects was comparable in the medical and surgical groups. Using a broad definition of viable myocardium as normal tracer uptake, reversible defect, or mild fixed defect, there were 14 + 4 segments/patient in the medical group and 15 f 5 segments/patient in the surgical group (p = NS). Prognosis: More cardiac deaths occurred in the medical group than in the surgical group during a mean follow-up of 31 months (range I to 8). Thus, 6 of 38 surgically treated patients (16%) died versus 16 of 47 medically treated patients (34%). The annual mortality rate was 6% in the surgical and 13% in the medical group. Actuarial survival analysis showed better eventfree survival in the surgical than the medical group (p = 0.056) (Figure 2). Of the events in the surgical group, 1 was due to perioperative death (operative mortality 2.6%). In the medical group, 9 deaths occurred in 20 patients with redistribution defects (45%) compared with 7 deaths in 27 patients with no redistribution (26%). The difference did not reach a statistically significant level, probably because of the fewer patients. Changes in left ventricular revaxularization: In
ejection
frock
after
coro-
26 patients, ejection fraction was remeasured after coronary revascularization. There was an improvement, from 29 f 10% before to 35 f 11% after operation (p = 0.0001). Ejection fraction was measured 14 f 19 days after coronary rcvascularization. However, early measurement may not be a reliable indicator of the degree of improvement. nary
with medical therapy. The previous 2 studies were performed with positron emission tomography using N-13 ammonia as a flow tracer and F-18-fluorodeoxyglucose as a metabolic marker of viability, whereas our study was performed with single-photon emission computed tomographic thallium imaging. Definition of Gob/e myocardium: Rahimtoola’ first used the word “hibernation” to describe myocardial dysfunction due to reduction in resting myocardial blood flow. In any given segment of the myocardium, the extent of normal myocardium, scar, hibernation, and stunning may determine whether an improvement after coronary revascularization will be realized. For global improvement in LV ejection fraction to occur, sufficient hibernating (or stunned) myocardium should be present. Incomplete revascularization, postoperative graft closure, or perioperative injury may explain some of the discordance between preoperative viability assessment and postoperative improvement in LV function. Assessment of myocardial viabiliv Various methods for assessing myocardial viability have been extensively reviewed.*- I* The method used in this study is rest-redistribution thallium imaging and is based on the principle that the initial tracer uptake is proportional to regional flow, whereas the 4-hour delayed images reflect myocardial viability. The presence of initial defects that show redistribution suggests the presence of myocardial hibemation. Previous studies have suggested that thallium-201 uptake is not affected by the presence of myocardial stunning’9,20; therefore, regions with initially normal uptake or only mild reduction in uptake (mild fixed defects) may represent either normal myocardium (with or without scar, respectively), or myocardial stunning (the differen-
DISCUSSION The important conclusion from this study is that patients with CAD and LV dysfunction due to viable myocardium are at high risk of having cardiac events with medical therapy (13%/year for a mean of 2!4 years of follow-up). The prognosis in a comparable group of patients was better with coronary rcvascularijlation (mortality 6%/year). In the medical group, the prognosis was worse in those with redistribution defects. These results are applicable to patients in whom there is evidence of a substantial amount of viable myocardium based on the results of rest-redistribution thallium imaging. On average, 70% of the myocardium was considered viable based on initial tracer uptake or the presence of redistribution in the 4-hour delayed images. The prognosis of patients with predominantly nonviable myocardium cannot be addressed from this study. Based on results of other studies, the prognosis of patients with LV dysfunction and nonviable myocardium is worse than that of patients in this study. 2~3Our results, therefore, are in agreement with those of Eitzman,2 Maddahi,” and their co-workers, showing improvement in survival in patients with viable myocardium with coronary revasculari7ation compared
CORONARY
Mantel-Cm
= 3
P = 0.056 7 0
6
12
16
Follow-up
24
30
36
42
46
(months)
FIGURE 2. Actuarial life-table analysis showing event-free survival in the medical and surgical groups. The prognosis was better in the surgical group.
ARIERY DISEASE/PROGNOSTIC
VAlUt OF RFSr REDISTRIBUTION IHAILIUM IVAGING
761
tiation is made by the presence or abscncc of regional wall motion abnormality). Study limitations: A limitation is bias in patient selection (i.e., these patients had predominantly viable myocardium). The treatment (medical or surgical) was not randomized and there could easily be a hidden bias in the surgical group, such as the suitability of coronary anatomy for revascularization. Neverthclcss, this criterion is a prerequisite in patient management and in the decision-making process for coronary revascularization. We did not have independent methods to assess myocardial viability; however, the intent of this study was not to compare different diagnostic modalities since these have been extensively reviewed and published. Clinical implications: The mortality rate is high in patients with CAD and LV dysfunction (ischemic cardiomyopathy) who have evidence of viable myocardium by rest-redistribution thallium imaging and are treated with medical therapy. If coronary anatomy is suitable for revascularization, these patients should undergo coronary artery bypass grafting.
1. Tillisch JH, Brunken K, Marshall R, Schwalger M, Mandclkom hl. Phclps M, Cheltxrt HR. Reversibility of cardiac wall motion ahnonnslity predicted hy positron tomography. N En,q/ J Med 1986;314:884 888. 2. Eirzman D, Al-Aouar Z. Kanter III., Vnm Dahl J. Kirsh U, Deeb GM. Schwuger .M. Clinical outcome of patients with advanced coronq ancry disease afier viability studies with positron cmision tomography. J Am Co// (brrliol 1992;2()%9-S6.5. 3. Maddahi J, Schclbcrt II. Hrunken R. DiCarh .M. Role of thallium-201 and PET imaging in evaluation nf myocardial viability and mmagcment of patients with core nary artery disease and left vcntriculnr dysfunction. J SIK/ Mrd 1994:35:707-71.5. 4. Irkandrian .4S. Hakki A. Kane SA. Gncl If’. .Mundth 1% Segal BL, Amenta A. Rest and redisttibution thallium-201 myncardial scintigraphy to predict irnprovemerit in left ventricular function after coronary anery bypa\\ Xrafting. ,\nl J Co).dial lYX3:51:1312-1316. 5. Ragosta M. Belier GA, Watson DD, Kaul S. Gimple LW. Quantitative planar rest-redistribution TI-201 imaging in detccrion of myncxdial viability and praliction of unprovement in left ventricular function after cnrnnary artery hypas\ surgev in patients with severely depressed left ventricular fun&m. Cir-rrrlu/ion 1993:87: 163(bl641.
6. Iskandrian /\S. Chae SC. I leo J. S~anbcrty CD, Wasserlcbcn V, Cave V. Indepcndcnt and incremental prnpnostic value of exercise single-photon cm&ion computed tomographic (SPECT) thallium imaging in coronary ancry direax. J Am (loll Cur-dm/ lYY3;22:665-670. 7. Rahimtool;~ SI 1.Coronary hypiw surgery fnr chronic angina. A perspective. Circulnrron 19X2:65:22.5 241. 0. Schclbcn IIR, Buxton 1~.Insights mm coronary awry diseare gained from metabolic imagmg. Cwrxlurton I9XX;7X:49&SOS. 9. IkSilva R. Yamamow Y. Rhodes CC;. Iida M. Nihoyannopoulos I’, Davies GJ, Lan~mertsma AA. Jones ‘I‘, Mwxi A. Prwpcrativc prediction of the outcome of coruna~ reva\culanation wing pootitron emission tomography. Circrtlarion 1992: 86: 1738.1742. 10. Marshall KC. Ttllisch JII. Phelps \fF., Huang SC, Canon R, Hewe F.. Schelbat HR. Identification and differentiation of rcrdnp myocardial ischemia and infarction in man with positron ermwon tomography F-IX labeled fluorodeoxyglucose and S- I3 ammonia. Circrrlorron 19X3;67:7&77X. 1 I. Dllsirian V. Rncco Tt’, Freedman UM. Leon MB, Bonow RO. Enhanced detecnon of ischemic hut Gable myocardium hy the reinjection of thalhum after stressredictribution imaging. N I.‘n,ql J M& 19903323:141-146. 12 Tam&i N, Ohtani II. Yamaqhita K. Magate Y. Yonekurc Y. Nohare R, Kumham H. Kawai C. Hirata K. Ban T. Konishi J. Mletabolic activit) in the arca of new lill-in after thallium-201 reinjection. Comparison with positron cm&inn tomography using fluorine- IX-denxyglucou. J h’rrcl Vcjd 1991:32:673-678. 13. Dilsirian V. Bonow RO. Current diagnostic techniques of awing myocardial viability in patierw wth hibernating and stunned myocardium. Circulafion 1993:87:1-20. 14. Groplcr RJ. Geltman EM, Samp~~thkumaranK. Perez JE, Schechtman KB. Convenmo A, Sokl HE. Hcrgmann SK, Siegel BA. Comparison of carhx-I I-acetate with fluorine- IX-flllorodenxyglucose fordelincatin:! viahlc myocardium by positron cm,s~,on tomography. J Am Co// Cardrol 1993~22: 1.587-1597. 15. Bonow RO, Dilsiuan V, Cuocolo A. Bacharach SL. Identification of viable myncardium in parientc with coronary artery diwse and left ventricular dyrfunclion: compariwn of thallium scintigraphy with reirtjection and PET imaging with IX-l;-fluorodeoxyglucl,\e. Cir-wlrrrion 1991;83:2&37. 16. Pierard I .A, De I .anrheiru CM. Henhc C, Rigo P. Kulbcnus NA. Identification of viable myocardium by echncardiography during dohutnmine infusmn in pahems with myocardial mfarction after thrombolytic therapy: comparisnn with positronem,ssmn tomography. J Am Co// Cardhl 1990:15: lO2l- 1031. 17. l+shbein JC. Lando V. Kanmatnse K, Ninomlya K, Mecrbaum S. FIshban MC, Swan HJC. Ganz W. Divergent effects of inntropic stimulation nn the ischcmic and severely depressed reperfured myoamlium. Crw~lotion 1982:66:397400. 18. Schclbcrt HR. Metabolic imaging 1n BI~SF myocardial viability. J NW/ MerJ 1994:3.5(suppl):x5-14.5. 19. Moore CA, Cannon J, Watwn DD, Kaul S. Bcllcr GA. ‘Thallium-201 kmetics in stunned m)ocardium characteri/sd hy severe post-ischemic systolic dycfuncrion. Cirdorion I ‘MH);XI : 1622- 1632. 20. Maddahi J, Cianj W. Ninomiya K, Havhida J. Fichbein .MC. Moudkar A, Buchbinder .U, Shah PK. Swan HJC, Berman DS. Myocardial salvage by intracoronary thrombolysis in evolvmg mywxrdial infarction: evaluation using intracornnary inJection of thallium-201. Anr llearr J 19x1: 102:6-74.
762
At’R 1 15, 1995
TtiE /WERlCAN
JOURUAI. 3- CARDIOLOGY”
VOL. /5
in
Giuseppe Gioia, MD, Joseph Powers, MD, Jaekyeong Heo, MD, and Abdulmassih S. Iskandrian, MD, with the technical assistance of Joseph Russell, CNMT, and David Cassel, BS The relation between the resence of viable myocarditit allium imaging and prognourn by rest-redistribution sis is not well defined. This study examined the prognostic value of rest-redistribution single-photon emission computed tomographic imaging with thallium-201 in patients with coronary artery disease (CAD) and left ventricular (LV) dysfunction. Patients were divided into 2 groups: group 1 patienk (n = 47) were treated medically and group 2 patients (n = 38) underwent corona revascularization. The 2 groups were comparable in it e extent of CAD and in LV ejection fraction. Thallium images showed normal tracer uptake in 1 group 1 and 3 group 2 patients, fixed defects in 26 group 1 and 18 group 2 patients, and both reversible and fixed defeck in 20 group 1 and 17 group 2 patients (p = NS). Based on analysis of 20 segmenk/patient, reversible
defects were seen in 4 +- 4 segmenk/patient in group 1 and 5 2 5 segmenk/patient in group 2 (p = NS). Viable myocardium (defined as normal tracer uptake, reversible defeck, or mild fixed defeck) was seen in 14 f 4 segments/patient in group 1 and 15 + 5 segmenk/ patient in group 2 (p = NS). During a mean follow-up of 31 months, there were 16 roup 1 (34%) and 6 group 2 (16%) deaths. The’annua B mortality rate was 13% in group 1 and 6% in group 2. Actuarial survival analysis showed better ‘survival in group 2 than in group 1 (p = 0.056). Thus, viable myocardium in patients with CAD and LV dysfunction is associated with poor prognosis with medical therapy. Coronary revascularization improves prognosis. (Am J Cardiol 1995;75:759-762)
he determination of presence or absence of viable myocardium in patients with left ventricular (LV) dysfunction and coronary artery disease (CAD) is important. Recent data derived mostly from studies using positron emission tomography suggest that coronary revascularization is associated with improvement in LV function, functional class, and survival in patients with viable myocardium, but not in patients with nonviable myocardium.‘-” Rest-redistribution thallium imaging has been used to detect the presence of viable myocardium.4,s In 1983, WCfirst reported the use of this technique to predict improvement in LV ejection fraction alicr coronary revascularization4 This study examines the prognosis of patients with CAD and LV dysfunction in relation to the results of rest-redistribution thallium imaging and the mode of therapy (medical vs coronary rcvascularization).
of extensive nonviable myocardium. Selection bias in the study patients may explain why most patients had evidence of viable myocardium (see later). Patients were not randomized for coronary revascularization or medical therapy; such a decision was made by each treating physician based on severity of symptoms, response to therapy, suitability of coronary anatomy for coronary revascularization, and patients’ preference of 1 treatment over the other. Also, results of thallium imaging may have altered the choice of therapy. Rest-redistribution thallium-201 imaging: After an overnight fast, a dose of 3 mCi (111 MBq) of thallium-201 was injected intravenously at rest. Initial images were obtained 20 to 30 minutes later and delayed images were repeated at 4 hours. Single-photon emission computed tomo@aphic imaging was obtained over a 180” anterior arc, with acquisition of 32 images using a 64 X 64 matrix at 40 seconds/image. Standard filtered back projection tcchnicluc was used for tomographic reconstruction using methods previously described from our laboratory.” Quantitative analysis of thallium images was performed on the basis of 20 segments/patient. Apical, mid, and basal short-axis tomograms were divided into 6 segments each, and a midventricular vertical long-axis tomogram was used to assess the antero- and inferoapical segments; thus, 20 segments/patient were evaluated. A segment was defined as normal if its regional activity was 280% of the apparent normal region in that image. Segments with 50% to 79% regional activity were clelined as mild to moderate defects. Segments with ~50% activity were defined as severe. Abnormal segments that demonstrated improvement on redistribution scans (partial or complete) were considered reversible defects. A normal segment was given a score of 0, mild defect a
T
METHODS Patient selection: The patients included in this study had CAD as seen with coronary angiography (250% diameter stenosis of 21 artery) and by LV dysfunction (ejection fraction ~50%). These patients presented with stable angina pcctoris or congestive heart failure (or both), but none had recent acute myocardial infarction or unstable angina pectoris. These are not consecutive patients because patients with severe LV dysfunction did not necessarily undergo coronary angiography if the restredistribution thallium imaging suggested the presence Frorr the Ph’lacelgh’a ieart IxtitLtc, P,es.iyterian MKIICUI Cmter. Phlladel hla, Jcmsylvoria. Ad s (es5 far repnts. Ahdulrvawh S skondxn, MD, P~lladel$ia Heart Instl:Jtc, Pwqtervx Medical Center, 5 I North 396 Street, Philadelphia, Pernsylvanla 19 100
CORONARY
ARTFRY CISEAS’/PYOGNOSlIC
‘JAII!t
OF ‘XSI-REDISTKIBUTICN
[I !AlllUM
WAGING
759
TABLE I Demographic
TABLE Ill Thallium Data in Study Patients
Data in Study Group Coronary Revoscularizotion (n = 38)
Men (%) &.vave myocardiol infarction Angina pectoris (%) Congestive heort failure (%) Systemic hypertension (%) Diabetes mellitus (%)
(%)
30 23 26 13 15 13
(79) (61) (68) (34) (39) (34)
Medical Therapy (n = 47) 34 30 26 16 18 16
Coronary Revosculorizotion (n = 38)
(72) (64) (55) (34) (38) (34)
Normol imoges (“A) Abnormal images (%) Fixed defects (%) Reversible defects (* fixed defects) (%) Left ventricular dilation (%) Increased lung thallium uptake (“Y)
None of the differences were statistically significant between the 2 groups.
I TABLE II Catheterization
Pulmonary
artery wedge
pressure
(mm Hgl Meon pulmonary artery pressure (mm Hg) ’ ’ left ventricular ejection fraction (%) Number of coronary arteries narrowed 250% in diameter 1 2 3
Medical Therapy (n = 47)
14 * 8
17*
21 +9
25 * 10
29*
10
0 10 28
(‘3) (92) (47) (45)
25 (66) 20 (53)
1 121 46 (98) 26 (55) 20 (43) 33 (70) 20 (43)
Coronary revascukwization: Surgical coronary revascularization was performed using standard cardiopulmonary bypass techniques with hypothermic arrest, aortic cross-clamping, cold potassium cardioplegia, and topical cold saline solution. The left internal mammary artery was used in 25 patients, and there were 3.1 f 1.6 grafts/patient. Follow-up information was obtained by telephone interview with the patient, family member, or patient’s primary physician.” Follow-up was complete in all patients. Only cardiac deaths were considered as events in this study. Sfatisticai analysis: Results are expressed as mean f SD. Statistical analysis was performed using the Student’s t test and chi-square test. A probability value ~0.05 was considered statistically significant. Differences in survival between the medical and revascularization groups were assessed using life-table survival analysis as previously described.6
Doto in the Study Group Coronory Revascularizotion (n = 38)
3 35 18 17
Medico1 Theropy (n = 47)
8
31 * 10
7 16 24
Values ore expressed us mean * SD
score of 1, and severe defect a score of 2. The total score was the sum of total segmental scores; the higher the score, the worse the thallium images. Lung thallium uptake was assessed visually. Gated blood pool imoging: In 26 patients, LV cjection fraction was measured before and after coronary revascularization using gated blood pool imaging. All studies were performed by 1 experienced observer, with a variability of <5’% on repeated measurements.
RESULTS The demographic data in the study group are listed in Table I. The medical and surgical groups were comparable in age (66 f 10 vs 63 + 14 years), other demographic data, and history of systemic hypertension and diabetes mellitus.
P:NS I l15
I
1 T
-iI
FIGURE 1. The extent of reversible and fixed defects in the medical (M) and surgical (S) 9”PS.
M S Reversible
760
M Mild
S fixed
THt AMEKICAN JOURNAI OF CARDIOLXY
M Severe
VOL. 75
S fixed
APRIL I .5. i Wl,
The cardiac catheterization data were also comparable in the 2 groups (Table II). Results of rest-redistribution thallium imaging are listed in Table III. The 2 groups were comparable in the perfusion pattern. The thallium score was 40 f 9 in the mcdical group and 40 f 11 in the surgical group (p = NS). The extent of the perfusion abnormality is shown in Figure 1. The number of reversible, mild fixed, and severe fixed defects was comparable in the medical and surgical groups. Using a broad definition of viable myocardium as normal tracer uptake, reversible defect, or mild fixed defect, there were 14 + 4 segments/patient in the medical group and 15 f 5 segments/patient in the surgical group (p = NS). Prognosis: More cardiac deaths occurred in the medical group than in the surgical group during a mean follow-up of 31 months (range I to 8). Thus, 6 of 38 surgically treated patients (16%) died versus 16 of 47 medically treated patients (34%). The annual mortality rate was 6% in the surgical and 13% in the medical group. Actuarial survival analysis showed better eventfree survival in the surgical than the medical group (p = 0.056) (Figure 2). Of the events in the surgical group, 1 was due to perioperative death (operative mortality 2.6%). In the medical group, 9 deaths occurred in 20 patients with redistribution defects (45%) compared with 7 deaths in 27 patients with no redistribution (26%). The difference did not reach a statistically significant level, probably because of the fewer patients. Changes in left ventricular revaxularization: In
ejection
frock
after
coro-
26 patients, ejection fraction was remeasured after coronary revascularization. There was an improvement, from 29 f 10% before to 35 f 11% after operation (p = 0.0001). Ejection fraction was measured 14 f 19 days after coronary rcvascularization. However, early measurement may not be a reliable indicator of the degree of improvement. nary
with medical therapy. The previous 2 studies were performed with positron emission tomography using N-13 ammonia as a flow tracer and F-18-fluorodeoxyglucose as a metabolic marker of viability, whereas our study was performed with single-photon emission computed tomographic thallium imaging. Definition of Gob/e myocardium: Rahimtoola’ first used the word “hibernation” to describe myocardial dysfunction due to reduction in resting myocardial blood flow. In any given segment of the myocardium, the extent of normal myocardium, scar, hibernation, and stunning may determine whether an improvement after coronary revascularization will be realized. For global improvement in LV ejection fraction to occur, sufficient hibernating (or stunned) myocardium should be present. Incomplete revascularization, postoperative graft closure, or perioperative injury may explain some of the discordance between preoperative viability assessment and postoperative improvement in LV function. Assessment of myocardial viabiliv Various methods for assessing myocardial viability have been extensively reviewed.*- I* The method used in this study is rest-redistribution thallium imaging and is based on the principle that the initial tracer uptake is proportional to regional flow, whereas the 4-hour delayed images reflect myocardial viability. The presence of initial defects that show redistribution suggests the presence of myocardial hibemation. Previous studies have suggested that thallium-201 uptake is not affected by the presence of myocardial stunning’9,20; therefore, regions with initially normal uptake or only mild reduction in uptake (mild fixed defects) may represent either normal myocardium (with or without scar, respectively), or myocardial stunning (the differen-
DISCUSSION The important conclusion from this study is that patients with CAD and LV dysfunction due to viable myocardium are at high risk of having cardiac events with medical therapy (13%/year for a mean of 2!4 years of follow-up). The prognosis in a comparable group of patients was better with coronary rcvascularijlation (mortality 6%/year). In the medical group, the prognosis was worse in those with redistribution defects. These results are applicable to patients in whom there is evidence of a substantial amount of viable myocardium based on the results of rest-redistribution thallium imaging. On average, 70% of the myocardium was considered viable based on initial tracer uptake or the presence of redistribution in the 4-hour delayed images. The prognosis of patients with predominantly nonviable myocardium cannot be addressed from this study. Based on results of other studies, the prognosis of patients with LV dysfunction and nonviable myocardium is worse than that of patients in this study. 2~3Our results, therefore, are in agreement with those of Eitzman,2 Maddahi,” and their co-workers, showing improvement in survival in patients with viable myocardium with coronary revasculari7ation compared
CORONARY
Mantel-Cm
= 3
P = 0.056 7 0
6
12
16
Follow-up
24
30
36
42
46
(months)
FIGURE 2. Actuarial life-table analysis showing event-free survival in the medical and surgical groups. The prognosis was better in the surgical group.
ARIERY DISEASE/PROGNOSTIC
VAlUt OF RFSr REDISTRIBUTION IHAILIUM IVAGING
761
tiation is made by the presence or abscncc of regional wall motion abnormality). Study limitations: A limitation is bias in patient selection (i.e., these patients had predominantly viable myocardium). The treatment (medical or surgical) was not randomized and there could easily be a hidden bias in the surgical group, such as the suitability of coronary anatomy for revascularization. Neverthclcss, this criterion is a prerequisite in patient management and in the decision-making process for coronary revascularization. We did not have independent methods to assess myocardial viability; however, the intent of this study was not to compare different diagnostic modalities since these have been extensively reviewed and published. Clinical implications: The mortality rate is high in patients with CAD and LV dysfunction (ischemic cardiomyopathy) who have evidence of viable myocardium by rest-redistribution thallium imaging and are treated with medical therapy. If coronary anatomy is suitable for revascularization, these patients should undergo coronary artery bypass grafting.
1. Tillisch JH, Brunken K, Marshall R, Schwalger M, Mandclkom hl. Phclps M, Cheltxrt HR. Reversibility of cardiac wall motion ahnonnslity predicted hy positron tomography. N En,q/ J Med 1986;314:884 888. 2. Eirzman D, Al-Aouar Z. Kanter III., Vnm Dahl J. Kirsh U, Deeb GM. Schwuger .M. Clinical outcome of patients with advanced coronq ancry disease afier viability studies with positron cmision tomography. J Am Co// (brrliol 1992;2()%9-S6.5. 3. Maddahi J, Schclbcrt II. Hrunken R. DiCarh .M. Role of thallium-201 and PET imaging in evaluation nf myocardial viability and mmagcment of patients with core nary artery disease and left vcntriculnr dysfunction. J SIK/ Mrd 1994:35:707-71.5. 4. Irkandrian .4S. Hakki A. Kane SA. Gncl If’. .Mundth 1% Segal BL, Amenta A. Rest and redisttibution thallium-201 myncardial scintigraphy to predict irnprovemerit in left ventricular function after coronary anery bypa\\ Xrafting. ,\nl J Co).dial lYX3:51:1312-1316. 5. Ragosta M. Belier GA, Watson DD, Kaul S. Gimple LW. Quantitative planar rest-redistribution TI-201 imaging in detccrion of myncxdial viability and praliction of unprovement in left ventricular function after cnrnnary artery hypas\ surgev in patients with severely depressed left ventricular fun&m. Cir-rrrlu/ion 1993:87: 163(bl641.
6. Iskandrian /\S. Chae SC. I leo J. S~anbcrty CD, Wasserlcbcn V, Cave V. Indepcndcnt and incremental prnpnostic value of exercise single-photon cm&ion computed tomographic (SPECT) thallium imaging in coronary ancry direax. J Am (loll Cur-dm/ lYY3;22:665-670. 7. Rahimtool;~ SI 1.Coronary hypiw surgery fnr chronic angina. A perspective. Circulnrron 19X2:65:22.5 241. 0. Schclbcn IIR, Buxton 1~.Insights mm coronary awry diseare gained from metabolic imagmg. Cwrxlurton I9XX;7X:49&SOS. 9. IkSilva R. Yamamow Y. Rhodes CC;. Iida M. Nihoyannopoulos I’, Davies GJ, Lan~mertsma AA. Jones ‘I‘, Mwxi A. Prwpcrativc prediction of the outcome of coruna~ reva\culanation wing pootitron emission tomography. Circrtlarion 1992: 86: 1738.1742. 10. Marshall KC. Ttllisch JII. Phelps \fF., Huang SC, Canon R, Hewe F.. Schelbat HR. Identification and differentiation of rcrdnp myocardial ischemia and infarction in man with positron ermwon tomography F-IX labeled fluorodeoxyglucose and S- I3 ammonia. Circrrlorron 19X3;67:7&77X. 1 I. Dllsirian V. Rncco Tt’, Freedman UM. Leon MB, Bonow RO. Enhanced detecnon of ischemic hut Gable myocardium hy the reinjection of thalhum after stressredictribution imaging. N I.‘n,ql J M& 19903323:141-146. 12 Tam&i N, Ohtani II. Yamaqhita K. Magate Y. Yonekurc Y. Nohare R, Kumham H. Kawai C. Hirata K. Ban T. Konishi J. Mletabolic activit) in the arca of new lill-in after thallium-201 reinjection. Comparison with positron cm&inn tomography using fluorine- IX-denxyglucou. J h’rrcl Vcjd 1991:32:673-678. 13. Dilsirian V. Bonow RO. Current diagnostic techniques of awing myocardial viability in patierw wth hibernating and stunned myocardium. Circulafion 1993:87:1-20. 14. Groplcr RJ. Geltman EM, Samp~~thkumaranK. Perez JE, Schechtman KB. Convenmo A, Sokl HE. Hcrgmann SK, Siegel BA. Comparison of carhx-I I-acetate with fluorine- IX-flllorodenxyglucose fordelincatin:! viahlc myocardium by positron cm,s~,on tomography. J Am Co// Cardrol 1993~22: 1.587-1597. 15. Bonow RO, Dilsiuan V, Cuocolo A. Bacharach SL. Identification of viable myncardium in parientc with coronary artery diwse and left ventricular dyrfunclion: compariwn of thallium scintigraphy with reirtjection and PET imaging with IX-l;-fluorodeoxyglucl,\e. Cir-wlrrrion 1991;83:2&37. 16. Pierard I .A, De I .anrheiru CM. Henhc C, Rigo P. Kulbcnus NA. Identification of viable myocardium by echncardiography during dohutnmine infusmn in pahems with myocardial mfarction after thrombolytic therapy: comparisnn with positronem,ssmn tomography. J Am Co// Cardhl 1990:15: lO2l- 1031. 17. l+shbein JC. Lando V. Kanmatnse K, Ninomlya K, Mecrbaum S. FIshban MC, Swan HJC. Ganz W. Divergent effects of inntropic stimulation nn the ischcmic and severely depressed reperfured myoamlium. Crw~lotion 1982:66:397400. 18. Schclbcrt HR. Metabolic imaging 1n BI~SF myocardial viability. J NW/ MerJ 1994:3.5(suppl):x5-14.5. 19. Moore CA, Cannon J, Watwn DD, Kaul S. Bcllcr GA. ‘Thallium-201 kmetics in stunned m)ocardium characteri/sd hy severe post-ischemic systolic dycfuncrion. Cirdorion I ‘MH);XI : 1622- 1632. 20. Maddahi J, Cianj W. Ninomiya K, Havhida J. Fichbein .MC. Moudkar A, Buchbinder .U, Shah PK. Swan HJC, Berman DS. Myocardial salvage by intracoronary thrombolysis in evolvmg mywxrdial infarction: evaluation using intracornnary inJection of thallium-201. Anr llearr J 19x1: 102:6-74.
762
At’R 1 15, 1995
TtiE /WERlCAN
JOURUAI. 3- CARDIOLOGY”
VOL. /5