Independent and Incremental Prognostic Value of 201Tl Lung Uptake at Rest in Patients With Severe Postischemic Left Ventricular Dysfunction
Background—An elevated 201Tl lung uptake after stress is related to an adverse prognosis.
Methods and Results—The functional and prognostic significance of resting 201Tl lung uptake was assessed in 124 consecutive patients with ischemic heart disease and ejection fraction ≤35% undergoing rest-redistribution tomography to evaluate myocardial viability. 201Tl lung uptake significantly correlated with pulmonary wedge pressure (r=0.66; P<0.01) and with a restrictive physiology by Doppler echocardiography (P<0.001). During a 13±13-month follow-up, 13 patients died and 23 patients required hospitalization as the result of worsening heart failure or nonfatal myocardial infarction (cumulative events rate 29%). Patients with events had a significantly higher 201Tl lung/heart ratio (L/H) (P<0.001). A L/H value >0.61 best separated patients with and without events (ROC area under curve 0.82). Event-free survival was significantly lower in patients with L/H >0.61 (P<0.001); L/H >0.61 (χ2=10.8; P<0.001) and a restrictive filling pattern (χ2=3.6; P<0.05) were independent predictors of events. The prognostic value of L/H was incremental over that obtained by clinical, echographic and Doppler data (global χ2=20.8).
Conclusions—In patients with severe postischemic left ventricular dysfunction undergoing rest-redistribution 201Tl imaging, an increased lung tracer uptake showed incremental prognostic value over clinical and other imaging findings, providing clinically useful risk assessment.
Heart failure remains a critical target for therapeutic intervention. The prospective identification of patients with systolic dysfunction at greatest risk is elusive, and residual myocardial viability is assessed to identify who will most benefit from coronary revascularization. Recently, noninvasive markers of diastolic dysfunction have been shown to herald an unfavorable outcome and yield incremental information.
At 201Tl myocardial scintigraphy, an elevated lung uptake of the tracer after stress is a marker of severe coronary artery involvement and adverse prognosis.1 2 3 The increased 201Tl lung uptake has been related to the extent of reversible ischemia and stress-induced increase in left ventricular (LV) end-diastolic pressure.4 5
The clinical meaning of this additional scintigraphic finding at rest 201Tl imaging is less clear. In a small group of patients with acute coronary syndromes, an increased resting lung uptake was associated with a higher rate of in-hospital heart failure and increased short-term mortality.6 The aim of this study was to assess the clinical correlates and the prognostic significance of an elevated lung uptake at rest 201Tl tomoscintigraphy in clinically stable patients with severe LV dysfunction.
We studied 124 consecutive patients with a history of ischemic heart disease and severe LV dysfunction (ejection fraction ≤35% at 2D echocardiography or angiography) who were referred to our nuclear cardiology laboratory from January 1994 to December 1996 for a 201Tl rest-redistribution study, to assess the presence of residual viability in dysfunctioning myocardial segments. All patients were clinically stable, with optimized cardiovascular therapy and in New York Heart Association (NYHA) functional class II or III. The study cohort included 30 (24%) patients with severe heart failure, admitted to our hospital for heart transplantation workup and who, after aggressive medical treatment, improved their functional class, thus being able to tolerate the supine position required for tomographic imaging. Patients with unstable angina, recent (<1 month) myocardial infarction, or overt heart failure were excluded. Recent (<3 months) coronary angiography was available in 86 patients and showed the presence of a >50% coronary diameter reduction of a major coronary artery in 76 patients (88%). Right heart catheterization data by Swan-Ganz catheterization performed within 48 hours of the 201Tl study were also available in 19 patients. The study protocol was approved by the local Ethics Committee for Human Research. Informed consent was obtained from all patients.
201Tl Imaging and Analysis
201Tl 3 mCi (111 MBq) was injected at rest, in fasting condition; cardioactive drugs were not withdrawn. Images were obtained 10 minutes and 4 hours after injection. The methods for imaging acquisition, reconstruction, and analysis have been previously reported.7 For the quantification of regional myocardial tracer uptake, a 20-segment LV model was adopted. A segmental 201Tl uptake <75% of peak myocardial activity was considered abnormal; the tracer uptake defect was classified as severe if <50% and mild-moderate if 50% to 75% of peak myocardial activity. A tracer defect was considered reversible when its abnormal 201Tl uptake on rest images increased >10% on redistribution images.7 The LV 201Tl distribution was also displayed as a polar map and normalized for peak myocardial activity; regions with tracer uptake ≤50% of peak uptake were considered “nonviable” and summed to obtain the total “nonviable” myocardial extent, expressed as a percentage of the LV surface.
The lung 201Tl uptake was assessed both qualitatively and quantitatively from the earlier acquisition. Three selected frames from 6° right anterior to 6° left anterior oblique projections were added to one composite image representing 150 seconds of data acquisition. The lung tracer uptake was qualitatively assessed by 2 experienced investigators unaware of clinical or hemodynamic data; all images were analyzed without background subtraction, after normalization for peak cardiac activity. By comparing the uptake in the lungs with the uptake in the myocardium and mediastinum, pulmonary uptake was scored on a 4-point grading system (0=lungs poorly visualized; 1=mild; 2=moderate and 3=marked lung uptake); a score ≥2 was considered abnormal. The interobserver agreement in grading the lung tracer uptake was 95%; discrepancies were resolved by consensus.
For the quantitative analysis, 2 regions of interest were drawn over the left lung and the area of maximal myocardial activity, and the tracer lung/heart uptake ratio (L/H) was calculated as (mean counts/pixel left lung)/(mean counts/pixel myocardium).8
Complete echocardiographic and Doppler ultrasound evaluation were performed in all patients, as previously reported.9 The left ventricle was divided into 13 segments, and systolic wall thickening and inward wall motion were visually assessed off-line by 2 experienced operators who were unaware of other information. In each segment, contraction was graded on a 4-point scoring system (from 0=normal to 3=dyskinetic) and a wall motion score index was calculated; in cases of discrepancies a consensus was reached. LV volumes were calculated from orthogonal apical views with the biplane area-length method, and the LV ejection fraction was derived. At Doppler examination of mitral diastolic inflow, peak early (E) and late (A) flow velocities, the peak E/A wave velocity ratio and deceleration time of early filling were measured.9 The restrictive filling pattern was defined as an E/A ratio ≥2 or the combination of an E/A ratio between 1 and 2 and a deceleration time ≤140 ms. Mitral regurgitation was evaluated by color Doppler flow imaging and scored as absent, mild, moderate or severe according to Helmcke et al.10
Follow-Up and Statistical Analysis
Follow-up was obtained at regular intervals in our outpatient clinic or through telephone interviews conducted by trained personnel, personal communication with the patient’s physician, and, in the case of new hospitalization, by reviewing the patient’s hospital records. For the purposes of this study, hard events signified cardiac death or nonfatal myocardial infarction; major events included hard events plus rehospitalization for heart failure.
Continuous data are reported as mean±SD. A Student’s t test for unpaired data or ANOVA was used to test differences between groups, with Bonferroni’s correction when indicated, and differences in rates of occurrence of categorical variables were compared by the χ2 test with Yates’ correction. Linear regression analysis was used to assess the correlation between L/H and hemodynamic, echocardiographic variables, and the extent of “nonviable” LV area. Receiver-operating characteristic (ROC) curves for the prediction of major events were generated with the use of individual L/H values; cut-points were generated at regular intervals and the best threshold was automatically identified as the value minimizing the expression [(1−sensitivity)2+(1−specificity)2]. The area under the curve was obtained according to Hanley and McNeil.11 Differences in event rates between patients with L/H above or below the cut-point value were analyzed with Kaplan-Meier survival curves and compared by means of the log-rank test.
A Cox logistic regression analysis was used to identify the baseline variables that were independently correlated to hard or major events among those significantly associated on univariate analysis. Patients who underwent surgery were included in the survival analysis up to the time of the revascularization or transplantation. All tests of significance were 2-tailed, and a P value of <0.05 was considered significant.
To assess the incremental prognostic information from the addition of demographic, clinical, echocardiographic, and scintigraphic variables, data analysis was also performed according to a modified stepwise procedure in which individual factors were included in the model in the same order in which they would be considered in the clinical practice (demographic and clinical data first, followed by echocardiographic variables, then Doppler findings, and finally the 201Tl lung uptake). The demographic and clinical variables considered were age, history of congestive heart failure, rales, and third sound; the echocardiographic and Doppler variables considered were LV volumes, ejection fraction, and the presence of a restrictive filling pattern. Increment in information of the model at each step was considered significant when the log-likelihood difference adjusted for differences in degrees of freedom associated with each model had a P value of <0.05.
The mean age of the study group was 62±12 years (range 33 to 79 years; 83% men) and the average LV ejection fraction was 23±6%; at coronary angiography, single-vessel and multivessel coronary artery diseases were detected in 16 (19%) and 60 (70%) patients, respectively. At 201Tl imaging, 5.1±3.4 myocardial segments had a tracer uptake <50%, and the extent of “nonviable” area at polar map was 28±15%. The average lung uptake score was 1.7±0.8 and the average L/H was 0.64±0.14. The significant relation between the 201Tl lung uptake at quantitative analysis and the corresponding visual score is reported in Figure 1⇓. The clinical, echocardiographic, and scintigraphic findings of the study cohort are reported in Table 1⇓.
201Tl Lung Uptake: Correlation With Functional and Hemodynamic Variables
At qualitative analysis, lung 201Tl uptake was elevated in 74 patients (60%) and normal in 50 (40%); the clinical and instrumental findings of the study cohort according to the presence of an elevated or normal lung 201Tl uptake are reported in Table 2⇓. At multivariate analysis, severe mitral regurgitation (χ2=6.8, P<0.05) and peak A-wave velocity (χ2=4.2, P<0.05) were independently correlated with an increased lung uptake.
A significant inverse relation was documented between L/H and peak flow velocity of A wave (r=−0.53; P=0.0005) (Figure 2⇓) or peak E/A flow ratio (r=−0.50; P=0.0007); no significant correlation was observed between L/H and LV volumes, ejection fraction, wall motion score index, E-wave peak velocity and E-wave deceleration time, or extent of nonviable myocardial area. When L/H and right heart catheterization data were compared, a significant correlation was found only between L/H and wedge pressure (r=0.66; P=0.008) (Figure 2⇓).
Patients were followed up for a period of 13±13 months, and follow-up data were obtained in all patients. There were 29 (23%) patients who underwent revascularization and 11 other patients (9%) who underwent heart transplantation. Among the remaining 85 (68%) medically treated patients, revascularization was not planned in 37 patients because of stenotic coronaries supplying an extensive nonviable LV area, and it was not performed in 14 patients because the target vessel was unsuitable for revascularization. In 34 other patients (27%) judged to be at high risk because of severely depressed LV function (ejection fraction <20% and LV end-diastolic volume index >120 mL/mq), additional invasive procedures were not scheduled. At follow-up, 1 patient had a new nonfatal myocardial infarction, 13 patients (10%) died, and 22 (18%) other patients required hospital admission for worsening congestive heart failure. Overall, 36 patients (29%) had major events: Clinical and instrumental findings in patients with and without unfavorable outcome are summarized in Table 3⇓.
On the basis of ROC analysis, an L/H value >0.61 best separated patients with and without major events at follow-up (area-under-curve=0.82; sensitivity 83%, specificity 74%) (Figure 3⇓). The clinical and instrumental findings of patients with versus without L/H >0.61 are reported in Table 4⇓. In patients with L/H >0.61, 31 (42%) had a major event, compared with only 5 (10%) patients with L/H ≤0.61 (P<0.001; relative risk 3.91, 95% CI 1.69 to 9.1). The Kaplan-Meier analysis showed that the 4 years survival free of major events was significantly lower in patients with L/H >0.61 than in those with L/H ≤0.61 (P<0.001) (Figure 4⇓). Survival without cardiac death (P<0.05) as well as without rehospitalization as the result of worsening heart failure (P<0.005) was also significantly lower in patients with increased tracer lung uptake (Figure 5⇓).
When the variables that significantly correlated with an adverse outcome at follow-up by univariate analysis were introduced in the logistic regression analysis, an L/H value >0.61 (χ2=10.8; P<0.001) and the presence of a restrictive filling pattern (χ2=3.6; P<0.05) were the sole independent predictors of major events.
Incremental Prognostic Value of Diagnostic Procedures
At the interactive stepwise procedure, the power of models developed at various steps in hierarchic order (clinical data; clinical and echocardiographic results; clinical, echocardiographic, and Doppler results; clinical, echocardiographic, Doppler, and 201Tl results) to predict major events is shown in Figure 6⇓. The global χ2 value achieved by the addition of the L/H value to the clinical, echocardiographic, and Doppler variables was significantly higher (global χ2=20.8), indicating that 201Tl lung uptake still added significant prognostic information.
An elevated lung 201Tl uptake on stress images is considered the result of an increased LV filling pressure caused by transient LV dysfunction.3 5 Other proposed mechanisms include increased lung transit time, lung water content, and vascular permeability.3 Thus, an elevated lung 201Tl uptake is a noninvasive marker of stress induced pulmonary congestion. This scintigraphic finding has been documented as a powerful predictor of extensive coronary artery disease and adverse prognosis.1 2
The functional and prognostic significance of an elevated lung 201Tl uptake at rest has been less investigated. A significant correlation between the amount of lung 201Tl uptake and wedge pressure was documented by Martinez et al12 in 16 patients with cardiomyopathy and congestive heart failure. A high lung 201Tl uptake was documented in 36% of patients with acute myocardial infarction by Jain et al6 ; it was associated with more severe LV dysfunction and higher event rates during hospitalization.
The present study is the first assessing the clinical and prognostic significance of resting lung 201Tl uptake in comparison with clinical, echocardiographic, and hemodynamic findings in a selected, large cohort of clinically stable patients with severe LV dysfunction. The resting lung uptake of the tracer significantly correlated with other indexes of increased LV filling pressure. Moreover, an elevated lung uptake proved a powerful predictor of adverse outcome, its prognostic information being incremental over that obtained by clinical, echocardiographic, and Doppler examination.
Lung 201Tl Uptake and Diastolic Function
Doppler analysis of mitral flow is a noninvasive technique widely used to assess LV diastolic function in different pathological settings.9 13 14 A close correlation between Doppler-derived diastolic function and invasively-derived hemodynamic parameters has been reported by several authors.13 14 15 Finkelhor et al14 correlated peak E- and peak A-wave velocities with L/H in 56 patients undergoing exercise perfusion scintigraphy: higher E-wave velocities were observed in patients with high L/H values. In the present study, severe mitral regurgitation and peak A-wave velocity were the sole independent predictors of increased 201Tl lung uptake, and a significant inverse linear relation between L/H and A was found. Peak A velocity reflects left atrial loading and systolic function: In the presence of an elevated LV stiffness, a reduced blood volume moves through the mitral valve, irrespective of atrial contraction, causing a reduction in peak A velocity and a rise in mean atrial pressure. This mechanism could also explain an increased L/H. In the subgroup of patients undergoing right heart catheterization, we also documented a significant correlation between L/H and wedge pressure; this result further supports the hypothesis that a high resting L/H reflects elevated LV filling pressure.12
Patients with severe LV systolic dysfunction are at high risk of death; assessment of diastolic function has been recently suggested to add information regarding their risk stratification. An abnormal relaxation pattern characterized by a restrictive physiology frequently occurs in patients with extensive damage and, in patients with depressed LV function, diastolic dysfunction may play an important role in determining clinical signs and symptoms. A significant correlation between hemodynamic indexes of pulmonary congestion and survival has been documented16 : This relation supports the link between a restrictive filling pattern at Doppler analysis and prognosis. Patients with a restrictive LV physiology after acute myocardial infarction were found to be in a particularly high-risk subgroup.9 17 Of note, in our study, 2 markers of diastolic dysfunction (an L/H >0.61 and a restrictive Doppler filling pattern) emerged as the sole independent predictors of poor prognosis. Since abnormalities of systolic and diastolic function frequently coexist, a marker of global LV dysfunction such as an increased 201Tl lung uptake may result a powerful marker of poor outcome. Unlike other studies that used a >0.50 L/H as a cutoff to identify high-risk patients undergoing stress 201Tl imaging, in our study, based on resting imaging, a higher L/H cutoff value (ie, 0.61) was found by ROC analysis to best discriminate patients at increased risk of major events. Differences in pulmonary mean transit time, pulmonary pressures, or vascular permeability at rest or during stress could explain the discrepancy in cutoff values. We also determined how well an increased resting 201Tl lung uptake performs when clinical and echo-Doppler findings have already been taken into account.
In the whole study cohort, roughly 75% of the left ventricle, on average, was viable. Since the LV ejection fraction was only 23%, this suggests that most of the viable segments were dysfunctional and presumably hibernating. For various reasons, however, only 23% of the patients underwent subsequent coronary revascularization. A consistent amount of myocardial viability in LV dysfunctioning regions is a powerful predictor of outcome18 19 20 . In patients with >7 viable segments out of 11 myocardial segments at preoperative rest-redistribution 201Tl imaging, Ragosta et al18 showed that mean LV ejection fraction significantly improved after coronary surgery. Pagley et al19 showed that the 201Tl extent of myocardial viability was the best predictor of transplant-free survival. Moreover, the extent of viable and potentially jeopardized myocardium has been shown a powerful predictor of mortality in medically treated patients.20 In contrast to these retrospective studies, in our study, which evaluated a more selective population with severe LV dysfunction, advanced functional class, and predominant symptoms of heart failure, the myocardial viability extent did not emerge as significant predictor either of hard or of major events at follow-up.
Our results were obtained in patients with ischemic heart disease and may not apply to patients with LV dysfunction of different causes. We found a significant correlation between the capillary “wedge” pressure and L/H, although invasive measurements were obtained in a limited number of patients and were performed within 48 hours of the scintigraphic study. If some degree of discordance in results is expected between invasive and noninvasive studies performed over separate days, any misclassification, however, would attenuate the results; so the true relations are likely to be even stronger than observed. We noninvasively assessed LV filling pressures by echocardiographic mitral inflow Doppler analysis; other accurate methods have been recently proposed.21 Finally, results of 201Tl imaging regarding the amount of LV viable tissue influenced the decision to perform coronary surgery in some patients; on the other hand, the degree of the thallium lung uptake, the topic of our study, played no role in the surgical decision.
Our data suggest that in patients with severe LV dysfunction undergoing 201Tl scintigraphy for the assessment of residual viability, an elevated tracer uptake in the lungs should also be considered by clinicians as having powerful prognostic significance. Abnormal L/H is easily and automatically detectable by standard 201Tl imaging and identifies a subgroup of patients with poor prognosis and in whom a very aggressive management seems warranted.
The authors are grateful to Rosemary Allpress, MA, for editorial assistance.
- Received March 14, 2000.
- Revision received May 5, 2000.
- Accepted May 19, 2000.
- Copyright © 2000 by American Heart Association
Gill JB, Ruddy TD, Newell JB, et al. Prognostic importance of thallium uptake by lung during exercise in coronary artery disease. N Engl J Med. 1987;317:1485–1489.
Marcassa C, Galli M, Cuocolo A, et al. Rest-redistribution thallium-201 and rest technetium-99 m-Sestamibi SPECT in patients with stable coronary artery disease and ventricular dysfunction. J Nucl Med. 1997;38:419–424.
Kahn JK, Carry MM, McGhie J, et al. Quantitation of post-exercise lung thallium-201 uptake during single photon emission computed tomography. J Nucl Med. 1989;30:288–294.
Helmcke F, Nanda NC, Hsiung MC, et al. Color Doppler assessment of mitral regurgitation with orthogonal planes. Circulation. 1987;75:175–183.
Ishida Y, Mesiner JS, Tsujioka K, et al. Left ventricular filling dynamics: influence of left ventricular relaxation and left atrial pressure. Circulation. 1986;74:187–196.
Ragosta M, Beller G, Watson D, et al. Quantitative planar rest-redistribution Tl-201 imaging in detection of myocardial viability and prediction of improvement in left ventricular function after coronary bypass surgery in patients with severely depressed left ventricular function. Circulation. 1993;87:1630–1641.
Pagley PR, Beller G, Watson D, et al. Improved outcome after coronary bypass surgery in patients with ischemic cardiomyopathy and residual myocardial viability. Circulation. 1997;96:793–800.
Di Carli MF, Davidson M, Little R, et al. Value of metabolic imaging with positron emission tomography for evaluating prognosis in patients with coronary artery disease and left ventricular dysfunction. Am J Cardiol. 1994;73:572–533.