Author + information
- Received August 11, 1998
- Revision received November 5, 1998
- Accepted December 23, 1998
- Published online April 1, 1999.
- ↵*Reprint requests and correspondence: Dr. Christopher L. Hansen, Section of Cardiology, Temple University Hospital, 3401 North Broad Street, Philadelphia, Pennsylvania 19140
The purpose of the study was to evaluate the relationship between elevated lung/heart ratio (LHR) and transient ischemic dilation (TID) after stress thallium-201 myocardial perfusion imaging and to provide further insight into the mechanism of cavity dilation.
Because both LHR and TID have been identified as adjunctive markers of severe coronary disease they should be found in the same patients. Although the mechanism of LHR has been defined, that of transient dilation has not.
We identified 4,618 consecutive patients undergoing maximal exercise perfusion imaging with thallium-201. Lung/heart ratio and a dilation index were derived and compared to each other and to relevant clinical parameters.
There was a very weak relationship between the LHR and dilation index (r = 0.15, p < 0.001). Defining a dilation index ≥1.10 and LHR ≥50% as abnormal revealed that 322 of the patients (7%) had TID only, 351 (7.8%) had LHR only and 40 (0.9%) had both. When compared to patients without these findings both TID and LHR had higher thallium stress defect and redistribution scores. When comparing subjects who had elevated LHR uptake to those who had TID, it was found that those with LHR were more likely to have had prior myocardial infarction (MI) (29% vs. 9%), coronary artery bypass grafting (22% vs. 8%), lower ejection fraction (34 ± 17% vs. 55 ± 11%) and had more evidence of ischemia based on thallium stress defect and redistribution scores. However, patients with cavity dilation had a higher frequency of positive electrocardiographic response (31% vs. 19%) despite lower scintigraphic markers.
Although pulmonary uptake and transient cavity dilation have both been associated with severe coronary disease, they have a very weak correlation, which, in combination with the different clinical parameters associated with each, suggests that they represent different pathophysiologic responses to exercise-induced ischemia. Our data support the hypothesis that TID represents transient subendocardial ischemia rather than physical dilation from increased end-diastolic pressure.
Thallium-201 (Tl-201) myocardial perfusion imaging is a technique that exploits the pathophysiology of coronary disease to establish the diagnosis. It most frequently demonstrates reversible perfusion defects in the distributions of obstructive coronary lesions. The adjunctive findings of increased pulmonary uptake of thallium, as measured by an elevated lung/heart ratio (LHR) and transient ischemic dilation (TID), have previously been reported and associated with severe coronary artery disease (1–6). Understanding the clinical context of these findings is an important foundation for recognition of the pathophysiologic processes involved and should allow a more accurate assessment of a patient’s condition.
The mechanism of elevated LHR is well understood and represents the effects of left ventricular decompensation with exercise. Decompensation leads to decreased cardiac output, which causes an increase in end-diastolic volume and pressure; the pressure is transmitted back to the pulmonary capillaries, resulting in leakage of thallium into the interstitial spaces of the lungs (2,7). The etiology of TID is less well defined; there are two possible mechanisms. The first, paralleling the mechanism of increased pulmonary uptake, would occur by increased end-diastolic pressure causing physical dilation of the chamber. A second mechanism would be diffuse subendocardial ischemia. This would make the ventricle appear larger and thinner on the initial images. With redistribution the subendocardial ischemia would resolve and the ventricle would regain its “normal” appearance.
If the mechanism for cavity dilation were, in fact, due to increased end-diastolic pressure, a strong correlation between cavity dilation and pulmonary uptake would be expected. A second reason to expect these markers to correlate is the fact that they have both been associated with severe coronary disease. In our laboratory, however, we have been surprised at the lack of correlation of these findings. The current study was undertaken to evaluate more formally the association between these findings and to see whether further inferences regarding the pathophysiology of cavity dilation could be made from the clinical context in which each of these findings occur.
We identified 4,618 patients undergoing maximal stress testing with single-photon emission computed tomography (SPECT) Tl-201 myocardial perfusion imaging over a 6-year period from December 1991 to December 1997. Patients were exercised on either the Bruce or modified Bruce protocols. They were injected with 3 to 3.5 mCi of Tl-201 at peak exercise and encouraged to exercise 1 min further. A 3-min anterior planar image of the heart and lungs was obtained approximately 10 min after exercise. The patients then underwent SPECT imaging using a single-headed camera (General Electric StarCam, Milwaukee, Wisconsin) over a conventional 180° orbit. The patients returned 3 to 4 h later for redistribution imaging.
The lung/heart ratio was calculated from the anterior planar image by dividing the average counts from regions of interest over the lung (avoiding the liver and the heart) and the portion of the left ventricle demonstrating the highest thallium activity. A LHR of greater than 50% was considered abnormal (8). The SPECT reconstruction was performed using commercially available software. Radial profiles were generated for the apical, mid- and basal short axis sections of both the stress and delayed images. Defects were quantitated against gender-based normal databases. A stress defect score was calculated by summing all regions where radial plot activity fell below the normal limits; defects were measured in units of standard deviations at the angle where they occurred and corrected for the radius of the short axis section to adjust for relative defect size in relation to the entire left ventricle. A redistribution score was calculated in an analogous manner from the redistribution images and the stress defect score. A size index was calculated from the total number of short axis sections and from the average radius of the apical, mid- and basal short axis sections.
Cavity dilation was calculated by taking the ratio of the average radius of the mid- and basal short axis sections for the stress images and dividing by the average radius of the corresponding short axis sections on the delayed images. Forty-two patients (22 men, 20 women) with a low probability of coronary disease were used to evaluate the range of cavity dilatation in a normal population using this technique.
For the purposes of this study, a history of myocardial infarction (MI) was recorded when the patient was known to have had a clinical event associated with a CK (creatine kinase) elevation greater than twice normal with a positive creatine kinase MB fraction or gave a history consistent with MI and was found to have pathologic Q waves on two adjacent leads on the electrocardiogram (ECG).
Echocardiographic findings were correlated by evaluating all patients who had echocardiography within two weeks of stress testing. Ejection fraction was visually estimated by experienced observers who had no knowledge of the results of the Tl-201 perfusion imaging. Left ventricular chamber size and wall thickness were measured using commercially available software.
Linear regression and ANOVA (analysis of variance) were performed using commercially available software (SPSS, Chicago, Illinois). Multiple comparisons of continuous variables were made by performing a one-way ANOVA and, when significant, differences between pairs were tested using the Bonferroni correction for the confidence limit. Multiple comparisons of categorical variables were made using the Kruskal-Wallis test and, when significant, pairs were compared using the Dunn procedure (9). An appropriately corrected p value <0.05 was considered significant for all tests.
The 42 normals had a mean heart/lung ratio of 31 ± 6.7% and a mean dilation index of 0.98 ± 0.043. In the 4,618 patients evaluated, both the LHR and dilation index showed heavily skewed distributions. We empirically chose a dilatation index of 1.10 to identify TID, which, incidentally, was the same number of standard deviations above the mean of the normals (2.8) as was the lung uptake threshold. Using these criteria, 322 of the patients (7.0%) had TID only, 351 (7.6%) had elevated LHR only and, surprisingly, only 40 (0.9%) had both. A scatter plot showing the relationship between dilation index and LHR is shown in Figure 1. Although the relationship was statistically significant (p < 0.0001) the low r value of 0.15 means that only 2% of the variation noted in the dilation index was related to change in the LHR.
An example of a patient showing both elevated LHR and TID is depicted in Figure 2. This was from a man with no prior cardiac history and a positive stress ECG. The LHR was 63%; the dilation index was 1.14; the left ventricular (LV) size index was 104; the stress severity score was 160 and the redistribution score was 135. Cardiac catheterization revealed severe triple-vessel disease.
Table 1compares the clinical parameters of patients with TID, elevated LHR, both and neither. Patients with elevated LHR were much more likely to have a history of prior MI and coronary artery bypass grafting (CABG) compared to those with TID and those with neither. Not surprisingly, patients with elevated LHR or TID had more severe thallium scintigraphic markers of ischemia compared to those without and patients with both had the most severe markers of all. It was interesting to note that although patients with elevated LHR had more severe thallium markers compared to TID, the patients with TID were more likely to have a positive ECG response for ischemia.
There were 994 patients who had echocardiography within two weeks of stress testing; the findings in these patients are summarized in Table 2. The patients with elevated LHR only had the lowest ejection fractions and largest LV chambers, suggesting that they represent the patient subgroup with the greatest amount of myocardial damage. It was interesting to note that patients with both TID and elevated LHR had chamber sizes and ejection fractions in between those with TID and elevated LHR only, suggesting that even though they represent the patient group with the greatest amount of ischemia (Table 1)they did not appear to have the greatest amount of myocardial damage.
Our findings agree with previous investigations that elevated LHR and TID are markers of more severe coronary disease (1–6). It is surprising, however, that two markers associated with severe disease have such a weak correlation; this strongly suggests that two separate pathophysiologic processes are responsible for these findings or that they reflect different time points in the natural history of coronary disease.
Possible mechanisms of TID
It is revealing that patients with TID were much more likely to have a positive ECG response for ischemia when compared to elevated LHR despite the fact that the thallium stress defect and redistribution scores were lower (Table 1). The association between subendocardial ischemia and ST depression is well known (10,11). A focal area of subendocardial ischemia would reduce counts in that area and be manifested as a mild defect. Diffuse subendocardial ischemia, when seen in multivessel disease, should symmetrically reduce activity everywhere in the ventricle and result in a more homogeneous perfusion pattern and a lower defect score. These findings strongly suggest that the occurrence of TID in our patients represents, at least in part, diffuse subendocardial ischemia. The fact that only a third of patients with cavity dilation had positive ECG responses would suggest either that the ECG is less sensitive to ischemia or that other factors may be involved.
Another way of explaining the weak correlation between TID and elevated LHR would be that they represent different modalities in which the compliance of the ventricle governs its response to severe exercise-induced ischemia. In both instances severe ischemia would cause systolic dysfunction associated with increased end-diastolic volumes. If patients with elevated LHR had lower compliance, the increased end-diastolic volume would result in higher end-diastolic pressure resulting in leakage of thallium from the pulmonary capillaries. If patients with TID had greater compliance, they might respond to increased end-diastolic volume by physically dilating, thus sparing the lungs from increased end-diastolic pressure. It is hard to prove or disprove this hypothesis as clinical markers of LV compliance during exercise have not been well established. Even the relationship of elevated LHR and TID with resting diastolic function is unclear; Mannting (12)reported that pulmonary uptake of thallium did not correlate with resting parameters of diastolic function derived from radionuclide ventriculography.
Our findings must be interpreted with regard to the protocol we employed. The SPECT imaging was not started until 15 min after exercise. Our measure of TID obtained from these images would reflect chamber dilation at the time of the imaging but would reflect subendocardial ischemia at the time of injection (peak exercise). Therefore, it is possible that many patients would have had physical dilation that resolved by the time SPECT imaging was performed. The fact that TID is frequently seen with Tc-99m sestamibi imaging an hour after exercise further supports diffuse subendocardial ischemia as the cause for TID. We cannot exclude myocardial stunning and persistent ventricular dilation in the small number of our patients (0.9%) who had both TID and elevated LHR.
Patients with elevated LHR had a more extensive history of coronary disease. They more frequently had a history of CABG, prior MI and their ejection fractions were lower; the thallium-derived size index and ECG end-diastolic volume were larger. This is reinforced by the fact that the severity of coronary artery disease (CAD) as measured by the thallium stress and redistribution scores was worse. It appears likely that they represent a group with severe disease and a poor prognosis. Despite the large number of patients studied, only a few patients had both TID and elevated LHR, making it difficult to demonstrate statistically significant differences. Nevertheless, they clearly had the most severe ECG and scintigraphic markers of ischemia (Table 1). Despite this, their ejection fractions are no worse, and could be slightly higher, than those patients with elevated LHR alone. This suggests to us that they represent a group very likely to benefit from revascularization.
Although it may be considered a limitation of the study that not all patients in this study were catheterized, it is obviously not feasible to catheterize all patients undergoing stress testing. We identified 772 patients who had catheterization within 60 days of stress testing. There was no significant difference in severity of coronary disease between patients with pulmonary uptake and cavity dilation. However, the stress defect scores were higher in this subgroup than in the group as a whole, and the difference between elevated LHR and TID was much smaller. Likewise, the redistribution scores were also much higher than in the group as a whole and there was no difference at all in redistribution scores between patients with elevated LHR and TID who were catheterized. This implies that there was, understandably, a significant referral bias affecting the subgroup that was catheterized, and thus the catheterization findings cannot be considered reflective of the group as a whole.
Weiss et al. (5)compared TID with LHR in 32 patients and concluded that TID was a superior marker of severe multivessel coronary disease. In their study they excluded patients who were unable to achieve 75% of age- and sex-predicted maximal heart rate where we have included all patients. Although the issue was not explored in their report (5), inspection of their Figure 5 suggests that there is no significant correlation between LHR and TID in the small group they studied.
We excluded the apical short axis section from the calculation of the cavity dilation index. When we included the apical short axis this did not appreciably change the mean but did increase the standard error. We believe this reflects the higher tracking error that occurs when performing radial plots on the relatively smaller apical short axis section.
There is a very weak relationship between LHR and TID, suggesting that these two recognized manifestations of severe CAD represent different pathophysiologic processes. Elevated LHR is much more likely to be seen in patients with prior MI, coronary bypass who have lower ejection fractions and increased chamber sizes; it likely represents more advanced disease. Conversely, TID is seen more often in patients with ST segment depression on their ECG despite the fact that quantitative analysis of their scans shows less severe defect and reversibility scores. Patients with both TID and elevated LHR have the most severe ECG and scintigraphic evidence for ischemia. The weak correlation of cavity dilation with LHR and the fact that TID has more ST depression despite less severe scintigraphic markers of CAD strongly suggest that TID reflects apparent dilation caused by diffuse subendocardial ischemia more than physical dilation from increased end-diastolic pressure.
- analysis of variance
- coronary artery bypass grafting
- coronary artery disease
- creatine kinase
- lung/heart ratio
- myocardial infarction
- single-photon emission computed tomography
- transient ischemic dilation
- Received August 11, 1998.
- Revision received November 5, 1998.
- Accepted December 23, 1998.
- American College of Cardiology
- Kushner F.G,
- Okada R.D,
- Kirshenbaum H.D,
- Boucher C.A,
- Strauss H.W,
- Pohost G.M
- Weiss A.T,
- Berman D.S,
- Lew A.S,
- et al.
- Canhasi B,
- Dae M,
- Botvinick E,
- et al.
- Rosner B
- Chou T.C
- Ellestad M.H