AIM OF THE STUDY ST-segment depression on exercise stress test (EST) is an independent predictor of future cardiac events. Nevertheless, in apparently healthy subjects without angina the occurrence of false positive results is frequent. Thallium myocardial imaging (TMI) may improve diagnostic and prognostic accuracy of EST. The aim of the present study was to assess the role of a normal exercise TMI for excluding a coronary artery disease in subjects with asymptomatic abnormal EST. METHODS Subjects referred for TMI from 1/1980 to 5/1991 with an abnormal EST and without history of ischemic, congenital, or valvular heart disease or abnormal resting ECG were included into the study. 137 subjects (98 men, 39 women), mean age 53 +/- 8 yrs (range 37-74 yrs) were enrolled and followed-up for 6.4 yrs (range 3-13 yrs). Clinical indications for EST were: atypical chest pain in 56 (41%) cases, check-up in 52 (38%) cases, sport activity in 29 (19%) cases. All subjects had a maximal symptom-limited EST. Abnormal EST was defined by a horizontal or downsloping > or = 1 mm or upsloping > or = 1.5 mm ST-segment depression at 0.08 sec. from J-point, in at least 2 leads. EST was discontinued for fatigue in 129 (94%) cases, for ST-segment depression > or = 3 mm in 8 (6%) cases. None had chest pain during EST. All subjects selected for the study had normal TMI. Criteria for normal TMI were homogeneous Thallium uptake on postexercise images and a normal washout in the delayed images by qualitative analysis. Planar images were obtained in 118 (86%) cases, and tomographic SPECT images in 19 (14%). RESULTS During the follow-up period no subject died for cardiac causes and only 9 subjects (1%/yr) had non fatal cardiac events: 4 (0.45%/yr) had a non fatal myocardial infarction (one subject had coronary angiography for postinfarction angina and subsequent 3 coronary bypass graft for multivessels disease), 2 subjects (0.2%/yr) became symptomatic for unstable angina (both had coronary angiography and subsequent PTCA for critical left main coronary artery stenosis) and 3 (0.34%/yr) developed stable angina (one had coronary angiography and subsequent bypass graft for a critical stenosis of left main coronary artery). Four further subjects died for non cardiac events. Comparing clinical data and TE results of subjects with and without coronary events, we found that some parameters were related to a higher incidence of cardiac events: hypertension (78% vs 31% respectively in subjects with and without cardiac events, p < 0.01), hypercholesterolemia (33% vs 4.7%, p < 0.01); > or = 2 conventional coronary risk factors (56% vs 17%, p < 0.02); and a slow regression of abnormal ST-segment depression during recovery (2.8 +/- 2 vs 1.5 +/- 1 min, p < 0.01). CONCLUSIONS In conclusion, in subjects without typical chest pain and with abnormal asymptomatic EST, a normal exercise TMI identifies subjects with very low risk of future cardiac events (1%/yr). Our data suggest that subjects with abnormal asymptomatic EST should be routinely submitted to exercise TMI.