The accuracy of pulmonary angiography in assessing resolution of experimental thromboemboli.

Quantification of pulmonary emboli has become important in assessing the results of thrombolytic therapy and in following the natural history of thromboemboli. This study evaluated angiography as a quantitative procedure by comparing sequential changes in emboli of known size and weight with alterations on angiograms. Twenty-four aged thrombi formed in vivo were embolized intact into 12 recipient dogs. Serial angiograms showed the location of each thrombus and allowed evaluation of its changes. Urokinase was administered to some animals to produce a spectrum of lysis. Quantification proved accurate when thrombolysis exceeded 80%. When less, angiograms over or underestimated lysis in 11 of 18 instances (61%). Postmortem dissection suggested four reasons for these discrepancies: (1) downstream movement impacted emboli and increased obstruction; (2) coiling of an embolus compromised more vessel area and increased obstruction; (3) circumferential flow of contrast material obscured some emboli; and (4) second emboli redistributed flow to an area with prior embolus, producing apparent improvement. These data suggest that quantifying changes in emboli by angiography are frequently misleading when less than 80% lysis occurs. The discrepancy, when present, usually results in overestimation of the amount of embolus remaining.