Pulmonary perfusion by iodine subtraction maps CT angiography in acute pulmonary embolism: comparison with pulmonary perfusion SPECT (PASEP trial)

Eur Radiol. 2020 Sep;30(9):4857-4864. doi: 10.1007/s00330-020-06836-3. Epub 2020 Apr 11.

Abstract

Objective: To assess the diagnostic accuracy of iodine map computed tomography pulmonary angiography (CTPA), for segment-based evaluation of lung perfusion in patients with acute pulmonary embolism (PE), using perfusion single-photon emission CT (SPECT) imaging as a reference standard.

Methods: Thirty participants who have been diagnosed with acute pulmonary embolism on CTPA underwent perfusion SPECT/CT within 24 h. Perfusion SPECT and iodine map were independently interpreted by 2 nuclear medicine physicians and 2 radiologists. For both modalities, each segment was classified as normoperfused or hypoperfused, as defined by a perfusion defect of more than 25% of a segment. The primary end point was the diagnostic accuracy (sensitivity and specificity) of iodine map for segment-based evaluation of lung perfusion, using perfusion SPECT imaging as a reference standard. Following blinded interpretation, a retrospective explanatory analysis was performed to determine potential causes of misinterpretation.

Results: The median time between CTPA with iodine maps and perfusion SPECT was 14 h (range 2-23 h). A total of 597 segments were analyzed. Sensitivity and specificity of iodine maps with CTPA for the detection of segmental perfusion defects were 231/284 = 81.3% (95% CI 76.4 to 85.4%) and 247/313 = 78.9% (95% CI 74.1 to 83.1%), respectively. In retrospect, false results were explained in 48.7%.

Conclusion: Iodine map CTPA showed promising results for the assessment of pulmonary perfusion in patients with acute PE, with sensitivity of 81.3% and specificity of 78.9%, respectively. Recognition of typical pitfalls such as atelectasis, fissures, or beam-hardening artifacts may further improve the accuracy of the test.

Key points: • Sensitivity and specificity of iodine subtraction maps for the detection of segmental perfusion defects were 81.3% (95% CI 76.4 to 85.4%) and 78.9% (95% CI 74.1 to 83.1%), respectively. • Recognition of typical pitfalls such as atelectasis, fissures, or beam-hardening artifacts may further improve the diagnostic accuracy of the test.

Keywords: Computed tomography angiography; Lung; Perfusion; Pulmonary embolism.

Publication types

  • Clinical Trial

MeSH terms

  • Acute Disease
  • Adult
  • Aged
  • Aged, 80 and over
  • Angiography, Digital Subtraction / methods*
  • Computed Tomography Angiography / methods*
  • Female
  • Humans
  • Iodine
  • Iodine Isotopes / pharmacology*
  • Lung / diagnostic imaging*
  • Male
  • Middle Aged
  • Pulmonary Embolism / diagnosis*
  • ROC Curve
  • Reproducibility of Results
  • Retrospective Studies
  • Tomography, Emission-Computed, Single-Photon / methods*

Substances

  • Iodine Isotopes
  • Iodine