Predictive Value of CT Pulmonary Angiography to Assess the Surgical Accessability for Pulmonary Endarterectomy in Patients with Chronic Thromboembolic Pulmonary Hypertension

Abstract

Purpose Currently, no radiological reference is available to predict the level of disease in patients with chronic thromboembolic pulmonary hypertension (CTEPH). The purpose of this study was to assess the predictive value of a new radiological classification of the level of disease in CTEPH patients based on CT pulmonary angiography (CTPA) using the surgical classification as a standard of reference. Methods We retrospectively evaluated 43 patients (mean age, 57 ± 16 years; 14 females) with CTEPH undergoing CTPA prior to surgery. The median time between CTPA and surgery was 77 days (range 1 to 248 days). Three chest radiologists, blinded to surgical results, independently classified the level of disease based on the most prox. thrombus. Radiological and surgical classification was scored as follows: L1 (L1) (main pulmonary artery [PA]), L2 (lobar PA), L3 (segmental PA), and L4 (subsegmental PA). L2 was subdivided into L2a (lobar PA) and L2b (lower lobe basal trunk). Fleiss kappa was calculated for interobserver variability. To assess the predictive value, “proximal disease” was defined as L1 and L2a and “distal disease” was defined as L2b, L3, and L4. Results Three radiologists classified L1 in 35%, 28%, and 21%, L2a in 51%, 49%, and 61%, L2b in 7%, 14%, and 7%, and L3 in 7%, 9%, and 12%, respectively. None rated L4. Interobserver agreement was k=0.55. All radiological classifications were within 1 level of surgical classifications. Considering surgical classification as the standard of reference, the sensitivity, specificity, and accuracy of CTPA in identifying proximal disease in a cohort of CTEPH patients is 89%, 70%, and 81%, respectively. Conclusion Our results show that CTPA is highly sensitive to predict the level of disease in CTEPH patients with a moderate interobserver agreement even for a more detailed categorization of the level of disease. This newly introduced imaging based classification may support surgeons’ decision-making and predict operative findings.