Background: Transpulmonary embolisation (TPE) using degradable starch microspheres (DSM) is a potential approach to treat pulmonary metastases. However, there is a paucity of detailed information on perfusion dynamics. The aim of this study was to establish a human ex vivo isolated lung perfusion (ILP) model to observe and evaluate the effects of DSM-TPE in a near-physiologic setting.

Methods: ILP was carried out on six surgically resected lung lobes. At baseline, computed tomography (CT), including CT perfusion imaging (CTPI), and histopathological sampling were performed (t30). DSM-TPE was initiated and increased stepwise (t45, t60, t75, and t90) to be followed by CT imaging, histopathological sampling, and pulmonary arterial pressure (PAP). After the last assessment (t90), alpha-amylase was injected into the pulmonary artery to allow for DSM hydrolysation and two additional assessments (t105; t120). Histopathological specimens were evaluated using a semiquantitative ordinal score. CTPI was used for time to peak (TTP) analysis.

Results: After DSM administration, PAP and TTP increased significantly: PAP slope 95% confidence interval (CI) 0.104−0.483, p = 0.004; TTP t30 versus t45, p = 0.046. After the addition of alpha-amylase, functional parameters reverted to values comparable to baseline. In histopathological samples, embolisation grades increased significantly until t90 (slope 95% CI 0.027−0.066, p < 0.001) and decreased after addition of alpha-amylase (slope 95% CI -0.060−0.012, p = 0.165),

Conclusions: The ILP model demonstrated successfully both the physiologic effect of DSM-TPE on human lungs and its reversibility with alpha-amylase. Thus, it can be used as a near-physiologic preclinical tool to simulate and assess later clinical approaches.