Objectives: To evaluate the use of diffusion-weighted imaging (DWI) for estimating infarcted splenic volume during partial splenic embolisation (PSE) using n-butyl cyanoacrylate (NBCA).
Methods: Twenty consecutive patients (57.2 ± 11.7 years) with hypersplenism underwent PSE. Intrasplenic branches were embolised using NBCA via a 2.1-French microcatheter aiming at infarction of 50 to 80 % of total splenic volume. Immediately after PSE, signal intensities (SI) of embolised and non-embolised splenic parenchyma were measured on DWI. Semi-automated volumetry (SAV) on DWI was compared with conventional manual volumetry (MV) on contrast-enhanced CT 1 week after PSE. Platelet counts were recorded before and after PSE.
Results: The SI on DWI in the embolised parenchyma decreased significantly (P < 0.01) to 24.7 ± 8.1 % as compared to non-embolised parenchyma. SAV and MV showed a strong correlation (r = 0.913 before PSE, r = 0.935 after PSE, P < 0.01) and significant (P < 0.01) reduction of normal splenic volume was demonstrated on both SAV (71.9 ± 12.4 %) and MV (73.6 ± 9.3 %) after PSE. Based on the initial SAV, three patients (15 %) underwent additional branch embolisation to reach sufficient infarction volume. Platelet counts elevated significantly (522.8 ± 209.1 %, P < 0.01) by 2 weeks after PSE. No serious complication was observed.
Conclusion: Immediate SI changes on DWI after PSE allowed semi-automated splenic volumetry on site.
Key points: • Partial splenic embolisation (PSE) is an important interventional technique for hypersplenism • Diffusion-weighted MR reveals an immediate decrease in signal in the embolised parenchyma • Such signal reduction permits semi-automated splenic volumetry on site. • This allows precise quantification of the amount of parenchyma infarcted, avoiding additional PSE.