The present study is a register-based observational study of an unselected consecutive patient cohort with pulmonary arteriovenous malformations (PAVMs) from a single national hereditary hemorrhagic telangiectasia and PAVM embolization center. The aim was to investigate the frequency of re-embolizations and the clinical outcome after embolization with the use of different embolization materials further, to define which PAVM morphology and size of feeding arteries that most often were re-embolized, and to estimate the clinical outcome of the patients including those that were re-embolized.

The population was included from 1996 until 2016 and was made up of a total of 136 patients with 322 PAVMs. Median follow-up was 38.3 (0.3–241 months).

The re-embolization rate was 9.3%. None of the PAVMs treated with detachable silicone balloons were re-embolized, while 4.5% treated with vascular plugs and 11.7% treated with coils were re-embolized (p=0.07). In total, 16/74 complex PAVMs were re-embolized compared with 14/248 simple PAVMs. In big-sized feeding arteries ≥ 6mm, 16/112 were re-embolized compared with 14/210 with smaller-sized feeding arteries. Out of the 30 re-embolized PAVMs, 23 resulted in a successful clinical outcome.

Our results suggest that standard coils probably should not be the first choice for embolization of PAVMs, and vascular plug alone or in combination with coils might be a better primary option for embolization in these patients.

Level 3A, non-randomized case controlled cohort/follow-up study.