The oral thrombin inhibitor dabigatran (Pradaxa) has been shown to be detectable in the vitreous and subretinal fluid after a single oral administration of 220 mg (Mulder et al. 2016). The maximum concentrations that were found were 8.5 ng/ml in subretinal fluid and 3.8 ng/ml in vitreous. An unexpected finding in this study was the 10 times higher vitreous concentration of a non-study patient, who used 150 mg dabigatran twice daily for atrial fibrillation. This finding in combination with the observation that the median vitreous concentration was the highest at our last time point and thus possibly still increasing led to the hypothesis that repeated administration of dabigatran might lead to higher intraocular levels. Therefore, we tested this hypothesis in patients on standard dabigatran therapy who were admitted to the Rotterdam Eye Hospital for retinal surgery.
We were able to include one male and two female patients, who were being treated with standard twice-daily dosages of dabigatran (see Table 1). One patient underwent surgery for a dropped nucleus after cataract surgery and two patients for a macular hole. One of these patients had a persistent macular hole and underwent repeat surgery after 5.5 weeks during which we collected a second sample. Collection of undiluted vitreous samples and analysis with LC-MS/MS were the same as described in our previous study (Mulder et al. 2016). The median concentration in vitreous was 8.0 ng/ml (range 5.6–19.5 ng/ml) and the median concentration in plasma was 100.4 ng/ml (range 95.8–159.1 ng/ml). These vitreous concentrations were at least twice as high as the mean vitreous concentration previously reported (1.9 ng/ml; Mulder et al. 2016).
Due to the relative rarity of patients on dabigatran therapy undergoing vitrectomy, we were able to include only three unique patients in 1 year of which one on two separate occasions. They are not patients with a rhegmatogenous retinal detachment (RRD) – like the previous study – but patients with a macular hole and a complication after cataract surgery. This difference in patient population might have influenced the measured vitreous concentrations. In contrast to patients with a macular hole, we assume that patients with RRD have a partially disrupted blood–retinal barrier which contributes to the influx of dabigatran in the vitreous. Therefore, these results show that even with an intact blood–retinal barrier, dabigatran is able to penetrate the vitreous as long as there is sufficient supply (i.e. repeated administration).
In addition, the presence of formed vitreous seemed to influence the concentration in the vitreous cavity. Patient 3 underwent two consecutive vitrectomy procedures in which we collected both vitreous and vitreous cavity fluid samples. The collection time was in both cases 3.5 hr after the last dose, but the concentration of dabigatran in the second sample was much higher than that in the first one (6.0 ng/ml versus 19.5 ng/ml). A possible explanation for this result is the lower viscosity of aqueous humour filling the vitreous cavity after the first vitrectomy (Gisladottir et al. 2009; Donati et al. 2014). As dabigatran is a hydrophilic compound, lower viscosity possibly results in higher diffusion in this medium than in the vitreous (Gisladottir et al. 2009). The same mechanism might have contributed to the high concentration we found in our previous study (25.8 ng/ml).
In conclusion, we have demonstrated that repeated use of the oral thrombin inhibitor dabigatran leads to higher vitreous levels of dabigatran than a single administration, even in patients with a supposedly intact blood–retinal barrier. This result increases the potential of dabigatran as a possible therapeutic agent in the modulation of PVR.

Additional Metadata
Persistent URL dx.doi.org/10.1111/aos.13348, hdl.handle.net/1765/97707
Journal Acta Ophthalmologica
Citation
Mulder, V.C, Kluft, C, van Etten, P.G, La Heij, E.C, & van Meurs, J.C. (2017). Higher vitreous concentrations of dabigatran after repeated oral administration. Acta Ophthalmologica, 95(4), e345–e346. doi:10.1111/aos.13348