Subretinal tissue plasminogen activator injection to treat submacular haemorrhage during age-related macular degeneration.

Abstract

We reviewed the records of patients who were admitted at the University Hospitals Leuven between November 2008 and July 2012 with submacular haemorrhage due to AMD and treated using 23-G vitrectomy with subretinal tPA (0.5–1 ml injection through a 41-G microcannula) and air tamponade to displace the haemorrhage downwards from the fovea. This treatment was followed by standard-of-care intravitreal anti-VEGF injection treatment. Inclusion criteria were as follows: unilateral submacular haemorrhage larger than one disc diameter, but smaller than the area between the two vascular arcades; and preserved vision capable of more than counting fingers in the fellow eye. The study included 74 eyes of 74 patients. The mean age was 79.65. A total of 27 patients were pseudophakic; all 47 phakic patients (64%) underwent combined cataract surgery. Most of the patients underwent surgery within a few days after presentation, but time of presentation at our department varied greatly. The mean delay between the onset of symptoms and the surgery was 22.68 days (1 day to 5 months). Follow-up visits were performed 1 day, 1 week and 6 weeks (mean follow-up: 53 days) after the surgery. The final follow-up visit varied from 18 days to over 4 years. At baseline, visual acuity (VA) ranged from 20/40 (0.3 logMAR) to light perception [mean VA: 20/800 (1.6 logMAR)]. After surgery, VA improved by 0.34 logMAR units at the 6-week follow-up visit, indicating an improvement of 23.5% (p < 0.0001). In 14 of the patients (19%), VA worsened, and 11 patients (15%) maintained their preoperative VA. A mean improvement of 0.24 logMAR units was recorded at the end of follow-up. The scatter chart in Fig. 1 displays the pre- and postoperative outcomes along the X- and Y-axes. VA tended to decrease with the length of follow-up; the final VA was 0.1 logMAR units worse than the 6-week postoperative VA. Delays in surgery are associated with significantly worse visual outcomes. There was a deterioration of VA of 0.003 logMAR units/day of delay at the 6-week follow-up (p = 0.040). No intra-operative complications were recorded, but adverse events occurred postoperatively. Two patients developed vitreous haemorrhage and hyphema that required reintervention. A total of seven patients suffered from retinal detachment: four of these patients presented 4 weeks after surgery, and the other three patients presented with proliferative vitreoretinopathy 8 weeks postoperatively. Two patients suffered from subchoroidal haemorrhage. Nine subjects presented recurrent subretinal haemorrhage during the follow-up period. As these new haemorrhages all developed more than 1 week postoperatively, they were likely caused by the underlying disease rather than the intervention. This study is limited by its non-comparative, non-randomized retrospective design and relatively short-term follow-up. The timing of the final follow-up visit was highly variable. Consequently, the results from the final follow-up are highly comparable to the results at 6 weeks (median: 55 days). The lack of a control group is another limitation of this study. However, our results compare favourably to those of other similar studies evaluating the natural course of submacular haemorrhage. A prospective case series by Cheung et al. (2013). indicated a median improvement in VA of 0.20 LogMAR units after 6 months. Similarly, retrospective data collection of 86 eyes revealed a 10.5% VA improvement after 6 months (Chen et al. 1999). In summary, the results of this study indicate that our surgical approach may be an effective treatment for submacular haemorrhage displacement in AMD patients. The procedure improved the visual recovery. However, the visual outcome is limited by the underlying macular pathology. Larger multicentre randomized controlled studies are warranted to determine the therapeutic effect of this surgical approach.