Song J. Kim, MD

Mac Grory B, Nackenoff A, Poli S, Spitzer SM, Nedelmann M, Guillon B, Preterre C, Chen CS, Lee AW, Yaghi S, et al. Intravenous Fibrinolysis for Central Retinal Artery Occlusion: A Cohort Study and Updated Patient-Level Meta-Analysis. Stroke. 2020;51:2018–2025.

Patients presenting with complete retinal artery occlusion (CRAO), often first evaluated by ophthalmologists, are advised to seek emergent neurological evaluation. On the receiving end, even when the patient arrives within the typical thrombolysis window of <4.5 hours, some vascular neurologists may demur on offering reperfusion therapy owing to : 1) lack of robust evidence with regards to outcomes, and 2) absence of a comprehensive ophthalmological evaluation in excluding other non-ischemic differentials. 

To address the above, Mac Grory and co-authors developed a protocol for a rapid evaluation of CRAO including a dilated ophthalmologic evaluation and other work-up, ruling out mimics such as optic neuritis, giant cell arteritis, and retinal detachment. Barring any contraindications, patients received tPa if CRAO was deemed the most likely diagnosis, with visual acuity of less than 20/200 in the affected eye and presenting within 4.5 hours of symptom onset. Over a 10-year time span, 112 were diagnosed with acute CRAO. They were subsequently included in a retrospective observational analysis on post-tPa visual outcomes, as well as an updated subject-level meta-analysis expanding upon the authors’ 2015 study.

Of 112 total patients with the diagnosis of CRAO, 25 (22.3%) received tPa. Recovery was reported dichotomously, defined as a final visual acuity of 20/100 or better and also reported as a quantifiable log-MAR adjusted scoring (0.3 on the scale indicates 3 lines on the Snellen chart). 44% of the tPa cohort (11 patients) achieved visual recovery vs. 13.1 % of those arriving at all times (p=0.003), and vs. 11.6% presenting within 4 hours who did not receive alteplase (p=0.03). Only one asymptomatic intracranial hemorrhage occurred.

The trend toward improved outcome in thrombolysis was seen in the updated patient-level meta-analysis incorporating the above patient data, as well as 3 cohort studies and one randomized trial. 37.9% of 67 patients treated with alteplase within 4.5 hours recovered visual acuity of 0.98 on the logMAR scale, compared with 17.7% visual recovery in untreated, known natural history of CRAO. In linear regression, earlier treatment correlated with a higher rate of visual recovery (P=0.01) after adjusting for age. There were 7 adverse events, two of which were symptomatic intracranial hemorrhage.

This small study has limitations owing to its retrospective design. First is the presence of selection bias; patients who received tPa arrived earlier, and with more dense vision loss (light perception vs. hand motion). A comparison between the two groups of patients all presenting within 4.5 hours of symptom onset — those receiving tPa and those not receiving tPa — would have helped elucidate any further background differences. Specific reasons for not administering tPa to the 43 patients arriving <4.5 hours were not reported. And because this study was underpowered to adequately detect large differences in outcome, perhaps a case-control series matched by the duration of symptoms, degree of vision loss at presentation may have eliminated confounding and highlighted the independent effect of tPa.

Regardless, the authors demonstrate that tPa administration in CRAO is safe, with low rates of bleeding complications. The trend towards recovery in patients receiving earlier administration of tPa is also in keeping with general principles of tPa therapy. Moreover, the particular strength of their work is in laying the groundwork for future prospective randomized trials by: 1) designing a systematic protocol that can be implemented across multiple institutions; 2) implementing a time cutoff that is in keeping with typical acute stroke management and findings from animal studies of reversible retinal ischemia; and 3) a clinically meaningful endpoint of visual outcome aligned with legal blindness, and a quantifiable measure thereof.

Altogether, these study results may prompt physicians to consider tPa in CRAOs until future randomized trial results become available. Further investigation of acute therapies in CRAOs, which leave patients with devastating outcomes, deserves much attention. Just as “time is brain” in administration of tPa in acute strokes, a randomized trial may prove that “time is vision” in reversing retinal ischemia.