Submacular hemorrhage (SMH) is a condition in which blood from the choroidal or retinal circulation accumulates between the neurosensory retina and the retinal pigment epithelium (RPE). Subretinal hemorrhage involving the macula is specifically called SMH.
Clinically, it may be difficult to distinguish between subretinal hemorrhage and subpigment epithelial hemorrhage. Generally, subpigment epithelial hemorrhage appears darker. OCT findings are useful for differentiation.
Hemorrhage outside the macula is less likely to affect vision, but subretinal hemorrhage in the macula can be removed by pneumatic displacement or vitrectomy before clotting, with expected visual improvement 12). If clotted, concomitant use of tissue plasminogen activator (tPA) is necessary 12).
QIs surgery always necessary for submacular hemorrhage?
A
It depends on the size, thickness, and cause of the hemorrhage. Small hemorrhages may improve with anti-VEGF therapy alone or observation 6). Large, thick hemorrhages often require surgical intervention. For details, see the section on Standard Treatment.
Fundus examination reveals a dark red hemorrhage in the macula. Fresh hemorrhage appears bright red, while old hemorrhage appears yellowish-white (due to dehemoglobinization) 8).
Fundus findings: In subretinal hemorrhage, retinal vessels can be seen over the hemorrhage. Subretinal pigment epithelium (sub-RPE) hemorrhage appears as a dome-shaped hemorrhagic pigment epithelial detachment (PED).
OCT findings: It appears as a highly reflective material beneath the retina4). The presence or absence of RPE elevation helps determine the layer of hemorrhage. Subretinal and sub-RPE components may coexist 4).
Hemorrhage size classification: Classified as small (less than 4 disc diameters), medium (4 disc diameters or more but within the vascular arcades), and large (extending beyond the vascular arcades) 5).
OCT is also useful for quantitative assessment of hemorrhage thickness and extent; a thickness greater than 100 μm is considered one criterion for surgery 5).
QHow to differentiate sub-RPE hemorrhage from subretinal hemorrhage?
A
Generally, sub-RPE hemorrhage appears darker than subretinal hemorrhage. OCT is useful for differentiation because the two types lie in different layers. However, subretinal hemorrhage may obscure the outer layer, potentially hiding sub-RPE hemorrhage.
Many patients with SMH associated with nAMD take anticoagulants or antiplatelet agents.
Weber et al. (2023) retrospectively analyzed 115 patients with nAMD-related SMH requiring surgery1). 72.2% were taking anticoagulants or antiplatelet agents, and the bleeding area in the medication group (mean 35.92 mm²) was significantly larger than in the non-medication group (mean 21.91 mm²) (p=0.001). The group taking vitamin K antagonists (VKA) had a larger bleeding area than the DOAC group (63.70 mm² vs 31.76 mm²; p=0.005) and had worse visual outcomes.
The indication for anticoagulants should be carefully evaluated in collaboration with a cardiologist 1).
The diagnosis of SMH is primarily based on fundus examination and OCT. Identifying the underlying disease is essential for determining the treatment strategy.
Observation with a slit-lamp microscope and indirect ophthalmoscope is fundamental. Dark red submacular hemorrhage is identified, and the size, thickness, and freshness of the hemorrhage are evaluated. If hemorrhage persists, it is often difficult to differentiate the underlying disease.
Used for longitudinal monitoring of treatment effects.
Newer generation devices such as SD-OCT and SS-OCT are recommended 10). Intraoperative OCT improves the accuracy and safety of subretinal injection 5).
Treatment of SMH is individualized based on the size, thickness, and duration of hemorrhage, underlying etiology, and the patient’s general condition 5). No established guidelines exist, and most evidence is based on case series and retrospective studies.
The timing of treatment intervention significantly affects visual prognosis. Animal experiments have shown that irreversible photoreceptor damage begins within 24 hours. Clinically, intervention within 7 to 14 days of onset is recommended; beyond 14 days, blood clot organization and photoreceptor loss limit visual recovery 5).
In traumatic SMH, pneumatic displacement within 30 days of onset is reported to provide optimal displacement effect and visual improvement 5). Motta et al. (2023) administered intravitreal tPA 0.25 μg + C₃F₈ 0.3 mL within 48 hours of injury and achieved complete visual recovery to 6/5 (equivalent to 1.2) at 3 months 2).
Iyer et al. (2021) reported a case of large, thick SMH associated with nAMD in which anti-VEGF monotherapy improved visual acuity from 20/400 to 20/30 and remained stable for 10 years 6). In another case of POHS-related SMH, visual acuity of 20/20 was maintained after 30 years with observation and intermittent anti-VEGF therapy6).
A subanalysis of the CATT trial showed that even in nAMD patients with hemorrhage occupying more than 50% of the lesion, anti-VEGF monotherapy achieved similar visual and morphological improvements as patients with less blood 6).
This method involves injecting an expansile gas (SF₆ or C₃F₈) into the vitreous and using prone positioning to displace the hemorrhage from the fovea5).
Indications: Small to medium-sized fresh SMH.
Procedure: After anterior chamber paracentesis of 0.3–0.5 mL, inject pure SF₆ 0.3 mL or 20% SF₆ intravitreally. Maintain face-down positioning for 5–7 days 5).
tPA combination: Intravitreal tPA 25–50 μg is often combined. However, whether intravitreally administered tPA passes through an intact retina to reach the subretinal space is debated.
Anti-VEGF combination: In SMH associated with nAMD, studies comparing anti-VEGF therapy with surgical hematoma displacement are ongoing 9).
For large SMH or thick hemorrhage, vitrectomy is performed 5).
Procedure: After 23G or 25G vitrectomy, inject tPA subretinally using a 38–41G needle from outside the macula (near the inferotemporal vascular arcade) 5). Typical tPA concentration is 100–500 μg/mL, with a total dose of 25–50 μg.
Combination therapy: A triple therapy is performed in which a “cocktail” of tPA + anti-VEGF + filtered air is injected into the subretinal space 5). The air helps move the hemorrhage downward, protects the fovea, and prolongs the contact time between the drug and the clot.
Tamponade: SF₆ or C₃F₈ gas is used. C₃F₈ is suitable for large hemorrhages. Silicone oil is used for retinectomy or complex cases 5)8).
In a randomized controlled trial (90 eyes), visual improvement at 6 months was similar between the vitrectomy + subretinal tPA + anti-VEGF group and the pneumatic displacement + intravitreal tPA + anti-VEGF group (vitrectomy group +16.8 letters vs PD group +16.4 letters). However, the vitrectomy group had a lower rebleeding rate (5% vs 15.8%).
In Japan, intravitreal gas injection or vitrectomy is performed for displacement of submacular hemorrhage 12). Intravitreal injection of anti-VEGF drugs and tPA (off-label use) may also be combined, but further discussion on indications is needed 12).
For subfoveal hemorrhage due to retinal arterial macroaneurysm, prompt removal of the hemorrhage is important for visual recovery. Intravitreal gas injection (SF₆ or C₃F₈, 0.2–0.8 mL) with prone positioning for 1–2 weeks, or vitrectomy for removal, is performed. For transient ocular hypertension, anterior chamber paracentesis or intravenous glycerin (Glycerol®) is used. In eyes with incomplete posterior vitreous detachment, gas injection carries risks of retinal tear, retinal detachment, and vitreous hemorrhage.
Subpigment epithelial hemorrhage is difficult to displace with gas tamponade and also difficult to remove by vitrectomy. Therefore, treating the underlying cause to prevent further bleeding is often a realistic option.
QWhat happens if treatment is delayed?
A
If more than 14 days have passed since onset, the clot becomes organized and irreversible damage to photoreceptors progresses, limiting visual recovery 5). However, even in chronic cases, reports have shown some functional improvement with subretinal endoscopic surgery 8) or retinectomy combined with RPE patch transplantation 5).
6. Pathophysiology and detailed mechanism of onset
Hemoglobin from lysed red blood cells releases hemosiderin, which produces reactive oxygen species via the Fenton reaction. This induces oxidative stress and apoptosis in photoreceptors 7). Ferritin, the final metabolite of iron, is toxic to the retina and promotes destruction of photoreceptors and retinal pigment epithelium5).
The blood clot interposed between the RPE and the neurosensory retina blocks bidirectional nutrient exchange. Nutrient supply from the RPE to photoreceptors is interrupted, leading to metabolic impairment and degeneration of photoreceptors 7).
Contraction of the fibrin clot exerts shear forces on the photoreceptor outer segments, causing detachment and degeneration of the outer segments.
In animal experiments, photoreceptor edema appeared 1 hour after subretinal autologous blood injection, and irreversible photoreceptor damage was observed at 24 hours. After 7 days, severe karyolysis of the outer nuclear layer occurred. Additionally, tight interdigitation between fibrin and photoreceptor outer segments was confirmed after 25 minutes, suggesting that mechanical and chemical damage to the photoreceptor layer occurs even when the hemorrhage is thin 6).
These damages begin within 24 hours of onset and lead to marked destruction of the outer retina within 7 days 8). The rationale for early intervention is based on these experimental findings.
However, clinically, not all cases progress to irreversible damage. When the hemorrhage is thin or the choroidal neovascularization is away from the fovea, visual acuity may recover with anti-VEGF therapy alone 6). Visual prognosis depends largely on the presence of choroidal neovascularization, the thickness and size of the hemorrhage, and whether the underlying disease is nAMD.
7. Latest research and future perspectives (reports at research stage)
The TIGER trial is a pan-European phase III randomized controlled trial targeting foveal SMH associated with nAMD. It compares standard anti-VEGF therapy with a surgical approach combining vitrectomy, subretinal tPA, and intravitreal gas 9). The results are anticipated as a study that fills the lack of large-scale prospective research in SMH management.
Chauhan et al. (2024) administered 60 μg of tPA, exceeding the conventional safe dose (25–50 μg), via subretinal injection at two sites using a 23G soft tip for traumatic extensive SMH (nearly total hemorrhagic retinal detachment) 3). One month postoperatively, visual acuity improved from hand motion to 20/80, and after silicone oil removal, to 20/60. No signs of retinal toxicity were observed.
Yokoyama et al. (2022) performed subretinal endoscopic surgery (SES) for chronic SMH (onset more than 3 weeks prior) due to polypoidal choroidal vasculopathy8). A 25G three-port system was inserted from the sclera into the subretinal space, and under endoscopic visualization, the SMH was directly removed and the polypoidal choroidal vasculopathy lesions (polyps and BVN) were coagulated. The SMH completely resolved, and macular retinal sensitivity improved. No anti-VEGF therapy was needed for 2 years postoperatively. Endoscopy directly confirmed that polyps and BVN were located within the RPE.
Pappas et al. (2021) reported a novel technique using foam evolution theory and the principle of biphasic absorption, in which multiple micro air bubbles are injected subretinally during vitrectomy to mobilize SMH along with tPA 7). In a 72-year-old male with nAMD-related SMH (3.5 disc diameters), more than 90% of the hemorrhage resolved after 2 weeks, and visual acuity improved from light perception to 20/70 at 5 months.
Iftikhar et al. (2025) administered alternating faricimab and aflibercept every 2 weeks after vitrectomy plus subretinal tPA for acute large SMH in a monocular nAMD patient 4). Visual acuity improved from 20/400 to 20/70 at 5 months. The possibility of individualized biweekly anti-VEGF administration in refractory cases was suggested.
Weber C, Bertelsmann M, Kiy Z, et al. Antiplatelet and anticoagulant therapy in patients with submacular hemorrhage caused by neovascular age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol. 2023;261:1413-1421.
Motta L, Ripa M, Theodorakis K, et al. A case of traumatic submacular hemorrhage treated with tPA and pneumatic displacement. Case Rep Ophthalmol. 2023;14:596-601.
Chauhan K, Narayanan R. A novel technique for extensive submacular hemorrhage using high-dose tissue plasminogen activator. Indian J Ophthalmol. 2024;72:921.
Iftikhar M, Hsu ST, Vajzovic L, Hadziahmetovic M. Acute submacular hemorrhage resulting from neovascular age-related macular degeneration in a monocular patient. J VitreoRetinal Dis. 2025;9(2):262-265.
Sarna M, Waszczykowska A. Submacular hemorrhage management: evolving strategies from pharmacologic displacement to surgical intervention. J Clin Med. 2026;15:469.
Iyer PG, Brooks HL Jr, Flynn HW Jr. Long-term favorable visual outcomes in patients with large submacular hemorrhage. Clin Ophthalmol. 2021;15:1189-1192.
Pappas G, Vidakis N, Petousis M, et al. An innovatory surgical technique for submacular hemorrhage displacement by means of a bioengineering perspective. Vision. 2021;5:23.
Yokoyama S, Kaga T, Kojima T, et al. Treatment of old submacular hemorrhage by subretinal endoscopic surgery and intraoperative subretinal endoscopic findings. Am J Ophthalmol Case Rep. 2022;25:101393.
Jackson TL, Bunce C, Desai R, et al. Vitrectomy, subretinal tissue plasminogen activator and intravitreal gas for submacular haemorrhage secondary to exudative age-related macular degeneration (TIGER): study protocol for a phase 3, pan-European, two-group, non-commercial, active-control, observer-masked, superiority, randomised controlled surgical trial. Trials. 2022;23(1):99. doi:10.1186/s13063-021-05966-3. PMID:35101110; PMCID:PMC8805308.