Premacular or preretinal hemorrhage and subhyaloidal hemorrhage were the commonly used synonyms for subhyaloidal and sub-ILM hemorrhages, Hemorrhage beneath the ILM is located within the neuroretina and the anatomically correct description should be sub-ILM hemorrhage4. The figures in some researches published before may be incorrect according to the anatomy definition and may be confused between subhyaloidal and sub-ILM hemorrhage.
Sub-ILM hemorrhages are caused by multiple reasons including Valsalva retinopathy, blood dyscrasias, Terson syndrome, and ocular trauma3 4. In this study, 4 cases were Valsalva retinopathy, 2 case was macroaneurysm, 1 case was diabetic retinopathy. Blood in sub-ILM hemorrhages spontaneously clears significantly slower than that of sub-hyaloid hemorrhages and sub-hyaloid hemorrhages can often be observed for spontaneous improvement. The catabolites in sub-ILM hemorrhage may lead to retinal toxicity and proliferative vitreoretinopathy3–5. In conclusion, the patients of sub-ILM hemorrhage with earlier intervention have better visual outcomes5.
Differentiating between the two types of hemorrhages clinically is a challenge. Some signs in fundus photography may be the clue to sub-ILM hemorrhage, Sub-ILM hemorrhages appear as sharply demarcated, round hemorrhages at the posterior pole and a glistening light reflex of ILM can often be observed2 3. Another sing is “double ring” with the “inner ring” caused by the sub-ILM hemorrhage and the “outer ring” caused by the sub-hyaloid hemorrhage13 18 19,but this may not be true in some cases (as case 7). Some authors emphasized “perimacular folds” may be seen with large sub-ILM hemorrhage12 and “Arcus retinalis" might be a novel clinical marker of sub-ILM hemorrhage20. Until now the only method to confirm the sub-ILM hemorrhage still remains intraoperative staining of the membrane overlying the hemorrhage and pathological approval10. OCT remains the non-invasive mainstay for accurate diagnosis1–4 10 11 20.
Normally ILM cannot be distinguished on OCT because of its thickness and anatomical characteristics. Foveal ILM has a thickness of approximately 100 nm, whereas the parafoveal ILM has a thickness of up to 3 µm25. The OCT features of sub-ILM hemorrhage is mentioned above. If there is coexistence of sub-ILM with sub-hyaloid hemorrhage or the posterior vitreous detachment is visible, the diagnosis is easy, otherwise OCT as a tool for diagnosis of sub-ILM hemorrhage is still difficult3–5.
The OCT features of sub-ILM hemorrhage after laser membranotomy has three main aspects, that is the change of ILM, the macular fovea and the outer retina.
The ILM is obviously thicken in sub-ILM hemorrhage. The ILM is straight and stiff in most cases but also can be undulant (case 7) after laser. The ILM can reattach to retina, or do not attach and form a cavity between the ILM and the retina after membranotomy (Fig. 1–3).
In this study, the ILM did not attach to retina and had the “cavity-structure” in 4 cases. This phenomenon was not less common. Zhou et al9 had demonstrate some reasons for this complication, such as large perforating puncture of the ILM and liquefaction of vitreous9. In our study, the average age of ILM fallen back group was 63 years old and the average age of ILM not fallen back was 30.5 years old, so we thought vitreous liquefaction might be the reason of ILM fallen back instead of “cavity-structure” formation.
The macular fovea might be normal contour after laser. But MH and epiretinal membrane could also be seen2 5 7 8 13 15 21 24. In this study, the macular fovea have normal contour in 3 cases, MH in 1 case and epiretinal membrane in 1 case. The macular fovea are covered by insufficient drainage of sub-ILM hemorrhage in 2 case. Several causes may contribute to the formation of MH. It could be induced by hemorrhage breaking through the neurosensory layers, traction of ILM to the fovea or toxicity of the long lasting blood 21 24. The primary closure rate of MH after sub-ILM hemorrhage seems to be low (57%) when compared with the rate for idiopathic MH (> 90%)14. In this the study (case 1) the MH is unsealed even after vitrectomy and silicone oil tamponade.
There were “peg-like” structure or “patchy-like” structure in outer retina in 4 cases of this study. Kumar V et al20 described a retinal yellowish-white arc (“arcus retinalis”) corresponding to the outer margin of sub-ILM hemorrhage. The “peg-like” structures on OCT are the result of localized traction on the retina by the ILM. Another reason may be exudates or hemorrhage in the outer retina derived from the sub-ILM hemorrhage (“patchy-like”) (Fig. 1).
Other OCT changes of sub-ILM hemorrhage after laser include hyperreflective spots in vitreous cavity and on the retina side of ILM (residual hemorrhage), the undulant surface of the retinal nerve fiber layer and retinoschisis at the attachment of ILM to retina (Fig. 2).
Laser is an effective method to treat sub-ILM hemorrhage but some attention should be paid to. Firstly, the best time of laser is no more than 21 days from the onset of hemorrhage7 9. The size and thickness (Cushing effect to prevent damage to the retina) of hemorrhage should be taken into count. Too small or too thin hemorrhage should avoid laser. Do not try to do laser If the lower boundary of the sub-ILM hemorrhage is near the central area of the macula. If the patient has frequent eye movement the laser should be avoided. Secondly, if sub-ILM combined with PDR, the laser membranotomy must be combined with photocoagulation in the lower periphery retina. Lastly, there is the possibility of immediate drainage or delayed drainage (1 day to 1 week). The hemorrhage with obvious fluid level is easy to be drained immediately, while hemorrhage without obvious fluid level is not easy to be drained. Even if the hemorrhage is drained into the vitreous cavity, the absorption is sometimes very slow which still affect the recovery of visual acuity and fundus observation (such as case 6 and 7) and vitrectomy may be applied.
The shortcoming of this study is limited cases and short time of follow-up, also there is no control groups of other types of laser or surgery. Prospective and randomized trials with large patient numbers will be required to validate the effectiveness and safety of 577nm laser as a treatment modality for sub-ILM hemorrhage.