|Home | About | Journals | Submit | Contact Us | Français|
Valsalva retinopathy was first described in 1972 by Thomas Duane as “a particular form of retinopathy, pre‐retinal and haemorrhagic in nature, secondary to a sudden increase in intrathoracic pressure.” Incompetent or no valves in the venous system of head and neck allow direct transmission of intrathoracic or intra‐abdominal pressure into the head and neck. Sudden elevation of venous pressure may cause a decompensation in the retinal capillary bed, with subinternal limiting membrane haemorrhages (Hg) that rarely may break through and become subhyloid or intravitreal 1. We report the histological findings of internal limiting membrane (ILM) peel in a case of Valsalva retinopathy.
A 41‐year‐old Caucasian male was referred to the vitreoretinal services with a spontaneous and sudden loss of vision in left eye for 3 weeks. There was no history of trauma or violent exertion but the patient had hay fever and had frequent episodes of sneezing. On examination his vision was 6/6 and hand movements in right and left eyes, respectively. Anterior segment examination was normal. Dilated fundoscopy revealed a dense vitreous haemorrhage in the left eye and normal fundus appearance in the right eye. Ultrasound echography revealed a posterior vitreous detachment, vitreous haemorrhage and a macular elevation in the left eye. Systemic examination was normal. Laboratory investigations showed normal complete blood count, prothrombin time and activated partial thromboplastin time. Blood pressure and urine analysis were normal. After discussions with the patient, a decision was made to perform a 20‐gauge three‐port pars plana vitrectomy. Intraoperatively, after core vitrectomy and removal of the vitreous haemorrhage, a sub‐ILM haemorrhage typical of Valsalva retinopathy was noted. ILM peel was performed without the assistance of dye, and the excised tissue was processed for histopathological assessment. Postoperatively, 3 months the patient's vision had improved to 6/6 unaided, with no secondary complications.
Histological examination of the excised tissue (fig 1A1A)) revealed that it contained convoluted ILM. The vitreous (smooth) surface of the ILM was free of cells but there was a cellular component in the specimen, and this component was on the retinal side (undulated surface) of the ILM (fig 1B1B).). The cellular component included a prominent multilayer aggregate of cells that was immunoreactive for cytokeratin 7 (fig 11 C), which is a marker of transdifferentiated retinal pigment epithelial (RPE) cells 4. These cells were negative for glial and neural markers. Nevertheless, glial and neural elements were present elsewhere in the specimen, again on the retinal rather than the vitreous surface of the ILM (fig 1D,E1D,E).). CD68pg‐positive macrophages were scattered through the specimen and there was also scattered pigment that was partly intracellular and partly extracellular. Perls (Prussian blue) staining confirmed that the pigment was a mix of melanin and haemosiderin (fig 1F1F).
The plane of retinal Hg in Valsalva retinopathy is sometimes difficult to determine, especially in the absence of PVD. Ocular coherence tomography (OCT) has been used to determine the exact location when the vitreous medium is clear and it is generally agreed that it is sub‐ILM in location. Following core and posterior vitrectomy, we could confirm that a sub‐ILM haemorrhage was present. The Hg was possibly a consequence of the patient's hay fever‐related sneezing that is thought to occur from a sudden rise in the intrathoracic pressure caused by a forceful exhalation against a closed glottis.
Therapeutic options in Valsalva retinopathy include conservative management, surgery (vitrectomy) and laser membranotomy. Epiretinal membrane (ERM) formation with ILM wrinkling has been reported 10 months after ND‐YAG membranotomy of Valsalva Hg.3 Histological examination of surgically removed ILM revealed the presence of haemosiderin within macrophages on the retinal side of the ILM and a fine glial ERM, resembling glial proliferation on the vitreous surface of the ILM 3. Our case also revealed haemosiderin on the retinal surface of the ILM, again confirming the sub‐ILM location of the haemorrhage, but instead of an ERM there was a mixed‐cell‐type proliferation on the retinal surface of the ILM. The sub‐ILM cells included trans‐differentiated RPE cells, and hence the proliferation had the histological appearances of a proliferative vitreoretinopathy (PVR)‐type membrane “beneath” the ILM. Presumably, the RPE cells had been attracted to this location by the sub‐ILM blood, since it is well established that RPE cells migrate to various blood components and can move through intact retina 5.
Intraretinal pathology in PVR is a well‐recognised pathological event, but usually the retinal changes resemble gliosis 6. Our case suggests that focal RPE proliferation, similar to that seen in PVR epiretinal membranes, can occur within the neuroretina, and specifically in a sub‐ILM location, by transmigrated RPE cells as a response to intraretinal haemorrhage. Such proliferation might prevent complete visual recovery after reabsorption of the retinal haemorrhage and justify early surgical intervention instead of routine observation or laser membranotomy.