Intravitreal Hemoglobin Spherulosis: A Rare Complication of Subretinal Hemorrhage

Hemoglobin spherulosis, a rare histological finding, represents a spherical degradation product of hemoglobin. It has been described in a single report as a vitreous hemorrhage originating from the subretinal space. We describe an immunosuppressed patient with chronic myelogenous leukemia (CML) and associated thrombocytopenia, with bilateral white intraretinal and subretinal hemorrhages, along with an overlying vitreous hemorrhage. Cytologic examination of a vitrectomy specimen demonstrated hemoglobin spherulosis. This represents only the second report of this phenomenon.

Three unusual sequelae of longstanding vitreous hemorrhage are cholesterolosis bulbi (synchisis scintillans), the formation of tubular vitreous cylinders, and hemoglobin spherulosis.¹ The last entity represents a spheroidal degradation product of erythrocytes and was first described by Grossniklaus et al² in an eye with a vitreous hemorrhage secondary to an idiopathic choroidal neovascular membrane. Spherulosis was so named because of its histologic resemblance to myospherulosis, an inflammatory condition characterized by saclike clusters of degenerating erythrocytes that have been physically altered by exposure to a foreign substance — typically an antibiotic-containing ointment.^3-5

The purpose of this article is to document the clinical and pathological correlation of hemoglobin spherulosis in an eye with bilateral white intraretinal and subretinal hemorrhages, along with an overlying vitreous hemorrhage of a patient with CML.

A 39-year-old African American man experienced decreased vision, with a greater degradation of vision in the right eye than in the left eye throughout one month. His medical history was significant for CML with associated anemia (hemoglobin level, 8.1 g/dL) and thrombocytopenia (22 × 10³/µL), though the disease was believed to be in remission. On ocular examination, his uncorrected visual acuity was counting fingers at 3 ft OD, and 20/40 OS. External and anterior segment examinations were unremarkable. Intraocular pressure was within normal limits, and there was neither iris rubeosis nor anterior chamber inflammation.

Dilated ophthalmoscopic examination in the right eye revealed prominent brown vitreous cells, which obscured the retinal details and prevented fundus photography. With indirect ophthalmoscopy, the optic disc margins appeared sharp and flat. Large white intraretinal and subretinal hemorrhagic infiltrates were visible in the mid periphery of the retina. Examination of the left eye was similar, though there was less prominent overlying vitreous hemorrhage (Figure 1).

The differential diagnosis included bilateral leukemic infiltrates, endogenous endophthalmitis, and/or possible retinal opportunistic infection. The patient underwent diagnostic and therapeutic pars plana vitrectomy in his right eye, with intravitreal injection of 1.0 mg of vancomycin hydrochloride, 2.25 mg of ceftazidime, and 5.0 µg of amphotericin B. Vitreous cultures for bacteria and fungi were negative, and cytology was negative for leukemic cells. Following vitrectomy, his visual acuity improved to 20/40 OD.

Both undiluted vitreous and the vitrectomy cassette were processed using the thin-prep technique, and the slides were stained with the Papanicolau and Diff-Quik (Dade-Behring, Deerfield, Ill) methods. Cytologic examination showed fragmented erythrocytes, ranging in diameter from 2 to 4 µm. Larger (10-12 µm) dark brown globules were observed, indicative of hemoglobin spherulosis (Figure 2). Rare inflammatory cells were identified. Neither tumor cells nor microorganisms were observed. Because of the relatively small amount of material, no additional studies (special stains and electron microscopy) could be performed.

McClatchie et al^6,7 in 1969 reported several East African patients with inflammatory nodules involving muscular and subcutaneous tissues. Histologically, the lesions represented granulomatous inflammation surrounding "bags" of spherules, each spherule being slightly larger than an erythrocyte. This condition was given the descriptive term myospherulosis.

Eight years later, Kyriakos³ used the term to describe a histologically similar inflammatory condition of the nose, paranasal sinuses, and middle ear. His work, as well as the work of of subsequent investigators^4,5 demonstrated the contents of the sac-like structures to be erythrocytes altered by a foreign substance, typically an antibiotic-containing ointment. Today, myospherulosis is often a surgical complication and has been described in numerous tissues, including the upper eyelid⁸ and caruncle.⁹

In 1988, Grossniklaus et al² described a unique hemoglobin degradation product recovered from a vitreous hemorrhage in an eye with an idiopathic choroidal neovascular membrane. They named the condition hemoglobin spherulosis because of its histologic resemblance to myospherulosis. The mechanism of the spherule formation is unknown, although they hypothesized that the initial subretinal location of the blood may have played a role.

Gass^10,11 has described a yellow-brown material in the vitreous following hemorrhagic detachment of the retina and retinal pigment epithelium. He proposed that this material was composed of hemoglobin rather than whole blood, and he noted that the material may also stain the iris stroma, resulting in heterochromia in a lightly pigmented patient. The mechanism of anterior migration of subretinal hemoglobin may be similar to that which causes corneal blood staining following a hyphema.

Since his initial publication, Grossniklaus has noted 2 additional cases, although these were not formally reported.² Our patient, therefore, represents only the second reported case of intravitreal hemoglobin spherulosis in the literature. It is significant that in our patient, the vitreous hemorrhage also seems to have originated in the subretinal space, which may explain the pathogenesis of the spherule formation. Our patient had at least 2 risk factors for the vitreous hemorrhage, including the underlying CML, along with thrombocytopenia. With the white intraretinal and subretinal hemorrhages, it was presumed that the intraocular hemorrhages had been present for an extended period.

As discussed previously, 3 uncommon complications of longstanding vitreous hemorrhage are cholesterolosis bulbi (synchisis scintillans), the formation of tubular vitreous cylinders, and hemoglobin spherulosis.² This last entity, which may be related to degenerative changes occurring to hemoglobin as it migrates from the subretinal space into the vitreous cavity, seems to be a poorly recognized phenomenon. Further studies are needed to support this hypothesis. The full ocular significance of this phenomenon is not fully known, though it may lead to a much more prolonged spontaneous clearance of hemorrhage from the vitreous cavity. We therefore recommend that vitrectomy specimens obtained from eyes with breakthrough subretinal hemorrhages be cytologically evaluated to rule out hemoglobin spherulosis.

Unfortunately, 6 months following the pars plana vitrectomy surgery, the patient experienced a relapse of his CML. The patient died due to complications of pneumonia. An autopsy was not performed.

Supported in part by a grant from the Retina Research Foundation, Houston, Tex, and by an unrestricted grant from Research to Prevent Blindness Inc, New York, NY.

Dr Schwartz is currently with the Department of Ophthalmology at the Medical College of Virginia Campus of Virginia Commonwealth University, Richmond.

Corresponding author: William F. Mieler, MD, Cullen Eye Institute, Baylor College of Medicine, 6565 Fannin St, NC-205, Houston, TX 77030 (e-mail: wmieler@bcm.tmc.edu).