Von Hippel–Lindau (VHL) disease is an autosomal dominant, cancer syndrome that is associated with cysts in multiple abdominal organs, pheochromocytoma, renal cell carcinoma and haemangiomas of the central nervous system and retina. It is well known that the retinal haemangiomas can lead to exudative or tractional retinal detachments, and these lesions have been treated with photocoagulation, diathermy, cryotherapy and brachytherapy (Kreusel et al. 1998). Vitrectomy has also been an optional treatment (McDonald et al. 1996); however, in some cases, a recurrence of the subretinal exudation and traction developed after the vitrectomy if the blood vessels continue to leak. These findings may lead to proliferative vitreoretinopathy (PVR).
Vascular endothelial growth factor (VEGF) is up-regulated in VHL disease by mutations in the VHL tumour suppressor gene, and this elevated level of VEGF probably causes an increase in the vascular permeability of the retinal haemangiomas (Los et al. 1997). An extravascular leakage of various growth factors might also be increased and contribute to the recurrence of the PVR.
This letter is based on two similar cases treated with intravitreal bevacizumab combined with vitrectomy, which prevented a recurrence of the PVR associated with VHL disease.
A 44-year-old woman with neurosurgical-established VHL disease was referred to our clinic for a recurrence of PVR in her left eye. Although she had had pars plana vitrectomy (PPV), pars plana lensectomy and encircling scleral buckling for the PVR, a second PPV with silicone oil (SO) tamponade had to be performed for a recurrence of the PVR 1 month later. Two weeks after these latter treatments, the PVR recurred, and the best corrected visual acuity (BCVA) was reduced to 20/100. At the initial examination in our clinic, haemangiomas were seen at the mid-peripheral retina. A proliferative membrane and a tractional retinal detachment were observed mainly in the temporal retina (Fig. 1). The PVR was graded as type CP 2–12. The surgery consisted of the removal of the SO, followed by PPV, endolaser, encircling scleral buckling and SO tamponade. At the end of the surgery, 1.25 mg of bevacizumab was injected intravitreally. The BCVA was 20/100 3 months after the operation without a recurrence of PVR. The SO was removed at this time (Fig. 2). Although the BCVA at 15 months improved to 20/50, the sizes of the haemangiomas were not reduced.
The other case with a 43-year-old man also received intravitreal injection of bevacizumab in the same manner and followed over 15 months. Although the visual acuity improved and there was no sign of recurrence of PVR, the sizes of the haemangiomas were not changed same as the former case.
Dahr et al. (2007) used intravitreal pegaptanib, a pegylated anti-VEGF, in patients with juxtapapillary or large peripheral angiomas secondary to VHL disease. Wong et al. (2008) also reported the effects of intravitreal injection of ranibizumab, a humanized recombinant monoclonal antibody fragment targeted against VEGF-A, for retinal capillary haemangioblastomas associated with VHL disease. Even though the patients received frequent injections, a regression of the tumour size or the exudation could not be seen. The authors concluded that an intravitreal anti-VEGF treatment as a monotherapy has only limited success for this disease.
Although vitrectomy alone might have helped the prevention of further exudation, only mild success has been reported in the use of vitrectomy in patients with VHL disease (McDonald et al. 1996). To increase the success rates of vitrectomy, cryotherapy, photocoagulation or brachytherapy might be useful as an adjunct.
Although the size of the haemangiomas was not reduced by the injection of bevacizumab in both of our cases, we believe that using bevacizumab as an adjunct with vitrectomy might suppress the exudation from the haemangiomas and decrease the incidence of a recurrence of a PVR, and thus, increase the success rate of vitrectomy for PVR in patients with VHL disease. But we have to mention that the effect of anti-VEGF drugs has a limited time of action and we could not conclude why recurrences were not observed over the 15 months in both of our cases.
We were not able to access the exact concentration of bevacizumab in our two cases. In silicone oil-filled eye, the concentration of the drug must be higher compared to the regular condition. There was, however, no sign of clinical toxicity of bevacizumab injected into SO for advanced proliferative diabetic retinopathy in a recent report. We think that injection of bevacizumab into SO was also tolerable in our cases.