To our knowledge this is the first documentation of increased vitreous VEGF levels in the setting of FAP. Some studies have suggested that the haemorrhages associated with FAP might be secondary to a weakening of the vessel wall as it is infiltrated with amyloid.3
Retinal neovascularisation, although previously reported, represents an uncommon complication of FAP. In a cohort of 37 patients with FAP caused by a Val30Met mutation with long term follow up, only one developed retinal neovascularisation.4
The levels of VEGF found within the vitreous samples of our FAP patients are raised to a level similar to that found in patients with inactive proliferative diabetic retinopathy (mean 698.2 pg/ml), and consistent with the presence of ischaemia within the posterior segment of these eyes.5
Recent studies of undiluted vitreous specimens from patients with non‐ischaemic ocular diseases, such as macular holes and epiretinal membranes in the absence of proliferative vitreoretinopathy, have found vitreous VEGF levels ranging from undetectable (defined as less than 62.4 pg/ml) to 109 pg/ml.5,6
Thus the markedly raised VEGF levels in our FAP patients suggest that VEGF upregulation and neovascularisation may play a role in the production of haemorrhage. Perhaps amyloid deposition in the retinal vessel walls serves as a barrier to oxygen delivery in the surrounding tissues, thus leading to hypoxia induced upregulation of VEGF.
Isolated Glu54Gly transthyretin mutation therefore represents a very rare TTR mutation. The Glu54Gly transthyretin mutation has been reported twice, once in isolation and once associated with a Gly6Ser transthyretin mutation.7,8
In our two FAP patients, and in the previously reported patient with an isolated Glu54Gly transthyretin mutation, the symptoms began in the late 20s. This is significantly earlier than the age of symptom onset in the reported patients with both the Glu54Gly mutation and the Gly6Ser transthyretin mutation, where vitreous involvement occurred in the middle 40s.7,8
The Gly6Ser transthyretin mutation itself has been described previously in a form of euthyroid hyperthyroxinaemia which is non‐amyloidogenic (no amyloid deposition occurs).9
Thus it had been postulated that the effect of Gly6Ser transthyretin mutation on the TTR structure does not contribute to TTR aggregation and therefore to amyloid deposition. Given the later onset of vitreous involvement in patients with both the Glu54Gly transthyretin mutation and the Gly6Ser mutation, it is possible that the Gly6Ser mutation may even be protective.7
Our two patients with an early onset phenotype in the absence of the Gly6Ser transthyretin mutation add evidence to this hypothesis.
The liver transplants which each patient received corresponded to a subjective decline in vision in both. This was probably secondary to continued amyloid deposition within the vitreous, a phenomenon which has been described previously.10
Although more than 90% of the mutant TTR is produced in the liver, and after liver transplantation its levels in the serum can be reduced to less than 1% of pretransplant levels, it continues to be produced within the eye by the retinal pigment epithelium, thus allowing continued ocular involvement.11
Both patients had chronic visual complaints of floaters and blurred vision several years before their liver transplants, and patient 2 had documented vitreous amyloidosis before transplantation, thus reducing the likelihood that the liver transplant itself was an aetiological factor in their early onset disease.
We report an aggressive FAP phenotype of early, severe vitreous involvement with retinal neovascularisation in the setting of a rare isolated Glu54Gly transthyretin mutation; and we provide what we believe to be the first description of an association between FAP and raised vitreous levels of VEGF, indicating an ischaemic aetiology for the neovascularisation seen in these patients.