Sir,

We have read with considerable interest the paper by Konstas et al.1 The authors concluded that latanoprost had a ‘similar safety than timolol as monotherapy in the treatment of exfoliation glaucoma (XFG)’. However, according to the paper's Table 3, there were significantly more cases of conjunctival hyperaemia in patients receiving latanoprost compared with those receiving timolol regimen. This can be reasonably explained because prostaglandins including latanoprost, as potent vasodilators, act directly on vascular smooth muscle, attenuating responses of vasoconstrictive stimuli, and enhancing microcirculation,2 thereby leading to conjunctival hyperaemia. The latter might be further aggravated by the fact that prostaglandins can promote angiogenesis.

Moreover, in Table 4 there was one case of upper respiratory tract infection in a patient treated with latanoprost. Although this systemic side effect has not reached statistical significance when compared with the timolol group, it can be attributed to latanoprost, because it has been shown that local administration of antiglaucoma drops results in systemic absorption through the nasal mucosa,3 and prostaglandins, released by cyclooxygenase (COX)-2 (one of the key isoenzymes in the production of prostaglandins), induce immune suppression,4 thereby leading to defective elimination of pathogens. Notably, disturbances of the immune system can induce disease in virtually any portion of the eye; examples include conjunctivitis, keratitis, keratoconjunctivitis, scleritis, uveitis, optic neuritis, and orbital inflammation. Therefore, patients under latanoprost treatment should be aware of potential systemic and/or topical infective side effects.

From another extreme viewpoint, although the authors have not administered latanoprost for an extended period of time, it should be noted that prostaglandins are implicated both in cell proliferation, and inhibition of immune surveillance; therefore these agents could favour a potential malignant growth. It is already known that COX-2 overexpression enhances prostaglandin synthesis, may inhibit apoptosis, increases invasiveness of malignant cells, and can be mutagenic and tumorigenic in vitro.5, 6 Tumour-derived COX-2 and prostaglandins may play an important role in antagonising the host immune response and therefore facilitating tumour growth and spread.4 Of note, Helicobacter pylori (H. pylori) infection, involved in the pathogenesis of glaucoma,3, 7 increases COX-2 and prostaglandin expressions, which might be one of the mechanisms leading to H. pylori-induced carcinogenesis. Moreover, prostaglandins promote angiogenesis, thereby contributing to invasiveness of a variety of tumours.8 In particular, prostaglandins enhance the proliferation and invasion of malignant cells via activation of major intracellular signal transduction pathways such as PI3K/Akt. Additionally, the synthetic machinery and receptors for prostaglandins, prominently expressed by T lymphocytes in tissues at the boundary of normal tissue with tumour cells, may play a central role in prostanoid-driven tumorigenesis. In view of all the above-mentioned data, ophthalmologists should consider and be aware of prostaglandin-induced infective side effects, and of their long-term, although still speculative and rare, oncogenic properties.