Abstract
Purpose: :
AMD was until recently untreatable, but several therapies have been developed, the most successful being the anti-VEGF Fab fragment ranibizumab (Lucentis). Treatment stabilises or improves visual acuity in most patients but a few continue to lose vision. Much of the genetic basis of AMD susceptibility is known. In a previous pilot study we tested whether genotypes for 3 SNPs associated with AMD could also predict response to treatment. We reported a consistent non-significant trend towards a better outcome with the high AMD risk allele for all 3. Here we report on findings in a second AMD cohort
Methods: :
202 AMD patients being treated with ranibizumab (0.5mg with a loading phase of 3 injections followed by treatment as required in a maintenance phase) were recruited, a blood sample was taken and visual acuity in ETDRS letters measured pre-treatment and again monthly thereafter for 6 months. Patients treated previously were excluded. Smoking history, age, sex and number of injections were recorded. Genotypes were obtained for rs11200638 near HTRA1, rs1061170 in CFH and rs1413711 in VEGF. Data were analysed with treatment response as both a binary (>5 letters improvement at 6 months vs ≤ 5 letter gain) and quantitative trait
Results: :
With response as a binary trait, smokers/ex-smokers responded better (75/55% >5 letters gained) than those who had never smoked (39%, p=0.01). The HTRA1 high risk allele was also a significant predictor of poor response as a binary trait, but this contradicts the trend we observed in the previous cohort. Pooled analysis of both cohorts showed no evidence for a significant effect. With response as a quantitative trait smokers and ex-smokers again showed better response (8.8/7.5 letters) than those who had never smoked (3.4 letters, p=0.02), but genotypes were not significantly associated with treatment outcome
Conclusions: :
After analysing a second larger AMD cohort we find no evidence that CFH, HTRA1/ARMS2 or VEGF SNP genotypes influence response to treatment with ranibizumab, but a history of smoking predicts a positive outcome. This does not exclude the possibility of genetic moderators of response to ranibizumab AMD therapy elsewhere in the genome, but a GWAS in a larger patient group would be required to detect such an effect. Ability to predict which patients will respond to treatment could save sight by redirecting poor responders to other therapies, as well as saving the cost and distress of unsuccessful treatment
Keywords: age-related macular degeneration • clinical (human) or epidemiologic studies: outcomes/complications • genetics