Verteporfin photodynamic therapy for neovascular age-related ...

68 Discussion of resultsThe VPDT study protocol specified that BCVA should be measured at 3-monthly intervals.We could not express the outcomes of treatment with VPDT in the study with the outcomesof treatment in the systematic review because of the loss to follow-up. Instead, we estimatedthe change in BCVA over time (modelling baseline BCVA as a covariate) and observed that at12 months it was not dissimilar to that reported in the key licensing trials, although we may haveunderestimated BCVA deterioration at 1 year because of attrition. Notably, for the participants inthe VPDT study who would have qualified for entry into the TAP study, the trajectory of changein BCVA was highly similar to that observed in the TAP trials. 4,30 This finding may be interpretedas providing some corroboration of the clinical effectiveness of the technology in terms of betterpreserved BCVA in the treated eye.Factors not related to treatment that influenced change in visualacuity over timeOverall in the VPDT cohort study, BCVA in the treated eye declined over time. This findingwas consistent with the outcomes observed in the key licensing trials. However, the size of thecohort study allowed us to examine with high statistical power the influence of a number ofbaseline covariates on the change in BCVA. Although treatment was associated with a lower rateof decline of BCVA, several other factors influenced change in BCVA. Those that contributedto deterioration included older age, poorer BCVA at commencement of treatment and being acurrent or ex-smoker. One factor was associated with a better outcome and this was having afellow eye with better vision than the treated eye.Although our findings are consistent with clinical wisdom, experience and intuition, this is thefirst study to quantify the effects of these factors on visual change. For example, the eyes of olderparticipants tended to deteriorate faster than those of younger participants with better BCVA.The magnitudes of the interactions between smoking status and vision in the fellow eye withtime, estimated here for the first time, are quite striking. Also, our finding of a better outcomewhen the treated eye is the better-seeing eye is important and has been overlooked in previousstudies. This finding is consistent with a previous report that suggested that an eye with nAMDdoes not achieve its full visual potential unless it is the better-seeing eye, 51 and with previousfindings of improvements in adult amblyopic eyes when vision in the fellow eye is lost. 60 Themodest size of the effect, and its consistency across conditions, suggests that it may arise from ashift in decision criterion. 61The number of verteporfin photodynamic therapy treatments administeredA striking feature of the VPDT cohort study was the much smaller number of treatments thatwere administered, an average of 2.3 and 0.4 treatments respectively in years 1 and 2, even thoughit was specified that ophthalmologists should retreat as in the TAP trials. By comparison, in theTAP trials 30 an average of 3.4 treatments were administered in the first year and 2.2 in the secondyear; these frequencies are not surprising because treatment was mandated if leakage was judgedto be present on an FA at the 3-monthly review visits. It is also notable that in the TAP trials theFAs had to be performed to standardised protocols and scrutinised by an accredited angiogramreadingcentre, thus ensuring consistency of interpretation for retreatment decision-making.Our findings suggest that ophthalmologists do not adhere to treatment algorithms that are usedin key licensing trials and that decisions to treat are influenced more by subsequent experiencegained from treating large numbers of patients. Thus, a matter of increasing unease is theapplicability to routine practice of the treatment protocols specified in pivotal licensing trialsand subsequent marketing authorisations. The need for regular review combined with invasiveand time-consuming imaging procedures followed by administration of treatment can imposesignificant burdens on already stretched health-care systems. Traditionally, these factors havenot been considered when implementing new therapies into routine practice. With VPDT,however, the question was raised about whether or not efficacy might be diminished if treatment