Portal de Periódicos da CAPES

Treatment patterns, visual acuity and quality-of-life outcomes of the WAVE study - A noninterventional study of ranibizumab treatment for neovascular age-related macular degeneration in Germany

2012; Wiley; Volume: 91; Issue: 6 Linguagem: Inglês

10.1111/j.1755-3768.2012.02493.x

1755-3768

Robert P. Finger, Peter Wiedemann, Francisca Blumhagen, Karin Pohl, Frank G. Holz,

Retinal and Optic Conditions

Purpose: To evaluate effectiveness, tolerability and safety of repeated intravitreal injections of 0.5 mg ranibizumab for the treatment of neovascular age-related macular degeneration in routine medical practice in Germany. Methods: A noninterventional study with 3470 patients treated in 274 medical centres according to German guidelines, with monthly intravitreal injections of 0.5 mg ranibizumab during upload (3 months) followed by a maintenance phase (9 months) with reinjections if medically indicated. Results: Mean injection rate was 4.34 (SE = 0.05; median = 3.0). Best-corrected visual acuity (BCVA) remained stable (mean change 0.02 LogMAR, SE = 0.01, p = 0.0169) and central retinal thickness (CRT) decreased (by −78.9 μm, SE = 2.95 μm, p < 0.0001). The NEI-VFQ 25 summary score showed a positive stabilization with a mean change of 0.73 (SE = 0.37, p = 0.0501) compared with baseline. Adverse events were documented for 6.5% of the patients with 3.9% of these events being classified as serious. Conclusions: The number of administered intravitreal injections of ranibizumab over the first year of treatment was very low but still achieved a stabilization of BCVA, a reduction in CRT and maintained vision-related quality of life. The management of patients with neovascular AMD in Germany needs to be improved to achieve better treatment results. Neovascular age-related macular degeneration (nv AMD) is the leading cause of blindness and severe visual impairment in all industrialized countries. Treatment with repeated intravitreal injections of anti-VEGF agents like ranibizumab has been shown to prevent further vision loss and even lead to a visual acuity improvement in some patients (Rosenfeld et al. 2006). In Germany, unlike other countries, the treatment recommendation for ranibizumab in accordance with the German approval and the German Summary of Product Characteristics advises a 3-month upload phase of monthly injections followed by a subsequent flexible maintenance regime (Pauleikhoff et al. 2011). Visual acuity and morphological findings are recommended to be monitored at monthly intervals following the upload. In case of deteriorating vision or the presence of signs of disease activity (haemorrhage, increase in exudates, oedema or increase in lesion area; leakage on fluorescein angiography (FLA) or central intra- or subretinal/subpigmentepithelial fluid on optical coherence tomography), further injections are recommended to prevent disease progression (Pauleikhoff et al. 2011). To achieve optimal long-term treatment outcomes, efficient monitoring of patients is of utmost importance during the flexible maintenance therapy regime as this allows for repeated intravitreal injections as required to achieve optimal treatment outcomes (Meyer et al. 2008). To depict the current patterns of nv AMD treatment with ranibizumab in routine clinical practice in Germany, and to collect effectiveness and safety data as required by German regulatory authorities after marketing of a new drug, this noninterventional study was initiated. Treatment was provided according to the summary of product characteristics (i.e. package insert, as highlighted previously), and the study protocol did not specify any alterations to the 'usual use pattern'. Additionally, vision-related quality of life (VRQoL) was evaluated over the study period of 12 months to characterize the impact of ranibizumab treatment on VRQoL in nv AMD in routine clinical practice in Germany. The noninterventional study WAVE (Lucentis in Wet AMD: Evaluation of Visual Acuity and Quality of Life) was conducted in accordance with section 4 paragraph 23 subparagraph 3 of the German Drug Law. Data were prospectively recorded from 1 March 2008 to 31 March 2010 in 274 ophthalmologic practices and outpatient departments of eye hospitals throughout Germany. Only retina specialists with appropriate qualifications and experience in intravitreal injections (IVI) and management of neovascular AMD took part in the study. In Germany, AMD patients are often managed by a general, community ophthalmologist who conducts the monitoring and refers to a retina specialist for intravitreal injections. Data were only collected at visits to managing retina specialists during the observation period of this study. All patients for whom the managing retina specialist recommended therapy with 0.5 mg ranibizumab based on the diagnosis of active nv AMD were eligible for this study. Owing to the noninterventional approach of the study, no further inclusion or exclusion criteria besides the specifications in the German Summary of Product Characteristics (i.e. package insert) were applied. Similarly, no additional examinations or testing other than the clinical routines were performed owing to the noninterventional design of this study. Written informed consent was obtained from all patients included in the study. Ethical approval was obtained from the IRB of the University Hospital Leipzig as principal site and from other sites as necessary under federal and state legislation. Best-corrected visual acuity (BCVA) was documented at baseline, after upload and at the final visit at the end of the observation period, as well as in the case of necessary reinjections. Because of the noninterventional design of the study, the process of visual acuity measurement was not predetermined by a WAVE trial protocol, and BCVA was measured as routinely done in the participating centres. In Germany, most commonly Snellen number charts are used at a distance of five metres, and visual acuity is tested with best possible correction. Therefore, the values were most often documented as decimal visual acuity and transformed into LogMAR visual acuity for the analysis (Holladay 1997). Implausible values were excluded from the analysis, and the last observation carried forward principle was not applied. Baseline findings of FLA were collected and classified into predefined categories for lesion position (subfoveal and extrafoveal), type of choroidal neovascularization (predominantly classic, minimally classic and occult) and activity (yes or no) by the retina specialists. Follow-up FLA results were not documented. Optical coherence tomography (OCT) was performed according to routine medical practice at participating sites and at the managing retina specialist's decision. As OCT imaging is not part of clinical routine in Germany, the process of obtaining OCT scans was not predetermined in the WAVE observational plan. If performed, OCTs could be documented at each visit. Following the German Summary of Product Characteristics of ranibizumab in the treatment of neovascular AMD, the observational plan was subdivided into two phases: the upload phase (upload) and the maintenance phase (maintenance). In the first 3 months, ranibizumab 0.5 mg is recommended to be injected monthly according to treatment recommendations, followed by a control visit 1–4 weeks after the last injection of the upload. During maintenance, the visual acuity of the patient is recommended to be controlled each month either by a general ophthalmologist or by the treating retina specialist. The documented therapy with ranibizumab followed routine medical practice and was solely based on therapeutic necessity as assessed by the retina specialist. Data were collected at each visit by the retina specialist, at the baseline visit, monthly visits during the 3-month upload phase, a control visit 1–4 weeks after the last injection of the upload phase, as well as in case of further reinjections during maintenance and at an optional final investigation at 12 months (Table 1). The tolerability was evaluated by the incidence and the profile of adverse events (AE), serious adverse events (SAE), adverse drug reactions and serious adverse drug reactions (SADR) occurring during the observation period based on the verbatim term coded in accordance with MedDRA® Version 11.0 [Medical Dictionary for Regulatory Activities; International Federation of Pharmaceutical Manufacturers and Associations (IFPMA, Geneva, Switzerland)]. Treating physicians elicited and reported AEs and SAEs as well as satisfaction and tolerability. Vision-related quality of life was recorded using the German NEI VFQ–25 in a standardized patient interview. The NEI-VFQ-25 has been evaluated for use in AMD and has been found to be generally reliable and valid (Marella et al. 2010; Revicki et al. 2010). The statistical evaluation was carried out using basic descriptive statistical methods. Insofar as statistical procedures were used, their results are to be understood as being descriptive and not confirmatory. The p values determined from t-tests and signed-rank tests for the pre–post comparisons (baseline to end of the upload phase/end of observation in the maintenance phase) have to be seen in their entirety and their pattern. No alpha adjustment was applied in the context of multiple statistical comparisons. The statistical evaluation was carried out using sas® Version 9.2 (SAS Institute, Cary, NC, USA). In total, the data of 3470 patients were analysed. Data until the first follow-up after the upload phase were available for 3124 patients, and data until the last follow-up after 12 months were available for 2587 patients. The mean documented observation time was 375.8 (SE = standard error; SE = 2.08) days (median = 388.0 days). The majority of patients (n = 2605; 75.1%) were treatment naïve, 11.7% (n = 405) had already received previous treatments with intravitreal injections. The CNV was located subfoveally in 2926 (84.3%) and extrafoveally in 326 (9.4%) patients, with a missing classification in 6.3% of cases. Of the CNVs classified, most were occult (n = 2092, 63.9%), followed by classic (n = 801, 24.5%) and minimally classic (n = 381, 11.6%) CNVs. The baseline characteristics for all patients including ophthalmological findings and risk factors are depicted in Table 2. Mean differences in BCVA showed a significant [p < 0.0001, 95% confidence interval: (−0.10; −0.07)] improvement from baseline (mean LogMAR 0.72, SE = 0.01; median = 0.70) to first follow-up after the upload phase (mean change in LogMAR −0.09; SE = 0.01; median change −0.10). During the maintenance phase, BCVA returned to baseline level (Fig. 1). Mean difference of visual acuity between baseline and month 12 was 0.02 LogMAR [SE = 0.01, median = 0; p = 0.0169, 95% CI: (0.003; 0.035), Fig. 1]. A summary of the results is given in Table 3. Mean change in best-corrected visual acuity (LogMAR) to baseline. 'n' denotes the number of patients receiving the respective injection. Error bars represent standard error of the mean. Control after upload was performed at 4 months, and final observation was performed at 12 months. Optical coherence tomographys were available for 871 (25.1%) patients at baseline, 679 (21.7%) patients at first follow-up after upload and 545 (21.1%) patients at the final visit. CRT (μm) showed a significant mean reduction from baseline (mean = 349.4 μm; SE = 4.31 μm; median = 325.0 μm) to first follow-up after upload of −98.6 μm [SE = 5.42 μm; median = −83.0 μm; p < 0.0001, 95% CI: (−109.2; −87.9)] and to final follow-up of −78.9 μm [SE = 7.31 μm; median = −65.0 μm; p < 0.0001, 95% CI: (−93.3; −64.5), Table 3 and Fig. 2]. Mean change in central retinal thickness (μm) to baseline. 'n' denotes the number of patients for whom Optical coherence tomographys were available. Error bars represent standard error of the mean. Control after upload was performed at 4 months, and final observation was performed at 12 months. An improvement of subjective vision was reported by 46.3% (n = 1447) of patients at first follow-up after upload and by 29.5% (n = 763) at the final follow-up. Patients' VRQoL using the NEI-VFQ-25 composite score showed a significant improvement from baseline (mean = 67.05; SE = 0.36; median = 70.63) to first follow-up after upload [mean difference 3.56; SE = 0.26; median = 2.80; p < 0.0001, 95% CI: (3.05; 4.08)]. At 12 months, VRQoL showed no change compared with baseline [mean difference between baseline and 12 months: 0.73; SE = 0.37; median = 0.60; p = 0.0501, 95% CI: (−0.0002; 1.4645), Table 3 and Fig. 3]. German version of the The National Eye Institute 25-Item Visual Function Questionnaire (NEI-VFQ–25); Mean change in vision-related quality of life (summary score) to baseline, Error bars represent standard error of the mean. In total, 505 AEs were reported for 225 (6.5%) patients, of these 137 patients (3.9%) experienced 320 AEs which were classified as serious. For more than half of the patients (n = 133; 3.8%), the AEs pertained to disturbances associated with eye disorders (Table 4). The rate of recorded strokes (n = 14, 0.4%) was clearly below the frequently discussed rate from the SAILOR-study of 1.2% (Boyer et al. 2009). During the course of the study, 42 patients (1.2%) died, mainly owing to neoplastic disease (n = 6; 0.2%), cardiovascular events (n = 5; 0.1%), general deterioration of health state (n = 4; 0.1%) and multi organ failure (n = 3, 0.1%). The number of patients treated at a particular injection time point during the observation period declined considerably after the upload phase (Fig. 4). In the subsequent flexible maintenance phase, only about one-third of the patients (n = 1222; 35.2%) received further ranibizumab reinjections. The time elapsed between the last injection of the upload phase and the first injection of the maintenance phase was on average 119.6 days (SE = 1.67 days; median = 112 days). Overall, the patients received an average of 4.34 (SE = 0.05; median = 3.0) injections, 2.95 (SE = 0.01) during upload and 1.39 (SE = 0.05) reinjections in the flexible maintenance phase. The mean interval between recommendation of treatment with ranibizumab by the managing retina specialist, that is, baseline visit, and the first injection was 23.8 days (SE = 0.41 days; median = 18.0 days). Reinjections were frequently administered in treatment clusters of 'triplets' (Fig. 4). With increasing age, the total number of injections and the number of reinjections decreased and the time between the upload and the first reinjection increased (Table 5). This decrease in the number of injections and increase in the delay between upload and the first reinjection was most marked in those over the age of 80. Numbers of patients who were followed up and number of patients who received intravitreal injections during the observation period (mean months from inclusion). Control after upload was performed at 4 months, and final observation was performed at 12 months. In this noninterventional study, repeated intravitreal injections of ranibizumab led to a moderate improvement of BCVA, a reduction in CRT and an increase in vision-related quality of life after upload. Both improvements in BCVA and VRQoL could not be maintained till last follow-up at 12 months with a small average number of injections (4.3/year). A number of prospective randomized clinical trials have shown that repeated intravitreal injections maintain or even improve visual function and lead to morphological improvements in neovascular AMD (Rosenfeld et al. 2006; Boyer et al. 2007, 2009; Regillo et al. 2008). Compared with the major phase III clinical trials, the WAVE study results are far behind treatment outcomes achieved under rigorously controlled clinical trial conditions. However, the results found in WAVE compare well with the PIER study: After 12 months, BCVA was on average the same as at baseline, which means the initial BCVA gain after upload could not be maintained (Regillo et al. 2008). Comparing our results to the PrONTO study (Fung et al. 2007), in which the improvement of visual acuity remained stable after the third injection throughout the maintenance interval, the better results are likely due to the higher rate of injections (5.6 in PrONTO vs. 4.3 in our study) during the first year. Contrary to WAVE, PrONTO was a clinical phase III trial in which strict adherence to the treatment protocol including regular, standardized examinations, OCTs and predefined reinjection criteria was enforced. Other studies with individualized, flexible treatment regimens have documented a decrease in visual acuity following the initial improvement seen after the upload phase, similar to our findings (Kumar et al. 2011). Within WAVE, only about one-third of patients (n = 1222; 35.2%) received further ranibizumab reinjections. Other studies documenting the use of intravitreal ranibizumab in daily clinical practice outside highly standardized phase III clinical trials found a similarly poor comparison of efficacy outcomes (Bandukwala et al. 2010). Clinician-determined retreatment after an upload phase appears to be less effective in improving BCVA compared with randomized controlled trials (Kumar et al. 2011). There are several reasons for this, amongst which a worse visual acuity and higher age at baseline as well as a considerable delay to treatment in routine clinical settings seem to be the most important (Schalnus et al. 2010; Mariani et al. 2011). Available evidence points towards baseline BCVA predicting treatment outcomes with poor baseline visual acuity determining poor treatment outcomes (Shona et al. 2011). Against this background, the WAVE sample with a relatively poor baseline BCVA, a higher age at presentation and a considerable delay to treatment would not be expected to do as well as patients in most clinical trials. The observed low number of injections and long time interval between upload and first reinjection in the PRN phase in those over the age of 80 may be due to a number of reasons, of which the most important are likely a suboptimal treatment response in the elderly, physical difficulties attending frequent and regular follow-ups and potentially disappointment on the patients' and/or providers' behalf about treatment outcomes and the need for ongoing treatment (Bloch et al. 2011; Lim et al. 2012). However, more in depth studies of these findings are needed to improve long-term treatment outcomes in particular in the elderly in Germany. Optical coherence tomography and central retinal thickness (CRT) assessments were available only for a small proportion of patients as this is not routinely provided in Germany. Based on this, no assumptions can be made as to whether a regular CRT assessment as part of patient monitoring would improve treatment outcomes. However, there seems to be increasing evidence that treatment outcomes may be better with stricter monitoring including OCT imaging (Spaide 2009). The observed outcomes indicate deficiencies in the management and treatment of AMD patients under real-life conditions in Germany. The dichotomy of treating retina specialist and monitoring general ophthalmologist may lead to delays in treatment and follow-up which predispose to worse treatment outcomes. The current German health insurance scheme requires all nonprivately insured patients (ca. 90%) to individually apply for reimbursement of treatment, unless there are agreements between insurers and healthcare providers in place. These applications usually encompass the next three injections of ranibizumab which explains the delay to treatment (processing the individual application) as well as the administration of triplets of injections. This pattern of service delivery – a dichotomy of treating physician and a highly bureaucratized reimbursement process – may partly explain the observed suboptimal treatment outcomes (Framme et al. 2012). The favourable safety profile of ranibizumab is well documented (Rosenfeld et al. 2006; Regillo et al. 2008; Boyer et al. 2009; Holz et al. 2011; Meyer et al. 2011). Ocular AEs occurred in only 3.8% of the patients, most of them transient and a consequence of the procedure. The numbers of strokes, cerebrovascular and cardiovascular events documented were well below the number reported in all phase III clinical trials (Holz et al. 2011). Similarly, a large review of over 140 000 patient records has found no increased risk of all-cause mortality, incident myocardial infarction, bleeding and incident stroke for ranibizumab treatment compared with PDT in AMD patients (Curtis et al. 2010). Strengths of WAVE include its sample size, which makes this the currently largest prospective study of the use of ranibizumab in nv AMD, documenting treatment patterns under real life conditions in Germany. Results of this study are confounded by its noninterventional nature which did not allow for implementation of study or treatment protocols, data acquisition protocols or monitoring visits. Therapy was administered in routine practice in accordance with the German federal approval of ranibizumab and the German Summary of Product Characteristics (i.e. package insert). Thus, comparisons of the data collected with those of controlled studies like ANCHOR, MARINA or PIER are limited. Currently, real-life data on the use of ranibizumab are collected within the multi-country, prospective LUMINOUS study, funded by Novartis. This may shed more light on barriers to optimal treatment in Germany as well as other countries. In conclusion, functional, morphological as well as patient-reported outcomes such as VRQoL can be maintained by a small number of ranibizumab injections (4.3/year) in nv AMD, avoiding further vision loss. However, these outcomes are overall much less favourable than what was achieved in the pivotal ranibizumab studies with fixed monthly follow-ups and monthly intravitreal injections or injections as needed. In this study, treatment outcomes are likely to be affected by deficiencies in the adherence to the recommendations of the Summary of Product Characteristics regarding regular (monthly) monitoring and the insufficient implementation of optimal diagnostic and retreatment criteria. Suboptimal treatment outcomes on this scale have not only a potentially devastating personal impact but also considerable cost implications. Thus, future studies should aim to gain insight into these factors to improve current management and treatment outcomes for patients with nv AMD in Germany. The WAVE study was funded by Novartis Pharma Germany. All authors were involved in conception and design, or analysis and interpretation of data, drafting the article or revising it critically for important intellectual content and final approval of the version to be published.