Vision-related function after low-dose transpupillary thermotherapy versus photodynamic therapy for neovascular age-related macular degeneration
Abstract.
Purpose: To compare the effects of low-dose transpupillary thermotherapy (TTT) and verteporfin photodynamic therapy (PDT) on patient-reported visual function using the National Eye Institute Visual Function Questionnaire 25 (NEI VFQ-25) in patients with occult neovascular age-related macular degeneration (AMD).
Methods: Patients were randomized to receive either low-dose TTT (and sham PDT) (n = 52) or PDT (and sham TTT) (n = 46). Patients were followed for 12 months with retreatment according to clinical assessment. The clinical outcome of this study has been recently reported. The NEI VFQ-25 questionnaire was administered at baseline and at 12 months.
Results: Forty-two patients (80.1%) in the TTT group and 37 patients (80.0%) in the PDT group completed the questionnaire at the 12-month follow-up. The mean change in the NEI VFQ-25 composite score was +1.2 for the TTT group (p > 0.05) and +0.7 for PDT group (p > 0.05). None of the subscale categories showed significant changes between treatment groups at 12 months. Subgroup analysis showed that NEI VFQ-25 scores were lower in patients treated in their better-seeing eye.
Conclusion: In this randomized study on patients with occult neovascular AMD, low-dose TTT and PDT appeared to be equally potent at stabilizing patient-reported visual function. However, the study was underpowered for this conclusion to be made firmly. Also, given the impressive results obtained with ranibizumab for all types of neovascular AMD, neither PDT nor low-dose TTT should be considered as first-line treatments.
Introduction
Neovascular age-related macular degeneration (AMD) characterized by choroidal neovascularization (CNV) is a primary cause of blindness in the Western world. Without treatment prognosis is poor, resulting in a rapid and progressive loss of vision and contrast sensitivity with profound impact on an individual’s ability to perform daily tasks such as reading, writing, driving and social functioning, leading to a progressive loss of independence.
The impact of vision on quality of life is often reflected inadequately by traditional clinical measurements of visual function. Discrepancies have been described; for instance, an improvement in visual acuity does not necessarily result in a better perception of visual function (Miskala et al. 2004). With the rapid development of new therapies for neovascular AMD it is therefore becoming increasingly important to assess the full impact of visual impairment on an individual’s quality of life. The National Eye Institute Visual Function Questionnaire 25 (NEI VFQ-25) was developed to measure a patient’s subjective assessment of vision-related functions (Mangione et al. 1998, 2001). The NEI VFQ-25 has been validated in patients with AMD and low vision with other causes such as age-related cataract, diabetic retinopathy, primary open-angle glaucoma and cytomegalovirus retinitis (Mangione et al. 2001).
Recent major clinical trials including the Age Related Eye Disease Study, the Submacular Surgery Trials and the MARINA (minimally classic/occult trial of the anti-VEGF antibody ranibizumab In the treatment of neovascular AMD) study have used and validated the NEI VFQ-25 to assess vision-related function (Clemons et al. 2003; Miskala et al. 2004; Chang et al. 2007).
We have recently performed a prospective, randomized study of low-dose TTT and PDT for occult CNV (Odergren et al. 2008). The main outcome of the study was that the proportion of patients with stabilized visual acuity (i.e. losing fewer than 15 letters at 12 months) was similar between the treatment groups [76.0% in the TTT group and 73.9% in the PDT group (p > 0.05)]. In this article, we describe the results of patient-reported vision-related function in the same set of patients.
Materials and Methods
Study design and patient selection
This investigation adhered to the tenets of the Declaration of Helsinki and was approved by the local ethics committee at the Karolinska Institute. All patients included in the study provided a signed informed consent form agreeing to participate.
Ninety-eight patients (98 eyes) were assigned randomly (in a 1 : 1 ratio) to PDT with verteporfin or to TTT. To be eligible for the study the patient had to meet the following inclusion criteria: (i) age 50 years or older; (ii) occult CNV under the geometric centre of the fovea [occult CNV was defined either as a fibrovascular retinal pigment epithelium detachment (PED) or as late leakage of undetermined source]; (iii) lesion size less than 5000 μm in the greatest linear dimension (GLD); (iv) best-corrected Early Treatment Diabetic Research Study (ETRDS) visual acuity (BCVA) corresponding to a Snellen equivalent of 20/200 to 20/25; and (v) clear ocular media.
Photodynamic therapy
Verteporfin (6 mg/m2) was infused intravenously for 10 min. Fifteen minutes after the start of infusion, diode laser treatment was performed (689 nm, 50 J/cm2 at an intensity of 600 mW/cm2) for 83 seconds, using a spot size 1000 μm larger than the GLD of the lesion. The GLD was measured from fluorescein angiograms. Sham PDT was given to all patients receiving TTT and was performed in the same way as PDT with the exception that a glucose infusion was given instead of verteporfin.
Transpupillary thermotherapy
A sham intravenous infusion of 5% dextrose in water was given for 10 min. Fifteen minutes after the start of the infusion, an infrared diode laser treatment at 810 nm (Iridex Corporation, Mountain View, California, USA) was performed for 60 seconds via a slit-lamp adaptor. The Mainster wide field lens was used with a standard 3 mm spot in the laser slit lamp, which produced a spot size of 4.41 mm diameter on the retina. The laser was selected to deliver 600 mW, corresponding to a power of 136 mW for a circular spot of 1 mm. Sham TTT was given to all patients receiving PDT and was performed in the same way as TTT with the exception that the laser power was turned off.
Synopsis of protocol
The NEI VFQ-25 was administered by trained study-site personnel at baseline and at the 12-month visit. In some cases the protocol was sent to the patients by mail. At baseline and at every follow-up (6, 12, 18, 36 and 48 weeks), assessment included BCVA, slit-lamp biomicroscopy of the anterior and posterior segments, fluorescein angiography and optical coherence tomography. If a patient showed active CNV leakage at the time of the regularly scheduled appointment, they were retreated using PDT (and sham TTT) or TTT (and sham PDT), as described earlier.
NEI VFQ-25 methods
The NEI VFQ-25 is a validated questionnaire, developed to measure vision-specific health-related quality of life. The interview instrument consists of 25 questions grouped into 12 subscales, consisting of one or more questions each. The subscales include general health and 11 vision-related subscales. The NEI VFQ-25 assesses difficulty with near vision activities, difficulties with distance activities, limitations in social functioning because of vision, role limitations because of vision, dependency on others because of vision, mental health issues because of vision, future expectations because of vision, driving difficulties, pain and discomfort, limitations of peripheral vision and colour vision. The subscale scores range from 0 (worst score) to 100 (best score). The score for each domain is a mean of the items included in that particular subscale,; the higher the score, the better the respondent’s quality of life. With normal vision and no vision-related quality of life limitation, the maximum score of 100 is achieved. The lower the score, the greater the vision disability. The NEI VFQ-25 interview was administered by trained study-site personnel.
Statistical analysis
The primary hypothesis was based on a difference between the treatment groups in the proportion of patients achieving the main outcome measure of the clinical study. Secondary outcome measures included mean changes in NEI VFQ-25 scores between baseline and 12-month follow-up. For power calculation we assumed that 50% of the patients treated with PDT and 80% of the patients treated with TTT would achieve the primary endpoint (loss of < 15 letters). With a statistical power of 90% and a p-value of < 0.05 we estimated that a total of 96 patients would be needed (MedCalc Software, Mariakerke, Belgium). Student’s t-test for independent data was used for statistical analyses.
Results
In this study 98 patients with occult CNV were randomly assigned to receive either low-dose TTT (n = 52) or PDT (n = 46). The number of patients completing the treatment for 12 months was 51 (98%) in the TTT group and 44 (96%) in the PDT group. The number of patients completing the quality of life questionnaire at 12 months was 42 (80.1%) in the TTT group and 37 (80.0%) in the PDT group.
Baseline characteristics for treatment groups were evenly matched, as described previously (Odergren et al. 2008).The mean change in the NEI VFQ-25 individual patient scores and the main subscale scores (near activity, distance activity and dependency) indicated a small, but not statistically significant, improvement in both treatment groups. For the overall composite score after 12 months of follow-up, the mean change was +1.2 for the TTT group (p > 0.05) and +0.7 for the PDT group (p > 0.05) (Table 1).
| NEI VFQ-25 subscale | PDT (baseline score, mean SD) (n = 46) | PDT (month 12 score) (n = 37) | TTT (baseline score, mean SD) (n = 52) | TTT (month 12 score) (n = 42) |
|---|---|---|---|---|
| Overall (composite) | 67.8 (11.3) | 68.5 (13.6) | 67.7 (10.7) | 68.9 (12.7) |
| General health | 50.7 (14.5) | 49.2 (15.3) (p = 0.65) | 48.6 (14.3) | 45.7 (15.8) (p = 0.60) |
| General vision | 51.7 (9.2) | 48.8 (8.6) (p = 0.56) | 52.0 (7.8) | 54.1 (8.4) (p = 0.53) |
| Ocular pain | 84.1 (13.5) | 83.9 (12.5) (p = 0.62) | 85.4 (11.1) | 83.0 (11.5) (p = 0.89) |
| Near activities | 51.1 (11.4) | 53.3 (16.4) (p = 0.57) | 53.6 (10.9) | 54.6 (15.9) (p = 0.72) |
| Distance activities | 56.9 (14.3) | 59.0 (20.4) (p = 0.56) | 58.5 (13.9) | 59.7 (19.1) (p = 0.65) |
| Social functioning | 75.0 (14.0) | 74.2 (15.6) (p = 0.77) | 72.2 (13.3) | 74.6 (14.1) (p = 0.48) |
| Mental health | 59.6 (10.1) | 57.8 (12.3) (p = 0.56) | 58.2 (9.7) | 58.7 (11.7) (p = 0.84) |
| Role difficulties | 61.4 (7.1) | 64.0 (11.2) (p = 0.59) | 64.9 (7.3) | 64.2 (10.3) (p = 0.77) |
| Dependency | 72.6 (7.8) | 75.0 (12.0) (p = 0.56) | 70.1 (8.4) | 75.2 (11.1) (p = 0.13) |
| Driving | ND | ND | ND | ND |
| Colour vision | 83.3 (13.6) | 83.3 (15.0) (p = 0.99) | 81.7 (11.3) | 84.3 (13.2) (p = 0.45) |
| Peripheral vision | 82.3 (12.3) | 85.3 (11.7) (p = 0.52) | 80.5 (11.9) | 80.3 (12.2) (p = 0.69) |
- SD, standard deviation; ND, not determined.
Because the NEI VFQ-25 questionnaire is a binocular test it was interesting to evaluate any differences between patients treated in the better- or worse-seeing eye. In particular, it was important to assess differences, if any, in the group of patients treated in their better-seeing eye. Almost half of the patients were treated in their better-seeing eye. These patients performed worst and subgroup analysis showed that VFQ-25 scores were markedly lower in these individuals (Fig. 1). However, there were no apparent differences between these subgroups when compared with each other.
Visual-related function in patients treated in better- and worse-seeing eyes, respectively. Overall National Eye Institute Visual Function Questionnaire 25 (NEI VFQ-25) composite score and subscale category scores for patients treated in the better- or worse-seeing eye are shown at baseline (0) and 12 months. The VFQ-25 questionnaire is a binocular test. Therefore, patients treated in their better-seeing eye are expected to perform worse and any differences between treatment groups may be more apparent. In the low-dose transpupillary thermotherapy group, 44.2% (n = 23) were treated in the better-seeing eye compared with 45.6% (n = 21) in the photodynamic therapy group. As expected, the VFQ-25 scores were generally lower in patients treated in their better-seeing eye. However, no differences were detected between any of the treatment groups.
Discussion
The NEI VFQ-25 questionnaire is a validated, sensitive and useful tool for assessing visual function in AMD patients. In this randomized study on patients with occult CNV we show that low-dose TTT and PDT similarly stabilized patient-reported visual function over 12 months (p > 0.05).
The power calculation for study size was solely based on visual acuity assumptions (Odergren et al. 2008). The main outcome measure, i.e. that low-dose TTT would prove superior to PDT at stabilizing BCVA, was not met in this study. Changes in NEI VFQ-25 scores were secondary outcome measures. Low-dose TTT and PDT behaved remarkably similarly for either stabilization of BCVA and NEI VFQ-25 scores. Based on this outcome it may have been preferable to design a non-inferiority study instead. However, such a study would have required approximately 300 patients per treatment group (Slakter et al. 2006).
Patient-reported visual function does not always correlate with visual acuity chart scores (Miskala et al. 2004). In the MARINA study, where intravitreal ranibizumab was compared with placebo in patients with neovascular AMD, a difference in visual acuity of 17.6 letters was reported between the treatment groups using the ETDRS chart (Rosenfeld et al. 2006). When the NEI VFQ-25 questionnaire was applied to the same set of patients, the mean improvement in overall composite scores was +5.6 for the ranibizumab group compared with −2.8 for the placebo group (Chang et al. 2007). Previous studies have shown that a 10-point change in the NEI VFQ-25 score corresponds roughly to a 15-letter change in visual acuity (Miskala et al. 2003). Thus, in the MARINA study there was no apparent difference between the NEI VFQ-25 score and visual acuity. In the present study we found that the NEI VFQ-25 score increased by 1.2 for the low-dose TTT group and by 0.7 for the PDT group. In the same set of patients, we previously reported a decrease in visual acuity of 7.2 letters for both of these treatment groups (Odergren et al. 2008). Thus, as with the MARINA study, our results on patient-reported visual function and visual acuity are in overall agreement.
The NEI VFQ-25 questionnaire is designed to test binocular vision. It is therefore important to take into account the impact of patients treated in their worse-seeing eye when evaluating the results (Cahill et al. 2005; Varma et al. 2006). It is likely that the performance of these patients will greatly influence the overall scores if the results are only analysed collectively. We found that more than half of the patients treated with either low-dose TTT or PDT were in fact treated in their worse-seeing eye. However, subgroup analyses showed that the patients treated in their better-seeing eye performed similarly (i.e. no change in VFQ-25 score at 12 months) compared with the group as a whole. In other words, treatment of the better-seeing eye with either low-dose TTT or PDT stabilized vision-related function in a similar manner.
No placebo group was included in the present study; therefore we can only speculate on whether either low-dose TTT or PDT treatment resulted in a meaningful change in patient-reported visual function. It has been suggested previously that a 10-point difference in NEI VFQ-25 score will result in a clinically meaningful change in quality of life (Chang et al. 2007). However, despite the lack of a placebo group it is unlikely that either treatment in the present study would have rendered differences in composite NEI VFQ-25 score of this magnitude. In the MARINA study (with a roughly similar cohort of patients consisting of occult and minimally classic CNV), the placebo group lost 2.8 points. If we assume that a placebo group would have performed similarly in our study, the difference between treated and untreated patients would have been approximately 4 points – well below the 10-point limit. On the other hand, in the MARINA study (where the difference in visual gain was highly significant between the ranibizumab and placebo groups), the difference in NEI VFQ-25 score was still only 8.4 points after 12 months, increasing to 10.9 points after 24 months (Chang et al. 2007). The additional gain after 24 months was caused by a further decrease in the placebo group. In our study, we have found that the stabilization in visual acuity is maintained after 24 months for both low-dose TTT and PDT (data not shown). It is thus possible that the overall gain in composite NEI VFQ-25 score would also increase after 24 months of treatment in our study. As mentioned earlier, it is also important to adjust for the impact of the patients treated in their worse-seeing eye on the overall score. In the MARINA study, less than half of the patients were treated in their better-seeing eye. Subgroup analysis for better- and worse-seeing eyes is not available for the MARINA study. However, it is reasonable to assume that the difference between the ranibizumab and placebo groups is in fact larger than reported and that the placebo group (when correcting for the positive impact of the patients treated in their worse-seeing eye) lost more than 2.8 points. Nevertheless, both our study and the MARINA study demonstrate the need for large differences in visual acuity scores in order to obtain a presumably meaningful gain in patient-reported visual function.
With the impressive results obtained with repeated intravitreal ranibizumab for all types of CNV, there is no doubt that ranibizumab is the current gold standard for managing neovascular AMD (Brown et al. 2006; Rosenfeld et al. 2006). However, in order to minimize cost and decrease the number of injections, combination therapy (e.g. ranibizumab and PDT) is currently investigated thoroughly in randomized trials. With its low cost and practical management, low-dose TTT should also be considered for combination treatment with anti-vascular endothelial growth factor agents in controlled randomized trials.
