Treatment of neovascular age-related macular degeneration with a variable ranibizumab dosing regimen and one-time reduced-fluence photodynamic therapy: the TORPEDO trial at 2 years
- First Online:
- Cite this article as:
- Spielberg, L. & Leys, A. Graefes Arch Clin Exp Ophthalmol (2010) 248: 943. doi:10.1007/s00417-009-1256-6
- 281 Views
The combination of verteporfin photodynamic therapy (PDT) and anti-angiogenics has been shown to be safe and efficacious in the treatment of choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD). The purpose of this study is to demonstrate long-term prevention of vision loss and improvement in best-corrected visual acuity (BCVA) after treatment with one-time reduced-fluence-rate PDT followed by administration of ranibizumab on a variable dosing regimen over 24 months in patients with neovascular AMD. Secondary outcome measures included the change in central macular thickness (CMT), reinjection frequency, and safety.
This prospective, nonrandomized, open-label, single-center study enrolled 27 consecutive patients (27 eyes) presenting at the Leuven University Eye Hospital with previously untreated, active neovascular AMD between September 2006 and January 2007. All patients were treated with one-time, reduced-fluence-rate verteporfin PDT, followed by intravitreal ranibizumab 0.5 mg on the same day. A second and third ranibizumab injection were given at weeks 4 and 8, respectively, after which patients were followed up monthly for 24 months. Additional treatment with ranibizumab was administered to eyes with active neovascularization as indicated clinically and on imaging studies. Retreatment was based on the following criteria: (1) presence of subretinal fluid (SRF), intraretinal edema or sub-retinal pigment epithelial fluid, as seen on OCT; (2) increase of CMT by >100 mm on OCT; (3) signs of active CNV leakage on fluorescein angiography; (4) new sub- or intraretinal hemorrhage; and (5) BCVA decreased of ≥5 letters on the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart. If any single criterion for reinjection was fulfilled, retreatment with ranibizumab was administered.
Twenty-five patients completed the 2-year study. Occult CNV was present in 64% and retinal angiomatous proliferative (RAP) lesions were present in 24% of the study eyes. The remaining three eyes had lesions classified as classic (one eye) or predominantly classic (two eyes) CNV. Month 24 data are available for 25 eyes (25 patients; age 55–86 years; mean 77; standard deviation (SD) = 7.2). Mean baseline VA was 58.6 letters (range: 35–70; SD = 8.4); 24-month VA was 66.2 letters (35–82; 12.7), not including one warfarin-treated patient who suffered vitreous hemorrhage. The mean visual acuity improved by 7.2 letters (p < 0.05) and the mean CMT decreased by 146 μm. VA improved >3 lines (15 letters) in 16%; improved 1–3 lines in 20%; remained within one line of baseline in 32%, decreased 1–3 lines in 16%, and decreased >3 lines in 16%. Losses of >3 lines were due to vitreous hemorrhage, geographic atrophy, fibrosis, and growth of an initially small CNV lesion. An average of 5.1 injections (range: 3–9) were administered during the first 12 months, and 7.1 injections (3–13) over 24 months. A total of 178 injections were performed; no systemic side-effects, uveitis, or choroidal collateral vascular damage were observed. Two patients were lost to follow-up.
Combined PDT and ranibizumab injection the same day was well tolerated in all patients. Eighty-four percent of patients had stable or improved vision at month 24.
KeywordsAge-related macular degenerationChoroidal neovascular membraneLucentis (ranibizumab injection)Photodynamic therapyVascular endothelial growth factor (VEGF)
Age-related macular degeneration is the leading cause of blindness and visual disability in patients aged >60 years in the Western world , and is becoming increasingly prevalent as this segment of the population grows. The most common cause of severe vision loss in AMD is CNV, also known as neovascular, exudative, or wet, AMD. This condition leads to a rapid, irreversible loss of vision if not promptly treated.
The introduction of inhibitors of vascular endothelial growth factor (VEGF) has drastically improved the prognosis of this devastating disease, and ranibizumab, which targets VEGF-A, has become the gold standard in its treatment. However, large randomized controlled trials have suggested that ranibizumab therapy requires sustained treatment regimens and frequent intravitreal injections. This intensive treatment represents a significant burden to both the patient and the health-care system. This burden, and the potential side-effects of intraocular administration, continues to drive research towards individualized dosing strategies and multimodal treatment combinations aimed at counteracting the complex pathogenesis responsible for CNV, and potentially decreasing the frequency of retreatment necessary for optimal results.
Photodynamic therapy (PDT) was the first treatment to prevent vision loss in neovascular AMD. However, when compared head-to-head in a randomized, controlled trial, ranibizumab has been shown to be significantly more efficacious than PDT in monotherapy [2, 3]. PDT has therefore been used less comprehensively, despite its favorable safety profile. However, it has received renewed interest as an adjunct to anti-VEGF treatments [4, 5]. Combination therapy has been explored as a way to decrease the number of intravitreal injections required in anti-VEGF monotherapy treatment regimens.
The Trial of One-Time Reduced-Fluence PDT and Evaluation-Based, As-Needed Ranibizumab for Neovascular Age-Related Macular Degeneration Using Optical Coherence Tomography (TORPEDO Trial) combined one-time, same-day reduced-fluence-rate PDT (rPDT) with same-day ranibizumab, followed by monthly ophthalmic evaluations with a variable dosing regimen of ranibizumab. The goal of the study was to determine whether the one-time addition of PDT would yield visual acuity results comparable to monthly ranibizumab in monotherapy while requiring fewer intraocular injections. This would minimize the risk of potential side-effects of the intraocular injections, as well as the costs and burdens of treatment.
Materials & methods
Patient selection and entry examinations
TORPEDO is a 2-year, open-label, prospective, single-center clinical study designed to investigate the efficacy and safety of one-time, reduced-fluence-rate PDT plus a variable dosing regimen with intravitreal ranibizumab in patients with neovascular AMD. The TORPEDO study is an investigator-sponsored trial supported by Novartis, Inc (sponsor’s protocol code number: CBPD952ABE02), approved by the Institutional Review Board of the University of Leuven, and registered with the European Union Drug Regulating Authorities Clinical Trials database (EudraCT number: 2006-003976-36).
Major eligibility criteria for enrollment in the TORPEDO study
Age 50 years or older
Presence of primary, active subfoveal CNV attributable to AMD, as detected on FA
CNV lesion in the study eye of ≤ 5,400 µm in the greatest linear dimension
Willing and able to return to the clinic for monthly visits
Prior treatment for neovascular AMD in the study eye
A long-standing CNV lesion at presentation
CNV with large hemorrhages at presentation
Angioid streaks or precursors of CNV in either eye attributable to other causes, such as ocular histoplasmosis, trauma, or myopia
Vitreous hemorrhage or history of rhegmatogenous retinal detachment or macular hole in the study eye
Any intraocular surgery within 6 months before screening, or intraocular surgery planned during the study follow-up period
History of uncontrolled glaucoma in the study eye
Refractive error of more than -6 diopters in the study eye
Active intraocular inflammation or ocular adnexal infection
Concomitant eye disease in the study eye
Cardiovascular, cerebrovascular, or peripheral vascular event within 6 months prior to screening
At the screening visit, a complete medical history and comprehensive ophthalmic evaluation were performed, including BCVA using the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart at 2 m, OCT (Stratus OCT and Cirrus HD-OCT, Carl Zeiss Meditec, Jena, Germany), color digital fundus photography using Topcon TRC-50DX retinal cameras (Topcon, Itabashi, Tokyo, Japan) and fluorescein angiography (also Topcon TRC-50DX). Initial CNV characteristics were determined by fluorescein angiography, and retinal morphology was determined by OCT using Stratus OCT (six radial lines and map) until August 2008 and Cirrus HD-OCT (macular cube and five-line raster settings) after August 2008. Because the Cirrus OCT measures retinal thickness ~45 μm greater than Stratus OCT , this difference was accounted for in the calculations, and 45 μm was subtracted from the measurements made on the Cirrus OCT.
The central 1-mm central macular thickness measurements were obtained from the macular thickness maps (Stratus) or macular cube (Cirrus) after the boundaries of delineation were confirmed to be the inner limiting membrane and the retinal pigment epithelium (RPE). In case of incorrect identification of the boundaries by the validated internal algorithm, the scan was repeated until the algorithm correctly identified the boundaries. Central macular thickness was defined as the distance between these boundaries. It did not include fluid under the RPE, otherwise known as a pigment epithelial detachment. Fluid in the macula was identified as cystic (intraretinal) or subretinal, as detected on OCT.
Fundus photography and OCT imaging were performed at baseline and monthly thereafter. Each follow-up examination consisted of BCVA measurement and a complete anterior and dilated ocular examination. The same examiner performed the BCVA and OCT examinations at each patient visit. Fluorescein angiography was performed at baseline and later whenever recurrence, persistence or leakage was suspected by the principal investigator (AL) based on OCT and clinical findings.
The major efficacy end point was the change in visual acuity at 2 years as compared to baseline measurements. Secondary outcome measures were change in CMT, reinjection frequency, and safety.
Initial treatment, administered within 28 days of screening, comprised one-time, reduced fluence verteporfin (Visudyne, Novartis, Basel, Switzerland) PDT and one intravitreal injection of 0.5 mg ranibizumab (Lucentis, Novartis, Basel, Switzerland) at baseline reduced-fluence verteporfin therapy was administered at a dosage of 6 mg/m2 infused intravenously over 10 min. Activating 689-nm light at a dose of 25 J/cm2 was started 15 min after start of infusion. The fluence was set at 600 mW/cm2 and was applied for a 42-s duration using the Opal Photoactivator (Coherent, Santa Clara, CA, USA).
Treatment and follow-up schedule
Additional treatment with ranibizumab was administered to eyes with active neovascularization as indicated clinically and on imaging studies. Retreatment was based on the following criteria: (1) presence of subretinal fluid (SRF), intraretinal edema or sub-retinal pigment epithelial fluid, as seen on OCT; (2) increase of CMT by >100 μm on OCT; (3) signs of active CNV leakage on fluorescein angiography; (4) new sub- or intraretinal hemorrhage; (5) BCVA decreased of ≥5 letters on the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart. If any single criterion for reinjection was fulfilled, retreatment with ranibizumab was administered.
Follow-up examinations and outcome measures
Follow-up examinations were performed at weeks 4 and 8, immediately before intravitreal ranibizumab administration, and every 4 weeks afterwards for 24 months from baseline.
The major outcome measurement in the TORPEDO study was ETDRS visual acuity letter scores. Secondary outcome measures were the change in central macular thickness (CMT), reinjection frequency, and safety.
Local and systemic adverse events were monitored throughout the study. For the purpose of analysis, ETDRS visual acuity data were converted into equivalent logarithms of the minimum angle of resolution (log MAR) values. Data were analyzed using the paired two-sample t-test for means. A p-value of less than 0.05 was considered to be statistically significant.
Between September 2006 and January 2007, 27 consecutive patients (27 eyes) presenting at the medical retina center of the Leuven University Eye Clinic were enrolled. A total of 25 eyes (25 patients; 17 female and eight male; mean age 77.4 ± 7.2 years (range: 55–86) were included in the analysis. Patient demographics, previous medical history, and FA characteristics are presented in Table 2. At baseline, the main presenting symptom was a decrease in BCVA, which was associated with metamorphopsia in all patients.
Patient characteristics at baseline
Sex - no. (%)
Race - no. (%)
Age - years
Age group - no. (%)
Number of letters as measure of visual acuity
58.6 ± 8.4
<62 - no. (%)
≥62 - no. (%)
Visual acuity (Snellen equivalent) - no. (%)
20/200 or worse
Better than 20/80 but worse than 20/40
20/40 or better
Type of choroidal neovascularization - no. (%)
occult with no classic lesions
Mean central macular thickness (mm)
OCT Lesion characteristics from baseline and at months 1, 3, 12 and 24 in neovascular AMD patients treated with PDT at baseline and ranibizumab at baseline, month 1, month 2 and later as needed
OCT lesion characteristics (n = 30)
Baseline n (%)
1 Month n (%)
3 Months n (%)
12 Months n (%)
24 Months n (%)
The treatment procedure was well tolerated. No clinical evidence of inflammation, uveitis, endophthalmitis, or ocular toxicity was observed. There was no choroidal collateral vascular damage detected on FA. One patient died of causes unrelated to the study treatment, and another patient was lost to follow-up. All other patients completed all follow-up visits. No study eyes had received previous treatment for AMD. No serious drug-related adverse events were observed. Further, there were no changes in lens status in any of the study eyes during the follow-up period.
Three months: visual acuity and OCT
Twenty-four months: visual acuity and OCT
The results of the TORPEDO trial indicate that the treatment of neovascular AMD with one-time, reduced-fluence PDT in conjunction with a variable, individualized ranibizumab dosing strategy leads to improvements in visual acuity and retinal morphology comparable to monthly ranibizumab in monotherapy. These results were obtained while administrating significantly fewer injections than in a fixed monthly treatment schedule. In this 24-month clinical trial, final visual acuity improved by an average of 7.2 letters as compared to baseline, and central retinal thickness decreased by an average of 146 μm. An average of 7.1 injections was administered over 24 months.
Monthly anti-VEGF treatment has been shown in large, randomized, controlled phase III trials (ANCHOR, MARINA) to improve vision in a majority of patients [4, 10, 11]. The benefits of ranibizumab apply to all angiographic subtypes of neovascular AMD and across all lesion sizes. Ranibizumab, which was designed specifically for intravitreal injection, has since become the gold standard for the treatment of neovascular AMD. However, the frequent injections required by this evidence-based approach place a significant burden on patients and health-care providers. A treatment regimen requiring fewer visits and especially fewer retreatments while providing equivalent efficacy is highly desirable. VEGF is an important molecule within the eye, and is not exclusively expressed by pathological neovascular tissue. A reduction in the number of anti-VEGF injections would lower the risk of potential disruption of physiologic VEGF-mediated ocular processes as well as possible ocular and systemic side-effects [12, 13].
Several studies have been carried out to test the efficacy of reduced treatment frequency of ranibizumab monotherapy. The phase IIIb PIER study administered ranibizumab monthly for the first 3 months followed thereafter by quarterly doses [14, 15]. The rationale for this regimen was based on evidence from phase I/II studies, which indicated that the pharmacodynamic activity of ranibizumab administered monthly for three doses might last 90 days. However, 1-year results showed a mean decrease in BCVA of 0.2 letters from baseline, as well as a mean increase in vascular leakage on FA and increases in mean retinal thickness once the quarterly dosing regimen began. The authors concluded that these results suggest that at least some of the subjects needed more frequent ranibizumab injections. However, it is likely that these subjects required not more frequent injections, but rather treatment to be administered only when leakage had occurred and intraretinal fluid was present.
This possibility was investigated in the PrONTO and SUSTAIN studies, which employed specific criteria, assessed monthly, to determine an optimal treatment regimen for each patient individually [15, 16]. Retreatment criteria in the PrONTO study were a loss of five letters in conjunction with fluid in the macula as detected by OCT, an increase in OCT central macular thickness of at least 100 μm, new-onset classic CNV, new macular hemorrhage, or persistent macular fluid detected by OCT at least 1 month after the previous injection of ranibizumab. This resulted in a greatly reduced injection frequency (5.6 and 9.9 injections over 12 and 24 months, respectively, including the three initial mandatory doses) compared to ANCHOR and MARINA, whose protocols called for standard monthly injections administered regardless of imaging or VA status at the time of evaluation. Significantly, the PrONTO study visual acuity results (improvement of 9.3 and 11.1 letters improvement at 12 and 24 months, respectively, compared to BCVA at baseline) were similar to the those achieved by the two phase III clinical trials, (11.3 and 7.2 letters, respectively), as well as those achieved by the current study (7.2-letter improvement). The SUSTAIN trial, which employed a protocol similar to that of PrONTO and included more patients, resulted in a mean VA gain of 6.7 letters and 5.3 injections at 12 months . The multi-center, randomized SAILOR (Safety Assessment of Intravitreal Lucentis for AMD) study also began with three mandatory monthly injections of 0.5 mg ranibizumab, but follow-up was quarterly and retreatment given based on similar criteria . This resulted in a mean 12-month VA gain of 2.3 letters, with a mean of 4.6 injections in 487 treatment-naive patients, suggesting that quarterly follow-up is insufficient to ensure optimal results.
Another strategy, known as “inject and extend,” aims to decrease both the number of retreatments and frequency of visits . In this strategy, patients are injected at each visit, but the scheduling of the upcoming visit is determined by the status of the macula at the current visit. If there is leakage, the next visit is scheduled sooner than if the macula had been “dry.” This strategy may minimize the number of clinic visits, alleviating some of the cost and burden of monthly follow-up and making it an attractive option for routine clinical practice. However, it is not clear whether this strategy minimizes the number of reinjections.
Regardless of the dosing and visit regimens, ranibizumab in monotherapy might not deliver the maximum possible benefit. The pathogenesis of CNV is complex, consisting of a cascade of retinal changes in which ischemia or hypoxia leads to the upregulation of growth factors, such as VEGF, and atrophy of the choroidal vasculature [27–29]. This causes inflammation, angiogenesis, and subsequently neovascularization . It is possible that treatment strategies might derive benefit by targeting various components of this cascade to achieve additive or even synergistic effects .
The two treatment modalities used in the TORPEDO trial have different mechanisms of action. Anti-VEGF therapy blocks angiogenesis, which is the growth of new blood vessels from existing vasculature. It also decreases the permeability of these capillaries, leading to a rapid reduction in retinal fluid and a corresponding improvement in visual acuity. The goal of PDT is to selectively occlude or destroy choroidal neovascularization, or the formation of completely new microvascular networks, while maintaining perfusion in the deeper physiological choroidal vessels as well as the overlying retinal tissue and vessels. This occurs via selective nonthermal protothrombosis of the choroidal neovasculature, in which the overlying neurosensory retina remains intact .
Although few patients experience VA improvements following PDT, it is associated with long-term stabilization of vision as well as a reduction in lesion activity. Further, in contrast to ranibizumab therapy, PDT offers a finite course of therapy; yearly treatment rates decline from about 3.5 treatments in the first year to a mean of about 0.1 in the fifth . However, PDT has been shown to induce a pro-angiogenic response, with upregulation of VEGF shortly after administration . This is associated with CNV recurrence and a need for repeated treatment . Combination therapy with anti-VEGF antibodies might prevent this pro-angiogenic response.
Several previous studies have investigated the combination of ranibizumab and PDT. In the phase I/II FOCUS study, all patients received PDT quarterly as needed, and either ranibizumab or sham injections, administered monthly [4, 35]. This produced two significant results. First, patients who received PDT plus ranibizumab experienced a mean improvement of 4.9 letters at month 12, while those in the PDT plus sham group lost 8.2 letters. Second, the addition of ranibizumab reduced the frequency of PDT from 3.4 sessions to 1.3.
Given the rapid elevation of intraocular VEGF levels after PDT , the administration of anti-VEGF treatment might have immediate benefits. Same-day treatment was administered in the TORPEDO study as well as in the phase I/II PROTECT study, which was specifically designed to provide proof of principle for the safety of verteporfin plus ranibizumab . Same-day PDT+ranibizumab was also shown to be safe in an experimental monkey model of CNV .
Reduced-fluence-rate PDT was used in the TORPEDO study in order to minimize the risk of excessive choriocapillary occlusion. This is in accordance with previous studies, which have suggested that studies using PDT in both monotherapy  and combination treatment  should use reduced-fluence PDT. Collateral vascular damage after PDT with higher fluence has been reported in both animal and clinical studies [39, 40], and although a recent study suggests that sPDT might be safer than was previously thought , others have shown a clear benefit of rPDT, especially for the occult CNV subtype . In our study, 66.7% of the CNV lesions were of the occult subtype.
Two ongoing trials will directly compare the combination of ranibizumab+PDT and monthly ranibizumab monotherapy: DENALI, and its European counterpart, MONT BLANC. Prior to these studies, this comparison has only been made using experimental choroidal neovascularization in the monkey . Husain et al. showed that treatment of experimental choroidal neovascularization with ranibizumab plus PDT combination therapy causes a greater reduction in angiographic leakage than PDT plus sham injection.
The results of a recent randomized trial by Potter et al. showed that the addition of reduced light dose PDT to an OCT-guided, anti-VEGF regimen significantly reduced the mean number of bevacizumab injections over 6 months from 5.1 (bevacizumab only) to 2.8 (bevacizumab + 25 J/cm2) and 2.5 (bevacizumab + 12 J/cm2). Patients returned monthly for possible retreatment with bevacizumab or combination therapy with a 3-month minimum interval between combination treatments. Although this study was powered to examine only the number of treatments, not visual acuities, these results support those of the TORPEDO study.
Our study suggests that reduced-fluence-rate PDT in conjunction with a variable dosing regimen of ranibizumab is safe and efficacious. No treatment-related problems were encountered. The decision to incorporate spectral domain (Cirrus) OCT during the study was based on the fact that spectral domain OCT offers far more detailed information on CNV lesion activity than time domain (Stratus) OCT. Indeed, spectral domain can detect disease activity not detected by time domain systems [7, 8]. Since this was a clinic-based study, in which the primary goal was improvement of visual acuity, and considering our belief that macular edema should be treated as soon as possible and prevented if possible, we switched from time domain to examination with spectral domain OCT as soon as the Cirrus machine became available. This may be considered a weakness in the study design, since no simple algorithm exists to convert results from one OCT format to another, particularly when retinal pathology is present. This is because the Cirrus calculates the thickness of the retina from the RPE to the inner limiting membrane (ILM), while the Stratus does so from the inner segment/outer segment junction (IS/OS) to the ILM. Thus, the Cirrus will be inherently thicker by the distance between the IS/OS and the RPE, which is approximately 45 mm, according to the manufacturer .
There was no collateral vascular damage or serious vision loss due to the treatment. Clinical implications of this study include the assurance that this treatment combination can be used safely, especially for those patients who cannot tolerate an increased injection frequency or who can only be present at infrequent visits. Further, there is no significant decrease in final mean visual acuity, despite fewer injections administered. A randomized trial is necessary to compare this combination of reduced-fluence-rate PDT plus ranibizumab with (1) standard-fluence rate PDT plus ranibizumab and with (2) ranibizumab alone. In the near future, a new paradigm may emerge wherein numerous therapeutic modalities are administered in customized combinations based on specific clinical and diagnostic findings.
The authors would like to thank Carol Heughebaert for her editorial assistance and Adrian Loehwing for her help with the statistical analysis. This study was supported by Novartis Inc, Basel, Switzerland, and the Department of Ophthalmology at the Leuven University Hospital, Leuven, Belgium. Dr. Leys has received research grants from Novartis Inc, has participated in competing scientific advisory boards, and has received honorarium and reimbursement for travel expenses from Novartis. We thank the following: those involved in design of study (A.L.); conduct of study (A.L., L.S.); data collection (L.S., A.L.); management (L.S., A.L.); analysis (L.S., A.L), and interpretation of data (L.S., A.L.); and preparation (L.S.), review and approval of the manuscript (L.S., A.L.). Before the initiation of the study, approval to perform the TORPEDO Study was obtained from the Institutional Review Board at the Leuven University Hospital. The study was registered with https://eudract.emea.europa.eu (no. 2006-003976-36). The study adhered to the tenets of the Declaration of Helsinki.