Tachyphylaxis during treatment of exudative age-related macular degeneration with ranibizumab
- Correspondence to Dr Maria Salling Eghøj, Department of Ophthalmology, Copenhagen University Hospital Roskilde, Køgevej 7-13, DK-4000 Roskilde, Denmark;
- Accepted 17 June 2011
- Published Online First 6 July 2011
Aim To determine whether tachyphylaxis occurs during treatment with ranibizumab (Lucentis, Genentech, Inc., South San Francisco, California, USA) for exudative age-related macular degeneration (AMD).
Design Retrospective review of cases.
Participants The treatment results of 1076 eyes (976 patients) treated with ranibizumab for exudative AMD was evaluated to identify patients with a potential tachyphylactic response. The participants had to have a minimum of 12 months follow-up.
Methods Tachyphylaxis was defined as a lack of response to the drug at the time of reactivation of choroidal neovascularisation (CNV) in patients who had responded to the initial treatment. The authors considered it a lack of response to ranibizumab if a decrease in vision and an increase in central retinal thickness (CRT) were observed despite repeated injections. Hence a stabilisation in vision and/or stabilisation in CRT during treatment were not considered tachyphylaxis, and other unfavourable responses such as a tear in the retinal pigment epithelium and therefore a decrease in vision during treatment were also not considered as tachyphylaxis. Every patient in this cohort who has had an injection-free interval after primary inactivation of CNV and who has received retreatment at a later stage was identified. In this population, those cases that did not respond to retreatment (tachyphylaxis) were identified and characterised.
Main outcome measures Number of patients who developed tachyphylaxis after treatment with ranibizumab.
Results 20 patients (2%) developed tachyphylaxis during their treatment.
Conclusion Tachyphylaxis can occur during the treatment of exudative AMD with ranibizumab. The precise mechanism for the development of tachyphylaxis is unclear. Both local and systemic factors might be involved.
Age-related macular degeneration (AMD) is the leading cause of irreversible vision loss among people older than 50 years in the USA and Western Europe.1 2 The most common reason for severe visual loss in patients with AMD is the development of a choroidal neovascularisation (CNV).3 4
The mechanism by which CNV forms is not known in detail, but both experimental studies and clinical trials have shown that vascular endothelial growth factor (VEGF) plays a major role in the formation of CNV.5 VEGF can be synthesised in substantial amounts by the retinal pigment epithelium; it results in the formation of CNV in rats if there is a rupture of Bruch's membrane and, furthermore, VEGF is readily detected in the vitreous humour of patients with exudative AMD.5–7 The potency and specificity of VEGF for vascular endothelial cells underscore VEGF's primary role in CNV development.5 8
Experimental data on the role of VEGF in CNV pathogenesis are supported by the results from clinical trials using anti-VEGF agents to treat exudative AMD.9 10 Ranibizumab (Lucentis, Genentech, Inc., South San Francisco, California, USA), which is a recombinant, humanised, monoclonal antibody antigen-binding fragment (Fab) that neutralises all biologically active forms of VEGF, has been shown to be effective in treating active CNV due to AMD in clinical trials. Thus, anti-VEGF agents have been widely used to treat patients with exudative AMD.
With the increasing use of ranibizumab for exudative AMD, a growing understanding of the long-term effects is emerging. It is clear that some patients do not respond to the treatment as anticipated, suggesting that other molecules (such as ICAM-1, E-selectin, CD44 and basic and acidic fibroblast growth factors) may play a role in CNV formation.8
However, reports have also emerged regarding potential tachyphylaxis with the use of bevacizumab (Avastin, Genentech, Inc.), another anti-VEGF agent used as an off-label treatment for exudative AMD.11 12 Tachyphylaxis is generally defined as a progressive decrease in therapeutic response after repetitive administration of a pharmacologically active substance.11–13
The development of tachyphylaxis during treatment with ranibizumab has not previously been reported, even though one study showed a possible decreasing but not tachyphylactic response after repeated use of ranibizumab.13
The purpose of this study was to investigate whether tachyphylaxis exists after intravitreal injection treatment of exudative AMD with ranibizumab.
In this study, we considered tachyphylaxis as an initial treatment response, defined as a complete resolution of intra/subretinal fluid and haemorrhages, after the initial treatment regime and subsequently a complete lack of response in the same patient after reactivation of the CNV. We considered it a lack of response to ranibizumab if a decrease in vision and an increase in central retinal thickness (CRT) were observed despite repeated injections. Hence a stabilisation in vision and/or stabilisation in CRT during treatment were not considered tachyphylaxis, and other unfavourable responses such as a tear in the retinal pigment epithelium, development of atrophy and/or large haemorrhages and therefore a decrease in vision during treatment were also not considered tachyphylaxis.
All eyes with signs of active exudative AMD initially received three treatments with intravitreal 0.5 ml ranibizumab 10 mg/ml every 4 weeks. Patients with vascularised pigment epithelial detachment were offered treatment, while patients with non-vascularised pigment epithelial detachment were not. Four weeks after the last injection, the patients underwent treatment assessment at the follow-up visit. Best-corrected visual acuity (BCVA) was measured using the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart and spectral-domain optical coherence tomography (OCT) imaging was performed using the Heidelberg Spectralis HRA+OCT (Heidelberg Engineering, Heidelberg, Germany). The Eyetracker software makes it possible to perform serial OCT scans and therefore makes available comparable data on CRT over time. An ophthalmic examination was also performed. If the examination revealed intraretinal/subretinal fluid on OCT or retinal haemorrhages, the patient received further treatment with intravitreal ranibizumab. If the examination did not show signs of activity of the CNV, treatment was deferred and the patient returned every 4 weeks for repeated evaluation (Pronto guidelines).9
In order to identify patients, we retrospectively reviewed our database of all patients receiving ranibizumab in the period of 2007–2010 and identified those diagnosed with exudative AMD. Patients with peripapillary CNV and those with signs of choroidal polypoidal vasculopathy on indocyanine green angiography were excluded. A total of 1076 eyes in 976 patients met these criteria.
Every patient in this cohort who has had an injection-free interval after primary inactivation of the CNV and who has received retreatment at a later stage was identified. In this population, those cases that did not respond to retreatment (tachyphylaxis) were identified and characterised. There had to be a minimum of 12 months follow-up.
Out of the 1076 eyes, we identified a total of 20 (2%) eyes that met the criteria for tachyphylaxis. None of the patients with tachyphylaxis received treatment in both eyes.
The mean age was 78 years (62–90) and all patients were Caucasians. There were 6 men and 14 women. None of the patients had received any prior treatment (anti-VEGF or photodynamic therapy) for exudative AMD.
The mean BCVA at the time of referral was 56 (35–80) ETDRS letters. This increased to 68 (35–95) at the stage where no signs of activity of the CNV were noted at treatment assessment visit. The mean number of injections necessary for inactivation of the CNV was 3.5 (3–6).
At baseline, the mean CRT was 358 μm (183–700 μm) decreasing to 244 μm (150–363 μm) at the stage of no activity.
Of all the patients, the initial lesion was classic in 33% of cases, occult in 56% of cases and both classic and occult in 11% of cases. In the tachyphylaxis group, the initial lesion was classic in 27% of cases, occult in 51% of cases and both classic and occult in 11% of cases. When comparing the tachyphylaxis group with the 20 comparable patients from the response group, the mean lesion size was 10.87 mm2 (0.8–49.5 mm2) and 10.95 mm2 (2.21–33.5 mm2), respectively.
The median period of injection-free interval was 2 months (2–18 months).
The mean BCVA at the time of reactivation was 62 (35–83) ETDRS, which decreased to 47 (17–70) despite repeated monthly injections.
The mean CRT at reactivation was 297 μm (151–539 μm), which increased to 358 μm (183–700) regardless of monthly injections.
The mean follow-up period was 24 months (13–37 months) and the mean number of total injections after reactivation was 13 (6–19).
The introduction of ranibizumab for the treatment of exudative AMD has changed the prognosis of the disease from a condition that would cause continuous inadvertent vision loss to social blindness (if both eyes are affected) if left untreated to a possible steady state or even vision gain.
However, some patients seem to have a decreased response after repeated treatments while other patients seem to not respond at all. No definite explanation for this so-called non-response has been put forward and no prognostic clinical characteristics have been identified.
Therefore, it can be difficult to determine when to stop treatment, if the treatment proves to be ineffective. One of the causes for the lack of efficacy of a treatment is tachyphylaxis, but it is still unknown whether tachyphylaxis does develop when using ranibizumab for the treatment of exudative AMD.11–13
In our study, we have identified 20 (2%) patients who have developed resistance to this treatment. At baseline, these patients had a comparable lesion composition and size with the responding group. Patients responded initially adequately to the treatment with resolution of the CNV, but with reactivation of the disease, they have shown no response to treatment with intravitreal ranibizumab.
This decrease in the bio-efficacy of ranibizumab may be caused by tachyphylaxis.
In general, there are several factors that influence the interaction between a pharmacological substance and the target tissue. Factors that play a role in the development of tachyphylaxis are drug administration frequency, dose, receptor expression on the target tissue and presence/development of antagonistic drugs/antibodies.
The treatment of exudative AMD today is a monotherapy with intravitreal ranibizumab, and many patients might be treated with ranibizumab for a long time resulting in a high number of injections, potentially increasing the risk of development of tachyphylaxis.14
One possible mechanism for the development of tachyphylaxis during the treatment of AMD could be the neutralisation of ranibizumab due to the formation of circulating antibodies. Even though ranibizumab is a humanised monoclonal antibody, it might be immunogenic and repeated injections might induce neutralising antibodies.10 This assumption is supported by previous studies, which have demonstrated that antibodies against ranibizumab do develop after intravitreal injections and that the levels of antibodies increase after repeated injections.10 This is also the case after intravitreal injections of bevacizumab.15 16 Patients with AMD who developed a clinical course suspicious of tachyphylaxis development after the administration of bevacizumab have been described, and the patients seem to have higher titres of circulating antibodies against bevacizumab.12 15
Another potential mechanism for the lack of sensitivity to treatment could be the desensitisation of the target tissue to the drug itself. This desensitisation could be caused by upregulation of the receptors on the target tissue so that the amount of intravitreal ranibizumab is not enough to shut down the process. Previous studies with bevacizumab have, however, shown that in cases of non-response this could not be overcome with increasing dosage.12 Another mechanism could be upregulation of the production of VEGF by macrophages, which have been implicated in the pathogenesis of CNV.17 Recent findings of increased density and activity of macrophages in surgically excised human CNV membranes from patients who had previously received treatment with intravitreal bevacizumab support this hypothesis.18
Another possible explanation could be that the reactivation of the CNV is driven by another pathway. Even though this has never been studied in a given patient, other potential molecules have been implicated in the development of CNV including ICAM-1, E-selectin, CD44 and basic and acidic fibroblast growth factors, and the development of drugs targeting these molecules could be a possible answer.5 6 19 This might be overcome by combining the treatment with other pharmaceuticals aimed at other pathways in the formation of CNV. Recent studies with bevacizumab have reported that tachyphylaxis was partially alleviated with the subsequent use of preservative-free triamcinolone, which suggests that the tachyphylaxis might be prevented by using pharmaceuticals with different modes of action.11
The previous studies reporting tachyphylaxis during the treatment with bevacizumab required continuous injections as an entry criterion.11 If tachyphylaxis increases with the number of injections, we would expect less development of tachyphylaxis when using the Pronto protocol than the continuous use of injections. Within the field of asthma, it has previously been reported that continuous treatment with β2-agonists has a negative effect on the treatment results compared with treatment as necessitated.20
We have shown in this report that approximately 2% of patients with exudative AMD fail to respond to anti-VEGF treatment, when being treated for a reactivation of the CNV. We propose that this is caused by tachyphylaxis. However, our study is purely observational and retrospective, hence it does not providing an exact mechanism for the development of this phenomenon. Both local and systemic factors might be involved, and a thorough understanding of these factors and their impact may help establish treatment strategies for this subgroup of patients.
Linked article 204685
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.