• Degeneration
• neovascularisation

Tachyphylaxis and the development of drug tolerance have both long been described as phenomena causing reduced drug efficacy. But is there a difference between the two?

Tachyphylaxis can develop quite quickly when drugs are used repeatedly over a short period, with no response occurring when the dosage is increased. However, efficacy can be restored if the medication is stopped for a short while.

In contrast, tolerance is characterised by a slow loss of efficacy over time. The effect of the drug can be improved if the dosage is increased or given over shorter time intervals, but efficacy is not restored if the treatment is halted temporarily.1

In the literature both words are used synonymously, for example in relation to anti-TNF-α biopharmaceuticals (infliximab) in chronic inflammatory diseases (including rheumatoid arthritis) where immunogenicity is suspected.2 In diabetes research where rapid tachyphylaxis at the level of gastric nervous activation was described3 and during treatment with salmeterol (a bronchodilator for chronic obstructive pulmonary disease), β2-receptor down-regulation was assumed to be a causative factor.4 Furthermore, tachyphylaxis is known to occur with chronic aerosol use in patients with asthma,5 and with drugs used for analgesia and anaesthesia.6

In the ophthalmic literature, tachyphylaxis/tolerance is reported with the use of selective α-2 receptor agonists (brimonidine) due to up-regulation of α-2 receptors because the basal norepinephrine level in the synaptic junctions is chronically depressed.7

The worldwide use of intravitreal application of anti-vascular growth factor (a-VEGF) in patients with neovascular age-related macular degeneration (AMD) and the realisation that regular applications over long periods of time are necessary to maintain vision in these eyes, has revealed the problem of tolerance/tachyphylaxy.

A possible mechanism causing decreased drug response in neovascular AMD could be alteration of the neovascular membrane including increased fibrosis acting as a resorption barrier, change of lesion type, chronic changes in the vessel wall and changes in important neighbouring structures such as photoreceptors or retinal pigment epithelium. The increased expression of VEGF due to increased macrophages in choroidal neovascularisation (CNV), increased expression of VEGF receptors, changes in signal transduction or a shift of the stimulus for CNV growth towards other growth factors is known as pharmacodynamic tolerance. A systemic immune response or the development of neutralising anti-bodies is known as pharmacokinetic tolerance.8 As bevacizumab and ranibizumab are both mouse-derived, humanised monoclonal antibodies, a systemic immune response can be expected with repetitive treatment.

Indeed, increased numbers of active antibodies over time were reported after ranibizumab treatment in the ANCHOR study. At baseline, immunoreactivity rates were 0.9% in the group receiving 0.3 mg ranibizumab, 0% in the group receiving 0.5 mg ranibizumab and 0.5% in the sham injection group. After 2 years in 4.4% of patients in the 0.3 mg group and 6.3% of those in the 0.5 mg group active antibodies were found in comparison to only 1.1% in the sham injection group.9

However, bevacizumab levels have also been detected systemically with intravitreal application.10 In 2007, two papers suggested for the first time possible tachyphylaxis/tolerance with chronic ranibizumab11 and bevacizumab treatment.12 The rationale for combination therapies with drugs with different modes of action such as, for example, intravitreal triamcinolone (IV-TRIAM) or photodynamic therapy, was suggested by Schaal12 and others.8 13 How frequently does tachyphylaxis/tolerance occur? Does it occur less frequently with ranibizumab than with bevacizumab or vice versa? In 2009, Forooghian and coworkers described a 10% rate (6/58) of tachyphylaxis/tolerance in eyes treated with bevacizumab with no response in the treated eyes when the dosage was increased.14 In this issue, Maria Salling Eghøj and Torben Lykke Sørensen from Denmark report a 2% incidence (20/1076 eyes) in 976 patients undergoing ranibizumab treatment for AMD over a 3-year period (see page 21). After an initial loading dose of three injections of 0.5 mg ranibizumab, patients were checked monthly to determine if the retina had become dry and the disease inactive (Pronto guidelines). ‘As needed’ treatment protocols are recommended and combination with other drugs is also suggested as a preventive measure.15

Do we need to differentiate more clearly between tachyphylaxis and tolerance so as to improve treatment strategies and respond better once the drug starts to fail? The options we have are: (1) to increase the dosage or shorten treatment intervals if tolerance has developed; (2) to pause treatment if tachyphylaxis has occurred; (3) to combine drugs with different modes of action; or (4) to switch to a similar drug with different properties (bevacizumab and ranibizumab differ in molecular size, affinity and absorption).

In their retrospective study, Gasperini16 and co-authors identify 26 eyes of 25 patients primarily treated with either bevacizumab or ranibizumab that showed an attenuated drug response (see page 14). The treatment was changed to the other anti-VEGF antibody and in 81% of cases, the switch to the other drug was successful in continuing to decrease fluid. Unfortunately, the number of all patients treated is not given, so that the percentage of the 25 described cases experiencing tachyphylaxis in their study is unclear. Sixteen eyes treated primarily with ranibizumab showed loss of response compared with 10 eyes treated primarily with bevacizumab. No difference was detected concerning the type of neovascularisation. Treatment intervals were 4 weeks for ranibizumab and 6 weeks for bevacizumab initially, followed by a PRN (as needed) strategy. The average follow-up was 13 months. Some eyes developed an attenuated response quickly after only two injections, others after 10 or more. Following the drug switch, some cases responded immediately to one or two injections, while some needed eight or more.16

Given that bevacizumab and ranibizumab are similar protein molecules and act at the same location, the fact that switching these drugs showed efficacy in such a high percentage of cases is surprising.

In general, drug tolerance can develop in many different ways and for different reasons. If we can determine the reason for loss of response to repeated treatment more precisely, our treatment strategy could become more tailored. Prospective designs, clearer definitions and comparable treatment strategies in ophthalmology will enable us to better ascertain drug response in both anti-VEGF treatment and chronic drug use.