To the Editor:
I believe Dr. Ron Gallemore's article [When Anti-VEGF Treatment Fails, Retinal Insider, March 2008] barely touches the surface of the anti-VEGF failure controversy. First, it is misleading to state that 95 percent of patients are helped by anti-VEGF medications. While this number is corroborated by ANCHOR and
Anti-VEGF failure can only be analyzed by utilizing simultaneous high-speed IVFA and ICG videoangiography. Classification can then be initially made into primary failures (those who don't respond to anti-VEGF treatment), and secondary failures (those who incompletely respond to treatment and those who initially respond and subsequently develop increased leakage).
Primary failure occurs with patients who have arteriolarization of neovascular capillaries, where smooth muscle and pericyte development produce vessels that are unresponsive to anti-VEGF medication, i.e., the pathological process is not "VEGF-driven." This condition is commonly found in patients with long-term NV, especially after PDT therapy. There is an associated fibrovascular component but not a frank disciform scar. Occasionally, a patient presents with an initial lesion that has this component. The OCT evidence of this process is easily seen with RPE thickening and more importantly, as a characteristic high-flow lesion on ICG.
Secondary failure is more common. It is found in patients who had successful anti-VEGF therapy but later develop mature vessels due to incomplete VEGF suppression. Another common cause is polypoidal vasculopathy, which is also not a VEGF-driven process. A last cause is peripapillary NV especially if it follows previous macular photocoagulation. The ICG typically will reveal these arteriolarized vessels.
Patients who have persistent leakage after anti-VEGF therapy are treated with triple therapy: ICG-directed PDT, intravitreal Kenalog and anti-VEGF therapy. ICG-directed PDT focuses the treatment on the origin of the NV and/or polypoidal lesion in a very limited manner (1.1-mm circle). By limiting the spot size and duration, acute vision loss has not occurred. Scott Cousins and I presented this procedure at the ASRS. It has no risks beyond those of normal intravitreal injections and has benefits of decreasing the frequency of treatment (average duration of treatment six months), drying the last vestiges of leakage, and improving the vision (proportional to duration of lesion—naïve lesions do very well).
Therefore, the exudative treatment algorithm should state that the initial step is to define the anatomy of the lesion by the appropriate angiography. Next, classify the lesion as VEGF-responsive/non-responsive, and last, employ proper treatment with constant monitoring as the vascular maturation process might demand a redeployment of strategies during the course of the disease.
Mark Nelson, MD
In reply:
I thank Dr. Nelson for his valued comments. Your discussion supports the fundamental concept of my review that combination therapies play a central role in the management of anti-VEGF failures. You also propose that all treatment failures should be evaluated with videoangiography with the goal of sub-classifying these lesions into primary and secondary failures. Specifically, you propose that primary failures are the result of arteriolarization of CNV feeder vessels,1,2 which are non-responsive to anti-VEGF therapy, while secondary failures are the result of lesions that have more mature vessels than other CNV and are less responsive to anti-VEGF therapy. In the end, you still propose a variation of combination therapy for the management of all such failures. The concept of sub-classifying treatment failures is intriguing and warrants further investigation, but even your classification scheme may prove to be a simplification of the complex nature of exudative AMD in general and treatment failures in particular.
Additional factors like the oxidative environment associated with the CNV (e.g., use of antioxidant supplements), the pressure differential associated with CNV flow, the surface area of the lesions and the presence of microvascular co-morbidities (diabetic retinopathy, LDL levels, etc.) may also play a critical role in characterizing anti-VEGF treatment failures. Heavy smokers, for example, appear to respond more poorly to all treatments, even combination therapies, and this "sub-classification" of treatment failures ("heavy smoker") may alter the prognosis.3 Some reports, including that of Dr. Cousins, have shown feeder vessel sensitivity to anti-VEGF treatment despite apparent arteriolarization of such vessels.4,5 Thus, not all lesions may be easily characterized. Your utilization of ICG-directed PDT treatment as part of your combination therapy regimen is also of interest. While I have utilized this approach in select patients, I'm not convinced that it is superior to low-fluence treatment of the entire CNV complex. In addition, identifying a definite focus to treat on ICG studies with or without videoangiography is not always possible. I do believe that refining the area of PDT application as we develop experience with combination therapy may prove to be important and may include a decrease in treatment margins, focal treatment of hot spots on ICG, and treatment of feeder vessels on videoangiography. Since treatment failures are relatively uncommon and a large randomized study of different treatment algorithms will be difficult to conduct, personal experience may be the most important information we have on managing these complex cases. At this time, when a rigorous regimen of anti-VEGF treatment administered every four to six weeks fails to control leakage and the patient is not eligible for a clinical trial, some form of combination therapy with anti-VEGF drug, Visudyne and steroid should be considered to stabilize or improve vision in these complex cases.
Ron P. Gallemore, MD, PhD
1. Yamamoto Y. Measurement of blood flow velocity in feeder vessels of choroidal neovascularization by a scanning laser ophthalmoscope and image analysis system. Jpn J Ophthalmol 2003;47:53-8.
3. Schmidt S, Hauser MA, Scott WK, Postel EA et al. Cigarette smoking strongly modifies the association of LOC387715 and age-related macular degeneration. Am J Hum Genet 2006;78:852-64.
4. Theodossiadis PG et al. Intravitreal bevacizumab for the treatment of feeder vessel of subfoveal choroidal neovascularization. Eur J Ophthalmol 2007;17:853-6.
5. Cousins SW. Combination Instrument Enhances Patient Care. Retinal Physician, Jan. 2008.
