Risk Factors for Geographic Atrophy in the CATT Trial

Kanish Mirchia, MBBS; Sophie J. Bakri, MD


September 19, 2014

In This Article

Study Findings

Regardless of dosing regimen, patients treated with ranibizumab had a 43% higher incidence of GA, which was a statistically significant increase compared with bevacizumab. The ranibizumab group had more complete resolution of fluid but similar final VAs as the bevacizumab treatment arm.

Examination of the dosing regimens for both drugs yielded interesting data. Specifically, monthly treatment (mean total injections, 22.5) with either drug showed a 59% increase in the formation of GA (P = .003) compared with the as-needed group (mean total injections, 13.1). This increased risk for atrophy is further supported by the fact that patients who were re-randomized at one year from monthly to as-needed injections had a lower incidence of GA in the second year (7%) compared with the first year (12.7%).

Baseline VA and presence of GA in the fellow eye showed one of the strongest associations with risk of developing GA (2.65 and 2 times increased risk, respectively). Low VA could reflect more damage to the photoreceptors and other retinal layers, thus increasing the risk for atrophy. Similarly, the presence of retinal angiomatous proliferans lesions was associated with a 1.69 times greater risk of developing GA. This could also be related to the presence of RPE detachments that resolve quickly with treatment, leading to RPE atrophy.

When analyzing the anatomical factors in the retina, intraretinal fluid (in the fovea or outside) increased the risk of developing GA by a factor of 2, whereas subretinal fluid or thicker sub-RPE was associated with a reduced risk. Persistent vitreomacular attachment also halved the risk of developing GA, which could be related to a simple physical phenomenon of supporting the retinal layers or could result from altered diffusion of the anti-VEGF agents. Blocked fluorescence, which probably represents fibrosis at the edge of choroidal neovascularization, was also associated with lower incidence of GA.

These study findings are also summarized in the Table.

Table. Factorsa That Alter the Risk for GA

Characteristic Risk Percentage P Value
Study drug   .02
Bevacizumab 1.00  
Ranibizumab ↑ 43%
Drug regimen   .003
As-needed 1.00  
Monthly ↑ 59%
Switched --
Baseline VA (study eye)   .007
20/25-20/40 1.00  
20/50-20/80 ↑ 66%
20/100-20/160 ↑ 70%
20/200-20/320 ↑ 165%
GA in fellow eye   .0003
None/questionable 1.00  
Present ↑ 107%
CNV lesion
Blocked fluorescence   .007
No 1.00  
Yes ↓ 51%
RAP lesion   .007
No 1.00  
Yes ↑ 69%
SRF thickness fovea   .006
0 µ 1.00  
> 0 to ≤ 25 µ ↓ 31%
> 25 µ ↓ 48%
SRT thickness fovea   < .0001
> 0 to ≤ 75 µ 1.00
> 75 to ≤ 160 µ ↓ 37%
> 160 to ≤275 µ ↓ 47%
> 275 µ ↓ 69%
Intraretinal fluid   .006
No fluid 1.00  
Fluid not in foveal center ↑ 80%
Fluid in foveal center ↑ 110%
Vitreomacular attachment   .04
No 1.00  
Yes ↓ 45%

CNV = choroidal neovascularization; GA = geographic atrophy; OCT = optical coherence tomography; RAP = retinal angiomatous proliferans; SRF = subretinal fluid; SRT = subretinal tissue; VA = visual acuity. a P < .05 in multivariate analysis.

Consistent with previous studies, CATT did not find a definitive relationship between the AMD risk alleles and GA. There was a protective trend for the TLR3 single-nucleotide polymorphism, but additional studies on larger populations are needed to confirm this effect.

The two-year results from the IVAN trial[1] did not demonstrate a difference in incidence between bevacizumab and ranibizumab for GA. Future studies in different population groups would need to include GA as an endpoint to have a clearer idea of the disease.

Histologic studies should also be included wherever feasible to better understand the RPE atrophy seen after anti-vascular endothelial growth factor (VEGF) therapy. This would help elucidate why the presence of sub-RPE fluid and increased subretinal tissue thickness were protective against the development of GA. Histologic evaluation or research models could also look into the mechanism by which ranibizumab promotes formation of GA and determine whether excessive drying or rapidity of drying of the retina is a factor.

Recent studies have indicated that VEGF has important functions in the normal milieu of the eye, including the maintenance of normal retinal vasculature and choriocapillaris. It is plausible that chronic inhibition of VEGF impairs normal housekeeping functions, which leads to development of retinal atrophy.

Strengths and Limitations of the Study

The strengths of this study are its large sample size and evaluation of an extensive set of risk factors within a well-defined protocol. One factor that could have introduced some error is the assessment of GA using a combination of color fundus photography and fluorescein angiography, instead of autofluorescence and spectral-domain optical coherence tomography, which are ideal to detect GA lesions.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: