Abstract
Purpose: :
To compare the efficacy of PERG and PhNR to detect early glaucoma.
Participants: :
some glaucoma cases had a normal PERG but abnormal PhNR and vice versa. An optimal linear combination of PERG and PhNR yielded an ROC area of 0.91.
Methods: :
We examined 30 eyes of 16 patients with glaucoma (12 POAG, 2 pigment dispersion, 2 PEX, mean MD -4.2±6.5, age 62±6 yrs) and 29 eyes of 17 normal subjects (age 56±6 yrs, mean MD 1.2±1.4). The PERG was evoked by 15 rps pattern-reversal of 0.8° and 16° checksize at full contrast. PhNR was evoked by red flashes of 0.01-4.0 cd·s/m² on a blue background of 10 cd/m².Results: Clear ERG and PERG records were obtained from all eyes. The identification of the PhNR was not always unique, it occurred around 70 ms. As expected, the PERG amplitude was sizably reduced in glaucoma for the 0.8° checks, less so than for 15° checks. The 0.8°/15° amplitude ratio reduced some of the interindividual variability. The PhNR was also sizably reduced in the glaucoma group; the group differences for PERG and PhNR were highly significant (p<0.001). Correlation of MD with the PERG ratio was significant: p=0.003, r=0.42. For the PhNR the correlation was not significant (p=0.09) and shallower (r=-0.25). Single-patient diagnostic efficacy was assessed via ROC analyses. For the PERG, the ratio performed best with an area-under-curve of 0.85; PhNR had 0.81. The most interesting finding was a dissociation (lack of correlation) of PERG and PhNR findings across
Conclusions: :
It is well known that both PERG and PhNR are affected by glaucoma. We here report on a dissociation of the two measures, suggesting that they represent different, yet unidentified disease mechanisms. Furthermore, this dissociation might allow for a substantial improvement of glaucoma diagnosis by combining PERG and PhNR results.
Keywords: electroretinography: clinical