We evaluated a large population of patients in the NMO spectrum, and compared FD-OCT findings of NMO and MS patients with and without a history of ON. Our data confirmed the ability of OCT to quantify axonal loss in these conditions. In MS + ON, a significant loss of RNFL compared to normals was observed with regard to overall average and all RNFL sectors except for the nasal, in agreement with previous studies reporting reduced average
1,2,19–22 and sectoral
2,5,20,23,24 RNFL thickness measurements in MS eyes with ON. We also found subclinical RFNL loss in MS eyes without ON history in accordance with previous studies using TD-OCT.
2,3,5,9,19,21,22,24,25
Our study also confirmed that MS patients have significantly reduced macular thickness measurements.
3,5,11,26 Interestingly, however, OCT abnormalities in MS patients were much more pronounced in macular than RNFL thickness measurements (
Fig. 3 and
Table 2). The ROC curve analysis also indicates greater discrimination ability for macular thickness than for RNFL thickness measurements in MS patients (
Table 3). Since studies evaluating OCT measurements indicate that macular thickness measurements are not more efficient than RNFL thickness measurements at detecting neural loss in different optic neuropathies,
27,28 we believe the greater discrimination ability observed for MS patients in our study is an indication that macular thickness measurements reflect the reduction not only of RGC, but also of other retinal layers. This interpretation is in accordance with recent data indicating that MS patients can present primary damage to retinal layers irrespective of RGCs, including the inner nuclear layer atrophy.
11,29
Our findings provide new information regarding the use of FD-OCT measurements in eyes of patients in the NMO spectrum. They also confirm the occurrence of greater loss of RNFL thickness in NMO than in MS patients (
Fig. 3,
Table 2), as observed in previous studies using TD-OCT technology.
3,6,30–33 These findings are supported by clinical findings of greater reduction in visual acuity and VF in NMO.
34 On the other hand, the results of our macular thickness measurements provided important new information. To our knowledge, only Ratchford et al. evaluated macular thickness measurements in NMO patients, and found a significant difference in average measurements compared to normals and MS patients.
3 In our study, however, while a significant reduction in all macular measurements was found when compared to normals, the comparison between NMO and MS patients yielded somewhat unexpected findings: while RNFL thickness measurements were significantly lower in NMO than in MS + ON, the two groups did not differ with regard to macular thickness measurements (
Table 2). We believe this finding is a further indication that in MS patients macular thickness reduction is the result of damage to retinal layers other than the RGC layer, as observed in previous histopathologic and OCT studies.
11,29
Our study also is important in that we compared MS and NMO patients by analyzing eyes without a history of ON. To do so, we evaluated 54 eyes of 28 patients with LETM (considered part of the NMO spectrum). Our data indicated that, like eyes affected with MS, eyes in the NMO spectrum also presented subclinical reduction of the RNFL. This finding differs from previous studies claiming that subclinical axonal loss occurs in MS but not in patients in the NMO spectrum—a fact that presumably would make it possible to differentiate the two conditions. Thus, while subclinical axonal loss on OCT in eyes of patients with MS without a history of ON is well documented, the possibility of subclinical axonal loss in eyes in the NMO spectrum is not acknowledged generally. Two previous studies found no retinal axonal loss in eyes with LETM, but only a small number of patients were investigated. Ratchford et al. found no significant reduction in the RNFL thickness of non-affected eyes of patients with NMO or LETM.
3 They suggested that OCT could help differentiate non-affected eyes of LETM or NMO patients from eyes of patients with MS, but only average RNFL thickness measurements were compared and the sample of NMO or LETM patients with non-affected eyes was small (8 and 17, respectively). Likewise, de Seze et al. found no subclinical RNFL loss in a sample of 8 non-affected eyes of patients with LETM and positive anti-NMO antibodies.
30 However, in a previous study of 17 LETM patients using TD-OCT, we found subclinical abnormalities in one quadrant and in one 30-degree segment of the disc.
35 In the present study, using a larger sample (28 patients, 54 eyes) and FD-OCT equipment, we found a significant reduction in RNFL thickness measurements in several parameters (
Table 2). Therefore, the presence of subclinical retinal axonal loss should not be used in the differential diagnosis of eyes with MS and eyes in the NMO spectrum.
Our current study also investigated the relationship between FD-OCT and SAP in MS and NMO patients using global and sectoral measurements. While several investigators have found a strong correlation between OCT measurements and visual acuity, contrast sensitivity, or frequency doubling perimetry,
2,5,7,10,14,22,31,36,37 few have investigated the relationship between OCT and SAP. In a sample of MS patients, Henderson et al. found a significant correlation between VF mean deviation in dB and average RNFL thickness or macular volume (
r = 0.31,
P = 0.038, and
r = 0.34,
P = 0.022, respectively).
5 Four other studies found correlation coefficients between average RNFL (but not macular) OCT measurements and global MD VF loss in MS patients ranging from 0.22 and 0.51.
7,24,25,36 Cheng et al. evaluated the relationship between average or quadrantic RFNL thickness loss and global or sectoral VF loss, and found a strong association between OCT measures and VF in all sectoral measurements evaluated, particularly between the inferior optic disc quadrant and the corresponding VF loss in the superotemporal area (
r = −0.65).
8 Macular thickness measurements were not investigated. Only one study evaluated the correlation between OCT measurements and VF loss in NMO patients, and reported a significant correlation (
r = −0.78), but provided no details on how the correlation was assessed.
30
In our study, significant differences in structure-function relationships were observed between MS and NMO, mainly in macular thickness measurements. Thus, when evaluating the relationship between RNFL thickness and VF loss, the correlation coefficients were slightly greater for eyes in the NMO spectrum (from 0.36–0.57, depending on the sector) than for MS eyes (from 0.10–0.36), matching the results of several studies in MS patients,
5,7,8,24,25,36 and one in NMO patients.
30 However, the difference between the two diseases was striking when investigating the correlation between VF and macular thickness measurements. The corresponding correlation coefficients were 0.36–0.52 for eyes in the NMO spectrum and 0.06–0.25 (frequently non-significant) for MS eyes. We believe the discrepancy may be ascribed to differences in retinal pathology between the two conditions. The presumption that reduced macular thickness in NMO patients is the direct result of RGC atrophy from axonal damage in the optic nerve would explain the similarities in correlation coefficients of the two types of measurements. In relapsing-remitting MS patients, however, recent studies indicate that macular thickness measurements are influenced by structures other than the RGCs.
11,29 Future studies using segmental analysis of FD-OCT macular thickness measurements and larger samples are necessary to confirm this finding.
In conclusion, our study shows that, while OCT-measured axonal loss usually is more severe in NMO patients, subclinical neuronal loss can occur in both diseases and, therefore, should not be considered in differential diagnosis. Axonal loss appears to be equally detectable in RNFL and macular measurements in NMO patients, but more evident in macular thickness measurements in MS patients. Accordingly, the two conditions differed significantly regarding the relationship between retinal axonal loss and VF deficit. Clinicians should take these differences into account when using OCT measurements for diagnosis and follow-up of MS or NMO patients.