In this index study, we aimed to characterize the various lesions seen clinically at different stages in PIC based on findings visualized on SD-OCT. The eye tracking capability of the Spectralis® SD-OCT has allowed for exact (in location) comparisons of lesions over time. The majority of lesions (24 of a total of 27) in our study, irrespective of clinical activity, showed involvement of the RPE. RPE changes included RPE elevation with sub-RPE signals and RPE discontinuity. In the first description of PIC lesions, Watzke et al. described serous retinal detachments occurring over choroidal lesions [1]. In our review of 27 lesions, the most commonly observed morphology of lesion on SD-OCT involved RPE elevation, with sub-RPE signals localized over an apparently intact BM. Furthermore, Watzke et al. described that serous RPE detachments occurred either at the first appearance of a new lesion or during recurrence [1]. Our findings are consistent with the clinical observations made by Watzke: clinically active patients in our study had lesions that demonstrated RPE elevation with sub-RPE signals.
Temporal SD-OCT changes were observed in all clinically active patients and also in seven lesions of two other patients who were not identified to be clinically active. Temporal SD-OCT changes in four of the five lesions in clinically active patients were very similar to the changes observed in seven lesions in the two clinically stable patients. There could be several explanations for such observation. One possible explanation is that the high-resolution property of SD-OCT makes it a sensitive imaging modality which may identify changes before they become clinically detectable. Such hypothesis is further supported by the fact that SD-OCT changes in three lesions of Patient 3 were visible up to 21 months before identification of clinical activity.
PIC is usually a benign disease but has the potential to cause severe visual loss if complicated by CNV and sub-retinal fibrosis: rates of up to 69% and 56%, respectively, were reported in our study of 77 PIC patients, and 75% of patients in a series of 12 patients had CNV [12, 13]. Essex et al. studied clinical outcomes of 271 eyes with typical and atypical PIC. Typical PIC included eyes with small lesions (less than or equal to a quarter of a disc diameter) located within the arcades, while atypical PIC included eyes with larger lesions located within and outside the arcades. No significant differences in demographics and clinical outcomes were identified between the two groups. They reported that presence of PIC lesions in fellow eyes of patients with unilateral CNV was a risk factor for development of CNV in the fellow eyes as well [14], hence emphasizing the importance of immediately identifying and treating any recurrence in disease activity. Previously, we have shown that the employment of IMT such as mycophenolate mofetil may help to decrease the frequency of attacks in recurrent PIC disease [4]. Our current results indicate that SD-OCT may play a role in early identification of recurrence of disease activity and development of new lesions.
An interesting phenomenon was noted while we were analyzing the SD-OCT scans. Patients 2, 3, and 7 had several lesions with RPE elevation and non-visibility of the overlying PRs. Review of latter scans of the same lesions revealed that the RPE elevation gradually resolved, and the PRs became visible. A possible explanation for this phenomenon may be that in the initial stages, there is elevation of RPE and compression of PRs. Resolution of the RPE elevation restored the PRs to their original position. The non-visibility of PRs persisted during follow-up in three lesions of Patient 5. Spaide et al. demonstrated loss of PRs in AZOOR complex of diseases [9]. They described that the recovery of PRs on SD-OCT was linked to the preservation of the ONL. Such prediction is possible in cases of isolated loss of PR as seen in Patient 5. However, in lesions with RPE elevation, visualization of the ONL and prediction of PR recovery based on its appearance is quite impossible.
In our scans, we were unable to identify changes in the choroid in all except one lesion. It was surprising because the first report from Watzke mentioned choroidal involvement of lesions, and eyes with PIC have been shown to have involvement of choriocapillaris on ICG [1–3]. The lack of choroidal involvement raised questions regarding the origin of the pathologic process in PIC: does the process begin in the choroid or does it involve the choroid in later stages? In Patient 3, lesion 9 began with an intact BM, but in later scans, the BM appeared to be disrupted, suggesting that in the initial stages, the pathological process was localized above the choroid, and only after disruption of the BM, it progressed into the choroid. However, there are limitations to such finding in our study. It has been reported that in order to optimally visualize the choroid, enhanced depth imaging (EDI) technique should be employed [15]. Our scans were intended to optimally view the retinal layers; therefore, we cannot make a definite conclusion about the extent of choroidal involvement based on OCT scans. Also, our study was a retrospective review, and it is difficult to draw definitive conclusions about the disease process.
The Japanese literature has previously reported TD-OCT findings of four patients with PIC disease [16, 17]. The investigators reported the formation of what they described as a “hump” at the site of the active PIC lesion and changes in reflectivity of the RPE-BM and choriocapillaris complex as the lesion progressed over time. It is difficult to compare their findings to ours as TD-OCT, at best, provides minimal detail on morphology of individual retinal layers. Stepien and Carroll described SD-OCT changes in an eye with PIC disease. At the time of clinical activity, there was homogenous outer retinal thickening overlying chorioretinal lesions which resolved 3 months after treatment. Although the descriptive terms used by them are not the same as the terms employed in our report, the SD-OCT images presented by them during development and resolution of disease activity are strikingly similar to ours [18].
The definitive pathophysiology of white dot syndromes has been much debated throughout the years [19]. SD-OCT may provide further evidence to distinguish among these disorders. Spaide et al. described lesion characteristics of 13 eyes with multifocal choroiditis with panuveitis (MCP) on SD-OCT [9]. Although they demonstrated loss of PRs, none of the eyes demonstrated RPE elevation as seen in our study of PIC patients. Could such finding be further evidence that PIC and MFC are different disease entities [20]?
One of the hypotheses for the pathogenesis of PIC is that the disease is secondary to the development of myopic cracks in BM. However, PIC is mostly seen in patients with moderate myopia. The assessment of integrity of BM was limited in some lesions due to the abnormal reflectivity of light by adjacent structures. However, when clearly visible, it appeared to be intact in the majority of the lesions in our study. Review of SD-OCT scans of Patient 3 demonstrated destruction of the BM on later scans when compared to earlier scans.
SD-OCT may be able to provide detailed structural characteristics of PIC lesions and may serve as a useful, non-invasive means of identifying disease activity in patients with PIC. RPE elevation is noted in many lesions, while BM and choroid are spared. PRs appear to be compressed during RPE elevation and resume clear visibility upon resolution of the RPE elevation. Therefore, SD-OCT may provide information on disease activity that is not detected clinically or on FA. Further systematic, prospective SD-OCT studies of PIC lesions will most likely provide additional insights about the clinical course of PIC, and may provide clinicians with a non-invasive imaging tool in the management of patients with PIC.