This is the first study in which the change in OCTA parameters based on the presence of OH in PD patients was investigated. PD patients had significantly lower retinal vessel densities than healthy controls, and the difference remained significant when comparing the PDOH + with control groups but was not significant in the PDOH − group. Furthermore, the amount of DBP change correlated with the vessel density in the DRCP central area. The results indicate the retinal vessel densities measured using OCTA are significantly associated with the presence of OH in PD patients.
Growing evidence shows that alterations of vasculature contribute to the development and aggravation of neurodegenerative diseases including PD7,13,14. OCTA is an emerging non-invasive technology used to investigate in vivo retinal microvascular alterations in various neurological disorders15. Robbinson and colleagues suggested that OCTA parameters, especially the inner ring of the macula, are a potential biomarker for the diagnosis of PD9. In recent studies, OCTA parameters were suggested a biomarker for diagnosis and associated with disease progression and cognitive decline16,17. In contrast, Rascuna et al. reported that PD patients showed no significant differences in OCTA parameters in DRCP or SRCP compared with healthy controls10. When the OCTA parameters in PD, essential tremor, and healthy controls were compared, differences were not observed among the three groups11. These inconsistent results indicate that a range of factors can affect the OCTA findings. Based on the present study results, the presence of OH is presumed a crucial factor associated with the OCTA parameters in PD patients. When the PD patients were divided into two groups based on the presence or absence of OH, the PDOH + group had lower intraretinal vessel densities in the SRCP and DRCP compared with the PDOH − and control groups. The results indicated intraretinal microvascular changes observed in PD can be associated with OH in addition to the direct association of PD, and OCTA can be useful for their detection.
OH is a common non-motor symptom in PD and induces recurrent episodic cerebral and retinal hypoperfusion3. Recurrent episodic cerebral hypoperfusion may result in cerebral microvascular and white matter damage. PD patients with OH had increased white matter hyperintensities on brain magnetic resonance imaging (MRI) compared with the patients without OH18, and chronic cerebral hypoperfusion was associated with microvascular pathology in the brain of PD mouse model19. In this context, OH induces retinal microvascular damage in addition to cerebral microvascular damage. In terms of retinal vessels, the microvascular damage may be associated with hypoperfusion induced by carotid artery stenosis as well as systemic hypoperfusion. For example, the OCTA metric changes are associated with the intradialytic hypotension episodes in chronic haemodialysis patients6. Furthermore, early changes in retinal microvasculature had a predictive value for the development of systemic vascular disorders20,21. However, because a control group with OH was not included in the present study, whether the changes are PD-specific results remains unknown. Further studies that include controls with OH and patients with other diseases that show OH (drug-induced OH, pure autonomic failure or multiple system atrophy) can aid in understanding the relationship among OH, PD and retinal microvascular damage.
In the present study, significant changes were found in superficial and deep retinal vessel density only in the central macular area, indicating the central retinal area is predominantly affected in patients with PD. The results are in agreement with previous studies in which the central retinal area was only or more greatly affected in PD patients9,12,22,23, indicating the retinal vessels in the central area are more vulnerable to hypoperfusion and disease. In addition, thinning of the macular inner retinal layers, an emerging biomarker of PD, reportedly occurs mainly in the parafoveal area in the early stage of PD24. This result supports the assumption that the central macular area is the most affected or vulnerable retinal area in PD patients24. Further research is needed to corroborate this hypothesis.
In addition, the increased reduction in DBP during the head-up tilt (HUT) test was significantly associated with lower vessel density in DRCP. The retinal artery consists of two parallel vascular networks. The superficial vascular plexus consisting of approximately 75-µm vessels supplied by the central retinal artery and smaller deep capillary plexuses (20 µm) supplied by vertical anastomoses from the superficial vascular plexus25,26. The smaller vessels in DRCP may be more sensitive to hypoperfusion induced by OH than SRCP 25. The relationship between the DBP and retinal vessel density is in agreement with previous studies in which the association between the DBP and hypoperfusion or neuronal damage was analysed. In previous studies, lower DBP was reportedly associated with the progression of normal tension glaucoma,27 exacerbation of cerebral hypoperfusion and brain atrophy28, and lower DBP also contributed to the development of dementia29. A similar mechanism could occur in the retina. In summary, both the SRCP and DRCP are affected by OH in PD. The DRCP is more likely to be affected and the reduction of DBP might play an important role in the microvasculature damage of the retina in PD patients.
Increasing evidence shows the importance of the retina as a potential biomarker of early diagnosis and prognostication in PD24. In recent studies, parafoveal inner retinal change was shown to be detected from the early stages of PD, followed by progressive atrophy of pRNFL and macula over time24. In the present study, although the intraocular vessel density was significantly reduced in PD patients compared with controls, the pRNFL and macular retinal thickness, which represent the degree of neuroaxonal damage, did not show a significant difference between PD patients and controls, even in the PDOH + group. This finding indicates intraretinal vascular dysfunction may occur primarily in PD rather than secondary to macular atrophy. Studies on Alzheimer’s disease, another type of common chronic neurodegenerative disease, revealed various morphological abnormalities in retinal blood vessels, including disturbed blood flow dynamics, pericyte loss and amyloid beta-protein deposition30–36. In studies on neuroinflammatory disorders, such as neuromyelitis optica spectrum disorder, vascular changes reportedly also occur prior to the development of optic neuritis37,38. However, vascular changes may also be attributed to an innate difference in the sensitivity of measurements. The sensitivity of vessel density measurements using OCTA were suggested to possibly be higher than measurements of structural changes using OCT even in optic neuropathies where microvascular changes occur secondary to axonal damage39,40.
The present study had several limitations. First, the HUT test was not performed in the control group; therefore, healthy controls with OH cannot be omitted but healthy controls who had any neurological signs including dizziness, headache or orthostatic symptoms were excluded. Second, the vascular risk factors were investigated; however, carotid artery stenosis, supine hypertension and smoking history were not included in the study. Third, vessel density was measured using OCTA in which the angiographic signal was based on movement. However, many other factors, such as blood flow velocity, morphology and alterations in the vascular endothelial barrier, can compromise the measurement of perfusion. Therefore, false-positive findings cannot be excluded due to technical and methodological issues. Finally, the OCTA measurements in this study included large blood vessels. The changes in vessel density found in the present study were, therefore, a combination of changes in both the microvasculature and major vessels.
In recent studies, the retina and PD were shown to be highly correlated, and the importance of retinal microvascular change in PD is increasing as a candidate biomarker for the development and progression of PD. Despite its importance, examination of microvascular damage in vivo in PD patients is very limited. In the present study, the microvasculature damage in PD patients based on the presence of OH was investigated and OH was a potentially critical factor associated with the microvascular damage in PD. The results showed that microvascular damage measured using OCTA occur prior to the development of macular thinning and pRNFL. Based on the results, OCTA can be a useful non-invasive method for detecting microvascular damage in PD patients.