Primary open-angle glaucoma (POAG), as a blindness-causing disease, is characterized by optic nerve injury and progressive visual field (VF) loss (1, 2). Among several risk factors for POAG, myopia, especially high myopia (HM), is an established risk factor proved by population-based studies (3–5). Previous literature have revealed that the risk of POAG in HM patients is six times higher than in those without HM, and HM is highly correlated with the progression of glaucoma (6, 7). However, the underlying role and mechanism of HM in glaucoma pathogenesis has not been fully established.

Optical coherence tomography angiography (OCTA) allows high-resolution retinal microvasculature imaging in an non-invasive way (8), and its application has facilitated further investigation of the effect of myopia on ocular microcirculation in POAG. Previous OCTA reports have shown that vessel density (VD) is related to the severity of VF damage in glaucomatous eyes with HM (9, 10). Peripapillary microvessel evaluation has also shown distinct value in detecting glaucoma damage in HM (11, 12). However, the existing research on this topic has focused on microvascular structural changes, and retinal vascular functional changes was rarely explored in glaucomatous eyes with HM.

Vasoreactivity, which is also known as vascular autoregulation, is the term used to refer to the intrinsic ability of humans to maintain sufficient and stable blood flow by self-adjustment of the vascular system under various metabolic conditions (13). Analyses of vasoreactivity are indispensable in investigations of POAG pathophysiology. Vasculature in the cerebrum and retina has been reported to have an impaired reactivity in patients with POAG (14–16). Considering that ocular blood flow is playing a key part in developing glaucomatous eyes with HM, determination of the role of retinal vascular reactivity, which reflects vascular function, in the effects of myopic factors on POAG is important.

During hyperoxia condition, the retinal vessels were supposed to contract and decrease the perfusion in healthy subjects, which was reflected by the decrease of vessel density in OCTA measurements (17, 18). We had previously reported an easy, safe, and reproducible method using OCTA and a hyperoxia test to detect retinal microvascular reactivity in healthy subjects (18), as well as identify impaired vascular response in the optic nerve head (ONH) of POAG patients compared with healthy controls (19). Here we examined the different vasoreactivity in POAG eyes with and without HM, and further explored the potential relationship between myopia and ONH vascular autoregulation in POAG patients. Our study would contribute to a better understanding of glaucoma pathophysiology.

In the study, we investigated retinal vasoreactivity in POAG eyes with and without HM, and determined the relationship between myopia and ONH vascular reactivity in glaucomatous eyes by using OCTA and hyperoxia tests. We found that POAG eyes with HM presented significantly impaired vascular autoregulation than eyes without HM. Moreover, we observed a strong significant linear correlation between ONH vasoreactivity and SE and AL in patients with POAG.

In our study, POAG patients with HM were younger than patients without HM. Moreover, probably due to the age difference between the groups, the BP of patients without HM was significantly higher than that of patients with HM, although patients with hypertension were excluded from the study. To confirm that the ONH vasoreactivity difference between the two groups was caused by HM and not by the differences in age and systemic blood flow variables, we performed linear regression to test the effects of these factors on the ONH vascular response. We found that none of the hemodynamic variables were associated with vascular response, and age showed no correlation with vasoreactivity in multivariate models (Supplementary Table 1). In addition, the regression models were also adjusted for age when investigating the relationship between myopia and vasoreactivity, and we found that the ONH vascular response was still significantly correlated with SE and AL after adjustment for age (Table 5). Therefore, the confounding effect of age and blood flow factors did not influence our findings on HM and vascular reactivity in glaucomatous eyes.

Peripapillary evaluations in glaucomatous eyes with HM are challenging due to structural variations such as a tilted disk and peripapillary atrophy (27). In OCT examinations, segmentation error is the main concern in HM eyes due to AL elongation and large peripapillary atrophy (28, 29). Segmentation errors in OCT may prejudice the precise assessment of peripapillary RNFL thickness and cause inaccurate glaucoma diagnosis (30, 31). In this respect, OCTA presents a unique advantage in peripapillary evaluation of HM eyes, since the visualization of retinal VD does not demand strictly accurate segmentation as much as the measurement of thickness does; consequently, it is less affected by HM-related anatomic variation (10). Previous studies indicated that peripapillary VD had stronger association with VF sensitivity than RNFL thickness did in glaucomatous eyes with HM, revealing that OCTA may be more useful in monitoring visual functional changes than OCT in HM eyes. However, these studies all focused on retinal vascular structure characteristics, and there have been no studies investigating vascular function alterations in high myopic glaucoma.

Vasoreactivity, which reflects the function of the vasculature, is an extremely critical aspect in research on POAG. Previous studies have used dynamic vessel analysis and laser Doppler flowmetry to detect decreased vasoreactivity in retinal vessels and cerebral vessels in patients with POAG (14, 32, 33). Our recent study also showed that ONH vascular reactivity was significantly impaired in eyes with POAG using OCTA (19). Furthermore, in the current study, we observed that POAG eyes with HM had worse vasoreactivity than eyes without HM. In POAG eyes without HM, RPC-VD still significantly decreased in response to hyperoxia provocation, which was consistent with our previous findings (19). However, in eyes with HM, the RPC-VD was not significantly reduced after hyperoxia in the whole image, and the VD differences between baseline and hyperoxia conditions were markedly lower than those in eyes without HM. These results imply that impaired ONH vascular reactivity may be involved in the pathophysiology and development of high myopic glaucoma, and retinal vascular dysfunction may aggravate the progression of POAG in eyes with HM.

We further examined the associations between refractive errors and ONH vascular response in patients with POAG. Linear regression was performed in all 48 patients with POAG, both with and without HM. Interestingly, we found that a lower ONH vasoreactivity was significantly correlated with a worse SE and elongated AL in both univariate and multivariate models after adjustment for age. This result indicates that refractive error may be an independent influencing factor for vascular reactivity in POAG. In accordance with our results, Lin et al. found that retinal VD decreased faster in POAG eyes with HM than in those without during longitudinal follow-up (34). Additionally, Li et al. reported a negative correlation between retinal VD and AL in myopic eyes, which is also consistent with our findings for retinal vasoreactivity (35). The mechanism underlying the effect of myopia on retinal vasoreactivity may result from the stretching of the microvascular network during AL elongation (35). This structural alteration may subsequently lead to or accompany microvascular dysfunction in glaucomatous eyes with HM. Therefore, AL elongation in myopic eyes resulted in not only vascular structural changes (decreased VD), but also functional changes in vascular autoregulation (impaired vasoreactivity), which should also be highlighted during the follow-up of high myopic glaucoma.

Our study also had certain limitations. First, the age and blood flow variables were significantly heterogenous between POAG eyes with and without HM, which may have affected our study outcomes. However, our findings proved that there was no correlation between these variables and ONH vascular reactivity, and we also made the adjustment when detecting the relationship between HM and vasoreactivity. Therefore, we believe that the confounding effect of the baseline factors was negligible in our study. Second, we did not exclude eyes using IOP-lowering medications, and some antiglaucoma agents may affect ocular microcirculation and vascular reactivity. However, recruitment of sufficient POAG patients while excluding those using any IOP-lowering eyedrops is difficult in clinical practice, and the number of eyes receiving antiglaucoma medication was comparable between the groups with and without HM in our study. Therefore, the application of these agents did not significantly affect our main findings. Third, to ensure data reliability, we only included high-quality OCTA images in the analysis of our study, which may limit the availability of the results in eyes with severe visual loss and poor-quality OCTA scans. Forth, we only included POAG eyes with and without HM, to evaluate retinal vasoreactivity in eyes with HM but without POAG is also important, and is needed in the future investigation. Finally, our sample size was relatively limited. Larger and longitudinal studies including more participants of various ages and ethnic backgrounds would provide more insights into ocular microcirculation in glaucomatous eyes with HM.

In summary, our study explored optic disk vascular reactivity in eyes with POAG and HM. In comparison with glaucomatous eyes without HM, those with HM showed significantly impaired vasoreactivity, and this impairment was significantly associated with worse SE and longer AL. Thus, our study revealed alterations in retinal vascular autoregulation in POAG eyes with HM, which may help improve our understanding of myopic glaucoma pathophysiology and contribute to research about HM in POAG development and progression. The feasibility and usefulness of OCTA in capillary functional analysis have also been highlighted and expanded in our study.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Institutional Review Board of the Eye and ENT Hospital of Fudan University, Shanghai, China (No. 2014043). The patients/participants provided their written informed consent to participate in this study.

XF designed the study, collected the data, performed the data analysis and data interpretation, and wrote the manuscript. HX, RZ, QS, and YS contributed to the data discussion. XK and TS contributed to the data discussion and reviewed the manuscript. All authors contributed to the manuscript revision, read, and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.