Sub-Tenon anesthesia for ocular surgery was introduced by Stevens (1992) and is widely used, due to its ease of administration, effectiveness, and safety profile (Adekola et al., 2018). The sub-Tenon’s block is employed in various procedures, including cataract, strabismus, and vitreoretinal surgery. This technique entails the transconjunctival administration of an anesthetic solution into the potential sub-Tenon space with a blunt canula. Its purpose is to deliver effective analgesia and in higher dosage akinesis, while reducing the risks associated with sharp needle techniques, such as globe perforation, retrobulbar hemorrhage, optic nerve injury, and accidental injection into the subarachnoid space (Kumar and Dodds, 2006; Kumar et al., 2005). Research has indicated that it is at least as effective as peribulbar anesthesia (Parkar et al., 2005; Guise, 2012; Antony et al., 2022).

Despite its advantages, concerns have been raised about the potential hemodynamic effects of the sub-Tenon technique on ocular perfusion (Coşkun et al., 2014; Kumar et al., 2011; Fry and Ring, 2008; Feibel and Guyton, 2003). Coşkun et al. (2014) suggested that sub-Tenon anesthesia decreased ocular blood flow significantly by using color doppler imaging. While various studies suggest that the rise in pressure is mainly a result of the volume effect caused by the local anesthetic agent being forcefully injected into the restricted space of the bony orbit (Fry and Ring, 2008). Larger volumes of injectate may lead to a more significant increase in IOP, but also factors such as the orbital volume, the tension in the extraocular muscles, orbital septum, and the conjunctival/Tenons incision acting as a pressure valve, influence this potential rise (Alwitry et al., 2001). Further, it is hypothesized that touching the arterioles with the tip of the blunt cannula could result in a spasm, which would cause vessels to close. This mechanism has been demonstrated by Lende et al. in an experiment with cats (Ellis, 1964). Finally, injecting anesthetic solution into the sub-Tenon space might lead to pharmacological vasoconstriction caused by the high local concentration of the anesthetic drugs (Creese et al., 2017; O’Day et al., 2019; Chen, 2017; Meyer et al., 1993; Watkins et al., 2001). The hypothesis is that the anesthetic drugs either inhibit endothelium-dependent relaxation of the ciliary arteries or attenuate the activity of the vasodilatory nerves (Hulbert et al., 1998). Hereby they alter the blood flow dynamics and decrease the pulsatile ocular blood flow (Rüschen et al., 2003). This may result in a temporary decrease in vessel density (VD) and perfusion density (PD) at the retinal microvascular level.

Due to concerns about retinal ischemia following sub-Tenon anesthesia, some publications have periodically reiterated recommendations to avoid its use in patients with compromised retinal perfusion (Rüschen et al., 2003). According to different authors this concern is especially relevant for patients with diabetic retinopathy or maculopathy (DM), and retinal vein occlusion (RVO) (Creese et al., 2017; Pianka et al., 2001; Enz et al., 2022). These conditions are characterized by impaired autoregulation, endothelial dysfunction, and structural microvascular loss, including capillary non-perfusion, increased retinal vascular permeability, and ischemia-related macular edema (Haydinger et al., 2023). Consequently, the hemodynamic response to sub-Tenon anesthesia may differ from that in eyes without pre-existing vascular disease, with greater pressure dependence and reduced ability to compensate for transient decreases in perfusion (Kohner et al., 1995). Any additional disruption of retinal perfusion could exacerbate pre-existing conditions, potentially leading to adverse visual outcomes (Kohner et al., 1995).

Optical coherence tomography angiography (OCTA) is a relatively new imaging modality that allows non-invasive visualization of the retinal microvasculature. OCTA can provide detailed maps of retinal blood flow and is an excellent tool for assessing retinal microvasculature (Gandhi et al., 2024). It is a non-invasive imaging technique that provides several benefits over conventional angiography methods, such as fluorescein (Kwiterovich et al., 1991) or indocyanine green (Yannuzzi et al., 2012) angiography. Being non-invasive, OCTA does not require the injection of a dye into the patient’s bloodstream, thereby eliminating the risks associated it: anaphylaxis, nausea, vomiting, rash, or skin and urine discoloration (Chua et al., 2024). However, image artifacts, due to movement, blinking, alteration of corneal surface refractive index or other irregularities, can affect the quality and accuracy of the OCTA scans, potentially leading to false-positive findings (Anvari et al., 2021).

This study aims to investigate the effects of sub-Tenon anesthesia on retinal microvasculature in patients with DM or RVO undergoing cataract surgery. By using OCTA, we assessed VD and PD before and after anesthesia, while also considering the impact of image’s quality on the reliability of our results.

This prospective, monocentric study included patients suffering from DM or RVO scheduled for cataract surgery at the Cantonal Hospital Aarau, Switzerland, and was approved by the Ethics Committee Northwest and Central Switzerland (EKNZ; approval ID 2019-01689). We excluded eyes with ocular ischemic syndrome; uncontrolled ocular hypertension (baseline IOP > 25 mmHg) or recent IOP-lowering procedures (<4 weeks); carotid artery stenosis; severe or unstable cardiovascular disease; severe or uncontrolled hypertension; hematological disorders affecting viscosity or oxygen-carrying capacity; active uveitis/infection; recent intraocular surgery (<3 months); media opacity other than cataract; or low-quality OCTA imaging. No formal a priori sample size calculation was performed because the study was exploratory. All procedures were performed in accordance with the Declaration of Helsinki and applicable national laws and regulations. The study population consisted of 10 patients diagnosed with DM and 10 with RVO.

Among the 20 patients included in the study, the mean age was 69.8 ± 10.2 years, with 11 females and 9 males. All patients completed the study without any significant adverse events related to anesthesia.

Even though sub- Tenon anesthesia is regarded as a safe and favored procedure, some cases of retinal ischemia have been reported over the years. Possible explanations of this complication are related to an increase in IOP, formation of a retrobulbar hemorrhage, or compression of blood vessel, while patients with compromised retinal perfusion or significant cardiovascular risk factors are considered high-risk (Kumar et al., 2011). Transient or permanent central retinal artery occlusions because of sub-Tenons anesthesia have been described in 1982 by Klein et al. (1982), in 2003 by Feibel and Guyton (2003), and in 2008 by Fry et al. Fry and Ring (2008). These occlusions were observed even in the absence of orbital hemorrhage, and a spasm of the central retinal artery—either pharmacologically induced or due to a compressive effect—was suspected (Feibel and Guyton, 2003). In contrast to the current study, four patients described by Klein et al. were diagnosed with either a severe hematologic disorder, including sickle cell retinopathy in two cases, or a vascular pathology, which encompassed diabetic retinopathy, carotid artery insufficiency associated with ocular ischemia, and neovascular glaucoma.

The initial results of the current study appeared to show a reduction of both VD and PD following anesthesia, which could have confirmed previous concerns about the hemodynamic impact of this technique (Coşkun et al., 2014; Enz et al., 2022). However, upon closer examination, the reduction in VD and PD was found to be associated with lower-quality OCTA images, which were likely a result of the intervention related artifacts such as surface irregularities caused by Betadine application, dry eye, instable tear film, presence of blood or mucus and debris around the injection site. A previous OCTA study examining the effects of sub-Tenon anesthesia similarly described a greater signal reduction in interventional than in controls eyes (Enz et al., 2022), while Czakó et al. (2019) and other researchers have already emphasized the influence of images quality on OCTA metrics and that it can potentially lead to misinterpretations. In the current study, once 3 low quality images were excluded, no significant reduction in VD or PD was any longer observed. What further supported that that the observed VD/PD reductions reflect measurement artefact rather than a true hemodynamic effect is that the Quality-adjusted DiD regression demonstrated that the apparent OCTA changes were largely driven by image-quality variation and were no longer present after adjustment.

The manifestation of OCTA artifacts during picture acquisition is a common and known challenge. A variety of studies has reported that the prevalence of OCTA artifacts can be up 89.4–97% (Holmen et al., 2020). Such artifacts can impair the reliability of the OCTA analysis since they might potentially misrepresent blood vessels and their morphology, mask certain pathologies in the vascular plexus of the retina and hereby lead to misinterpretations. There are different reasons and types of OCTA artifacts. Most commonly seen are motion artifacts due to saccades, blinking, or head motion that lead to disruptions in the image’s continuity despite eye-tracking mechanisms (Lauermann et al., 2018); further common artifacts appear due to signal attenuation factors usually as a result of an element obstructing the light transmission (such as cataracts, floaters, debris in the tear film) (Anvari et al., 2021); and finally segmentation errors (Lauermann et al., 2018). However, the presence of OCTA artifacts does not necessarily result in the total exclusion of the image. When the images are manually controlled and evaluated, to assess the extent and the potential scale of the artifact’s impact, they can still be eligible for further analysis (Holmen et al., 2020). For this purpose, in this study for each eye multiple images were taken, those with the least relevant artifacts and the best quality were further analyzed, in cases where no good images could be obtained for either the intervention or the control eye the patient was totally excluded from the study.

Our study found no evidence of negative hemodynamic effects on the retinal microvasculature in patients with pre-existing perfusion deficits, which is particularly reassuring for ophthalmic surgeons who favor sub-Tenon anesthesia for ocular procedures. The absence of significant changes in IOP or systemic BP further supports the safety of sub-Tenon anesthesia, even in patients with compromised retinal microvasculature. Our findings are consistent with the broader literature indicating that, sub-Tenon anesthesia has minimal or no impact on IOP (Enz et al., 2022; Pianka et al., 2001; Alwitry et al., 2001). To our knowledge this is the first study examining the effects of sub-Tenon anesthesia on eyes with compromised retinal perfusion. In addition, our study highlights the importance of image quality when interpreting OCTA results. Artifacts can lead to false-positive findings, which could be mistakenly attributed to real physiological changes.

Despite being a prospective study, we acknowledge several limitations and shortcomings. A principal limitation is the modest sample size, despite the extended recruitment period. This might have reduced the precision and have limited the power of the study to detect small perfusion changes. Recruitment was constrained by the additional testing burden, which impacted theatre workflows, and by the high appointment load commonly faced by patients with DM and RVO (e.g., frequent intravitreal injections), reducing willingness to undergo extra imaging. Furthermore, several initially enrolled patients were excluded because high-quality, artefact-free OCTA images—particularly in the intervention eye—could not be obtained. A further limitation of this study was that the type of cataract and the degree of nuclear sclerosis were not systematically documented. As these parameters are not routinely assessed or recorded in our clinical practice, they were unavailable for analysis. Consequently, it was not possible to evaluate the influence of specific cataract characteristics on image quality, including OCTA signal strength and overall image reliability.

Furthermore, glaucoma—particularly normal-tension glaucoma—was not an exclusion criterion. This may represent an additional limitation, as these patients may exhibit reduced perfusion of end organs, including the optic nerve head and retinal microvasculature. Additionally, the scan dimensions were not standardized in our OCTA protocol. Therefore, some investigators acquired 3 × 3 mm field of view images, while others used 6 × 6 mm fovea-centered scans. This inconsistency was only identified after study completion during data evaluation. In total 7 patients received a 6 × 6 mm field of view measurement, two of which were later on excluded from the study due to too low image quality. For 6 × 6 mm scans, analysis was restricted to the central 3 mm (perifoveal 3–6 mm ring excluded) to improve comparability with 3 × 3 mm data. As 3 × 3 mm OCTA generally affords higher spatial resolution and sensitivity for detecting microvascular changes (Ho et al., 2019), the 6 × 6 mm subset may have reduced spatial sensitivity. Notably, each fellow control eye was acquired with the same scan size as its intervention eye, maintaining within-subject comparability. Patients with glaucoma and the OCTA scan field size have been identified in the raw dataset provided in the Supplementary material. Further, although imaging was performed immediately pre-block and again within minutes post-block to capture acute vasomotor/pressure effects, the post-intervention scans were acquired at 7.7 ± 5.5 min, which may have coincided with time-dependent normalization of intraocular pressure and perfusion. Finally, some eyes in both the DM and RVO groups had macular edema at the time of cataract surgery and, consequently, during the study imaging. This factor was not analyzed separately, which may have influenced our results.

In conclusion, our study did not identify any significant reduction in VD or PD following sub-Tenon anesthesia when accounting for the OCTA image quality. These findings support the further use of sub-Tenon anesthesia even in patients with retinal perfusion deficits such as those with DM and RVO. Additionally, the study highlights vulnerability of OCTA metrics immediately after periocular anesthesia. Larger-scale research and optimized imaging techniques may help provide more insight into any potential subtle hemodynamic effects of this anesthesia technique.