4.1. Hyphema
Hyphema is the most common complication causing early postoperative visual decline after CB-MIGS. When the IOP temporarily drops below the episcleral venous pressure postoperatively, reflux bleeding can result in hyphema. In the absence of intraoperative iris trauma or damage to the angle vessels, postoperative hyphema serves as evidence that the outflow pathway beyond the collector channels has been connected to the episcleral veins. This is also considered a positive indicator of IOP reduction.
Hyphema is not limited to AIT or micro-implantable devices; it is also a common complication of ABiC, where partial incision of the trabecular meshwork during the expansion of Schlemm’s canal can lead to bleeding. The reported incidence of hyphema varies depending on the procedure used. For iStent and iStent injections, the incidence ranged from 1.1–22.2%, whereas it ranged from 1.9–6.5% for Hydrus. Procedures, such as the KDB, have reported an incidence of 0–34.9%, whereas the TOM has a wider range of 4.7–95.0%. For TMH, the incidence is 14–41%, while that for GATT ranges from 1.0–50.6%. Similarly, the incidence of ABiC ranges from 1.0–20.0%.
Visual acuity naturally reduces, whereas hyphema persists postoperatively. However, in most cases, bleeding resolves spontaneously within approximately 1 week, and visual acuity returns to baseline [
34,
59,
68,
71,
78,
80,
103,
115].
In rare cases, massive or prolonged bleeding may require anterior chamber irrigation because such conditions can lead to elevated IOP or adverse effects on the corneal endothelium. Reports indicate that anterior chamber irrigation was required in 0–9.0% of cases following AIT [
21,
84,
88,
92,
93] and in 0–10.9% of cases following GATT [
102,
105,
112,
114,
116,
205]. However, no previous studies have identified this as a direct cause of permanent visual impairment. However, MIGS aims to promote early visual recovery, and bleeding should ideally be avoided or, if it occurs, should remain minimal and resolve spontaneously within a short period.
Regarding the speed of visual recovery in CB-MIGS alone, Tanito et al. reported postoperative visual outcomes in 159 eyes treated with TMH alone [
92]. The preoperative logMAR BCVA was 0.11 ± 0.39. Postoperatively, the BCVA was 0.30 ± 0.68 at 1–2 weeks (P < 0.0001), 0.15 ± 0.48 at 1 month (p = 0.02), and 0.07 ± 0.29 at 3 months (p = 0.83). These results indicate that while visual acuity recovered to levels close to the preoperative values after 1 month, a statistically significant difference remained at that point. In contrast, reports on iStent and iStent injections suggest that, unlike AIT, these procedures do not require an incision of the trabecular meshwork. Consequently, early postoperative complications, including hyphema, are less frequent [
130,
133,
134], and early postoperative visual recovery is often expected [
133].
In a comparative study of TMH, Harano et al. evaluated 78 eyes of 39 patients who underwent simultaneous cataract surgery with TMH in one eye (TMH group) and iStent injection of W in the contralateral eye (injection W group) [
136]. The incidence of hyphema in the inject W group was 5%, which was significantly lower than that in the TMH group (26 %; P = 0.025). Additionally, the BCVA in the injection W group was significantly better at 2 weeks (P = 0.005) and 1 month (p < 0.0001) postoperatively, whereas no significant difference in BCVA was observed between the two groups at 3 months postoperatively.
The same research group also reported in a comparative study of first-generation iStents and TMH that anterior chamber hyphema scores were significantly lower in the iStent group on postoperative days 1, 2, and 3 [
137]. Compared with TMH, the implantation of the iStent and iStent injection W was associated with less postoperative hyphema, leading to faster visual recovery. These characteristics make them particularly suitable for patients requiring early postoperative visual recovery to facilitate a quick return to social and occupational activities.
The frequency and severity of hyphema vary not only between different surgical techniques but also considerably within the same technique across various reports. This discrepancy can be attributed to three main factors. First, the definition of hyphema has not been standardized. Although many studies define hyphema as the formation of a blood niveau in the anterior chamber [
88,
92,
93], some reports, although not explicitly stated, may include coagulated blood clots in their definition. Furthermore, hyphema does not always present as blood niveau; it can also manifest as individual red blood cells floating in the anterior chamber. This flutter-type hyphema is challenging to quantify and is often omitted from reports. However, it is believed to have an impact on visual acuity reports. Sugihara et al. reported that when flutter-type hyphema was included, hyphema was observed in 97% of cases after TMH with a nasal incision alone and in 100% of cases after TMH with both nasal and temporal incisions [
97].
Second, differences in the extent of the incision may influence the frequency and severity of hyphema. In procedures like TOM and KDB, it is common to incise or excise the trabecular meshwork over a range of approximately 90–120 degrees through a single corneal port [
20,
21,
26,
56,
207]. In contrast, TMH with curved hooks allows for a broader incision range. Incisions may cover 90–240 degrees using one or two corneal ports [
88,
89,
97,
207,
208], whereas GATT typically involves a 360-degree incision of the trabecular meshwork [
100,
101,
105,
112].
Reports of hyphema persisting beyond 1 week postoperatively or even beyond 1 month requiring additional anterior chamber irrigation are more prominent in studies involving 360-degree incisions than in those involving 120-degree incisions [
102,
105,
112,
114,
116]. However, reports of anterior chamber irrigation have also been noted in studies on TOM, TMH, and KDB [
21,
84,
88,
92,
93], making it challenging to compare techniques based on retrospective studies.
Several studies have compared the incidence of hyphema based on different incision ranges when using TMH for AIT. Okada et al. compared the 120-degree and 180-degree incisions, whereas Sugihara et al. and Mori et al. compared the 120-degree and 240-degree incisions. In all these studies, no significant differences in IOP reduction were observed between the groups. However, broader incision ranges were associated with a higher incidence of hyphema [
97,
208,
209]. It is important to consider the potential for delayed visual recovery when choosing surgical techniques involving extensive TM incisions.
Third, differences in IOP settings at the time of wound closure may influence the degree of reflux bleeding from the episcleral veins. Although no studies have directly compared the incidence of hyphema based on IOP levels at closure, the difference in hyphema incidence between ab externo and ab interno trabeculotomy techniques may provide some insights.
Kanda et al. compared the incidence of hyphema between AIT and exLOT and found that hyphema occurred in 33% of cases with EIT and 11% with AIT, with the incidence being significantly lower in AIT (P < 0.01) [
96]. The higher incidence of hyphema in EIT may be attributed to the use of a sinusotomy, which facilitates aqueous humor filtration into the external environment, making achieving a low IOP more likely and predisposing patients to reflux bleeding. By contrast, AIT may reduce the likelihood of reflux bleeding from the episcleral veins by setting a higher IOP during hydrosealing at wound closure.
As IOP settings at the time of closure vary depending on the surgeon's preferences and individual cases, this factor may contribute to differences in the incidence of hyphema in CB-MIGS. To reduce the frequency of hyphema, the surgeons routinely aimed to complete a procedure with relatively higher IOP settings at the time of wound closure.
In GATT combined with cataract surgery, Loayza-Gamboa et al. reported a low incidence of hyphema (9.3 %) [
200], whereas Cubuk et al. reported a much higher incidence (89.1 %) [
209]. Cubuk et al. attributed the high frequency of hyphema to the omission of filling the anterior chamber with viscoelastic material at the end of the procedure [
209], suggesting that this step may help prevent postoperative hyphema [
100]. Additionally, recent studies have identified exfoliation glaucoma (XFG) as a risk factor for increased hyphema incidence [
84,
210,
211]. Many patients with glaucoma are elderly and often receive anticoagulants. While some studies have found no association between anticoagulant and hyphema incidence [
84,
211,
212], Chihara et al., in their study of exLOT, reported a significant difference in hyphema incidence between patients who discontinued the anticoagulant therapy and those who did not. They further noted that patients who continued anticoagulant therapy experienced prolonged hyphema required more anterior chamber irrigation, and showed differences in early safety and complications compared with those who discontinued the therapy [
213].
However, they also reported no significant difference in IOP 1 year postoperatively between the two groups. Given the potential risk of ischemic systemic diseases such as stroke or myocardial infarction, which could affect life expectancy, the decision to discontinue anticoagulant therapy remains unclear.
4.4. Complications Potentially Leading to Severe Visual Impairment
Although severe and permanent visual loss due to CB-MIGS complications is extremely rare, it can still occur. Kaplowitz et al. reported that [
219] the incidence of hypotony < 5 mmHg after TOM was 0.09% [
175,
220]. Most cases are believed to result from intraoperative errors, leading to angle disinsertion. In cases where macular folds persist, treatment options, such as injection of viscoelastic agents into the anterior chamber or laser therapy, may be considered [
221].
In addition, the risk of endophthalmitis, a complication often associated with filtration surgery, is extremely low in CB-MIGS performed using small corneal incisions. One reported case of postoperative endophthalmitis caused by
Enterococcus faecalis was detected through culture, with a meta-analysis indicating an incidence of 0.01% [
222]. However, at our facility, we experienced one case of endophthalmitis [
18]: a patient with severe atopic dermatitis, associated lagophthalmos, and a history of multiple intraocular surgeries. Fortunately, following vitrectomy and antibiotic therapy, the patient's vision was recovered, and the IOP was well controlled.
Patients with atopic dermatitis are highly likely to harbor resistant bacteria, such as methicillin-resistant
Staphylococcus aureus or methicillin-resistant
Staphylococcus epidermidis, in their conjunctival sacs [
223]. Therefore, careful monitoring for postoperative endophthalmitis is essential.
Similarly, suprachoroidal hemorrhage, another serious complication of filtration surgery, is exceedingly rare in CB-MIGS, with a reported incidence of 0.01% [
224]. In a recent study, Waldner et al. compared the outcomes of superior and inferior hemi-section GATT in 297 eyes and reported two cases of suprachoroidal hemorrhage. However, both cases were rapidly resolved with appropriate management [
114].
4.5. Delayed-Onset Hyphema
Delayed-onset hyphema is a complication that may occur > 2 months after surgery, often resulting in mild visual decline. Patients typically present with sudden blurring of vision after an extended postoperative period. On slit-lamp examination, floating red blood cells (flutter-type hyphema) are often observed in the anterior chamber, and gonioscopy frequently reveals mild hyphema localized inferiorly.
To date, 12 cases of delayed-onset hyphema occurring > 2 months after TOM have been reported [
225]. Other cases have also been documented, including one case occurring 6 months after GATT [
100], one 13 months after iStent implantation [
226], and one 4 years after TMH [
227]. The proposed mechanisms for these events include an increase in episcleral venous pressure due to physical exertion or sudden decompression caused by ocular compression during sleep [
225].
However, delayed-onset hyphaema following CB-MIGS is generally mild and self-limiting. Management typically involves observation rather than aggressive intervention. If a transient mild increase in IOP is observed, the addition of IOP-lowering medications is recommended.
Given that the distal outflow pathways beyond the collector channels connected to the episcleral veins remain functional, recognizing the potential for delayed-onset hyphema, even in the long-term postoperative period, is essential.