In this study, the KDB significantly improved IOP and drop scores, regardless of whether the preoperative IOP was high or low. In particular, a 54% survival rate was observed at 1 year in patients with high preoperative IOP. These results suggest that the KDB may have a relatively high IOP-lowering effect in patients with preoperative IOP > 20 mmHg, as previously reported [21]. Even in patients with preoperative IOP < 20 mmHg, there was a significant difference in IOP compared with the preoperative IOP. In comparison with the preoperative eye drop score, the median drop score was − 1, indicating a reduction of > 1 drop in many cases. Even when the IOP is relatively low, aggressive indications may be considered when the goal is to decrease the ophthalmic score. Linear regression analysis of the IOP change between the preoperative period and 12 months postoperatively revealed a significant association between the preoperative and postoperative day 1 in the amount of IOP change. Patients with high preoperative IOP and patients with high IOP on the first postoperative day may possibly have a certain degree of IOP reduction at 12 months postoperatively. Linear regression analysis of the change in drop score between preoperative and 12 months postoperative showed significant involvement of age and preoperative drop score. As the regression coefficient estimates were negative, patients may have been older and more surgeries may have been performed to reduce the number of eye drops in response to poor eye drop compliance and other factors. Additionally, the postoperative eye drop score may be higher in patients with high preoperative eye drop scores. However, no significant relationship was identified between the preoperative IOP and drop scores (P = 0.1).
The incision range was grouped by quartile range and analyzed; however, no significant difference was found in the IOP and drop scores in each group In a previous report using optical coherence tomography with Tabectome™, no significant relationship was found between the incisional area and IOP reduction [23]. This is due to the fact that aqueous humor flows through the TM, SC, and collector channels and then flows into the episcleral vein [11]. Although the TM and SC are located around nearly the entire angle of the corneal perimeter, there are only a limited number of collector channels, and even if the outflow resistance of the TM is reduced, the collector channels may become a bottleneck, limiting the degree of IOP reduction. Therefore, even if the TM outflow resistance is lowered, the drop in IOP is limited. Therefore, theoretically, a significant difference in IOP reduction may occur between a very small area and a 360° TM incision. In fact, many reports on 360° TM incisions have been conducted, some with favorable results from the viewpoint of complications [24, 25]. In contrast, a report comparing the results of suture LOT 360° incision with those of a KDB found no significant difference in IOP reduction [26]. Based on these results, the KDB may not necessarily require extremely large incisions although may be an option for extensive incisions only in cases where reoperation is difficult due to age or background.
The analysis classified according to preoperative MD values showed favorable results regardless of the degree of glaucoma progression, which is consistent with previous reports [20]. Outflow-tract surgery may also be indicated in cases of severe visual field defects. Another report that found no significant difference between cataract surgery alone, and combined KDB and cataract surgery in glaucoma patients with moderate-to-mild visual field impairment and good IOP control [17]. In cases with good IOP control, cataract surgery alone may be acceptable, except in cases where there is a strong objective to reduce the use of drops, whereas in cases with poor IOP control, the use of the KDB may be recommended even in cases of moderate-to-severe visual field impairment. However, a spike may occur at a certain rate, and the spike risk should be fully explained.
A significant association was found between the preoperative drop scores, postoperative day 1 IOP, and axial length regarding the occurrence of spike. The results for postoperative day 1 IOP were similar to those previously reported [27]. In patients whose condition is controlled using multiple medications preoperatively, eye drops are expected to strongly suppress aqueous humor production and promote outflow. Generally, eye drops are discontinued postoperatively. The timing of spike occurrence varies from case to case; however, it is thought that spike occurs when the timing of recovery of the individual's own aqueous humor postoperatively coincides with a decrease in aqueous humor outflow for some reason that has increased after surgery. Notably, most of the patients who had spikes were able to reduce their IOP with medication within 1–3 months, and only a few patients required reoperation owing to persistently high IOP caused by spikes. For patients with high preoperative IOP scores, frequent observation during the first month after surgery is recommended. Previously, the length of the ocular axis is also reportedly involved in the occurrence of spikes after cataract surgery [28]. Another report suggested that high myopia may lead to prolonged inflammation in the anterior chamber [29], the outflow resistance of the collector channel in the sclera may be increased due to the irregular shape of the eye [30], and that the depth of the anterior chamber may be deep enough to leave behind the viscoelastic material used during the surgery. Although an increase in IOP is often observed on postoperative day 1 in cases where viscoelastic material is left behind, multivariate analysis of the occurrence of spikes in our study showed no collinearity between ocular axis length and postoperative day 1 IOP; both variables were significantly different on multivariate analysis (P = 0.02, P < 0.001; respectively). Further studies on the effects of anterior chamber inflammation and ocular shape are warranted.
The limitations of this study include the fact that it was a single-center study; therefore, future studies should be conducted with a larger number of patients. Missing values were supplemented by last observation carried forward. The incision range was treated as an ordinal variable using the interquartile range rather than a continuous variable because it is a subjective finding of the surgeon based on the surgical record and may lack reliability. Therefore, quantitative evaluation of the incision range is insufficient.