Exact Matching of Trabectome-Mediated Ab Interno Trabeculectomy to Conventional Trabeculectomy with Mitomycin C Followed for Two Years

Purpose: We used exact matching for a highly balanced comparison of ab interno trabeculectomy (AIT) with the trabectome to trabeculectomy with mitomycin C (TRAB). Methods: 5485 patients who underwent AIT were exact-matched to 196 TRAB patients by baseline intraocular pressure (IOP), number of glaucoma medications, and glaucoma type. Nearest-neighbor-matching was applied to age. Success was defined as a final IOP of less than 21 mmHg, IOP reduction of at least 20% reduction from baseline, and no secondary surgical interventions. Outcomes were measured at 1, 3, 6, 12, 18, and 24 months. Results: 165 AIT could be matched to 165 TRAB. The mean baseline IOP was 22.3+/-5.6 mmHg, and the baseline number of glaucoma medications was 2.7+/-1.1 in both groups. At 24 months, IOP was reduced to 15.8+/-5.2 mmHg in AIT and 12.4+/-4.7 mmHg in TRAB. IOP was lower than baseline at all visits (p<0.01) and lower in TRAB than AIT (p<0.01). Glaucoma medications were reduced to 2.1 +/- 1.3 in AIT and 0.2 +/- 0.8 in TRAB. Compared to baseline, patients used fewer drops postoperatively (p<0.01) and more infrequently in TRAB than in AIT (p>0.01). Secondary surgical interventions had the highest impact on success and became necessary in 15 AIT and 59 TRAB patients. Thirty-two challenging events occurred in TRAB and none in AIT. Conclusion: Both AIT and TRAB reduced IOP and medications. This reduction was more significant in TRAB but at the expense of four times as many secondary interventions.


Introduction
Trabeculectomy (TRAB), first performed in patients in 1961 by Sugar [1] and made more effective by the addition of mitomycin C in 1990 [2] , continues to play a central role in surgical glaucoma treatment. However, postoperative challenges and complications can occur in up to 77-78% of patients [3][4][5][6] . This concern led to the development of non-penetrating and microincisional glaucoma surgeries (MIGS) [7] . One of the MIGS, ab interno trabeculectomy (AIT), is now often used both as a primary [8] and a secondary surgery [9] due to its proven, relative effectiveness in a range of glaucoma severity, including severe glaucoma [10][11][12] . AIT can also be performed after a failed trabeculectomy [13] or tube shunt [14] . This is surprising because TRAB allows aqueous humor to bypass the conventional outflow system, causing Schlemm's canal and collector channels to atrophy [15,16] .
In theory, the IOP that AIT can achieve is limited by the episcleral venous pressure of 8 mmHg [17] , yet most studies report IOPs around 15 mmHg. In contrast, TRAB can achieve very low IOPs, even below episcleral venous pressure, and eliminate drops [18,19] . Hypotony [20,21] is a feared consequence of too much pressure reduction. Phako-AIT has been compared to Phaco-TRAB in a randomized controlled trial but was stopped, with only 19 patients enrolled due to lack of clinical equipoise: TRAB was more complication prone than AIT, yet IOPs were similar [22] . There is scant data that directly compare TRAB to AIT. TRAB is a filtering surgery that drains aqueous humor into a subconjunctival reservoir, the bleb, while AIT enhances conventional outflow along its physiological route by removing outflow resistance at the level of the trabecular meshwork.
To address this unmet need, we applied an advanced method in statistics, exact matching [23][24][25] , and created nearly identical pairs of AIT and TRAB based on IOP and number of topical glaucoma medications while using nearest neighbor matching for age. This design allowed us to compare the fate of highly similar eyes after being treated by these two distinctly different methods. page 3 of 20

Study design
The study protocol was approved by the local ethics committee of the University of Würzburg, Germany, and performed in accordance with the ethical standards set forth in the 1964 Declaration of Helsinki and the Health Insurance Portability and Accountability Act. Because of its retrospective nature, informed consent was waived. All patients included in the study underwent either ab interno trabeculectomy using the Trabectome (AIT) or trabeculectomy with 0.02% mitomycin C (TRAB). An indication for surgery was an above-target IOP despite maximally tolerated topical treatment, as determined by a glaucoma specialist, or a need to reduce glaucoma medications despite stable IOP due to eye drop intolerance. We used data from the Trabectome Study Group database [13,26]  The decision to continue glaucoma drops was at the discretion of the treating specialist, as was the decision to advance to a different glaucoma surgery.

Statistics
Data were described as frequency, percentage, mean±SD, median, and range. Continuous and categorical variables were compared with the Mann-Whitney U test and chi-squared test. Using exact matching , both groups were matched using preoperative IOP, glaucoma medications, type of glaucoma, and using nearest neighbor matchin g for age [27] . Each unit in AIT was matched using exact matching to all possible control units in TRAB. Whereas nearest neighbor page 4 of 20 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020. 10.27.20165944 doi: medRxiv preprint matching selected the best match based on the distance to the value in AIT. P-values of less than 0.05 were considered statistically significant. Mean±SD was used to express continuous variables. Statistical analyses were performed using R [28] . A Kaplan-Meier plot was generated.

Surgical technique
AIT was performed as described before [29] . Briefly, a 1.6 mm uniplanar, temporal clear corneal incision was created. Under direct gonioscopic visualization, the tip of the Trabectome handpiece (MST MicroSurgical Technology, Redmond, WA, USA) was inserted into Schlemm's canal. The TM was ablated counterclockwise, followed by clockwise ablation with a total length of about 120 degrees [30][31][32] . Ablation was started with the power set to 0.8 mW and increased as necessary. The handpiece was withdrawn from the anterior chamber.
In TRAB, the eye was rotated downward with a traction suture, as described before [33] .  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint

Results
Using exact matching , 165 AIT were paired with 165 TRAB. The baseline IOP and glaucoma medications in AIT and TRAB were identical. IOP was 22.3±5.6 mmHg, and medications were 2.7±1.1. The demographic characteristics are presented in Table 1. AIT patients had an average age of 50±15 years, significantly younger than TRAB 67±11 (p<0.01). Despite their younger age, 60% of AIT patients had prior surgeries, many more than TRAB, with only one prior In AIT, the medication reduction was small and only became statistically significant after three months (after three months p<0.05).

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020. ; An additional intervention was needed in 15 (9%; 45% of all failures) of AIT and 59 (36%; 98% of all interventions) of TRAB (Table 3). In AIT, 15 (9%) patients (45% of all interventions) received a tube shunt. In TRAB, one patient needed a secondary glaucoma surgery. However, 24

Discussion
This study aimed to compare IOP and glaucoma medication reduction by two distinctly different glaucoma surgeries, AIT and TRAB, commonly used as primary surgical treatments. We wanted this comparison to be agnostic to disease severity and instead generate a highly balanced comparison, strictly focusing on IOP and medications in our exact matching algorithm. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020.  [15,16] . This concern did equally not materialize in two studies that found that there was a reasonable IOP reduction by AIT after failed tubes [14] and trabeculectomies [13] . page 8 of 20 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020. Limitations of our study include its retrospective nature. This shortcoming is countered by exact matching , a robust automated statistical method that creates a highly balanced comparison and reduces confounding [59,60] . Although the algorithm discarded AIT data, we were fortunate to be able to afford this and create identical IOP and glaucoma eye drop matches with the much more limited number of TRAB so that only 31 were lost. Had we increased the number of matching variables to improve the precision of matching, many more patients would have been excluded, reducing the study sample size and variability of the patient population [60,61] . Consequently, we applied nearest neighbor matching to age, and AIT patients ended up with a younger average age of 50±15 years compared to TRAB with 67±11 years. Although this age difference is considerable, it would have put AIT only at a slight disadvantage of approximately 0.5 mmHg as predicted by our IOP calculator [42] .
Another limitation is that we ignored the individual target IOP to focus on IOP and page 9 of 20 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint medications. Using a common definition of success, which considered any secondary surgical intervention [62,63] , AIT had a higher success rate than TRAB. However, many trabeculectomists would not consider needling or revisions a complication as long as the target pressure is achieved eventually. Our TRAB achieved an IOP reduction equal to that in the TVT study but with fewer medications, while our AIT appeared to be similar to the tubes in that study [64] .
Our study confirmed that TRAB remains a good choice for patients with low IOP targets and need to reduce drops. This procedure requires patients to accept a high chance of postoperative challenges. Analysis of AIT patients matched to the relatively low preoperative IOP of TRAB patients confirm that AIT is a low-risk procedure that achieves physiological IOPs.
Fortunately, TRAB and AIT do not exclude each other as AIT can be performed after TRAB has failed [13] and vice versa [65] .

Compliance with Ethical Standard
Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the University of Würzburg and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Funding:
No funding was received for this research.

Potential Conflict of Interest:
The authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. page 14 of 20 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint Tables   Table 1   Table 1  . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020.  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020. page 20 of 20 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 28, 2020. ; https://doi.org/10.1101/2020.10.27.20165944 doi: medRxiv preprint