Abstract
Objective
To investigate the outcomes of aqueous shunt exposure repair and to identify risk factors for recurrent exposure after surgical repair.
Methods
This is a retrospective interventional case series of consecutive patients who underwent repair of an exposed aqueous shunt. Demographic and clinical data were extracted from the patients’ records and survival analysis was employed to determine the probability of survival of the repair without further exposure. Cox proportional regression analysis was utilised to identify potential risk factors for failure of the repair surgery.
Results
Seventy-six eyes of 76 patients were included in the study. The probability of survival without further exposure was 76.2% (CI 66.9–86.8%), 72.7% (CI 62.8–84.1%) and 54.7% (40.5–73.7%) at 1,2 and 4 years, respectively. No eye in which the tube was repositioned in the anterior chamber or in the sulcus (n = 9; 11.8%) developed a recurrence of the exposure. A shorter interval from the time of tube insertion to the repair surgery (HR 2.87 [CI 1.11–7.39]; p = 0.03; reference longer interval), a history of previous surgical revision (HR 3.06 [1.15–8.12]; p = 0.02; reference no prior revision) and the use of a human pericardial donor patch at the time of revision (HR 3.43 [1.16–10.13]; p = 0.03; reference other than pericardium) was associated with an increased risk of exposure recurrence.
Conclusion
Aqueous shunt exposure revision is associated with reasonable mid and long term success. A shorter interval from tube insertion to exposure revision, previous revisions and the use of a human pericardium patch were associated with increased risk of failure.
Subject terms: Optic nerve diseases, Eye manifestations, Risk factors, Outcomes research, Surgery
Introduction
There has been an increased interest in aqueous shunt (glaucoma drainage device or tube shunt) implantation over the last 2 decades in parallel with publication of surgical outcomes of a number of randomised trials, including the Tube versus Trabeculectomy (TVT) Study [1], Ahmed Baerveldt Comparison (ABC) Study [2], Ahmed Versus Baerveldt (AVB) Study and the more recent Primary Tube Versus Trabeculectomy Study (PTVT) [3, 4]. Analysis of glaucoma surgical activity among Medicare beneficiaries between 1994 and 2012 revealed a five-fold increase in the number of aqueous shunts implanted in the United States between these dates (from 2356 in 1994 to 12021 in 2012) [5]. At the same time, there was an almost equal increase in the number of tube revision procedures recorded (from 404 in 1994 to 1884 in 2012).
Tube erosion through conjunctiva is a recognised complication of aqueous shunt implantation and if left unrepaired, may lead to complete extrusion of the implant or endophthalmitis [6].
In randomised surgical trials of aqueous shunt surgery such as TVT, ABC, AVB and PTVT, the percentage of exposure has been reported between 1% and 5% after 5 years of follow-up [1–4]. However, data on the outcome of exposure repair are scarce. The purpose of this study was to investigate the survival of aqueous shunt exposure repair without further exposure and to identify risk factors for recurrence of exposure.
Subjects and methods
Study design and eligibility
This was a noncomparative, single-armed, retrospective interventional study of consecutive patients who had undergone repair of an exposed aqueous shunt in a subspecialist glaucoma service between 2010 and 2018.
Patients 16 years of age or older at the time of revision surgery, were identified from a computer search using 2 different internal databases: Galaxy (search codes: “revision of aqueous shunt” and “revision of tube”) and PAS (search code: “other specified operations following glaucoma surgery”; search sub-codes: “graft to sclera”, “allograft to organ”, “resuture to organ” and “revisional operation”). Duplicate entries (i.e. identical cases identified by both the databases) were removed and the remaining entries were manually checked in the electronic medical record system (OpenEyes) to confirm the procedure.
Patients who had undergone revision for a reason other than exposure and those who had the shunt removed during the revision procedure were excluded.
Eligible patients were therefore those 16 years of age or older, who underwent revision surgery for an exposed shunt during the study time period and if sufficient data on clinical details before shunt implantation, during shunt implantation, during revision surgery and follow-up were available. For the purpose of this study we considered sufficient to have data available on visual acuity (VA), intraocular pressure (IOP), number of drops and type of surgery. If a patient had bilateral aqueous shunt exposure, we only included the first eye which developed the exposure.
Data retrieval
Case notes of those deemed eligible were retrieved. Demographic details, as well as clinical data before shunt implantation, surgical implantation technique, revision surgical technique and follow-up data were recorded in a spreadsheet. Demographic information included age at revision surgery, gender and ethnicity. Baseline clinical data before aqueous shunt insertion included laterality, diagnosis, prior glaucoma surgery, lens status, IOP, VA and number of glaucoma medications. Clinical data on shunt insertion surgery comprised date of surgery, grade of surgeon, shunt type, implantation quadrant, adjunctive use of Mitomycin C (MMC), donor patch type, method of tube occlusion, technique of tube fixation to sclera and conjunctival closure technique. Revision surgery data included reason for revision, date of first revision, number of further revisions, time until further exposure, IOP, VA and number of glaucoma medications at most recent follow-up available. If a patient had further shunt exposure repair, the clinical review when the new exposure was noted was considered the last follow-up.
VA was recorded in Snellen format from the notes and then converted in logMAR for statistical analysis.
Outcome and statistical analysis
The primary outcome was recurrence of erosion. A successful outcome was failure to develop further erosion at last follow-up.
Survival without further erosion was determined by Kaplan-Meier analysis.
Cox proportional regression analysis was performed to identify covariates associated with a higher hazard for recurrence. R version 4.0.3 was used for statistical analysis with a p < 0.05 set as significant.
Results
Out of 5462 cases of tube surgeries performed in the considered timeframe, 76 patients (mean age 59 ± 17) fulfilled the inclusion criteria and were included in the study (overall percentage of tube exposure: 1.39%). Before aqueous shunt insertion the median (range) VA was 0.47 (−0.08 to 3.5) logMAR and mean ± SD IOP was 27 ± 10.4 mmHg with a median (range) of 3.5 topical drugs (1–5). Twenty-nine patients (38.1%) were taking systemic acetazolamide (Table 1).
Table 1.
Baseline characteristics before aqueous shunt implantation.
| No. patients/eyes | 76/76 |
| Age at revision (years, mean ± SD) | 59 ± 17 |
| Gender | |
| Female | 28 (36.8%) |
| Ethnicity | |
| White | 46 (60.5%) |
| Black | 15 (19.7%) |
| Asian | 14 (18.5%) |
| Other | 1 (1.3%) |
| Laterality | |
| Right | 34 (44.7%) |
| Diagnosis | |
| Secondary glaucoma | 53 (69.7%) |
| POAG | 20 (26.4%) |
| PACG | 2 (2.6%) |
| NA | 1 (1.3%) |
| Lens status | |
| Pseudophakic | 42 (55.3%) |
| Phakic | 20 (26.3%) |
| Aphakic | 12 (15.8%) |
| NA | 2 (2.6%) |
| Previous surgery (other than uncomplicated phaco + IOL) | |
| Yes | 41 (53.9%) |
| No | 33 (43.5%) |
| NA | 2 (2.6%) |
| Pre-BGI insertion IOP (mean ± SD) | 27 ± 10.4 mmHg |
| No. classes of topical glaucoma medications pre-BGI insertion (median; range) | 3.5; 1–5 |
| Acetazolamide pre-BGI insertion | |
| No | 36 (47.4%) |
| Yes | 29 (38.1%) |
| NA | 11 (14.5%) |
| VA pre-BGI insertion (logMAR, median; range) | 0.47; −0.08–3.5 |
Details of the original shunt implantation procedure are listed in Table 2. The majority of the original implantation procedures (n = 46; 60.5%) were performed by consultant (attending) level surgeons. Most (n = 61; 80.3%) were 350 mm2 Baerveldt implants (BGI-101–350, Johnson & Johnson Vision, Santa Ana, CA) and the most common placement was supero-temporal (n = 61; 80.3%). Adjunctive Mitomycin C (MMC) was applied under the equatorial Tenon’s in the region of the aqueous shunt plate in almost three quarters of the cases (n = 55; 72.4%), most frequently 0.4 mg/ml (n = 45; 59.2%) for 3 min (n = 31; 40.8%) on sponges. The vast majority of the implants (n = 70; 92.1%) were occluded with a rip-cord, frequently combined with an external ligature, either nylon or vicryl.
Table 2.
Surgical details of original aqueous shunt implantation.
| Tube type | |
| BGI-101-350 | 61 (80.3%) |
| BGI-103–250 | 7 (9.2%) |
| AGV | 3 (3.9%) |
| Molteno | 1 (1.3%) |
| NA | 4 (5.3%) |
| Plate position | |
| Supero-temporal | 61 (80.3%) |
| Supero-nasal | 3 (3.9%) |
| Infero-nasal | 5 (6.6%) |
| Infero-temporal | 3 (3.9%) |
| NA | 4 (5.3%) |
| Tube cover | |
| Tutoplast | 36 (47.4%) |
| Sclera | 25 (32.9%) |
| Cornea | 5 (6.6%) |
| Scleral flap | 2 (2.6%) |
| Tenon | 1 (1.3%) |
| None | 2 (2.6%) |
| NA | 5 (6.6%) |
| Plate-limbus distance (mm) | |
| 10 mm | 23 (30.3%) |
| Less than 10 mm | 7 (9.2%) |
| More than 10 mm | 9 (11.8%) |
| NA | 37 (48.7%) |
| MMC | |
| Yes | 55 (72.4%) |
| No | 15 (19.7%) |
| NA | 6 (7.9%) |
| Plate fixation | |
| Prolene | 33 (43.5%) |
| Nylon | 9 (11.8%) |
| NA | 34 (44.7%) |
| Tube fixation | |
| Prolene | 28 (36.9%) |
| Nylon | 12 (15.8%) |
| None | 2 (2.6%) |
| NA | 34 (44.7%) |
Most of the implants (n = 32; 42.1%) were placed so that the anterior edge of the plate was situated at least 10 mm from the limbus and fixated with polypropylene sutures at the level of the plate (n = 33; 43.5%) and the tube (n = 28; 36.9%). Only a minority of the cases (n = 7; 9.2%) had the plate sutured at less than 10 mm from the limbus, presumably for a difficult access to the equatorial space due to patient anatomy. Typically a donor human pericardium patch graft (Tutoplast, Innovative Ophthalmic Products, Costa Mesa, California, USA) was used to cover the tube portion near the limbus (n = 36; 47.4%). About a third of the cases received a scleral patch graft (n = 25; 32.9.%), 5 patients (6.6%) received a cornea patch graft and the rest underwent different tube cover techniques.
Tube venting slits were documented in only 5 cases (6.6%).
Fibrin glue (Tisseel, Baxter AG, Vienna, Austria) was used to secure the patch and conjunctiva in 23 patients (30.3%) and conjunctival closure was achieved with a combination of 10/0 nylon and 8/0 vicryl continuous sutures, with (n = 44; 57.9%) or without fibrin glue (n = 18; 23.7%). Interrupted 10/0 nylon sutures were used to secure the conjunctiva to the limbus (two wing sutures and one horizontal mattress suture) and continuous 8/0 vicryl sutures were placed to suture the side radial cuts of the conjunctiva.
Seventeen patients (22.3%) underwent a surgical revision for reasons other than exposure (i.e. conjunctival dehiscence or tube malposition in the AC) before the onset of the exposure.
Exposure repair details are reported in Table 3. The majority of exposures occurred at the level of the tube (n = 73; 96.1%). The median time (range) between aqueous shunt insertion and exposure repair was 688 days (5–6494).
Table 3.
Exposure and exposure repair details.
| Exposure type | |
| Tube exposure | 73 (96.1%) |
| Plate exposure | 1 (1.3%) |
| Patch exposure | 1 (1.3%) |
| Stent exposure | 1 (1.3%) |
| Patch used for exposure repair | |
| Tutoplast Human Pericardium | 36 (47.4%) |
| Sclera | 22 (28.9%) |
| Cornea | 5 (6.6%) |
| Fascia lata | 1 (1.3%) |
| NA | 12 (15.8%) |
| Glue used for revision | |
| Yes | 24 (31.6%) |
| No | 39 (51.3%) |
| NA | 13 (17.1%) |
| Tube resiting in the anterior chamber or in the sulcus | 9 (11.8%) |
| Further exposures | 23 (30.3%) |
| Tube removal | 9 (11.8%) |
| Devastating adverse events | |
| Endophthalmitis | 2 (2.6%) |
| Eviscerations | 1 (1.3%) |
| Time from BGI insertion to revision (days - median; range) | 688; 5–6494 |
| IOP at last follow-up (mmHg) | 13.5 ± 5.8 mmHg |
| Topical glaucoma medications at last follow-up (median; range) | 1; 0–5 |
| Acetazolamide at last follow-up | |
| No | 74 (97.4%) |
| Yes | 2 (2.6%) |
| VA at last follow-up (logMAR, median; range) | 0.6; 0–3.5 |
Repair technique
During surgical repair, after careful and extensive conjunctival dissection, a pericardial patch graft was employed in 36 cases (47.4%), whereas 22 patients (28.9%) were revised with donor sclera (Table 3). The remaining cases had either cornea or fascia lata. Fibrin glue was used to secure the patch and assist in conjunctival closure in 24 cases (31.6%). Nine patients (11.8%) underwent repositioning of the tube: 8 in the anterior chamber and 1 in the sulcus.
Serious adverse events occurred in 13 patients (17.1%). In 9 cases (11.8%) the tube was eventually removed: in 4 cases for lack of adequate tissue cover, in 2 cases for accidental tube amputation during revision, in 2 cases for plate exposure and in 1 case for persistent hypotony.
Three patients (3.9%) had infection-related complications: one tube exposure with tube related endophthalmitis treated medically before exposure repair (VA was hand motions at last follow-up compared with 2/60 before tube insertion); one had a suspected tube related endophthalmitis after the second tube exposure repair (VA 6/12 at last follow-up); one underwent evisceration for a Fusarium keratitis which developed after a superficial keratectomy. One case (1.3%) developed corneal decompensation for flat AC with loss of LP at last follow-up.
The median (range) follow-up time after exposure repair was 540 (15–3090) days. Survival without further exposure was 76.2%, 72.7% and 54.7% of the cases at 1,2 and 4 years respectively (Fig. 1). Median VA at last follow-up was significantly lower than VA prior to revision (0.6; 0–3.5 vs 0.53; 0.1–3.5; p = 0.02). Median IOP at last follow up was 13.5 ± 5.8 mmHg. Four patients (5.2%) had IOP higher than 21 and underwent further glaucoma surgery.
Fig. 1.
Kaplan-Meier survival curve of follow-up (months) without re-exposure.
Twenty-three patients (30.3%) developed further exposure requiring surgical repair after a median (IQR) of 28 (266) days. At the last follow-up their median VA (range) was 0.25 (0–2.9) and their mean IOP was 13 ± 4.8 mmHg on a median (range) of 1 medication (0–5). Fourteen of them (60.8%) were pseudophakic and 9 of them (39.1%) had previous ocular surgery other than phacoemulsification (7 trabs, 1 tube and 1 scleral buckle for repair of retinal detachment).
On univariate regression analysis, a shorter interval (less than 620 days, i.e. the median) from aqueous shunt implantation to exposure (HR 2.4 ± 0.4; p = 0.04; reference longer interval), a diagnosis of POAG (HR 2.5 ± 0.4; p = 0.02; reference POAG), a history of surgical revision prior to exposure (HR 2.4 ± 0.4; p = 0.04; reference no prior revision) and the use of donor human pericardium as a patch graft (HR 3.2 ± 0.5; p = 0.01; reference other than pericardium) were associated with an increased risk of further exposure after repair. In the multivariate regression model a shorter interval from tube insertion to exposure (HR 2.87 ± 0.4; p = 0.03; reference longer interval), history of revision prior to exposure (HR 3.06 ± 0.4; p = 0.02; reference no prior revision) and use of a human pericardial patch graft (HR 3.4 ± 0.5; p = 0.03; reference other than pericardium) was associated with an increased risk of exposure (Fig. 2).
Fig. 2.
Multivariate Cox analysis of risk factors for failure of exposure repair. Pericardium refers to Tutoplast Human Pericardium.
No case where the tube was repositioned in the anterior chamber (n = 8) or in the sulcus (n = 1) experienced further exposure (n = 9; 11.8%). Lack of repositioning however did not come up as a risk factor for recurrence of exposure in the regression model. This might be related to the small number of patients who had their tube repositioned. The follow-up after revision of those who had the tube repositioned was similar to that of the other patients (median ± SD: 540 ± 363 vs 540 ± 687 days; p = 0.86).
Discussion
Aqueous shunt exposure is a complication that needs to be addressed promptly to prevent the risk of tube related endophthalmitis.
Several risk factors for the development of tube exposure have been proposed, including ocular surgery prior to or concomitant with aqueous shunt implantation, Black, White and Hispanic ethnicities, younger age, older age, female gender, diabetes, neovascular glaucoma, type of tube cover, inferior placement of the device, number of glaucoma medications and topical steroid use [7–13]. Despite the potential role of MMC as a risk factor for exposure, there is no available evidence of an increased risk for exposure or for recurrence of exposure after repair in cases where MMC was employed. When repairing an exposed tube, several options are available. The most common approach is to undermine and dissect eroded conjunctiva, suture the tube securely to sclera to prevent excessive movement, place a fresh patch graft and close the conjunctiva on top.
Different approaches, involving the use of amniotic membrane, buccal mucous membrane and conjunctiva pedicle flap have been proposed more recently with encouraging results [14–18].
These represent options to supplement conjunctiva when it is in short supply, but realistically, still require adjunctive patch grafting.
Although many exposure repair techniques have been described there are few reported data on the outcomes of revision and specifically risk factors associated with recurrence of the exposure.
Huddleston et al. first reported about success rates of aqueous shunt revision for exposure and risk factors for recurrance [19]. They reported a median survival of exposure repair of 56 weeks and 43% required further revisions for exposure, with 13% who needed aqueous shunt explantation. Diabetes mellitus was found to be associated with 3-fold increase in the risk of failure of the revision in the Cox regression analysis whereas Black ethnicity, number of pre-shunt glaucoma eyedrops, previous glaucoma laser procedures and aqueous shunt surgery combined with other ocular surgery were found to be associated with higher risk of further revision in the stepwise regression analysis.
Thompson et al. retrospectively analysed the outcome of tube revision in 94 patients and found that 44% of the cases re-exposed, with 56.1% requiring explantation [20]. The median time between the first and second revision was 21.7 weeks. Caucasian ethnicity and the use of non-scleral patches during primary revision were found to increase the re-exposure risk by more than twofold.
Our results are consistent with these findings, in that we observed the use of human pericardium at the time of revision as a predictor of re-exposure (HR 3.43; 95% CI 1.16–10.13; p = 0.03; Fig. 2). This might be because it is thinner than sclera, or for other mechanical or immunological reasons.
In this regard, Sheha [21] et al. prospectively analysed the host-tissue integration of pericardium patch grafts compared to amniotic membrane-umbelical cord patch grafts. They analysed the patch thickness with anterior segment OCT and found a significantly higher rate of early (43% vs 12%; p = 0.002) and late (28% vs 5%; p = 0.007) graft thinning in the pericardium group compared to the amniotic membrane-umbelical cord. Although the rate of exposures were not statistically different after a mean ± SD follow-up of 29 ± 8 months (5% vs 2% in the pericardium and amniotic membrane-umbelical cord group, respectively; p = 0.54), a thinner patch graft might entail a higher risk of tissue erosion in the long term.
Lankaranian et al. [22] observed a considerably reduced rate of exposure using a double rather than a single layer pericardial graft (0% vs 16%; p = 0.002). Unfortunately, due to the retrospective nature of our analysis, we were unable to identify in which patients a single or double layer of pericardium was employed at the time of repair. This technique modification might mitigate the higher risk of exposure recurrence from the use of pericardium.
The choice of the type of patch reflected the personal preferences of the surgeon. Consequently, it was not possible to identify in our cohort of patients the specific reason why a type of patch was preferred over another, both at the time of the original surgery and at the time of exposure repair.
In terms of the patch fixation technique the commonest choice was suturing (40.8%), followed by fibrin glue (31.6%).
Weinreb et al. recently reported on outcomes of aqueous shunt revisions for a variety of causes, with tube exposure repair representing 41% of the cases [23]. Survival at 5 years was 35% for tube exposure repair, which had twice the risk of failure as compared to revision for repositioning of the tube. The authors found that a different surgeon performing the revision compared to the index surgery, longer time from onset of the symptoms to the repair procedure, right laterality and suspected infection before revision were all associated with higher risk of failure of the revision on stepwise regression analysis.
Bouris et al. recently found a survival of 44% at 36 months following tube revision for exposure [24].
Low et al. reported the results of aqueous shunt revision for exposure with a technique involving autologous buccal mucous membrane and corneal patch graft to cover the exposed area. They achieved success in 81% of the erosion revisions with a mean follow-up of 1.7 years [25]. They observed time from aqueous shunt insertion to repair to be the factor most associated with recurrent erosion (5.2 vs 0.84 years in the success and failure group respectively; p < 0.01).
We also found that a shorter interval from original tube surgery to aqueous shunt revision for exposure was linked to a greater risk of re-exposure (HR 2.87 ± 0.4; p = 0.03; Fig. 2). Logically one might expect longer duration to exposure to be associated with higher success: if the tube was more prone to erode, one would expect it to do so sooner rather than later.
In our study, patients with a history of surgical revision for different reasons (i.e. conjunctival dehiscence and tube malposition) prior to exposure, had a greater risk of re-exposure (HR 3.06 ± 0.4; p = 0.02). We would speculate that prior surgery will have resulted in greater conjunctival shortening and likely result in a higher risk of re-exposure.
Liu et al. recently found a higher number of tube exposure events in cases of wet AMD receiving anti-VEGF injections compared with dry AMD (2.1 vs 1.3; p = 0.01) [26]. The authors postulated that the use of anti-VEGF might interact with the quality of the tissues overlying the tube. Data on the use of anti-VEGF was not collected in our analysis so we could not test this potential association and it is difficult to understand what the potential mechanism might be.
In summary, patients with aqueous shunt exposure who underwent exposure repair surgery showed a reasonable survival rate of 76.2%, 72.7% and 54.7% at 1,2 and 4 years respectively. Recurrence of exposure developed in 30.2% of cases and in 11.8% the tube was explanted. A longer interval from the time of tube insertion to the onset of exposure, a history of previous surgical revision, and the use of a human donor pericardial patch graft during revision rather than other type of patch graft, were associated with a higher risk of further exposure.
Interestingly, no patients in whom the tube was repositioned developed further exposure after revision. This seems most likely due to the reduced external course of the tube, as re-positioning is usually performed to place the tube in the anterior chamber or sulcus more posteriorly than the original entry. Another potential factor in repositioning might be if the angle of entry were less acute after revision, resulting in a lower risk of dolphining of the tube.
These findings should be considered while counselling a patient about an exposure repair procedure and while planning the surgical technique.
We observed 2 cases of tube-related endophthalmitis (2.6%), although only one had a clear clinical presentation whereas the other was only suspicious of early endopthalmitis, rather than a definitive diagnosis. This percentage might seem high, however it reflects the perceived higher risk of this complication in patients with aqueous shunt exposure. The other serious adverse event was fungal keratitis that required evisceration. This latter case was unlikely related to the exposure repair surgery and developed in an eye with a history of corneal disease.
The main limitation of this study is its retrospective design. Some data were missing from the case records or not clearly recorded and were therefore excluded. The implantation and repair procedures were performed by different surgeons with different techniques. Although this might be considered a limitation, it also reflects real life and improves the generalisability of our findings. Another limitation, which is frequent in retrospective time to event analyses, is the different follow-up length among the cases. However, the use of Kaplan-Meier survival analysis compensates for this variability. Moreover, retrospective case series have limited interpretation in terms of risk factors due to bias and confounding, hence the results of our study need to be corroborated with larger, prospective studies in the future.
In conclusion, aqueous shunt exposure is difficult to manage, running the risk of infection, re-exposure or loss of IOP control with limited options if the tube is removed. This supports prior studies that recurrent exposure is relatively common but offers the encouraging finding that re-exposure can be reduced by repositioning the tube and using thicker and potentially more durable patches than pericardium.
Summary
What was known before
Tube exposure has a reported incidence between 1 and 5% at 5 years in the major relevant trials.
It can lead to devastating complications such as endophthalmitis.
There is a paucity of data available on tube exposure repair outcomes and on risk factors for recurrence of tube exposure after repair surgery.
What this study adds
Tube exposure repair surgery has reasonable outcomes.
Risk factors for recurrence of tube exposure after repair were: a shorter interval from the original tube insertion to exposure, history of revision surgery prior to tube exposure and the use of human pericardial patch graft.
The results of this study may help to better consult patient prior to tube exposure repair surgery and to inform surgeons on potential risk factors for failure of this type of surgery.
Author contributions
All the authors have made substantial contribution to the design, drafting, revision and final approval of the work. CG acquired and analysed the data for the work.
Data availability
The data that support the findings of this study are located in Moorfields Eye Hospital and are available from the authors upon reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Gedde SJ, Schiffman JC, Feuer WJ, Herndon LW, Brandt JD, Budenz DL. Tube versus trabeculectomy study group. treatment outcomes in the tube versus trabeculectomy (TVT) study after five years of follow-up. Am J Ophthalmol. 2012;153:789–803.e2. 10.1016/j.ajo.2011.10.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Budenz DL, Feuer WJ, Barton K, Schiffman J, Costa VP, Godfrey DG, et al. Ahmed baerveldt comparison study group. postoperative complications in the Ahmed baerveldt comparison study during five years of follow-up. Am J Ophthalmol. 2016;163:75–82.e3. 10.1016/j.ajo.2015.11.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Christakis PG, Kalenak JW, Tsai JC, Zurakowski D, Kammer JA, Harasymowycz PJ, et al. The Ahmed versus baerveldt study: five-year treatment outcomes. Ophthalmology. 2016;123:2093–102. 10.1016/j.ophtha.2016.06.035. [DOI] [PubMed] [Google Scholar]
- 4.Gedde SJ, Feuer WJ, Lim KS, Barton K, Goyal S, Ahmed II, et al. Primary tube versus trabeculectomy study group. postoperative complications in the primary tube versus trabeculectomy study during 5 years of follow-up. Ophthalmology. 2022;129:1357–67. 10.1016/j.ophtha.2022.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Arora KS, Robin AL, Corcoran KJ, Corcoran SL, Ramulu PY. Use of various glaucoma surgeries and procedures in medicare beneficiaries from 1994 to 2012. Ophthalmology. 2015;122:1615–24. 10.1016/j.ophtha.2015.04.015. [DOI] [PubMed] [Google Scholar]
- 6.Gedde SJ, Scott IU, Tabandeh H, Luu KK, Budenz DL, Greenfield DS, et al. Late endophthalmitis associated with glaucoma drainage implants. Ophthalmology. 2001;108:1323–7. 10.1016/s0161-6420(01)00598-x. [DOI] [PubMed] [Google Scholar]
- 7.Gedde SJ, Herndon LW, Brandt JD, Budenz DL, Feuer WJ, Schiffman JC. Tube versus trabeculectomy study group. postoperative complications in the tube versus trabeculectomy (TVT) study during five years of follow-up. Am J Ophthalmol. 2012;153:804–814.e1. 10.1016/j.ajo.2011.10.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bains U, Hoguet A. Aqueous drainage device erosion: a review of rates, risks, prevention, and repair. Semin Ophthalmol. 2018;33:1–10. 10.1080/08820538.2017.1353805. [DOI] [PubMed] [Google Scholar]
- 9.Muir KW, Lim A, Stinnett S, Kuo A, Tseng H, Walsh MM. Risk factors for exposure of laucoma drainage devices: a retrospective observational study. BMJ Open. 2014;4:e004560. 10.1136/bmjopen-2013-004560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Trubnik V, Zangalli C, Moster MR, Chia T, Ali M, Martinez P, et al. Evaluation of risk factors for glaucoma drainage device-related erosions: a retrospective case-control study. J Glaucoma. 2015;24:498–502. 10.1097/IJG.0000000000000034. [DOI] [PubMed] [Google Scholar]
- 11.Koval MS, El Sayyad FF, Bell NP, Chuang AZ, Lee DA, Hypes SM, et al. Risk factors for tube shunt exposure: a matched case-control study. J Ophthalmol. 2013;2013:196215. 10.1155/2013/196215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Van Hoefen Wijsard M, Haan M, Rietveld E, van Rijn LJ. Donor sclera versus bovine pericardium as patch graft material in glaucoma implant surgery and the impact of a drainage suture. Acta Ophthalmol. 2018;96:692–8. 10.1111/aos.13721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Al-Beishri AS, Malik R, Freidi A, Ahmad S. Risk factors for glaucoma drainage device exposure in a middle-eastern population. J Glaucoma. 2019;28:529–34. 10.1097/IJG.0000000000001220. [DOI] [PubMed] [Google Scholar]
- 14.Godfrey DG, Merritt JH, Fellman RL, Starita RJ. Interpolated conjunctival pedicle flaps for the treatment of exposed glaucoma drainage devices. Arch Ophthalmol. 2003;121:1772–5. 10.1001/archopht.121.12.1772. [DOI] [PubMed] [Google Scholar]
- 15.Grover DS, Merritt J, Godfrey DG, Fellman RL. Forniceal conjunctival pedicle flap for the treatment of complex glaucoma drainage device tube erosion. JAMA Ophthalmol. 2013;131:662–6. 10.1001/jamaophthalmol.2013.2315. [DOI] [PubMed] [Google Scholar]
- 16.Guajardo JM, Lim KS. Long-term safety and efficacy of conjunctival pedicle graft revision combined with repeated pericardium allograft for exposed glaucoma drainage devices. J Glaucoma. 2018;27:910–3. 10.1097/IJG.0000000000001060. [DOI] [PubMed] [Google Scholar]
- 17.Einan-Lifshitz A, Belkin A, Mathew D, Sorkin N, Chan CC, Buys YM, et al. Repair of exposed ahmed glaucoma valve tubes: long-term outcomes. J Glaucoma. 2018;27:532–6. 10.1097/IJG.0000000000000951. [DOI] [PubMed] [Google Scholar]
- 18.Ainsworth G, Rotchford A, Dua HS, King AJ. A novel use of amniotic membrane in the management of tube exposure following glaucoma tube shunt surgery. Br J Ophthalmol. 2006;90:417–9. 10.1136/bjo.2005.084905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Huddleston SM, Feldman RM, Budenz DL, Bell NP, Lee DA, Chuang AZ, et al. Aqueous shunt exposure: a retrospective review of repair outcome. J Glaucoma. 2013;22:433–8. 10.1097/IJG.0b013e3181f3e5b4. [DOI] [PubMed] [Google Scholar]
- 20.Thompson AC, Manjunath V, Muir KW. Risk factors for earlier reexposure of glaucoma drainage devices. J Glaucoma. 2017;26:1155–60. 10.1097/IJG.0000000000000821. [DOI] [PubMed] [Google Scholar]
- 21.Sheha H, Tello C, Al-Aswad LA, Sayed MS, Lee RK. Outcomes of the shunt tube exposure prevention study: a randomized clinical trial. Ophthalmol Glaucoma. 2019;2:392–401. 10.1016/j.ogla.2019.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lankaranian D, Reis R, Henderer JD, Choe S, Moster MR. Comparison of single thickness and double thickness processed pericardium patch graft in glaucoma drainage device surgery: a single surgeon comparison of outcome. J Glaucoma. 2008;17:48–51. 10.1097/IJG.0b013e318133fc49. [DOI] [PubMed] [Google Scholar]
- 23.Weinreb S, Cardakli N, Jefferys J, Quigley H. Long-term functional outcomes of glaucoma tube shunt revision surgery. Ophthalmol Glaucoma. 2019;2:383–91. 10.1016/j.ogla.2019.08.002. [DOI] [PubMed] [Google Scholar]
- 24.Bouris E, de Gainza A, Barsegian A, Caprioli J. The success rate of glaucoma drainage device revision. J Glaucoma. 2023;32:489–96. 10.1097/IJG.0000000000002217. [DOI] [PubMed] [Google Scholar]
- 25.Low SA, Rootman DB, Rootman DS, Trope GE. Repair of eroded glaucoma drainage devices: mid-term outcomes. J Glaucoma. 2012;21:619–22. 10.1097/IJG.0b013e3182447d83. [DOI] [PubMed] [Google Scholar]
- 26.Liu KC, Gomez-Caraballo M, Challa P, Asrani SG. Recurrent tube erosions with anti-vascular endothelial growth factor therapy in patients with age-related macular degeneration. Ophthalmol Glaucoma. 2020;3:295–300. 10.1016/j.ogla.2020.04.005. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are located in Moorfields Eye Hospital and are available from the authors upon reasonable request.