Improvement of the loss of corneal endothelial cells by relocating the tip of the Baerveldt® implant tube from the anterior chamber to the vitreous cavity

It has been reported that corneal endothelial cells (CEC) decrease faster when the tip of the Baerveldt glaucoma implant (BGI) tube is inserted into the anterior chamber than into the vitreous cavity. We investigated whether surgically relocating the tip of the BGI tube from the anterior chamber to the vitreous cavity could reduce CEC loss. This was a single facility retrospective cohort study. The inclusion criteria were the CEC density less than 1500 cells/mm2 and the CEC reduction ratio was greater than 10%/year. The subjects were 11 consecutive patients that could be followed > 12 months after relocation surgery. All patients were undergone vitrectomy, and the tip of tube was inserted into the vitreous cavity from the anterior chamber. We compared the intraocular pressure (IOP), reduction slope of CEC density and annual reduction rate of CEC density before and after relocation surgery. We calculated the annual reduction ratio in comparison with the preoperative CEC density (%/year). The mean of period between the Baeveldt with anterior chamber insertion surgery and the relocation surgery was 33.8 ± 15.0 months. The mean of follow-up after relocation surgery was 21.8 ± 9.8 months. The relocation surgery did not significantly change IOP (p = 0.974). The mean preoperative and postoperative IOP were 13.1 ± 4.5 and 13.6 ± 4.3 mmHg. The reduction ratio of the CEC density was 15.4 ± 6.7 (%/year) before relocation surgery and significantly slower at 8.3 ± 6.5 (%/year) after relocation surgery (p = 0.024). Two patients resulted in bullous keratopathy after relocation surgery. Relocating the tip of BGI tube from the anterior chamber to the vitreous cavity could reduce CEC loss.


Introduction
It is well known that filtration surgeries for glaucoma decreases corneal endothelial cells (CEC). Various devices are used for glaucoma filtration surgery, and some of studies were reported that these surgeries decreased CEC. Higashide et al. [1] reported that trabeculectomy decreased 9.3% after 2 years. Arimura et al. [2] reported that Ex-press decreased 18.0% after 2 years. Nassiri et al. [3] reported that Ahmed glaucoma valve decreased 11.52%, and Molteno decreased 12.37% after 2 years. Gillmann et al. [4] Abstract Purpose It has been reported that corneal endothelial cells (CEC) decrease faster when the tip of the Baerveldt glaucoma implant (BGI) tube is inserted into the anterior chamber than into the vitreous cavity. We investigated whether surgically relocating the tip of the BGI tube from the anterior chamber to the vitreous cavity could reduce CEC loss. Patients and methods This was a single facility retrospective cohort study. The inclusion criteria were the CEC density less than 1500 cells/mm 2 and the CEC reduction ratio was greater than 10%/year. The subjects were 11 consecutive patients that could be followed > 12 months after relocation surgery. All patients were undergone vitrectomy, and the tip of tube was inserted into the vitreous cavity from the anterior chamber. We compared the intraocular pressure (IOP), reduction slope of CEC density and annual reduction rate of CEC density before and after relocation surgery. We calculated the annual reduction ratio in comparison with the preoperative CEC density (%/year). Results The mean of period between the Baeveldt with anterior chamber insertion surgery and the relocation surgery was 33.8 ± 15.0 months.
reported XEN decreased 14.3% after 2 years. Tan et al. reported that Baerveldt with anterior chamber insertion decreased 13.6% after 3 years.
The Baerveldt® glaucoma implant (BGI) (Johnson & Johnson Vision, Santa Ana) is a useful option in the treatment of refractory glaucoma [5]. There are two types of BGI device: one is designed to be inserted into the anterior chamber, and the other is designed to be inserted into the vitreous cavity. It has been reported that insertion of the BGI into the anterior chamber leads to faster decrease in CEC density than insertion into the vitreous cavity [6][7][8][9]. Rapid corneal endothelial cell loss in the case of anterior chamber insertion could cause corneal decompensation.
Although such these studies have been reported, there are no reported results of surgery relocating the tip of a tube from the anterior chamber into the vitreous cavity. We investigated whether this relocation surgery could suppress the CEC reduction rate.

Patients
This was a single facility retrospective cohort study. We examined consecutive patients who had already undergone BGI surgery with anterior chamber insertion and were showing rapid reduction of CEC. Relocation surgery was performed for the cases with significant CEC loss and a high risk of developing bullous keratopathy. The inclusion criteria were the CEC density less than 1500 cells/mm 2 and the CEC reduction ratio was greater than 10%/year. We excluded one case in which the tip of the tube was in contact with the cornea, leading to bullous keratopathy and keratoplasty surgery. The subjects selected for analysis were 11 consecutive patients (11 eyes) who had been followed at least 12 months after undergoing surgery to replace the tip of the tube of BGI from the anterior chamber to the vitreous cavity at Toyama University Hospital between December 2016 and February 2019. We inserted Bearveldt into anterior chamber for 68 eyes during the follow-up period. During this period, 4 eyes developed bullous keratopathy. We experienced such cases and we performed the relocation surgery.
We compared the intraocular pressure (IOP) and reduction rate of CEC density before and after the relocation surgery. The procedures used conformed to the tenets of the Declaration of Helsinki. All patients agreed to treatment with written consent. The research protocol was approved by the Institutional Review Board of the University of Toyama.

Surgical techniques
All relocation surgeries were performed by one glaucoma surgeon (N.T.). In all cases, after retrobulbar anesthesia, a pars plana vitrectomy was performed with EVA (DORC; Zuidland, The Netherlands) simultaneously by a vitreous surgeon before relocation of the BGI. All patients had undergone cataract surgery and IOL was inserted.
All patients were inserted Baerveldt tube (BG350-101) into anterior chamber. (Fig. 1A) The conjunctiva was incised to expose the tube that had previously been inserted in the anterior chamber (Fig. 1B). The conjunctiva on the BGI plate was not incised. The tip of the tube was removed from the anterior chamber (Fig. 1C), and the insertion wound was closed with a 10-0 nylon suture (Mani, Utsunomiya, Japan) (Fig. 1D). All patients underwent vitrectomy to prevent tube occlusion with vitreous body (Fig. 1E). The tip of the tube was then trimmed and inserted into the pars plana at 3.5 mm from the limbus through a 20-ga. V-lance (Mani). The tip of tube was inserted into vitreous cavity (Fig. 1F). When the insertion of tube into vitreous cavity, we did not use ligature. The exposed part of the tube was then covered with a suitable amount of preserved donor sclera (Fig. 1G). The donor sclera was sutured to the host sclera at each corner with 9-0 nylon sutures (Mani). The conjunctiva was sutured with 9-0 absorbable sutures (Mani) so that the sclera would not be exposed (Fig. 1H).

Post-operative protocol
Topical antibiotics and steroids were used as postoperative treatments, and were gradually reduced over the 4-week period following surgery. Topical nonsteroidal anti-inflammatory drugs were used for 3 months. Anti-glaucoma medication was used when the glaucoma specialists (N.T.) considered it necessary after the surgery.

Measurement of IOP and CEC
IOP was measured with Goldmann applanation tonometer (GAT). The timing of the IOP measurement was not specified; we simply used the IOP value at the time of consultation. We used an EM-4000 specular microscope (Tomey Corp., Nagoya, Japan) for the measurement of CEC density. We measured only the central (not peripheral) CEC density and calculated the value of the density automatically. We measured the IOP and CEC density at least every 6 months after the BGI surgery with anterior chamber insertion. We measured the IOP and CEC density at 1, 3, 6 and 12 months after relocation surgery.
We evaluated the loss of CECs in two ways. One is calculated how the CEC density would decrease over the course of a year. We plotted the value of CEC density at each period and calculate the density of CEC reduction value per year with the regression line (cells/mm 2 /year). The other was calculated the reduction ratio in comparison with the preoperative CEC density (%/year). The preoperative CEC density was defined as the value of CEC density measured just before the surgery.

Statistical analysis
All statistical analyses were performed using the JMP Pro 14 software program (SAS, Cary, NC). We used the Wilcoxon signed-rank sum test to compare the change of CEC, IOP and the number of glaucoma medications with same patients, and we used the student-t test to compare the reduction ratio between preoperative and postoperative relocation surgery. The threshold for significance was p < 0.05.

Ophthalmic data
The patients' characteristics are summarized in Table 1. We analyzed the cases of 11 eyes of 10 males and 1 female. The mean ± standard deviation values and range for all 11 eyes were as follows: age at the time of relocation surgery, 62.6 ± 16.6 years. All patients had undergone glaucoma surgeries before BGI surgery. Table 2 summarized the comparison between BGI into anterior chamber insertion and the relocation surgery. The mean of follow-up period from the anterior chamber BGI insertion surgery to the relocation surgery, 33.8 ± 15.0 months; period after relocation surgery, 21.8 ± 9.8 months.

Postoperative IOP
The BGI surgery with insertion onto anterior chamber could decreased IOP from 34.9 ± 7.4 to 13.1 ± 4.5 mmHg (p < 0.0001). The mean of IOP before and after relocation surgery were 13.1 ± 4.5 and 13.6 ± 4.3 mmHg, respectively. The relocation surgery did not significantly change the mean of IOP (p = 0974). The mean of number of glaucoma medications before the BGI surgery with insertion onto anterior chamber was 3.7 ± 0.6 and after BGI surgery was 2.7 ± 1.2 drops. The mean of number of glaucoma medications before the relocation surgery was 2.7 ± 1.2 and after relocation surgery was 3.3 ± 0.6 drops.

Changes in the CEC density
The CEC densities before and after relocation surgery for all cases are plotted in Fig. 2. CEC density before anterior chamber BGI insertion surgery, 2003 ± 699 cells/mm 2 and CEC density decreased 1225 ± 472 cells/mm 2 during only 33.8 months.
The mean of preoperative and postoperative (at the final follow-up) CEC density was 1225 ± 472 and 1153 ± 514 cells/mm 2 . The reduction value of CEC was significantly improved from 294 ± 130 to 53 ± 89 cells/mm 2 /year (p = 0.0002). The reduction ratio after BGI surgery into anterior chamber insertion 15.4 ± 6.7%/year. The reduction ratio after relocation surgery 8.3 ± 6.5%/year. The relocation surgery could significantly reduce the CEC loss (p = 0024).

Complications
There were no cases of endophthalmitis or hypotony maculopathy. There was one case with tube occlusion. The tip of tube was covered with uvea at pars plana. We peeled off the uvea tissue using a 25-ga. V-lance (Mani) to release the blockage. Two cases resulted in bullous keratopathy. The CEC density immediately before the relocation surgery were only 521 and 1281 cells/mm 2 . These patients did not wish to undergo a Descemet's stripping automated endothelial keratoplasty and was followed up conservatively.

Discussion
Our present analyses demonstrated that the reduction ratio of CEC density decreased from 15.4 ± 6.7  [1][2][3][4]. However, this study recruited the cases with faster CEC reduction, so it is difficult to compare the results simply. It was reported that the mean percentage of CEC losses at 1 years were 4.5-13.1% after BGI surgery with anterior chamber insertion, and 1.8-2.1% after BGI surgery with vitreous cavity insertion [6][7][8][9]. Regarding CEC protection, BGI surgery with insertion into the vitreous cavity might be better. However, insertion into the vitreous cavity is not always advisable because it requires a vitrectomy, which could cause various complications [10]. In our study, there were no complications associated with vitrectomy.
In our study, we targeted cases with a rapid decrease in CEC, so the preoperative reduction ratio was as high as 15.4%. Several theories have been proposed as the mechanism for CEC loss subsequent to insertion of a drainage device into the anterior chamber: (1) the tube and the CEC might come into physical contact when the eyes are moved [11]; (2) jet water flow may occur due to aqueous reflux in the tube [12]; and (3) the tube stimulates iritis and the production of inflammatory cytokines [13].
It has been reported that the closer the distance between the tip of the tube and the CEC, the more likely the CEC are to decrease [8]. Gillmann et al. [4] reported that the risk factor for CEC loss was the presence of a shallow anterior chamber for XEN surgery. We did not measure the distance between the tip of the tube and the CEC. All patients had already undergone the cataract surgery before first Bearveldt surgery (anterior chamber insertion). We suspect the relocation surgery could return the flow of aqueous humor to physiological conditions. By eliminating the factors that affect CEC, it may be possible to minimize the loss of CEC. As long as the tube remains in the anterior chamber, it is thought that the CEC density will continue to decrease.
The annual reduction ratio of CEC after relocation surgery was 8.3%. These results were higher than those of previous reports of BGI insertion into the vitreous cavity. One reason might be that all patients had a history of other glaucoma surgery, 9 to 11 had undergone filtration surgery (trabeculectomy and Express). Glaucoma filtration surgery is known to accelerate CEC loss [1,14,15]. All patients underwent cataract surgery and vitrectomy, and we speculate that these surgeries could affect CEC loss. Bourne et al. [16] reported that cataract surgery decreased 10% of CEC after 1 year. Emo et al. [17] reported that combined cataract surgery and vitrectomy decreased 9.5% of CEC after 3 months. Koushan et al. [18] reported a 3-month decrease in CEC by 13.9% after vitrectomy.
The limitations of our study are that it is retrospective, so that our results could be subject to various biases and without comparative arm. All patients had undergone a prior cataract surgery, and most had undergone surgery for glaucoma. Furthermore, all patients underwent simultaneous vitrectomy with the relocation surgery. Intraocular surgeries, especially trabeculectomy, are known to promote CEC loss and might have affected our outcome. Although Fig. 2 The CEC density before and after the relocation surgery pseudo-exfoliation and secondary glaucoma are known to show rapid CEC loss, our study mixed different types of glaucoma. We did not consider the effects of peripheral anterior synechia, the distance between the tip of the tube and the corneal endothelium, or glaucoma medications. We could not measure the CEC density at the exact same place every time. Since a single surgeon performed all the surgery, the surgical outcomes might not be universal. Our study had a small sample size and short-term follow-up period. However, it should be noted that there was a significant difference in the reduction rate of CECs after BGI relocation surgery.
In conclusion, relocating the BGI tube tip from the anterior chamber to the vitreous chamber might reduce the rate of CEC loss. This relocation surgery might decrease serious complications such as corneal decompensation.
Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Akio Miyakoshi, Yuuki Honda and Naoki Tojo. The first draft of the manuscript was written by Yuuki Honda and Naoki Tojo and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Toyama University.
Consent to participate Informed consent was obtained from all individual participants included in the study.

Content and publish
Our manuscript do not contain any individual person's data in any form.

Employment
No employment by any organization that may gain or lose financially through publication of this manuscript.