Article Text
Abstract
Background/Aims To analyse graft detachments prior to rebubbling, the influence of rebubbling on the postoperative outcome after Descemet membrane endothelial keratoplasty (DMEK) and the need for rebubbling on the contralateral eye.
Methods In this retrospective cohort study, out of 1541 DMEKs, optical coherence tomography scans and clinical records of 499 eyes undergoing rebubbling after DMEK at the University Hospital of Cologne, Cologne, Germany, were examined. Main Outcome measures were (a) number, localisation and size of graft detachments; (b) influence of rebubbling/s on postoperative outcome after 12 months; and (c) rebubbling risk of the contralateral eye after DMEK.
Results Mean number of detachment areas was 2.02±0.9. Mean lateral diameter of all detachments was 4534.76±1920.83 μm. Mean axial diameter was 382.53±282.02 μm. Detachments were equally distributed over all regions of the cornea. Best spectacle corrected visual acuity ( BSCVA) after 12 months was 0.197±0.23 logarithm of the minimum angle of resolution, endothelial cell density (ECD) was 1575.21±397.71 cells/mm2 and mean central corneal thickness (CCT) was 566.37±68.11 μm. BSCVA, CCT, ECD or endothelial cell loss of all rebubbled patients were not influenced by the number of rebubblings or the time between DMEK and rebubbling. Of the rebubbled patients, which received a DMEK subsequently on the other eye, 193 (58.8%) also received a rebubbling, which was significantly higher, when compared to the overall rebubbling rate of 32.3% (p=0.000).
Conclusions The overall number of rebubblings has no influence on the postoperative outcome after DMEK, if a rebubbling becomes necessary. Patients who received a rebubbling on one eye have an elevated risk for a rebubbling on the fellow eye.
- Cornea
- Imaging
- Treatment Surgery
Statistics from Altmetric.com
INTRODUCTION
Descemet membrane endothelial keratoplasty (DMEK) is a safe and frequently performed type of lamellar keratoplasty to treat diseases of the corneal endothelium, for example, in Fuchs endothelial corneal dystrophy or pseudophakic bullous keratopathy after cataract surgery.1– 5 Over the past several years, many topics regarding the technique of DMEK have been discussed controversially, including dyes for the graft preparation, injection cartridges, stripping techniques or the s-stamp technique to confirm correct graft orientation.6–9
One of the most discussed and least standardised aspects is the treatment of graft detachment with a reinjection of an anterior chamber tamponade, such as room air or 20% sulfur hexafluoride (SF6) gas.10 11 Regarding the best way to handle graft detachments, different approaches have been postulated, from treating every detectable detachment of the Descemet’s membrane–endothelium (DME) graft regardless of size and location to treating only very large detachments, affecting the optical axis of the patients or causing discomfort of the patient through epithelial oedema.12–15 In previous studies, we have shown that, if carefully investigated, nearly all patients develop subtle, detectable DME graft detachments without the general need for further treatment.16 Moreover, we have shown that patients who were treated by a rebubbling had both larger detachments and more areas of detachment at the same time, compared with patients who did not receive a rebubbling.17 Additionally, it has been shown that the size and number of graft detachments after DMEK can be reduced by using SF6 gas, which was introduced into DMEK surgery by Güell et al. 10 17 Furthermore, we have shown in another study that rebubbling has a significant impact on endothelial survival after DMEK.18 However, there are no exact data regarding the characteristics of the detachments immediately before rebubbling. It is also unclear how the time point and the relative and overall number of the rebubbling procedure influence the postoperative outcome after DMEK and how many patients also need rebubbling in the fellow eye.
MATERIAL AND METHODS
In this retrospective single-centre study, we analysed anterior segment optical coherence tomography (AS-OCT) scans (slit-lamp-OCT or Spectralis OCT, Heidelberg Engineering, Heidelberg, Germany) and clinical records of the patients who underwent one or more rebubblings after DMEK as a consequence of postoperative graft detachment out of overall 1254 patients (1541 eyes), who underwent DMEK at the University Hospital of Cologne, Department of Ophthalmology between July 2011 and August 2017. Read-out parameters were (a) number, location, and axial and lateral diameter of graft detachments; (b) influence of the time period between surgery and rebubbling as well as the overall number of rebubblings on the postoperative outcome after 12 months; and (c) rebubbling rate of the fellow eye after subsequent DMEK. Moreover, the donor characteristics of the grafts, which where rebubbled, were analysed regarding endothelial cell density (ECD), donor age, donor sex and culture time. Patients were excluded if they were suffering from glaucoma or had a previous graft rejection.
Indications for rebubbling were generally defined as previously published. They were (a) progressive detachments which affected more than one-third of the graft size, with either (b) surface irregularities due to corneal oedema or (c) corneal oedema. which reduces visual acuity in addition to not improving graft detachments. SF6 gas was used for rebubbling only for very large detachments, for purely inferior detachments and for large open iridectomy. Surgery and rebubblings were performed by two experienced DMEK surgeons (CC and BB). DMEK surgery was performed as previously described.7 No viscoelastic device was used. Transplants were prepared immediately before surgery and marked using the Bachmann mark.19 20 No prestripped tissue was used. Room air or 20% SF6 gas was used as an anterior chamber tamponade. Remaining anterior chamber gas fill at the end of surgery was 80%. After surgery and after rebubbling, the patients were positioned in a supine position, until the tamponade of the anterior chamber had resorbed at the pupillary level.
Image analysis
OCT images of whole volume scans or single axial scans were collected and manually analysed as published earlier.21 For exact measurements, the Heidelberg Eye Explorer software was used, and detachments were measured using the integrated measuring tool (Heidelberg Engineering). We analysed the location (central cornea within 4 mm vs periphery of the cornea; upper, middle or lower third, complete detachment) of the graft detachments according to the scheme published earlier (figure 1).17 Moreover, we measured the lateral and axial diameters of the largest detachment. In the case of detachments that were extending over two or more different contiguous locations, they were registered separately for each area of interest.
Statistical analysis was performed using commercially available statistics software (SPSS Version 22, IBM Corp, Armonk, NY, USA). Differences in frequencies were calculated using the Fisher’s exact test. For the analysis between two groups, the Mann-Whitney U test was used. For comparison between different groups, the Kruskal-Wallis test was used. P values ≤0.05 were defined to be statistically significant. SD is given in brackets. Due to the number of analysis, no correction for multiple testing was performed.
All research was in conformity to the Declaration of Helsinki. The ethics commission of the University of Cologne approved the retrospective analysis and prospective data collection (EF-number 14–373).
RESULTS
In the analysis, 624 rebubbling procedures of 499 eyes (430 patients) (age 70.93±9.83 years; range: 39–91 years) were included (table 1).
Of these eyes, 286 (57.3%) were from female patients and 213 (42.7%) from male patients. A total of 240 (48.1%) eyes received a pseudophakic DMEK, 236 (47.3%) underwent a triple-DMEK procedure and 23 (4.6%) a phakic DMEK procedure. In 391 (78.6%) eyes, air was used as initial anterior chamber tamponade during DMEK, and in 108 (21.4%) eyes, it was SF6 gas. In 264 (52.9%) procedures left eyes and in 235 (47.1%) right eyes were rebubbled. The majority of the eyes received one rebubbling (398 eyes,79.8%), 83 (16.6%) eyes received two rebubblings, 13 (2.6%) eyes received three rebubblings, 4 (0.8%) eyes received four rebubblings and 1 (0.2%) eye received five rebubblings after the same DMEK surgery. The overall number of rebubblings did not differ significantly comparing the use of SF6 gas (1.24±0.53; range: 1–4) or air (1.25±0.57; range: 1–5) as initial anterior chamber tamponade (p=0.953). There were no significant differences between pseudophakic, phakic or triple-DMEK procedures (p=0.129) regarding the overall number of rebubblings.
Mean time until the first rebubbling after DMEK was 7.91±8.5 days (range 0–92 days, n=499) and time between the first and the second rebubbling was 9.56±8.35 days (range 1–43 days, n=101). Air was used for anterior chamber tamponade in 413 rebubbling procedures (83.6%), and SF6 gas was used in 81 rebubbling procedures (16.4%). Mean interval between initial surgery and the first rebubbling was significantly longer in eyes that received SF6 than in eyes that received air as initial anterior chamber tamponade (13.46±9.78 days for SF6 vs 6.38±7.42 days for air; p=0.000). Of the patients analysed in this study, 329 (65.9%) also received DMEK surgery on the fellow eye. Of these patients, 193 (58.8%) received a rebubbling on the fellow eye. Compared to the overall rebubbling rate of 32.3% (n=499), the probability for the need of a rebubbling was significantly elevated (p=0.000) when the first eye had already gotten a rebubbling. Relative rebubbling rate in this study was 48.5% (n=308) for air and 14.7% (n=108) for SF6 (compare table 2 for graft characteristics of the subgroups SF6 and air). These differences did not differ between both surgeons. Of the rebubbled patients initially operated on by surgeon 1, 62 patients were also operated on the other eye. Of these 62 patients, 32 were also rebubbled on the partner eye (51,6%). Of the rebubbled patients initially operated on by surgeon 2, 267 patients were also operated on the other eye. Of these 267 patients, 161 were also rebubbled on the partner eye (60.3%). There was no statistically significant difference between those values (p=0.246).
Characteristics of the graft detachments
For the exact analysis of the detachment characteristics by AS-OCT, 442 patients (70.8%), which had evaluable OCT images taken immediately before the rebubbling, were included (table 3).
Mean number of visible overall detachments was 2.02±0.96 (range: 0–5). Mean number of detachments before the first rebubbling compared to detachments before subsequent rebubblings was significantly lower (2.08±0.98; vs 1.78±0.87; p=0.013). However, the overall lateral diameter of the detachments was significantly larger when comparing these groups (4415.07±1871.02 μm vs 5085.62±2059.47 μm; p=0.007). Mean lateral diameter of all detachments before rebubbling was 4534.76±1920.83 μm (range: 1014–10 560 μm). Mean axial diameter of all the detachments was 382.53±282.02 μm (range: 7–2784 μm), but in case of a further rebubbling, the mean axial diameter after the first rebubbling was significantly larger than the initial detachment (353.64±265.69 μm vs 517.26±316.84 μm; p=0.000).
Regarding the location of the detachments, 325 (73.53%) eyes showed detachments in the upper third of the cornea, 298 (67.42%) showed detachments in the mid-third and 353 (79.86%) showed detachments in the inferior third. Nearly all eyes had peripheral detachments (435 (98.42%)), whereas 159 (34.16%) showed central graft detachments. Twenty-eight (6.33%) eyes had graft detachments involving all areas of the graft (all thirds as well as graft centre and periphery).
Donor characteristics of rebubbled DMEK grafts
Donor ECD was 2703.9±229.0 cells/mm2 (range: 2200–3800 cells/mm2, n=496). Mean donor age was 68.25±11.05 years (range: 18–90 years; n=493). Of the rebubbled grafts, 161 (39.1%) were of female donors and 251 (60.9%) were of male donors (n=412). Female donor grafts were slightly more often rebubbled than male grafts (1.3±0.5 vs 1.21±0.5, p=0.042). Culture time of the grafts in warm and cold culture was 15.67±6.07 days (range: 4–35 days; n=488).
Outcome of all rebubbled patients after 1 year
The overall best spectacle corrected visual acuity (BSCVA) in logarithm of the minimum angle of resolution after 12 months of all rebubbled patients was 0.13±0.15 (n=203). Patients were excluded from this analysis if they were suffering from extracorneal diseases. Mean ECD was 1575.21±397.71 cells/mm2 at this time and mean central corneal thickness (CCT) was 566.37±68.11 μm after 12 months. There was no influence of the overall number of rebubblings on BSCVA, CCT, and ECD or endothelial cell loss after 1 year. For single values for each subgroup depending on the number of rebubblings, see table 4. There was no difference in the postoperative outcome after 1 year regarding BSCVA, CCT and ECD or endothelial cell loss if patients were rebubbled within the first 7, 14 or 21 days after DMEK. No differences regarding the postoperative outcome (visual acuity, central corneal thickness, ECD) were observed between the SF6 or the air group after 2 and 3 years, but slightly significant differences were observed after 12 months in BSCVA and CCT (table 5).
DISCUSSION
In this study, we investigate a large group of patients, who received one or more rebubbling after DMEK due to detachments of the DME graft, for the first time. So far, this topic is controversially discussed and no studies have been published, analysing a large number of rebubbled patients.22
In a previous study, we showed that patients who had received SF6 gas as anterior chamber tamponade showed significantly smaller and less graft detachments than patients who received room air.17 Comparing these two groups in our study, statistically significant differences, such as time until rebubbling, were found. In addition, the graft detachments in the patients analysed here who required rebubbling were larger in patients who received SF6 as anterior chamber tamponade than in patients who received air. This could be explained by the fact that due to the long remaining gas in the anterior chamber, in the case of SF6 tamponade, a rebubbling is performed later than in the case of air tamponade. This could mean that the decision whether to rebubble a patient depends on the initial anterior chamber tamponade. Also, the resorption of the tamponade has an impact on the surgeon’s decision to rebubble or not. However, the time between surgery and first rebubbling was significantly higher in patients who received SF6, which is most likely explained by the longer resorption time of SF6. Moreover, in a previous study, we could show that most detachments were localised in the inferior third or the periphery of the cornea.17 Interestingly, there was no preference regarding the location of the detachments analysed in this new study. The detachments were distributed equally in all investigated regions of the cornea (centre vs periphery; upper, medial and inferior third of the cornea), which could be a selection artefact, as detachments in the superior parts of the transplant are rare. Looking at the size and total number of graft detachments as a function of the number of rebubblings performed, it is noticeable that, while it initially increases in terms of its maximum extent, the total number of detachments decreases as the number of rebubblings increases. This could probably mean that existent detachments reorganise and become confluent, meaning they become less in number but larger in size. This hypothesis would include another theory we published earlier, namely the process of graft plasticity.16 Graft plasticity means that there is no steady state of graft attachment or detachment but a dynamic process of reattachment and detachment of the graft, influenced by different factors. These factors might be (a) type and amount of anterior chamber tamponade (eg, SF6 gas or air), (b) compliance of the patient and supine position, (c) postoperative intraocular pressure and (d) type of DMEK surgery (phakic, pseudophakic, triple-DMEK) and configuration of the anterior chamber (eg, depth, synechia, glaucoma tubes). Moreover, the concentration of the used SF6 gas (eg, 5% vs 20%) could have an important impact on the rebubbling behaviour.23 Furthermore, the technical skills of the surgeons might be essential, as younger and inexperienced surgeons could have an increased graft detachment rate. The learning curve of the surgeon was previously described as crucial for the outcome of the patients.20 24 In this study, we also investigated that for 193 (58.66%) of the 329 (65.9%) patients, who received rebubbling and DMEK on the fellow eye later, another rebubbling was necessary. This could lead to the conclusion that the necessity for a rebubbling is elevated if DMEK surgery is performed on the fellow eye, when compared to a normal population. It is unclear, whether this elevated rebubbling rate is a consequence of the individual postoperative behaviour, compliance of the patients or individual wound healing conditions. This hypothesis could be supported by the fact that the contralateral eye of a patient is usually operated on by the same surgeon.
Generally, rebubbling rates were reported ranging from around 10–40% of the patients. The influence of the rebubbling on endothelial cell loss is unclear.11 25 26 In this context, our results are contrary to an earlier study on this topic. In a study published by Feng et al, the authors conclude that, regarding the endothelial cell loss, there was no difference between patients who had no rebubbling and those who had only one rebubbling.22 Additionally, they showed that patients who received more than one rebubbling had significantly more endothelial cell loss than the other two groups. This result could be due to the size of the different study groups or varying rebubbling techniques by the different surgeons.
Finally, the following conclusions could be drawn in this study:
Rebubbled graft detachments were of no preferred location.
Were on average large large detachments of around 4.5 mm of lateral and 350 μm of axial diameter.
The time point of rebubbling coincides with the resorption of the anterior chamber tamponade (around 1 week with air, around 2 weeks with SF6 gas).
Patients who had a rebubbling on one eye might have an elevated risk for a rebubbling when they get DMEK on the other eye.
This implies that, if a patient already was treated by a rebubbling after DMEK on one eye, the fellow eye should be investigated very carefully within the early postoperative period.
Nevertheless, the study has the following limitations. First, this is only a retrospective analysis of rebubblings also subjectively indicated by the respective surgeon and not a prospective study. Second, the results could be distorted by the fact that two different surgeons operated on the patients evaluated here. Third, DMEKs were included where both SF6 and air were used as anterior chamber tamponade. Fourth, the number of patients with SF6 and with air was not identical.
In this study, we could not show that the overall number of rebubblings as well as the time period between surgery and rebubbling have an influence on the postoperative outcome (endothelial cell, CCT, BSCVA). However, we previously published that rebubbling in itself increases the risk for endothelial cell loss.18 Nevertheless, this result should be interpreted with great caution, as the number of patients treated with a larger number of rebubblings or a longer period between DMEK and rebubbling is very small and may have very little statistical significance.
REFERENCES
Footnotes
Contributors Planning: SS, KK. Conduct: SS, KK, PS, MM, AH, JF, SSC, TH, BB, CC. Reporting: SS, MM, KK, CC, BB, JLG. All the authors provided significant contribution to planning, conduct and reporting of this study.
Funding This study was supported by DFG FOR 2240 ‘(Lymph) Angiogenesis and Cellular Immunity in Inflammatory Diseases of the Eye’ (SS, CC; www.for2240.de); EU COST BM 1302 ‘Joining Forces in Corneal Regeneration’ (CC, SS); EU STRONG FP7 (CC); EU EFRE NRW (BB, SS); and EU Horizon 2020 Arrest Blindness (CC).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request.
Linked Articles
- At a glance