Identification of the preoperative and perioperative factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty: A retrospective cohort study.

Low postoperative endothelial-cell density (ECD) plays a key role in graft failure after Descemet-membrane endothelial keratoplasty (DMEK). Identifying pre/perioperative factors that predict postoperative ECD could help improve DMEK outcomes. This retrospective study was conducted with consecutive adult patients with Fuchs-endothelial corneal dystrophy who underwent DMEK in 2015-2019 and were followed for 12 months. Patients underwent concomitant cataract surgery (triple-DMEK) or had previously undergone cataract surgery (pseudophakic-DMEK). Multivariate analyses assessed whether: patient age/sex; graft-donor age; preoperative ECD, mean keratometry, or visual acuity; triple DMEK; surgery duration; surgical difficulties; and need for rebubbling predicted 6- or 12-month ECD in the whole cohort or in subgroups with high/low ECD at 6 or 12 months. The subgroups were generated with the clinically relevant threshold of 1000 cells/mm2. Surgeries were defined as difficult if any part was not standard. In total, 103 eyes (95 patients; average age, 71 years; 62% women) were included. Eighteen eyes involved difficult surgery (14 difficult graft preparation or unfolding cases and four others). Regardless of how the study group was defined, the only pre/perioperative variable that associated significantly with 6- and 12-month ECD was difficult surgery (p = 0.01, 0.02, 0.05, and 0.0009). Difficult surgery also associated with longer surgery duration (p = 0.002). Difficult-surgery subgroup analysis showed that difficult graft dissection associated with lower postoperative ECD (p = 0.03). This association may reflect endothelial cell loss due to excessive graft handling and/or an intrinsic unhealthiness of the endothelial cells in the graft that conferred unwanted physical properties onto the graft that complicated its preparation/unfolding.

Introduction

In 1998–2002, endothelial keratoplasty became a feasible alternative to penetrating keratoplasty (PKP) for corneal endothelial disorders such as Fuchs endothelial corneal dystrophy (FECD) and moderate bullous keratopathy (BK). At this timepoint, Melles introduced deep lamellar endothelial keratoplasty (DLEK). This procedure involves (i) dissecting off a posterior lamellar disc from the diseased recipient cornea; (ii) inserting a folded donor disc consisting of posterior stroma, Descemet membrane, and endothelium via a self-healing tunnel incision; and (iii) unfolding the disc and appending it to the recipient cornea with an air bubble [1-3]. This method was then further refined by Melles and others [4-10], thus generating first Descemet stripping endothelial keratoplasty (DSEK), then Descemet stripping automated endothelial keratoplasty (DSAEK), and most recently, Descemet membrane endothelial keratoplasty (DMEK), where the donor disc consists only of Descemet membrane and endothelium [10, 11]. Multiple studies then showed that compared to PKP, endothelial keratoplasty leads to more predictable refractive outcomes, increased tectonic stability, faster postoperative visual rehabilitation, and overall better and faster visual recovery. This reflects in part the absence of graft sutures, which can induce ocular surface complications and high or irregular astigmatism [11-15].

A common concern regarding all keratoplasty techniques is that they induce a steep short-term decrease in endothelial cell density (ECD) that eventually slows to low rates [16-20]. For example, in DMEK, 30–40% of endothelial cells are lost in the first 12 months, after which there is a stable annual loss of 7% [21]. This is significant because the Cornea Donor Study and its secondary analyses showed that endothelial cell loss (ECL) predicts long-term graft failure after PKP [22-25]. This reflects the fact that loss of endothelial cells can compromise the ability of the cornea to control stromal hydration and therefore maintain corneal transparency [26]. A meta-analysis of 10 studies suggests that PKP associates with more ECL than DSAEK [27] while other meta-analyses suggest that DSAEK and DMEK are similar in terms of ECL [28-31]. This further supports the transition from PKP to endothelial keratoplasty, particularly to DMEK, since it associates with significantly better postoperative best spectacle-corrected visual acuity (BSCVA) than DSAEK. This superiority may reflect the fact that DMEK grafts lack the posterior stroma that is found in DSAEK grafts [28-31], which could generate irregularities and incongruent stromal collagen fibers in the interface between the host stroma and donor [32]. Moreover, DMEK yields fewer high order aberrations [33-35] and may even be safe for regrafting after DSAEK has yielded poor visual results [36, 37]. In addition, studies show that DMEK associates with promising medium-term (8 years) endothelial survival [38] and faster visual rehabilitation than other keratoplasty techniques [39-42].

With the aim of further improving DMEK outcomes, several studies have comprehensively sought to identify the preoperative and perioperative factors that associate with low ECD after DMEK (S1 Table) [20, 43–48]. However, there are significant variations between these studies in terms of the predictive factors that have been found: for example, older patient age associates with worse ECL in two comprehensive studies [20, 45] but not in five others [43, 44, 46–48]. To further clarify this issue, we conducted a retrospective cohort study to determine which of a large range of preoperative and perioperative factors can predict ECD at 6 and 12 postoperative months.

Materials and methods

Study design and ethics

This retrospective single-center cohort study was performed in the Ophthalmology Department of Metz-Thionville Regional Hospital Center, Grand Est, France and was approved by the Ethics Committee of the French Society of Ophthalmology (IRB 00008855). All procedures conformed to the principles of the Declaration of Helsinki. All patients were informed before surgery that their surgery-related data might be used for research purposes. All consented to this possibility. The consent procedure was conducted in accordance with the reference methodology MR-004 of the National Commission for Information Technology and Liberties of France (No. 588909 v1).

Patient selection

The study cohort consisted of all consecutive adult (≥18 years) patients with FECD who underwent DMEK between October 2015 and October 2019 and who were followed up for at least 12 months. Patients either underwent concomitant cataract surgery (defined as triple-DMEK) or had previously undergone cataract surgery (defined as pseudophakic-DMEK). The surgery was usually deemed necessary when the BSCVA reached 0.3 logMAR. Exclusion criteria were prior eye surgery other than cataract surgery, indications other than FECD, the operative eye had important retinal or optic nerve diseases or amblyopia, and intraoperative complications.

Preoperative and postoperative testing

Before and 6 and 12 months after surgery, all patients underwent a macular OCT Scan (NIDEK, Tokyo, Japan) to detect retinal anomalies, specular microscopy (NIDEK CEM-530; NIDEK, Tokyo, Japan) to measure ECD, and measurement of BSCVA (in Monoyer scale, followed by conversion into logMAR).

Graft preparation, surgical techniques, postoperative treatment, and follow-up

When surgery was planned with the patient in consultation, inferior iridotomy with a Nd:YAG Laser (Laser ex-Super Q; Ellex Europe, Medical Quantel, Cournon-d’Auvergne, France) was conducted to prevent pupillary block during and after DMEK surgery.

All surgeries were performed by an experienced surgeon (JMP). General anesthesia was used for all but two patients, who had contraindications and therefore underwent surgery under a peribulbar block (a 50:50 mixture of ropivacaine 7.5 mg/mL and lidocaine 2% delivered with a 23-gauge Atkinson canula).

All grafts were issued by either of two French regional tissue banks (Nancy or Besançon). They were stored in organ culture medium (Eurobio) at 31°C until transplantation and had the requested ECD (>2400 cells/mm2), which was measured by the eye bank. To prepare the grafts for DMEK, they were first trephined with a Hanna trephine (Busin Punch 17200D 8mm single use; Moria SA, Antony, France) and the Descemet membrane-endothelium complex was dissected off with a disposable curved nontoothed forceps (Single Use Tying Forceps Curved 5mm Platform 17501; Moria SA, Antony, France) under a microscope. The resulting roll was stained with Trypan Blue (VisionBlue, 0.5-ml syringe; D.O.R.C. Dutch Ophthalmic Research Center, Zuidland, Netherlands), marked with a "F" on its stromal side, and inserted into a D.O.R.C injectable system (30G Curved cannula for air injection; D.O.R.C. Dutch Ophthalmic Research Center, Zuidland, Netherlands).

Paracentesis was performed using a 2.2-mm blade (Securityblade BD, Xstar 2.2-mm, 45 degrees, 37822; Beaver-Visitec International, Waltham, USA) and a Worst 15 blade (ophthalmic knife 15 degrees; ALCON, Ruel Malmaison, France). A 9-mm central descemetorhexis was then performed under sterile air with an inverted Sinskey Price hook (Single Use Price Reverse Hook Sim 17302; Moria SA, Antony, France) and an inverted spatula (90th single use Spatula 17303; Moria SA, Antony, France). The main incision was enlarged to 4 mm and the graft was inserted into the anterior chamber via the D.O.R.C injectable system. Two 27-gauge Rycroft cannulas were manipulated to center the graft and orient it such that the rolls of the scroll faced upward. To keep the graft in place, a sterile air bubble was injected underneath the graft. If the epithelium seemed even slightly pathological, it was removed. A 10–0 Nylon suture was placed at the main incision.

If cataract surgery was indicated, it was conducted before DMEK using the standard supracapsular "garde à vous" technique [49] with implantation of a Zeiss CT Asphina 409MV intraocular lens. The refractive target in triple DMEK was a residual myopia of about -0.50 to -1.00 D to compensate for the expected hyperopic shift induced by DMEK surgery [50, 51].

The postoperative regimen consisted of drops composed of 0.1% dexamethasone, neomycin, and polymyxin B (Maxidrol; ALCON, Rueil Malmaison, France) four times daily for 4 weeks. It was then gradually tapered and followed with a 0.1% fluorometholone-tapering (Flucon; ALCON, Rueil Malmaison, France) regimen for at least 12 months. Patients who underwent triple DMEK also received non-steroidal anti-inflammatory drops (indomethacin) (Indocollyre; Chauvin, Montpellier, France) four times daily for 4 weeks.

The patients were hospitalized for 3 days in the Ophthalmology Unit and monitored by visits 1, 7, and 15 days and 1, 3, 6, and 12 months after surgery.

Patients who developed cystoid macular edema and loss of visual acuity were treated with oral acetazolamide (Diamox®; Sanofi, Gentilly, France; one 250 mg tablet 3 times/day for one month) and NSAID (indomethacin 0.1%; Chauvin, Montpellier, France; 4 times/day for one month).

Patients with allograft rejection (defined as a line of retro-descemetic precipitates) were treated for one month with combined treatment composed of an anti-inflammatory corticosteroid ointment (Sterdex®, which contains dexamethasone with oxytetracycline; Thea, France; Clermont-Ferrand, 2 times/day) and an anti-inflammatory corticosteroid eyedrop (Maxidrol; ALCON, Rueil Malmaison, France; 12 times/day for one week then 8, 6, and 4 times/day for the second, third, and fourth week, respectively).

Rebubbling was performed immediately if anterior segment OCT (NIDEK; Tokyo Japan) showed that more than a third of the graft had detached or the graft detachment was threatening the visual axis [52]. Rebubbling was conducted under topical anesthesia (0.4% oxybuprocaïne hydrochloride; Thea, Clermont-Ferrand, France) using sterile air and an operating microscope. If a fourth rebubbling session was needed, the graft was considered to have failed and the patient was scheduled for repeat keratoplasty.

Preoperative, perioperative, and postoperative variables

The following data were obtained from the prospectively maintained clinical database: patient age and sex, donor age, graft ECD before and 6 and 12 months after surgery, BSCVA before and 6 and 12 months after surgery, preoperative mean anterior keratometry, triple DMEK, difficult surgery, surgical time, and number of rebubbling sessions. DMEK surgery was labelled as "difficult" if any part of it was not standard. Examples include difficult graft dissection, graft defects, abnormal graft elasticity, difficulties unfolding the scroll in the anterior chamber, or any condition hindering the normal course of surgery or requiring more maneuvers than usual. Surgical time was defined as the time taken from starting graft preparation to placing the suture.

Statistical analysis

All data were expressed as mean and standard deviation (range), as medians (IQR: 10; 25; 75; 90), or n (%). The relationships between 6- and 12-month ECD in the whole cohort and patient, graft, and operative variables (including difficult surgery) were determined with simple and multiple linear regression models. All 10 variables examined in the univariate analyses were considered in the multivariate analyses because the sample size was sufficient for this number of factors. The 6- and 12-month cohorts were also each divided into two subgroups depending on whether their postoperative ECD was <1000 or ≥1000 cells/mm2. This threshold was selected because several ophthalmology societies consider it to be a risk threshold of corneal decompensation [53, 54]. The high/low postoperative ECD subgroups were then compared in terms of preoperative/perioperative variables by Mann-Whitney U test (quantitative variables) or Fisher’s exact test (qualitative variables) and then with logistic regression. Subgroup analysis was also conducted to determine whether eyes with different types of surgical difficulties differed from eyes with non-difficult surgery in terms of postoperative ECD: here, Kruskal-Wallis test followed by Mann-Whitney U test was used. Student’s t-test was used to evaluate whether rebubbling associated with mean anterior keratometry. All statistical analyses were conducted using the Statview 4.5 statistical package (Abacus Concepts, Inc., New Jersey, USA). P values <0.05 were considered significant.

Results

In total, 141 eyes underwent primary or triple DMEK during the study period. Of these, 38 were excluded because the indication was not FECD (n = 19) or there was a history of an ocular intervention other than cataract surgery (n = 11), another corneal, retinal, or optic nerve disease (n = 4), or intraoperative complications (n = 4; two extreme brunescent cataracts that led to high effective phaco time, and two eyes that were converted to DSAEK due to extensive progression of stromal edema between the last visit and the surgery). Thus, 103 eyes of 95 patients were included in the study (Fig 1).

Fig 1

Flow chart showing patient disposition in the study.

Patient and eye characteristics

Average patient age was 71 years and 62% were women. Preoperative mean anterior keratometry was 44 D. Average graft-donor age was 76 years. Triple DMEK was conducted in 40 eyes and pseudophakic DMEK in 63. Average surgery duration was 35 min and 18 (17%) surgeries were deemed difficult. The average durations of the difficult and not difficult surgeries were 40.3±10.1 and 33.2±8.2 min, respectively (p = 0.002). The causes of difficult surgery were difficult graft dissection (n = 7), difficult unfolding of the graft in the anterior chamber (n = 7), anterior segment hazards (iridocorneal synechiae and two cases that required premature postoperative YAG treatment) (n = 2), air bubble under the iris during surgery (n = 1), and a patient on peribulbar block moving during surgery (n = 1). Average graft ECDs before and 6 and 12 months after surgery were 2557, 1403 (45% ECL), and 1244 (51% ECL) cells/mm2, respectively. Preoperative BSCVA was 0.64 logMAR on average, and surgery yielded excellent visual outcomes: average BSCVAs at 6 and 12 months were 0.09 and 0.05 logMAR, respectively. Rebubbling was conducted in 34% of cases. Most only required a single rebubbling session (25% of all eyes). More than one rebubbling session was needed in 9% of all eyes. None of the eyes required more than three rebubbling sessions (Table 1). Cystoid macular edema and graft rejection were observed in three and one patients, respectively. In total, 10% of the grafts failed at 12 months and the patients underwent repeat keratoplasty.

Table 1

Demographic and clinical characteristics of the patients and eyes undergoing Descemet membrane endothelial keratoplasty.

Variable	n (%) or mean ± standard deviation
No. of eyes (patients)	103 (95 patients)
Patient sex, F/M	64/39
Age, years
Patient age at surgery	70.5 ± 9.0 (48.0–90.0)
Graft donor age	75.5 ± 11.0 (30.0–99.0)
Mean anterior keratometry, D	43.93 ± 1.54 (39.87–48.61)
ECD, cells/mm2
Preoperative (graft)	2557 ± 200 (2000–2950)
6 postoperative months	1403 ± 429 (535–2500)
12 postoperative months	1244 ± 416 (509–2400)
BSCVA, logMAR
Preoperative	0.64 ± 0.29 (0.30–2.00)
6 postoperative months	0.09 ± 0.12 (-0.10–0.50)
12 postoperative months	0.05 ± 0.10 (-0.20–0.40)
Surgery and postoperative complications
Triple procedure	40 (39%)
Difficult surgery	18 (17%)
Surgery time, min	34.5 ± 8.9 (20.0–60.0)
Rebubbling
None	68 (66%)
One	26 (25%)
More than one	9 (9%)

BSCVA, best spectacle-corrected visual acuity; ECD, endothelial corneal density; F, female; M, male.

Relationships between 6- and 12-month endothelial cell density in the whole cohort and preoperative and perioperative variables

Simple and multiple linear regression analyses with the whole cohort showed that 6-month ECD associated strongly with difficult surgery (p = 0.005 and 0.01, respectively) (Table 2). This association was also present at 12 months (p = 0.014 and 0.02, respectively) (Table 3). We then categorized the difficult surgeries as graft dissection difficulties (n = 7), graft unfolding difficulties (n = 7), and other difficulties (n = 4), and compared them to the 85 DMEK eyes that did not involve surgical difficulties. Kruskal-Wallis test showed that the four groups differed significantly in terms of 12-month ECD (p = 0.03). Mann-Whitney U tests then showed that graft dissection difficulties associated with significantly lower median 12-month ECD than non-difficult DMEK (945 vs. 1331 cells/mm2; p = 0.01). Graft unfolding difficulties also tended to associate with lower 12-month ECD (800 vs. 1331 cells/mm2; p = 0.07). The other complications (anterior segment hazards, air bubble below iris, and patient movements) did not affect 12-month ECD (1500 vs. 1331 cells/mm2) (Fig 2). Kruskal-Wallis test also showed that the non-difficult and three difficult subgroups tended to differ in median 6-month ECD (1470 vs. 1009 vs. 1000 vs. 1025 cells/mm2) but these differences did not achieve statistical significance (p = 0.06) (Fig 2).

Table 2

Simple and multiple linear regression analysis of the relationship between endothelial cell density at 6 postoperative months and preoperative patient or graft or operative variables.

	Simple regression							Multiple regression
Variable	B	SE	t	p-value	R2	95% CI		B	SE	t	p-value	95% CI
						Lower	Upper					Lower	Upper
Patient age, years	1.93	4.77	0.40	0.69	0.002	-7.53	11.39	-0.38	5.06	-0.10	0.94	-10.48	9.52
Patient sex	16.62	87.60	0.19	0.85	365.4E-6	-157.15	190.39	22.83	86.99	0.30	0.79	-145.60	197.46
Donor age, years	3.68	3.85	0.95	0.34	0.009	-3.97	11.32	5.36	3.86	1.42	0.17	-2.19	13.07
Mean anterior keratometry, D	14.53	27.64	0.53	0.60	0.003	-40.30	69.36	11.87	28.63	0.42	0.68	-44.50	68.65
Preoperative ECD, cells/mm2	0.39	0.21	1.86	0.07	0.033	-0.03	0.81	0.27	0.23	1.18	0.25	-0.18	0.71
Preoperative BSCVA, logMAR	168.72	148.10	1.14	0.26	0.013	-125.06	462.51	77.24	161.34	0.51	0.63	-235.30	398.49
Triple procedure	-111.97	86.48	-1.30	0.20	0.007	-283.52	59.59	-105.38	100.81	-1.10	0.30	-308.00	89.06
Difficult surgery	-309.31	107.59	-2.88	0.005	0.076	-522.73	-95.89	-317.90	121.47	-2.66	0.010	-557.36	-80.66
Surgery time, min	-4.80	4.77	-1.01	0.32	0.010	-14.26	4.66	3.72	6.02	0.66	0.54	-7.79	15.58
Any rebubbling	-66.51	65.28	-1.02	0.31	0.010	-196.01	62.99	-69.15	67.94	-1.02	0.31	-204.10	65.79

All data are presented as slope unstandardized coefficient (B) and standard error (SE) of the regression. Statistically significant relationships between ECD and preoperative/perioperative variables (p<0.05) are bolded. The R2 for the multiple regression model was 0.14, p = 0.15 (ANOVA).

BSCVA, best spectacle-corrected visual acuity; ECD, endothelial cell density.

Table 3

Simple and multiple linear regression analysis of the relationship between endothelial cell density at 12 postoperative months and preoperative patient or graft or operative variables.

	Simple regression							Multiple regression
Variable	B	SE	t	p-value	R2	95% CI		B	SE	t	p-value	95% CI
						Lower	Upper					Lower	Upper
Patient age, years	0.91	4.78	0.19	0.85	3.60E-04	-8.57	10.39	-1.09	4.97	-0.02	0.83	-10.95	8.78
Patient sex	27.36	85.63	0.32	0.75	0.001	-142.52	197.24	35.53	83.43	0.04	0.67	-130.18	201.24
Donor age, years	5.63	3.72	1.51	0.13	0.022	-1.76	13.01	7.05	3.69	0.19	0.06	-0.28	14.37
Mean anterior keratometry, D	35.42	26.74	1.33	0.19	0.017	-17.63	88.48	25.96	27.38	0.10	0.35	-28.43	80.35
Preoperative ECD, cells/mm2	0.27	0.21	1.30	0.20	0.016	-0.14	0.68	0.15	0.22	0.07	0.49	-0.28	0.58
Preoperative BSCVA, logMAR	141.75	144.39	0.98	0.33	0.01	-144.72	428.21	93.16	153.04	0.06	0.54	-210.84	397.17
Triple procedure	-151.89	83.47	-1.82	0.07	0.032	-317.49	13.70	-128.83	96.04	-0.15	0.18	-319.59	61.93
Difficult surgery	-270.34	107.80	-2.51	0.014	0.059	-484.22	-56.46	-279.74	118.13	-0.25	0.02	-514.39	-45.09
Surgery time, min	-7.46	4.60	-1.62	0.11	0.026	-16.58	1.66	1.14	5.69	0.03	0.84	-10.17	12.45
Any rebubbling	-97.30	63.14	-1.54	0.13	0.023	-222.57	27.98	-79.61	65.79	-0.13	0.23	-210.29	51.07

All data are presented as slope unstandadized coefficient (B) and standard error (SE) of the regression. Statistically significant relationships between ECD and preoperative/perioperative variables (p<0.05) are bolded. The R2 for the multiple regression model was 0.16; p = 0.08 (ANOVA).

BSCVA, best spectacle-corrected visual acuity; ECD, endothelial cell density.

Fig 2

Comparison of surgically difficult and uncomplicated DMEK cases in terms of ECD at 6 and 12 months.

The surgically difficult cases were divided according to the type of difficulty (difficult graft preparation, unfolding, or other). The data are presented as box-whisker plots showing the median, first and third quartiles, and maximum and minimum values. Kruskal-Wallis analyses showed that the four groups differed significantly in terms of 12-month ECD (p = 0.03). Mann-Whitney U-tests showed that 12-month ECD was significantly lower in the difficult graft preparation cases (p = 0.01) and tended to be lower in the difficult graft unfolding cases (p = 0.07) compared to the group with no surgery difficulties.

In relation to the other variables, preoperative ECD tended to associate with postoperative ECD at 6 months on univariate analysis (p = 0.07) but not on multivariate analysis (p = 0.25) and not at 12 months (p = 0.20 and 0.50, respectively). Other variables did not associate with postoperative ECD, including rebubbling (Tables 2 and 3).

Comparison of patients with low and high postoperative ECDs in terms of preoperative and perioperative variables

We then divided both the 6- and 12-month cohorts into two subgroups by applying a postoperative ECD cut-off of 1000 cells/mm2. This value was selected because several ophthalmology societies consider it to be a risk threshold of corneal decompensation [53, 54]. Compared to patients with a high 6-month ECD (≥1000 cells/mm2), patients with a low 6-month ECD (<1000 cells/mm2) tended to associate with a higher surgical difficulty rate (32% vs. 14%, p = 0.06 and 0.05 on univariate and multivariate analysis, respectively) (Table 4). This trend became significant at 12 postoperative months (37% vs. 8%, p = 0.001 and 0.0009, respectively) (Table 5).

Table 4

Comparison of patients with 6-month ECD <1000 and ≥1000 cells/mm2 in terms of preoperative and perioperative variables.

	6-month ECD <1000 cells/mm2	6-month ECD ≥1000 cells/mm2	Univariate*	Multivariate**
Variables	n = 22	n = 81	p-value	OR (95% IC)	p-value	R
Age, years	71.1 ± 0.5	70.4 ± 9.1	0.83	0.97 (0.91–1.04)	0.39	0.000
Female sex	16 (73%)	48 (59%)	0,32	0.54 (0.17–1.66)	0.28	0.000
Donor age, years	73.9 ± 9.3	76.0 ± 11.5	0.18	1.03 (0.98–1.08)	0.24	0.000
Mean ant. keratometry, D	43.85 ± 1.53	43.96 ± 1.55	0.91	1.00 (1.00–1.01)	0.85	0.000
Preop. ECD, cells/mm2	2487 ± 139	2576 ± 210	0.03	1.43 (0.17–12.42)	0.17	0.000
Preop. BCVA, logMAR	0.61 ± 0.18	0.65 ± 0.31	0.98	0.96 (0.65–1.43)	0.74	0.000
Triple procedure	12 (55%)	28 (35%)	0.14	0.55 (0.25–1.21)	0.16	-0.004
Difficult surgery	7 (32%)	11 (14%)	0.06	0.41 (0.12–1.41)	0.05	-0.128
Surgery time, min	36.1 ± 7.2	34.0 ± 9.3	0.11	0.26 (0.07–1.02)	0.51	0.000
Rebubbling			0.35	1.03 (0.95–1.10)	0.14	-0.043
None	12 (55%)	56 (69%
One	7 (32%)	19 (23%)
More than one	3 (14%)	6 (7%)

The data are presented as mean ± standard deviation or n (%). For logistic regression, data are presented as odds ratios (OR), 95% confidence interval (CI), and p-value.

*The two postoperative ECD subgroups were compared in terms of preoperative/operative variables by using Mann-Whitney U test (quantitative variables) or Fisher’s exact test (qualitative variables).

**Logistic regression model. The R2 for the multiple regression model was 0.13.

Statistically significant differences between the subgroups are bolded.

Ant., anterior; BSCVA, best spectacle-corrected visual acuity; ECD, endothelial corneal density; preop., preoperative.

Table 5

Comparison of patients with 12-month ECD <1000 and ≥1000 Cells/mm2 in terms of preoperative and perioperative variables.

	12-month ECD <1000 cells/mm2	12-month ECD ≥1000 cells/mm2	Univariate*	Multivariate**
Variables	n = 30	n = 72	p-value	OR (95% IC)	p-value	R
Age, years	71.6 ± 7.8	70.1 ± 9.4	0.49	0.96 (0.90–1.02)	0.19	0.000
Female sex	20 (67%)	44 (61%)	0.66	0.82 (0.29–2.29)	0.70	0.000
Donor age, years	74.5 ± 9.2	76.0 ± 11.7	0.22	1.02 (0.98–1.07)	0.32	0.000
Mean ant. keratometry, D	43.68 ± 1.47	44.04 ± 1.57	0.49	1.08 (0.75–1.55)	0.69	0.000
Preop. ECD, cells/mm2	2528 ± 158	2569 ± 214	0.25	1.00 (1.00–1.00)	0.98	0.000
Preop. BSCVA, logMAR	0.60 ± 1.18	0.66 ± 0.32	0.76	2.14 (0.28–16.30)	0.46	0.000
Triple procedure	16 (53%)	24 (33%)	0.08	0.35 (0.11–1.14)	0.08	-0.093
Difficult surgery	11 (37%)	6 (8%)	0.001	0.09 (0.02–0.38)	0.0009	-0.271
Surgery time, min	36.5 ± 7.09	33.6 ± 9.47	0.03	1.02 (0.95–1.10)	0.53	0.000
Rebubbling			0.16	0.50 (0.23–1.10)	0.08	-0.089
None	16 (53%)	52 (72%)
One	10 (33%)	15 (21%)
More than one	4 (13%)	5 (7%)

The data are presented as mean ± standard deviation or n (%).

*The two postoperative ECD subgroups were compared in terms of preoperative/operative variables by using Mann-Whitney U test (quantitative variables) or Fisher’s exact test (qualitative variables).

**Logistic regression model. The R2 for the multiple regression model was 0.18.

Statistically significant differences between the subgroups are bolded.

Ant., anterior; BSCVA, best spectacle-corrected visual acuity; ECD, endothelial corneal density, Preop., preoperative.

Patients with a lower 6-month postoperative ECD also had significantly fewer endothelial cells in the graft before surgery on univariate analysis (2487 vs. 2576 cells/mm2, p = 0.03) but this was not observed on multivariate analysis (p = 0.15) (Table 4) or at 12 postoperative months (2528 vs. 2569 cells/mm2, p = 0.25 and 0.98 on univariate and multivariate analysis, respectively) (Table 5). Patients with a lower 12-month postoperative ECD also underwent significantly longer surgery on univariate analysis (36.5 vs. 33.6 min, p = 0.03) but this was not observed on multivariate analysis (p = 0.53) (Table 5) or at 6 postoperative months (36.1 vs. 34.0 min, p = 0.11 and 0.51 on univariate and multivariate analysis, respectively) (Table 4).

Other variables did not associate with postoperative ECD in this analysis, including rebubbling (Tables 4 and 5).

Relationship between mean anterior keratometry and rebubbling

In a side analysis, we observed that when we divided the patients according to whether they underwent no rebubbling or any rebubbling, rebubbling associated with significantly smaller mean anterior keratometry values (i.e. flatter cornea) compared to no rebubbling (43.44±1.41 vs. 44.19±1.55 D; p = 0.02).

Discussion

The present cohort study searched for pre/perioperative factors that shape the postoperative ECD after DMEK. We observed with two statistical methods that difficult surgery associated significantly with lower postoperative ECD, particularly at 12 months. DMEK surgery was deemed difficult if there was any condition that hindered or prolonged the normal surgical course. Univariate analyses also showed that low preoperative graft ECD and longer surgery associated with low preoperative ECD at 6 and 12 months, respectively. However, these relationships were not observed at the other postoperative timepoint or on multivariate analysis. All other factors, namely, patient age and sex, donor age, preoperative mean anterior keratometry, preoperative BSCVA, triple procedure, and rebubbling, did not associate with postoperative ECD.

To date, seven retrospective cohort studies have, like us, recently (2016–2021) used multivariate analysis to comprehensively assess the ability of a range of pre/perioperative variables to predict postoperative ECD/ECL [20, 43–48] (S1 Table). The findings of these studies, and other studies that are less comprehensive and/or focused instead on the relationship between pre/perioperative variables and rebubbling/graft detachment, are as follows. It should be noted that the results of these studies often agree poorly: some find certain relationships while others do not. This may reflect, at least in part, differences between the study populations in terms of surgical practice, variable measurement, postoperative timepoint, and study/statistical design. It is also possible that significantly-associated pre/perioperative variables relate only indirectly to the fundamental endothelial cell variable(s) that ultimately shape(s) the postoperative density and function of these cells [55, 56].

Association between surgical difficulty and ECL

In our study, surgical difficulty associated consistently with lower ECD at two postoperative timepoints. The main causes of surgical difficulty were problematic graft dissection and unfolding, which respectively associated significantly (p = 0.01) and tended to associate (p = 0.07) with lower ECD at 12 months. We speculate that surgical difficulty reduced postoperative ECD after DMEK either because the greater graft handling directly damaged the endothelial cells and/or because the intrinsic unhealthiness of the endothelial cells in the graft conferred unwanted physical properties onto the graft that complicated its preparation and/or placement against the recipient cornea. This notion is supported by several other studies. First, two of the seven comprehensive studies examined the influence of difficult surgery on ECL [44, 45]. One, which was on a cohort of 351 eyes, found that intraoperative complications such as unfolding difficulties associated significantly with higher ECL at 48 months (Estimate 6.50, 95% confidence intervals 0.59–12.41, p = 0.032) [45]. However, the second comprehensive study, which was on 2485 eyes, did not find that intraoperative complications predicted ECL [44]. Second, Schrittenlocher et al. found that ECL depended on the learning curve, being higher in the early period [57]. Moreover, a retrospective study on 169 eyes showed that severe unfolding difficulties associated with significantly higher ECL at 3, 6, and 12 months compared to eyes with no or milder unfolding problems [58]. In addition, a retrospective study on 28 eyes observed a positive correlation between ECL and DMEK graft unfolding time [59]. A similar study on 93 eyes did not observe this association but this may reflect the fact that all surgeries were uncomplicated [60].

Third, others have noted that surgical difficulties, including poor graft decentration [61, 62], anterior band layer fragments [63], and incomplete removal of host Descemet membrane [64], associate with increased graft detachment. Moreover, a study using prospective transplant registry data on 752 DMEK-treated eyes showed with multivariate analysis that surgical difficulty was a risk factor for both rebubbling (Odds Ratio = 2.28, 95% confidence intervals = 1.20–4.33, p = 0.012) and early graft failure (Odds Ratio = 2.93, 95% confidence intervals = 1.42–6.04, p = 0.012) [65]. These observations are notable because graft detachment (or rebubbling, a surrogate measure of graft detachment) associated repeatedly with ECL in three of the five comprehensive studies that have studied this variable [20, 45, 46] (S1 Table). Other studies have also observed this relationship in DMEK [38, 65–68]. However, we did not observe that rebubbling predicted ECL in our study. This may reflect in part the liberal use of rebubbling in our center. Other studies have also failed to detect a relationship between rebubbling and ECL [47, 48, 69].

Association of patient-related variables with ECL

Patient age was not a predictor of ECL in our study or in five comprehensive studies [43, 44, 46–48] but two comprehensive studies [20, 45], a prospective registry study of 752 eyes [65], and a retrospective study of 66 eyes [70] observed that older patient age associated with more ECL. Dunker et al. proposed that older patients may be more prone to ECL because they find it more difficult to remain lying in the supine position after surgery; this is supported by the fact that graft detachments most often develop inferiorly [65].

We found that patient sex did not associate with ECL. This was also observed by all seven comprehensive studies [20, 43–48].

Our study focused on FECD cases but two comprehensive studies [20, 45] and other studies on cohorts with mixed indications found that severe FECD, or BK rather than FECD, associate with higher ECL [38, 71]. While three other comprehensive studies did not observe this association [43, 44, 48], the aqueous humor of eyes with BK and advanced FECD have high concentrations of proinflammatory cytokines; it has been proposed that this inflammation may damage the endothelial cells in the graft [17].

We observed that patients with low 6-month ECD had significantly lower preoperative ECD, although this was only noted on univariate analysis at 6 months. Nonetheless, two other comprehensive studies have found on multivariate analysis that lower preoperative graft ECD associates with higher ECL [45, 47]. However, this was not observed in a third comprehensive study [48] or a retrospective analysis of 857 eyes [72]. Preoperative ECD may have only a limited impact on ECL if it exceeds a certain threshold [26, 73] (2400 cells/mm2 in our case).

Preoperative visual acuity did not associate with ECL in our study, three comprehensive studies [20, 44, 47], or a prospective study on 108 eyes [68]. It also did not associate with graft detachment in a prospective case series study on 30 eyes [63].

Association of donor age and sex with ECL

Donor age did not associate with ECL in our study and in four comprehensive studies [43, 45, 46, 48]. This was also observed elsewhere [72]. However, several other studies have suggested that older donors yield wider scrolls that are easier and faster to unroll and may lead to less ECL [59, 74]. Indeed, in our experience, grafts from young donors are generally more difficult to dissect due to high graft elasticity, thus potentially causing difficulties during surgery. It has also been observed that higher preoperative ECDs associate with wider scrolls [59]. Therefore, it is now widely accepted that older donors with high ECD are preferable for DMEK.

We did not study the relationship between donor sex and ECL but the three comprehensive studies that did look at this variable did not observe an association [43, 45, 46].

Association of surgery duration and triple DMEK/preoperative lens status with ECL

Our 12-month univariate analysis showed that longer surgery duration associated with greater ECL. While there was limited correlation between difficult surgery and surgical duration (r = 0.3, Pearson’s correlation test), this finding is compatible with our definition of difficult surgery (any condition hindering the normal course of surgery or requiring more maneuvers than usual). It is also consistent with our finding that difficult surgery associated with longer surgery duration than not-difficult surgery. We did not find other studies on the relationship between ECL and surgery duration. It should be noted that the relatively low correlation between difficult surgery and surgical duration in our study ruled out the possibility that difficult surgery and surgery time were confounding each other in the multivariate analyses.

Triple DMEK (versus DMEK-alone) did not associate with ECL in our study, two comprehensive studies [44, 48], and three other studies [71, 75, 76]. However, other studies have found that triple DMEK associates with worse ECL [77] and graft detachment [78]. The reason for this discrepancy is unclear but it may reflect differences between studies in terms of whether DMEK-alone was conducted on phakic or pseudophakic eyes: Siebelmann et al. noted that DMEK on pseudophakic eyes associated with worse rebubbling than triple DMEK and phakic DMEK [79]. Indeed, the comprehensive study of Peraza-Nieves et al. also noted that phakic DMEK associated with better ECL than pseudophakic DMEK [20], although this was not observed in three other comprehensive studies [43-45] or a retrospective analysis of 62 eyes [80]. It has been suggested that pseudophakia may associate with high levels of proinflammatory cytokines in the aqueous humor [81].

Association between mean anterior keratometry and rebubbling

Interestingly, our side analysis showed that an arched mean anterior keratometry tended to associate with less rebubbling. To our knowledge, the contribution of anterior keratometry to DMEK outcomes has not been studied previously. We speculate that this association may reflect the easier adherence of the DMEK graft (once it is unfolded in the anterior chamber) to a steep cornea rather than to a flat one.

Study limitations

The present study has several limitations. First, its retrospective design may lend itself to selection and information bias. However, the data were recorded prospectively. Second, it is a monocentric study. However, all surgeries were conducted by one experienced surgeon, which may limit confounding due to surgeon differences. Third, postoperative ECD was measured at the cornea center with specular microscopy: a recent study found that such measurements may not accurately reflect postoperative ECD throughout the entire graft [82]. This could also partly explain the inconsistencies between studies in terms of the relationships between postoperative ECD and pre/perioperative variables. Fourth, since all patients underwent surgery for FECD, our results cannot be extrapolated to DMEK for other endothelial disorders. It will be of interest to determine whether ECL after pseudophakic BK associates with similar pre/perioperative variables. Fifth, it is possible that patient variables that associate with unusual ocular anatomy (e.g. high myopia or previous vitrectomy) promote surgical difficulty and ECL. While these variables were rare or not routinely collected in our cohort, they warrant further research.

Conclusions

Overall, the refractive and endothelial outcomes of our case series were excellent, which supports the notion that DMEK is the surgery of choice for FECD [83] provided that there is no stromal scarring [84]. However, difficulties during surgery may cause lower ECDs at 6 and 12 months. This highlights the importance of meticulous surgical approaches in DMEK and close supervision during the learning curve.

Summary of the published studies that comprehensively examined a range of pre/perioperative variables for their ability to predict endothelial cell density/loss after DMEK surgery.

(DOCX)

Click here for additional data file.

25 Oct 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Hidenaga Kobashi, M.D., Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please amend your current ethics statement to address the following concerns:

a) Did participants provide their written or verbal informed consent to participate in this study?

b) If consent was verbal, please explain i) why written consent was not obtained, ii) how you documented participant consent, and iii) whether the ethics committees/IRB approved this consent procedure.

3. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

"Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

Additional Editor Comments:

The reviewers and editor have completed their assessments of your manuscript

Factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty(PONE-D-21-31908)

and would like to publish it in the journal once you have responded to the referees' comments (enclosed below). In the cover letter with the revised manuscript, please indicate how each of the reviewers' suggestions was addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Authors describe the Factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty. Although it’s an interesting study, there’re many points that should be addressed.

This may reflect endothelial cell loss due to excessive graft handling and/or an intrinsic unhealthiness of the endothelial cells in the graft that conferred unwanted physical properties onto the graft that complicated its preparation/unfolding.

1) Authors should discuss after dividing into the factors of graft, and those of patients, which means DMEK for eyes after vitrectomy, or with high myopia, or more. The method that authors defined include many confounding factors. At least, authors should be combine or discuss these two factors of difficult graft preparation or unfolding simultaneously. (Line 37, 182)

2) Please let readers know the definition of the surgical time. Do you mean the time from the starting of graft preparation by surgeons, or the patient’s procedure? Does it include the graft preparation time? If so, your surgical time should be diminished by the duration time of graft preparation. Do you use pre-stripped tissues by eyebanks, or non-stripped tissues?

3) Regarding regression analysis, there is no description of R value.

4) The evaluating factors of simple, and multiple regression analysis seem to be same. How did you decide these evaluating factors with evidence? Authors should consult the statistician.

5) Difficult surgery might be the confounding factor of surgery time. Please consider getting rid of confounding factors from them.

6) “We then divided both the 6- and 12-month cohorts into two subgroups by applying a postoperative ECD cut-off of 1000 cells/mm2. This value was selected because it is considered a risk threshold of chronic corneal edema [Line 271-172]” Eyes with low ECD about 1000 have no corneal edema. Please let me know the reason why you decide the ECD of 1000 was the cut-off point clearly.

7) Please consider showing the dataset.

Reviewer #2: In this study, the authors evaluated the risk factors of endothelial cell loss in Fuchs endothelial corneal dystrophy patients post DMEK. The statistical analysis was well performed, and the authors' findings provide important data on the risk factors of endothelial cell loss post DMEK.

Minor Comments:

1. Please enlist the services of a professional English editing service, as numerous sentences in the manuscript are not suitable for use in an academic research paper.

2. Title: The words "Factors that predict …" need to be revised, as the wording is not fully suitable.

3. Line 30: The authors stated "pseudophakic-DMEK or triple-DMEK". Although I understand what the authors wish to state, the wording needs to be revised in order to make it suitable for use in an academic research paper and so that all readers can easily understand.

4. Lines 50-56: This is a 'run-on' sentence that involves multiple improperly connected independent clauses (i.e., complete sentences). Thus, it needs to be revised into 2-3 sentences for improved readability.

5. Lines 104-107: "primary DMEK (if their affected eye was pseudophakic) or triple-DMEK (phacoemulsification followed by DMEK)". Once again, "primary DMEK and triple DMEK" needs to be revised for improved clarity.

6. Line 126: Please indicate the organ culture medium that was used and the storage protocol/status, including the temperature.

7. Table 1: "64//39 (62)". What does "62" indicate?

8. Table 4: "Sex 16 (73%)". Please indicate what "73%" is.

9. Line 315: The statement "difficult surgery" is too vague and needs to be explained in greater detail in order to better clarify the authors' intended meaning.

10. The Discussion section is divided into too many subheadings. For example, Lines 417-420 include only 2 sentences. Thus, please reduce the number of subheadings in the Discussion section for improved readability.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: renamed_86e1b.docx

Click here for additional data file.

26 Jan 2022

Point-by-Point Responses to the reviewers

We are very grateful for all reviewers for their thorough review of our paper and their thoughtful comments. We have addressed all points to the best of our ability, as follows.

Reviewer #1

Authors describe the Factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty. Although it’s an interesting study, there’re many points that should be addressed.

This may reflect endothelial cell loss due to excessive graft handling and/or an intrinsic unhealthiness of the endothelial cells in the graft that conferred unwanted physical properties onto the graft that complicated its preparation/unfolding.

1) Authors should discuss after dividing into the factors of graft, and those of patients, which means DMEK for eyes after vitrectomy, or with high myopia, or more.

Reply: To address this comment, we looked at the patients with and without vitrectomy and high myopia, as follows.

Since there were only two eyes that had undergone previous vitrectomy, we did not conduct further analyses on this patient variable.

We compared the patients with and without high myopia (defined as axial length >25 mm) in terms of any surgical difficulty, difficult graft preparation alone, difficult graft unfolding alone, and ECD at 6 and 12 months. It should be noted that the high myopia case size was small (n=12) and the axial length data were missing in 26 of the 103 cases. As shown in the table below, univariate analysis revealed that the patients with high myopia were significantly more likely to experience difficult graft unfolding (p=0.045). However, high myopia did not associate with greater postoperative ECD loss at 6 months (p=0.56) or 12 months (p=0.84). In the literature, one study has noted that eyes with a *shorter* axial length associate with greater ECL after DMEK (Borroni et al. 2017) but two other studies did not observe an association between axial length and post-DMEK ECL (Inoda et al. 2020; Lekhanont et al. 2021). Given this variability in the literature and the sample size limitations in the univariate analysis below, we decided not to add these data to the paper. Instead, we added the following text to the Study Limitations section:

“Fifth, it is possible that patient variables that associate with unusual ocular anatomy (e.g. high myopia or previous vitrectomy) promote surgical difficulty and ECL. While these variables were rare or not routinely collected in our cohort, they warrant further research.” Pages 23–24, lines 457–459.

Variables No myopia High myopia Univariate p-value

Axial length <25 mm Axial length ≥25 mm

N = 65 N = 12

Any surgical difficulty 8 (12%) 4 (33%) 0.09

Difficult graft preparation 3 (5%) 1 (8%) 0.50

Difficult graft unfolding 3 (5%) 3 (25%) 0.045

6 postoperative months 1423 ± 406 (690-2500) 1352 ± 503 (700-2180) 0.56

12 postoperative months 1258 ± 397 (509-2400) 1261 ± 505 (688-2180) 0.84

Borroni D, Ferronato M, Krumina Z, Parekh M. Importance of Axial Length and Functional Corneal Endothelial Cells in Descemet Membrane Endothelial Keratoplasty. Cornea. 2017 Dec;36(12):e35-e36.

Inoda S, Hayashi T, Takahashi H, et al. Factors associated with endothelial cell density loss post Descemet membrane endothelial keratoplasty for bullous keratopathy in Asia. PLoS One. 2020;15(6):e0234202. Published 2020 Jun 11.

Lekhanont K, Pisitpayat P, Cheewaruangroj N, Jongkhajornpong P, Nonpassopon M, Anothaisintawee T. Outcomes of Descemet Membrane Endothelial Keratoplasty in Bangkok, Thailand. Clin Ophthalmol. 2021 May 31;15:2239-2251.

1a) The method that authors defined include many confounding factors. At least, authors should be combine or discuss these two factors of difficult graft preparation or unfolding simultaneously. (Line 37, 182)

Reply: The “difficult surgery” variable in Tables 1–5 refers to all 18 cases of difficult surgery, which included seven difficult graft dissection cases, seven difficult graft unfolding cases, and four other cases (two anterior segment hazard cases, one air bubble below the iris case, and one patient movement case). Only in Figure 2 did we examine the three main types of difficult surgery cases separately (graft dissection, graft unfolding, other).

To address this comment, we altered the Abstract as follows:

“Eighteen eyes involved difficult surgery (14 difficult graft preparation or unfolding cases and four others). Regardless of how the study group was defined, the only pre/perioperative variable that associated significantly with 6- and 12-month ECD was difficult surgery (p=0.01, 0.02, 0.05, and 0.0009). Difficult surgery also associated with longer surgery duration (p=0.002). Difficult-surgery subgroup analysis showed that difficult graft dissection associated with lower postoperative ECD (p=0.03). This association may reflect endothelial cell loss due to excessive graft handling and/or an intrinsic unhealthiness of the endothelial cells in the graft that conferred unwanted physical properties onto the graft that complicated its preparation/unfolding.” Page 3, lines 38–46.

We also changed the Statistics section to make clear that the multivariate analyses were conducted with difficult surgery (not difficult surgery subgroups) and that a separate univariate analysis was conducted with the three difficult surgery subgroups:

“The relationships between 6- and 12-month ECD in the whole cohort and patient, graft, and operative variables (including difficult surgery) were determined with simple and multiple linear regression models. The 6- and 12-month cohorts were also each divided into two subgroups depending on whether their postoperative ECD was <1000 or ≥1000 cells/mm2. This threshold was selected because several ophthalmology societies consider it to be a risk threshold of corneal decompensation [53,54]. The high/low postoperative ECD subgroups were then compared in terms of preoperative/perioperative variables by Mann-Whitney U test (quantitative variables) or Fisher’s exact test (qualitative variables) and then with logistic regression. Subgroup analysis was also conducted to determine whether eyes with different types of surgical difficulties differed from eyes with non-difficult surgery in terms of postoperative ECD: here, Kruskal-Wallis test followed by Mann-Whitney U test was used.” Pages 9–10, lines 189–201.

2) Please let readers know the definition of the surgical time. Do you mean the time from the starting of graft preparation by surgeons, or the patient’s procedure? Does it include the graft preparation time? If so, your surgical time should be diminished by the duration time of graft preparation. Do you use pre-stripped tissues by eyebanks, or non-stripped tissues?

Reply: We dissect the grafts just before the actual surgical procedure. This is indicated by the text in the Methods:

“To prepare the grafts for DMEK, they were first trephined with a Hanna trephine (Busin Punch 17200D 8mm single use; Moria SA, Antony, France) and the Descemet membrane-endothelium complex was dissected off with a disposable curved nontoothed forceps (Single Use Tying Forceps Curved 5mm Platform 17501; Moria SA, Antony, France) under a microscope...” Page 7, lines 130–134

The variable “surgical time” includes the time needed to prepare the graft. Since the term “difficult surgery” includes graft preparation difficulties, we specified both terms in the “Preoperative, perioperative, and postoperative variables” section of the Methods:

“DMEK surgery was labelled as "difficult" if any part of it was not standard. Examples include difficult graft dissection, graft defects, abnormal graft elasticity, difficulties unfolding the scroll in the anterior chamber, or any condition hindering the normal course of surgery or requiring more maneuvers than usual. Surgical time was defined as the time taken from starting graft preparation to placing the suture.” Page 9, lines 182–186.

Please note that the definition of the term “difficult surgery” was moved from the “Graft preparation, surgical techniques, postoperative treatment, and follow-up” Methods section to comply with a request of another reviewer.

3) Regarding regression analysis, there is no description of R value.

Reply: We have provided the R2 values in Tables 2–5 along with other statistics for the simple and multivariate regression analyses.

4) The evaluating factors of simple, and multiple regression analysis seem to be same. How did you decide these evaluating factors with evidence? Authors should consult the statistician.

Reply: All statistics were conducted by a hospital biostatistician, who is also an author (AE). The same 10 factors that were analyzed in the simple regression analyses were also included in the multiple regression analyses because the sample size was sufficient for this number of factors. This point was added to the Statistics section:

“The relationships between 6- and 12-month ECD in the whole cohort and patient, graft, and operative variables (including difficult surgery) were determined with simple and multiple linear regression models. All 10 variables examined in the univariate analyses were considered in the multivariate analyses because the sample size was sufficient for this number of factors.” Pages 9–10, lines 189–193.

5) Difficult surgery might be the confounding factor of surgery time. Please consider getting rid of confounding factors from them.

Reply: The correlation coefficient for difficult surgery and surgery time was 0.3 (Pearson’s test). This suggests that there is not a collinearity problem in the multiple regression analyses. Nonetheless, to check, we repeated the multiple regression analyses (i) with surgical difficulty but not surgical time and (ii) with surgical time but not surgical difficulty. The results were essentially the same:

(i) Surgical difficulty associated/tended to associate with low ECD at six months (whole cohort, p=0.011; high/low preoperative ECD subgroups, p=0.05) and especially at 12 months (p=0.02 and 0.0005) (the equivalent p values in the multiple regression analyses with both difficult surgery and surgical time in Tables 2–5 were 0.010, 0.05, 0.02, and 0.0009, respectively).

(ii) Associations between surgical time and ECD were not observed at six months (p=0.79 and 0.92) or 12 months (p=0.49 and 0.47) (the equivalent p values in the multiple regression analyses with both difficult surgery and surgical time in Tables 2–5 were 0.54, 0.51, 0.84, and 0.53, respectively).

To address this point, we added the following text to the Discussion section on surgical time:

“Our 12-month univariate analysis showed that longer surgery duration associated with greater ECL. While there was limited correlation between difficult surgery and surgical duration (r=0.3 , Pearson’s correlation test), this finding is compatible with our definition of difficult surgery (any condition hindering the normal course of surgery or requiring more maneuvers than usual). It is also consistent with our finding that difficult surgery associated with longer surgery duration than not-difficult surgery. We did not find other studies on the relationship between ECL and surgery duration. It should be noted that the relatively low correlation between difficult surgery and surgical duration in our study ruled out the possibility that difficult surgery and surgery time were confounding each other in the multivariate analyses.” Page 22, lines 418–426.

6) “We then divided both the 6- and 12-month cohorts into two subgroups by applying a postoperative ECD cut-off of 1000 cells/mm2. This value was selected because it is considered a risk threshold of chronic corneal edema [Line 271-172]” Eyes with low ECD about 1000 have no corneal edema. Please let me know the reason why you decide the ECD of 1000 was the cut-off point clearly.

Reply: We chose the 1000 cells/mm2 threshold because the American Academy of Ophthalmology indicates that corneas with ECDs below this value are at risk of corneal decompensation [New Ref 53]. The Japanese Corneal Society also considers ECDs of 500–1000 to indicate corneal endothelial damage that could endanger corneal transparency [New Ref 54]. Other studies have therefore utilized this threshold [e.g. Refs A–C below]. To address this, we have:

(i) Added the following text to the Statistics section of the Methods, removed ref 53, and replaced it with new ref 53 and ref 54:

“This threshold was selected because several ophthalmology societies consider it to be a risk threshold of corneal decompensation [53,54].” Page 10, lines 194–196.

(ii) Altered the original Results text to explain why we used this threshold, removed ref 53, and replaced it with new ref 53 and ref 54:

“We then divided both the 6- and 12-month cohorts into two subgroups by applying a postoperative ECD cut-off of 1000 cells/mm2. This value was selected because several ophthalmology societies consider it to be a risk threshold of corneal decompensation [53,54].” Page 15, 279–280.

New ref 53: Corneal endothelial photography. Three-year revision. American Academy of Ophthalmology. Ophthalmology. 1997 Aug;104(8):1360-5. PMID: 9261327

New ref 54: Kinoshita, Shigeru, et al. "Grading for corneal endothelial damage." Nippon Ganka Gakkai Zasshi (2014): 81-83.

Ref A: Li Y, Fu Z, Liu J, Li M, Zhang Y, Wu X. Corneal Endothelial Characteristics, Central Corneal Thickness, and Intraocular Pressure in a Population of Chinese Age-Related Cataract Patients. J Ophthalmol. 2017;2017:9154626. doi:10.1155/2017/9154626

Ref B: Chen HC, Huang CW, Yeh LK, Hsiao FC, Hsueh YJ, Meir YJ, Chen KJ, Cheng CM, Wu WC. Accelerated Corneal Endothelial Cell Loss after Phacoemulsification in Patients with Mildly Low Endothelial Cell Density. J Clin Med. 2021 May 24;10(11):2270. doi: 10.3390/jcm10112270.

Ref C: Hayashi K, Yoshida M, Manabe S, Hirata A. Cataract surgery in eyes with low corneal endothelial cell density. J Cataract Refract Surg. 2011 Aug;37(8):1419-25. doi: 10.1016/j.jcrs.2011.02.025. Epub 2011 Jun 17. PMID: 21684110.

7) Please consider showing the dataset.

Reply: Unfortunately, the datasets generated during the current study cannot be made publicly available according to French Law No. 2018-493 of June 20, 2018 on the protection of personal data (The General Data Protection Regulation (Regulation (EU) 2016/679) (GDPR: article 9). However, they are available from the corresponding author on reasonable request. This data processing is compliant with a baseline methodology reference methodology (MR-004) for which the our hospital signed a compliance commitment on October 8, 2018.

Reviewer #2

In this study, the authors evaluated the risk factors of endothelial cell loss in Fuchs endothelial corneal dystrophy patients post DMEK. The statistical analysis was well performed, and the authors' findings provide important data on the risk factors of endothelial cell loss post DMEK.

Minor Comments:

1. Please enlist the services of a professional English editing service, as numerous sentences in the manuscript are not suitable for use in an academic research paper.

Reply: One of us (YZ) is a native English speaker with a PhD in the medical field and 25 years of experience in medical and scientific writing and editing. Her website can be viewed on https://www.scimeditor.com/

She has made all changes requested and is grateful to the reviewer for pointing out where the language could be improved or made more precise/readable.

2. Title: The words "Factors that predict …" need to be revised, as the wording is not fully suitable.

Reply: The title has been changed to be more descriptive (underlined texts were added):

“Identification of the preoperative and perioperative factors that predict postoperative endothelial cell density after Deescement membrane endothelial keratoplasty: a retrospective cohort study” Page 1, lines 2–5.

3. Line 30: The authors stated "pseudophakic-DMEK or triple-DMEK". Although I understand what the authors wish to state, the wording needs to be revised in order to make it suitable for use in an academic research paper and so that all readers can easily understand.

Reply: It is important to distinguish between (i) DMEK that immediately follows phacoemulsification (triple-DMEK), (ii) DMEK on eyes that have previously undergone phacoemulsification (what we called pseudophakic-DMEK), and (iii) DMEK on eyes that have not undergone previous phacoemulsification and are not undergoing it during the index surgery (none of our patients fell into this category). Since the latter two surgical regimens are often insufficiently distinguished in the literature, we specified that the patients who did not undergo concomitant phacoemulsification had already previously undergone this procedure. To ensure that this is clear, the terms “triple-DMEK” and “pseudophakic-DMEK” have now been explicitly defined in both the Abstract and Methods section.

Abstract: “This retrospective study was conducted with consecutive adult patients with Fuchs endothelial corneal dystrophy who underwent DMEK in 2015–2019 and were followed for 12 months. Patients either underwent concomitant cataract surgery (triple-DMEK) or had previously undergone cataract surgery (pseudophakic-DMEK).” Page 3, lines 28–32.

Methods: “The study cohort consisted of all consecutive adult (≥18 years) patients with FECD who underwent DMEK between October 2015 and October 2019 and who were followed up for at least 12 months. Patients either underwent concomitant cataract surgery (defined as triple-DMEK) or had previously undergone cataract surgery (defined as pseudophakic-DMEK).” Page 6, lines 106–109.

4. Lines 50-56: This is a 'run-on' sentence that involves multiple improperly connected independent clauses (i.e., complete sentences). Thus, it needs to be revised into 2-3 sentences for improved readability.

Reply: The sentence has been simplified as follows:

“In 1998–2002, endothelial keratoplasty became a feasible alternative to penetrating keratoplasty (PKP) for corneal endothelial disorders such as Fuchs endothelial corneal dystrophy (FECD) and moderate bullous keratopathy (BK). At this timepoint, Melles introduced deep lamellar endothelial keratoplasty (DLEK). This procedure involves (i) dissecting off a posterior lamellar disc from the diseased recipient cornea; (ii) inserting a folded donor disc consisting of posterior stroma, Descemet membrane, and endothelium via a self-healing tunnel incision; and (iii) unfolding the disc and appending it to the recipient cornea with an air bubble [1,2,3]. This method was then further refined by Melles and others...” Page 4, lines 51–58.

6. Line 126: Please indicate the organ culture medium that was used and the storage protocol/status, including the temperature.

Reply: This information (Eurobio medium at 31°C) has been added to the Methods section. Page 7, line 129.

7. Table 1: "64//39 (62)". What does "62" indicate?

Reply: It was intended to indicate “62% of the patients are female”. We have deleted it since it is unnecessary.

8. Table 4: "Sex 16 (73%)". Please indicate what "73%" is.

Reply: It refers to “Female sex” (i.e. 73% of the patients with ECD of <1000 cells/mm2 were female). The word “Female” was added to Tables 4 and 5.

9. Line 315: The statement "difficult surgery" is too vague and needs to be explained in greater detail in order to better clarify the authors' intended meaning.

Reply: The term “difficult surgery” was defined in the Methods as:

“DMEK surgery was labelled as "difficult" if any part of it was not standard. Examples include difficult graft dissection, graft defects, abnormal graft elasticity, difficulties unfolding the scroll in the anterior chamber, or any condition hindering the normal course of surgery or requiring more maneuvers than usual.” Page 9, lines 182–185.

However, this text was originally located in the “Graft Preparation, surgical techniques, postoperative treatment, and follow-up” Methods section. To ensure that the reader can more easily find this definition, this text was moved to the Methods section that is now called “Preoperative, perioperative, and postoperative variables”.

Moreover, the following text was added to the beginning of the Discussion:

“The present cohort study searched for pre/perioperative factors that shape the postoperative ECD after DMEK. We observed with two statistical methods that difficult surgery associated significantly with lower postoperative ECD, particularly at 12 months. DMEK surgery was deemed difficult if there was any condition that hindered or prolonged the normal surgical course.” Page 18, lines 326–330.

10. The Discussion section is divided into too many subheadings. For example, Lines 417-420 include only 2 sentences. Thus, please reduce the number of subheadings in the Discussion section for improved readability.

Reply: The subheadings in the Discussion were reduced to six (from nine). Patient-related variables (age, sex, indication, and preoperative ECD and BSCVA) were combined into one section while surgery duration was added to the Triple-DMEK section. Section headings and references were altered to reflect these changes.

Submitted filename: Point-by-Point Response to the Reviewers.docx

Click here for additional data file.

10 Feb 2022

Identification of the preoperative and perioperative factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty: a retrospective cohort study

PONE-D-21-31908R1

Dear Dr. Perone,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Hidenaga Kobashi, M.D., Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Congratulations, the reviewers and academic editor agreed the publication in Plos One.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The data presented looks proper.　This paper was well revised.　I would like to expect you to perform further study.

Reviewer #2: I appreciate that authors revised the manuscript based on the comments. I believe that the manuscript was improved and it will be beneficial for the readers in this research field.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: TAKAHIKO HAYASHI

Reviewer #2: No

14 Feb 2022

PONE-D-21-31908R1

Identification of the preoperative and perioperative factors that predict postoperative endothelial cell density after Descemet membrane endothelial keratoplasty: a retrospective cohort study

Dear Dr. Perone:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Hidenaga Kobashi

Academic Editor

PLOS ONE