Purpose
To report the incidence and risk factors of elevated IOP following deep anterior lamellar keratoplasty (DALK).
Design
A retrospective case series.
Methods
A retrospective study investigating the 5-year incidence of raised IOP following DALK cases performed from 2004 to 2008 in a tertiary center. Patients with less than 6 months of follow-up were excluded. Elevated IOP was defined as IOP >21 mm Hg.
Results
An episode of elevated IOP occurred in 36.1% of cases (44/122 cases), 11.4% (n = 5) occurring within the first week. The average duration of raised IOP was 48.9 (SD: 65.5) days. Causes included pupil block from air, swollen grafts, and corticosteroid response. Surgical intervention to lower IOP was required in 3 cases (6.8%). In multivariate analyses, the use of Olopatadine 0.1% or cyclosporine eye drops before DALK (OR = 14.51, 95% CI = 1.43–147.23) and use of topical prednisolone acetate 1% compared with dexamethasone 0.1% post DALK (OR = 4.79, 95% CI = 0.73–31.52) were associated with higher rates of elevated IOP post DALK. At 5 years post DALK, 3 of 71 cases (4.48%) developed de novo glaucomatous field defects, and 1 case with pre-existing glaucoma had progression of glaucomatous field defect.
Conclusions
DALK was associated with a significant incidence of transiently elevated IOP postoperatively, but had a low incidence of de novo glaucoma at 5 years in our study. Risk factors for raised IOP post DALK included the use of topical prednisolone acetate 1% compared with dexamethasone 0.1%, and the use of Olopatadine 0.1% or any concentration of cyclosporine eye drops prior to DALK.
Penetrating keratoplasty (PK) has traditionally been the standard surgical management for corneal opacification. Recently, improvements in surgical instrumentation have led to increasing popularity of lamellar techniques such as deep anterior lamellar keratoplasty (DALK), which involves removing the stroma while leaving the host endothelium and Descemet membrane (DM) behind. The advantages of DALK over PK are the retention of the recipients’ endothelial cells, greater resistance to globe rupture, faster visual rehabilitation, avoidance of endothelial rejection and complications of an open-sky PK, and earlier discontinuation of topical corticosteroids.
Glaucoma following PK is a serious problem leading to endothelial cell loss, graft failure, and visual loss. The etiology of post-PK glaucoma is multifactorial and may be related to iatrogenic injury of the angle, collapse of the trabecular meshwork, postoperative inflammation or hyphema, vitreous in the angle, pupil block, corticosteroid response, and peripheral anterior synechiae. Preoperative glaucoma is also a risk factor for post-PK and post-DSAEK glaucoma.
Few studies have examined the incidence of elevated intraocular pressure (IOP) and glaucoma after DALK. As expected, compared to PK, they report a lower incidence of elevated IOP (1.3%–6%) and glaucoma (0%) following DALK. A search on PubMed, EMBASE, and CENTRAL using a time frame of 2000–2016, without any language restrictions, using keywords “DALK,” “risk factors,” “glaucoma,” and “intraocular pressure” found no previous articles reporting risk factors for elevated IOP following DALK. Our study aimed to describe the incidence and risk factors of elevated IOP following DALK in our center.
Methods
Inclusion criteria included all patients who had DALK procedures performed in the Singapore National Eye Centre between February 6, 2004 and December 20, 2008 (n = 148 eyes). We excluded cases with previous PK done in the same eye (n = 7) and cases with less than 6 months of follow-up (n = 19), leaving 122 eyes that fit the inclusion criteria (84.7%), including 6 patients with bilateral grafts and 110 patients with unilateral grafts. A retrospective review of patients’ medical records was carried out. The study received Singhealth Institutional Board approval prospectively.
Variables
We collected data on sociodemographic characteristics including age, sex, and ethnicity. Preoperative data collection included Snellen visual acuity (VA), slit-lamp examination for the presence of narrow angles, vertical cup-to-disc ratio (CDR), lens status, IOP, indication for surgery, coexisting ocular disease, and previous ocular surgery. Intraoperative data collected included the primary surgeon, additional procedures, and complications.
Owing to the retrospective nature of this study, IOP measurements were not masked as they were taken by the surgeon. IOP was measured each visit with a noncontact air-puff tonometer. If the air-puff measurement was higher than 21 mm Hg, the IOP would be rechecked by the surgeon using the Goldmann applanation tonometer (GAT) (Haag-Streit, Bern, Switzerland) as a first choice. If the cornea was too irregular for an accurate measurement using the GAT, then a Tonopen was used. At least 2 GAT readings or 3 Tonopen readings with the indicator against the 5% level were taken and if the difference was >2 mm Hg, a further reading was taken and averaged.
The DALK surgeries were performed by 7 corneal surgeons (the majority by D.T.T. and J.S.M.), using the modified Anwar technique or a manual layer-by-layer predescemetic technique.
Modified Anwar Technique for Deep Anterior Lamellar Keratoplasty
The modified Anwar technique has been previously described. In brief, it involved partial trephination using a Hanna trephine followed by manual dissection of the stroma to a depth of approximately 50%, leaving 150–200 μm of residual stroma. This initial dissection allowed the needle or DALK air injection cannula to be placed more consistently close to the DM, enabling a higher rate of big-bubble separation of DM. A 27 gauge needle or DALK air injection cannula (Rycroft Air Injection Cannula; ASICO, Illinois, USA) attached to an air-filled 5-mL syringe was inserted bevel-down into the paracentral cornea, parallel to the corneal surface, and advanced 2–3 mm. Air was injected to create a cleavage plane between the DM and posterior stroma. On successful attainment of a big bubble reaching the trephination margins, a slit was created in the posterior stroma to break the big bubble and gain access to the DM, and the remnant posterior stromal tissue was removed. If a bubble was not obtained, the procedure was converted to a manual technique.
Manual Technique of Deep Anterior Lamellar Keratoplasty
The manual technique involved performing anterior lamellar dissection freehand using a crescent blade. Deep lamellar dissection was attempted to within 100 μm of DM but baring of DM was not attempted.
In both the big-bubble and manual technique, the donor cornea was punched on a Hanna trephine; stripped of DM, endothelium, and epithelium; and sutured onto the recipient bed with 10-0 nylon monofilament suture. A bandage contact lens was placed to aid re-epithelialization. In the majority of cases, there was no oversizing of the donor graft. In cases of extreme ectasia, donors were undersized by 0.25 mm to reduce postoperative high myopia.
Postoperative Care
All cases received topical antibiotics and corticosteroids postoperatively (including prednisolone acetate 1%, prednisolone acetate 0.12%, or dexamethasone 0.1%), depending on the surgeon’s clinical judgment. The topical corticosteroid chosen was given initially at a 3-hourly interval, and was continued for 6–8 months or longer according to clinical response, tapered to once a day before discontinuation. Sutures were removed between 6 and 18 months postoperatively, depending on the VA and astigmatism. Data on Snellen VA and IOP were collected at 6 months, 12 months, 3 years, and 5 years. We also collected data on Humphrey visual field (HVF) findings, postoperative glaucoma diagnosis and management, other complications, and ocular surgery undertaken after DALK surgery.
Eyes without pre-existing glaucoma were defined based on the European Glaucoma Society Guidelines : (1) no IOP of >21 mm Hg, and no IOP-lowering medication use; (2) no glaucomatous optic neuropathy (no CDR >0.6 with a glaucomatous visual field defect or focal neuroretinal rim defects); (3) no previous glaucoma laser or filtration procedures; and (4) no history of glaucoma.
A glaucoma-related diagnosis was defined as patients with established glaucoma, ocular hypertension, or glaucoma suspects.
The primary outcome was an elevated IOP post DALK, defined as IOP >21 mm Hg. Secondary outcomes included development of glaucoma and worsening of pre-existing glaucoma.
Statistical Analysis
Data were analyzed using SPSS version 22 (SPSS Inc, Chicago, Illinois, USA). The χ 2 and Fisher exact test were conducted for categorical comparisons. Quantitative analysis was conducted with the independent Student t test and 1-way analysis of variance. A P value <.05 was considered significant. The proportions of cases developing elevated IOP were charted using Kaplan-Meier survival analysis. Logistic regression analysis was used to assess risk factors for elevated IOP in univariate and multivariate models.
Results
There were 122 cases performed for optical (91.0%), tectonic (0.8%), and therapeutic (8.2%) indications ( Table 1 ). Of these, there were 13 (10.7%) with glaucoma-related diagnoses, including 3 (2.5%) with pre-existing glaucoma, for which 2 cases were on 1 glaucoma medication each and 1 had previous trabeculectomy; 1 with ocular hypertension who was on 1 glaucoma medication; and 9 disc suspects.
| Indication for DALK | Number of Patients With or Without Raised IOP (%) (N = 122) | Number of Patients With Any Raised IOP Post DALK (%) (N = 44) |
|---|---|---|
| Optical | ||
| Keratoconus | 39 (32) | 16 (36.4) |
| Lattice dystrophy | 4 (3.3) | 0 (0) |
| Granular dystrophy | 4 (3.3) | 0 (0) |
| Macular dystrophy | 2 (1.6) | 2 (4.5) |
| Crystalline dystrophy | 2 (1.6) | 0 (0) |
| Interstitial keratitis | 9 (7.4) | 3 (6.8) |
| Chemical injury | 7 (5.7) | 2 (4.5) |
| Stevens-Johnson syndrome | 3 (2.5) | 2 (4.5) |
| Others | 41 (33.6) | 15 (34.1) |
| Tectonic | ||
| Corneal melt | 1 (0.8) | 1 (2.3) |
| Therapeutic | ||
| Active infectious keratitis | 10 (8.2) | 3 (6.8) |
The mean age was 36.6 years (standard deviation [SD]: 19.1), and 57.5% were male. There were 47 (38.5%) Chinese, 25 (20.5%) Indian, 15 (12.3%) Malay, and 35 (38.7%) of other ethnicity. None had narrow angles. Cases with or without glaucoma-related diagnoses had comparable preoperative IOP of 14.3 mm Hg (SD: 6.0) and 13.7 mm Hg (SD: 6.4) ( P = .73), respectively ( Table 2 ). The median sizes of the donor and recipient button were both 8 mm (range 7–10 mm). There were 30 cases with intraoperative microperforations requiring a combination of conversion to manual dissection (n = 15), intracameral injection of air (n = 12), sealing of the microperforation with fibrin glue (n = 7), and intraoperative stromal patching (n = 8). None were converted to PK. Postoperative complications included 13 cases with DM detachments, for which 2 cases resolved spontaneously, and the remaining required a combination of air bubbling (n = 9), graft resuturing (n = 2), and fibrin glue (n = 1). There were 2 cases of graft rejection, both successfully treated with topical corticosteroids. There were 3 cases of graft failure, related to fungal keratitis, limbal stem cell deficiency, and endothelial failure, respectively. There were 56 cases (45.9%) that underwent further ocular surgeries following DALK ( Table 3 ).
| All Cases (Eyes) (N = 122) | Cases With Glaucoma-Related Diagnosis | P Value | Cases Without Glaucoma-Related Diagnosis | P Value | |||
|---|---|---|---|---|---|---|---|
| Without Raised IOP Post DALK (N = 9) | With Raised IOP Post DALK (N = 4) | Without Raised IOP Post DALK (N = 69) | With Raised IOP Post DALK (N = 40) | ||||
| Age, mean (SD) | 36.6 (19.1) | 47.56 (26.5) | 44.8 (19.1) | .854 | 36.3 (18.9) | 33.9 (16.9) | .509 |
| Sex, male (%) | 70 (57.4) | 7 (77.8) | 1 (25) | .217 | 36 (52.2) | 26 (65) | .193 |
| Race (%) | |||||||
| Chinese | 47 (38.5) | 0 | 2 (50) | .032 | 28 (40.6) | 17 (42.5) | .926 |
| Malay | 15 (12.3) | 0 | 1 (25) | 10 (14.5) | 4 (10) | ||
| Indian | 25 (20.5) | 6 (66.7) | 1 (25) | 11 (15.9) | 7 (17.5) | ||
| Others | 35 (38.7) | 3 (33.3) | 0 | 20 (29) | 12 (30) | ||
| BCVA, mean (SD) | 0.98 (0.7) | 1.11 (0.7) | 0.49 (0.3) | .049 | 0.96 (0.7) | 1.05 (0.8) | .518 |
| Presence of narrow angles | 0 | ||||||
| Vertical CDR, mean (SD) | 0.4 (0.2) | 0.69 (0.1) | 0.63 (0.2) | .437 | 0.36 (0.1) | 0.34 (0.11) | .530 |
| IOP, mean (SD) | 13.7 (6.4) | 13.4 (6.5) | 16.3 (4.9) | .463 | 13.8 (7.4) | 13.5 (4.3) | .862 |
| Size of donor button, mean (SD) | 8.03 (0.4) | 8.2 (0.5) | 7.9 (0.4) | .250 | 8.1 (0.5) | 7.9 (0.4) | .355 |
| Size of recipient bed, mean (SD) | 8.02 (0.5) | 8.3 (0.5) | 7.8 (0.5) | .134 | 8.1 (0.4) | 7.9 (0.4) | .221 |
| Repeat graft | 6 | 0 | 0 | NA | 3 (4.3) | 3 (7.5) | .667 |
| Lens status | 116 (95.1) | 9 (100) | 4 (100) | 1.0 | 66 (95.7) | 37 (92.5) | .667 |
| Previous intraocular surgery (%) | 15 (12.3) | 3 (33.3) | 0 | .497 | 7 (10.1) | 5 (12.5) | .756 |
| Other ocular disease (%) | 18 (14.8) | 2 (22.2) | 1 (25) | 1.0 | 11 (15.9) | 4 (10) | .385 |
| Pre-existing glaucoma diagnosis (%) | 4 (3.3) | 2 (22.2) | 2 (50) | .530 | 0 | 0 | NA |
| Pre-existing glaucoma treatment (%) | 10 (11.3) | 2 (22.2) | 2 (50) | .530 | 3 (4.3) | 3 (7.5) | .667 |
| Additional intraoperative procedures (%) | 8 (6.6) | 1 (11.1) | 0 | 1.0 | 4 (5.8) | 3 (7.5) | 1.0 |
| Complications (%) | 30 (24.6) | 1 (11.1) | 0 | 1.0 | 17 (58.6) | 12 (41.4) | .541 |
| BCVA (logMAR) (6 months) | 0.48 (0.5) | 0.59 (0.6) | 0.57 (0.2) | .950 | 0.44 (0.5) | 0.51 (0.5) | .491 |
| BCVA (logMAR) (12 months) | 0.39 (0.4) | 0.33 (0.3) | 0.35 (0.2) | .859 | 0.39 (0.5) | 0.40 (0.5) | .964 |
| BCVA (year 3) (logMAR) | 0.65 (0.5) | NA | 0.5 (.14) N = 2 | NA | 0.55 (0.4) | 0.86 (0.7) | .161 |
| BCVA (year 5) (logMAR) | 0.34 (0.4) | NA | 0.32 (0.23) N = 4 | NA | 0.38 (0.4) | 0.29 (0.4) | .435 |
| IOP (6 months), mean (SD) | 15.9 (4.7) | 14.8 (1.5) | 15.3 (6.1) | .855 | 14.8 (3.2) | 18.1 (6.2) | .004 |
| IOP (12 months), mean (SD) | 14.8 (4.5) | 14.5 (1.5) | 19.3 (9.3) | .463 | 13.5 (3.9) | 16.7 (4.5) | <.001 |
| IOP (year 3), mean (SD) | 14.4 (3.9) | NA | 18.3 (4.9) N = 3 | NA | 13.7 (4.0) | 14.9 (3.4) | .142 |
| IOP (year 5), mean (SD) | 15.2 (5.5) | NA | 17.5 (3.4) N = 4 | NA | 13.6 (2.9) | 17.0 (7.7) | .014 |
| Duration of raised IOP postoperatively, mean (SD) | 48.9 (65.6) | NA | 79.5 (109.9) | NA | NA | 45.2 (60.2) | NA |
| Steroid use duration, mean (SD) | 18.7 (23.4) | 6 (3.4) | 21.5 (27.1) | .336 | 18.9 (23.7) | 20.1 (24.3) | .813 |
| Type of topical steroid use (%) | |||||||
| Prednisolone acetate 1% | 71 (63.4) | 5 (83.3) | 4 (100.0) | 1.0 | 34 (54.0) | 28 (71.8) | .04 |
| Prednisolone acetate 0.12% | 7 (6.3) | 0 (0) | 0 (0) | 3 (4.8) | 4 (10.3) | ||
| Dexamethasone 0.1% | 34 (30.4) | 1 (16.7) | 0 (0) | 26 (41.3) | 7 (17.9) | ||
| Presence of glaucomatous HVF findings at 5 years (%) | 4 (3.3) | 0 | 2 (50) | 0 | 2 (5.0) | .133 | |
| CDR at 5 years, mean (SD) | 0.45 (0.17) | NA | 0.73 (0.15) | NA | 0.39 (0.1) | 0.49 (0.2) | .034 |
| Presence of glaucoma within follow-up period (%) | 3 (2.5) | 0 | 1 (25) | 0 | 2 (5.0) | .133 | |
| Duration after DALK when glaucoma diagnosis made (mo), mean (SD) | 39.7 (25.5) N = 3 | NA | 48 (NA) N = 1 | NA | NA | 35.5 (34.6) N = 2 | NA |
| Number of glaucoma eye drops post DALK, mean (SD) | 1.6 (0.8) | 1 (NA) N = 1 | 2.3 (1.2) | NA | NA | 1.56 (0.8) N = 2 | NA |
| Duration of glaucoma eye drops post DALK (months) | 24.4 (27.7) N = 32 | 5.5 (2.1) N = 2 | 31.5 (25.3) | .243 | NA | 24.8 (28.9) N = 26 | NA |
| Surgical interventions for glaucoma post DALK (%) | 4 (3.3) | 0 | 0 | NA | 0 | 4 (10) | .016 |
| Other complications post DALK (%) | 30 (24.6) | 0 | 1 (25) | .308 | 16 (23.2) | 13 (32.5) | .289 |
| Development of other ocular diseases after DALK (%) | 18 (14.8) | 1 (11.1) | 1 (25) | 1.0 | 6 (8.7) | 10 (25) | .020 |
| Other ocular surgeries post DALK (%) | 41 (33.6) | 0 | 2 (50) | .077 | 21 (30.4) | 18 (45) | .126 |
| Graft failure (%) | 3 (2.5) | 0 | 0 | NA | 1 (1.4) | 2 (5.0) | .553 |
| Surgeries | Number of Eyes |
|---|---|
| Phacoemulsification and intraocular lens implant | 14 |
| Graft resuturing | 12 |
| Intracameral injection of air | 10 |
| Graft refractive surgery | 7 |
| Repeat DALK | 4 |
| Anterior chamber reformation | 3 |
| Descemet membrane stripping automated endothelial keratoplasty | 1 |
| Amniotic membrane transplant | 1 |
| Penetrating keratoplasty | 1 |
| Removal of sutures under general anesthesia | 1 |
| Synechiolysis | 1 |
| Levator palpebrae superioris recession and lateral canthopexy for poor lid closure | 1 |
Overall, 36.1% (n = 44) developed elevated IOP, with 11.4% (n = 5) occurring within the first week. The median time to onset was 60 days (range = 1–2190 days). The cumulative probability of elevated IOP post DALK was 30.3% in the first year, 33.6% in the second year, 34.4% in the third year, 36.1% in the fourth year, and 36.1% in the fifth year ( Figure ). Causes of elevated IOP included pupil block from air fill (3/44; 6.8%), swollen grafts (2/44; 4.5%), corticosteroid response (37/44; 84.1%), and secondary open-angle glaucoma (2/44; 4.5). Topical corticosteroids used included prednisolone acetate 1% (63.4%), prednisolone acetate 0.12% (6.3%), and dexamethasone 0.1% (30.4%). The median duration of corticosteroid use was 10 months (range = 1–163 months).
