OUTCOMES OF DMEK – STAGED PROCEDURE VS COMBINED: SYSTEMATIC REVIEW & META-ANALYSIS

To compare the clinical outcomes of Descemet membrane endothelial keratoplasty (DMEK) in combination with, before or after cataract surgery

Systematic review and meta-analysis

All studies in English language published in peer-reviewed journals, which assessed the outcomes of DMEK in combination with (DMEK triple; group 1), before (phakic DMEK; group 2) or after cataract surgery (pseudophakic DMEK; group 3) in patients with Fuchs endothelial corneal dystrophy (FECD) were included. Studies with mixed indications were included but only results of FECD patients were collected and results of any other indication, keratoplasty or aphakic patients were excluded. Primary outcome: Mean gain in best corrected LogMAR visual acuity (BCVA). Secondary outcomes: Graft detachment, rebubbling rates, graft rejection, graft failure and endothelial cell loss.

A total of 12 studies (number of eyes, n = 1932) were included; of these five studies were exclusively for DMEK triple (n=696), one for phakic DMEK (n=286), two for pseudophakic DMEK (n=950) and remaining four compared between two of these three categories. At 6 months, there was a mean gain in logMAR BCVA of 0.34+0.04 in group 1, 0.25+0.03 in group 2, and 0.38+0.03 in group 3. When compared, the difference was significant between groups 1 & 2 (Chi²=11.47; p=0.0007) and between groups 2 & 3 (Chi²=35.53 p<0.00001); however, there was no significant difference between groups 1 & 3 (Chi² = 2.20; p=0.14). At 12 months, there was mean gain in LogMAR BCVA of 0.52+0.05 in group 1 and 0.38+0.06 in group 3, difference being statistically significant (Chi2=14.04; p=0.0002). There were no studies reporting BCVA at 12 months in group 2. Rebubbling rate was 15% in group 1 (n=430), 4% in group 2 (n=275) and 10% in group 3 (n=883), the difference being statistically significant (p=0.001). Graft detachment rate was 31% in group 1 (n=180), 8% in group 2 (n=286) and 13% in pseudophakic DMEK group (n=696), difference again being statistically significant (p=0.001). However, there was no significant difference between DMEK triple and pseudophakic DMEK groups in terms of graft rejection, graft survival rates and endothelial cell loss at 12 months; this comparison could not be made for phakic DMEK group as the follow-up duration was only six months.

Mean gain in vision in DMEK triple group was comparable to staged pseudophakic DMEK at 6 months; however, was significantly better at 12 months. Although rebubbling rate and graft detachment rates were highest in DMEK triple compared to other groups, there was no significant difference in overall graft rejection, graft survival rates and endothelial cell loss. However, the results should be interpreted with caution due to high risk of study design bias and unequal distribution of number of eyes in each group for different outcomes.