To compare vision, intraocular forward light scatter and corneal backscatter between deep lamellar endothelial keratoplasty (DLEK) and penetrating keratoplasty (PK) for endothelial dysfunction.

A randomized clinical trial.

Thirteen eyes (12 patients) were randomized to DLEK with a 9 mm scleral incision, and 15 eyes (14 patients) were randomized to PK. The primary outcome was high-contrast best spectacle-corrected visual acuity (BSCVA) at 12 months after surgery; intraocular forward light scatter and corneal backscatter were measured at one, three, six, and 12 months after surgery.

BSCVA at 12 months was 0.34 ± 0.16 logMAR (logarithm of the minimum angle of resolution) for DLEK and 0.25 ± 0.21 logMAR for PK (P = .23; minimum detectable difference at 12 months was 0.23 logMAR). The change in postoperative forward light scatter after DLEK correlated with the change in BSCVA (r = −0.66; P < .001; n = 11). Corneal backscatter was higher after DLEK than after PK at three and six months in the anterior third (P ≤ .005), at one through 12 months in the middle third (P < .001), and at one through six months in the posterior third (P ≤ .02) of the cornea. Backscatter after DLEK did not return to normal through 12 months (P < .001).

BSCVA was similar at one year after DLEK and PK. Improvement in BSCVA after DLEK correlated with decreasing forward light scatter. Increased backscatter after DLEK originated not only from the posterior cornea (interface) but also from the host cornea, which might limit visual outcomes after posterior lamellar keratoplasty.