Results of viscobubble deep anterior lamellar keratoplasty after failure of pneumatic dissection

Aim To report the outcomes of bubble formation obtained by means of intrastromal injection of ophthalmic viscoelastic device (OVD) after failure of pneumatic dissection for deep anterior lamellar keratoplasty (DALK). Methods DALK was performed in 140 keratoconic eyes of 130 patients by injecting air and OVD only secondarily, after pneumatic dissection had failed; the bubble formation rates after air and OVD injection were recorded; complications, best spectacle-corrected visual acuity (BSCVA) and corneal tomographic parameters were evaluated 3 months, 6 months and 12 months postoperatively, as well as after complete suture removal. Results Air injection created a big bubble in 106/140 eyes (75.71%); OVD injection was not attempted in 4 eyes (perforation during cannula insertion n=2; air bubble burst n=2) and created a big bubble in 28 of the remaining 30 eyes (93.33%, 20% of the total). Manual dissection was required in 2/30 eyes (6.66%, 1.42% of the total) after failed OVD-assisted dissection. Deep folds, interface opacity and reduced BSCVA were noted in both eyes after failed OVD-assisted dissection. BSCVA was statistically better after pneumatic-assisted than after OVD-assisted dissection (P 0.01) only up to 3 months postoperatively; no statistically significant differences were recorded between the two techniques at later examinations. Conclusion Intrastromal injection of OVD after failed pneumatic dissection increases considerably the success rate of bubble formation (from 75.71% to 95.71% in our series); however, when bubble formation fails, infiltration of OVD into the residual stroma makes manual dissection particularly challenging and causes severe interface haze resulting in poor visual outcomes.