One of the cases that we always remember as good was the first one, not because it solved a singular problem, but because it made us change our behavior before the corneal edema in a radical way. It was a patient with an operated eye of penetrating keratoplasty (PK), with whom he was happy until his other eye was operated on for endothelial lamellar keratoplasty (ELK or DSAEK) and 3 weeks later he achieved a visual acuity with correction (BCVA) of 1.0. From that moment on, the result of the PK was considered bad and we adopted the ELK as the technique of choice for almost all corneal edema. This first ELK was performed in 2007 in the right eye (RE) of a patient with cataract and significant corneal edema secondary to Fuchs' dystrophy, which limited its BCVA to 0.05. His left eye (LE) had been operated on PK plus phacoemulsification and arcuate incisions 5 years before and his BCVA was 0.6 with -4 -3 @ 140° (Figure 1).

A combined surgery of ELK plus phacoemulsification was performed, with insertion of the graft by traction with a suture. The cornea acquired transparency, the BCVA recovered very quickly (Figure 2) and stabilized 2 months after the intervention at 1.2 with +1.5 -1 @ 150° (Figure 3). The corneal topography showed the clear advantages in terms of the regularity of the anterior face, and therefore in terms of the quality of vision, in the RE with respect to the LE (Figure 4).

After 8 years of follow-up, this patient has not had any acute rejection episode but an annual endothelial loss of 11% up to 2013. From that moment on, we keep him with 2 drops of corticosteroid per day and this loss seems to have stopped. Its refraction and BCVA remain stable. The patient is much more satisfied with the ELK than with the PK, for the better and faster visual recovery he obtained. We are too, for the safety to injuries, the ease of replacement if the transplanted endothelium becomes insufficient, because it has avoided revisions, removal of sutures, astigmatism control, etc. Transplanting only the endothelium had already shown us its advantages with the ELK technique (endothelial lamellar keratoplasty) but the DSAEK has imposed indisputably for its simplicity and results. Today we recognize it as one of the great advances in ophthalmological surgery in recent years, as it has radically changed the prognosis of corneal edema.

This case of endothelial lamellar keratoplasty (ELK or DSAEK) has the interest of showing that it is possible to obtain a good result despite a complicated surgery. This is a 71-year-old patient with edematous (bullous) keratopathy after cataract surgery in his left eye (LE). The examination showed diffuse corneal edema, correct posterior chamber pseudophakia, intraocular pressure (IOP) = 10 mmHg, and visual acuity (VA) of counting fingers at 2 m. The fundus of the eye could not be observed clearly, and the patient lacked a medical history of interest.

It was programmed for ELK, with the usual tissue preparation technique (ALTK system from Moria) and the cornea of ​​a 69-year-old donor, with an endothelium of 2,356 cells/mm², preserved in Eusol for 6 days. Its central pachymetry was 525 μm. We practiced lamellar keratotomy with a 350 μm head, without removing the epithelium, at a slow pace and with the artificial anterior chamber pressurized to the just "pre-edema" level. At the end of the passage of the microkeratome we observed a small central eyelet on the bed (Figure 1). After staining with trypan blue, we verified that there was no loss of endothelial tissue, only of the posterior stromal layer and a rupture in Descemet’s membrane (DM). We therefore decided to continue with the surgical procedure. With care and some difficulty, as we did not have any pressure in the anterior chamber, we marked the diameter of the graft on the stromal side and transferred it to the base of the trephine, which in this case was 8.25 mm.

In the recipient, an inferior iridectomy was performed with scissors and forceps under miosis with acetylcholine. Descemetorhexis under continuous infusion of air. When passing the posterior lamina to Busin's spatula, we took care to slide it so that the DM flap tended to unfold without touching the metal (Figure 2). We inserted the graft by traction with Busin’s forceps under saline solution. When performing the pneumopexy we made sure that the broken DM did not remain folded on itself and we finished the procedure with the usual technique. In the postoperative period, a lack of adherence was observed in the area of ​​rupture, confirmed with OCT (Figure 3). After 7 days we re-injected air and achieved apposition of the graft and clearing of the cornea. The evolution was favorable and at 5 months his corrected VA was 0.75. In biomicroscopy with backlighting (Figure 4) the break in DM is perfectly observed, in the form of an elongated "S". In the OCT (Figure 5) the rupture and the central defect in the stromal lamina of the donor can be seen. The slit shows a transparent cornea (Figure 6).

This case of a classic complication of ELK (DSAEK)¹, as is the perforation of the donor bed during the microkeratome cut, allows us to comment on two aspects of this technique. The measured thickness of the donor cornea, 525 μm, was just at the limit, according to the nomogram for the ALTK system, between using the 350 μm head with epithelium – what we did – and the 300 μm – what we should have done –. We paid the price in attempting to obtain a too thin of a disk, with this complication. On the other hand, this case shows that it is possible to use a tissue with a small perforation, although it tends more to detachment. The management of DM without stroma already hinted that the DMEK was coming to stay.

A 58-year-old woman with an eye history of high myopia and chronic simple glaucoma poorly controlled with medical treatment. She was operated on by combined phacoemulsification surgery and intraocular lens implant associated with canaloplasty in the right eye. In the early postoperative period, he developed a haemorrhagic detachment of the Descemet’s membrane (DM) that evolved with the formation of a blood clot in the lower paracentral area of ​​the cornea (Figure 1). Given the persistence of the blood supply 6 weeks later, it was referred to our unit to assess its possible surgical treatment.

After unsuccessfully attempting to wash and aspirate the clot through a small incision in the DM, we performed a sectorial descemetorhexis of 5 mm in diameter, which allowed the clot to be completely eliminated (Figure 2). We started treatment with corticosteroid and hypertonic eye drops and decided to observe the response, waiting for a possible coating of the denuded area by migration of neighbouring endothelial cells. Although the picture improved in the following weeks, a zone of residual oedema persisted that caused a significant visual loss, poorly tolerated by the patient. We decided to perform an endothelial lamellar keratoplasty (ELK or DSAEK), with a circular descemetorhexis of 8.5 mm in diameter and the placement of a thin graft (about 80 μm) (Figure 3). The immediate postoperative period was uneventful, with excellent adherence and centred lenticule. No alterations in intraocular pressure were recorded and visual quality improved significantly.

Canaloplasty – cannulation, visualization and placement of a microcatheter in the Schlemm’s canal – has proven effective in the surgical treatment of glaucoma¹. There have been reports of detachments of DM associated with this technique, although it is rare that these are haemorrhagic in nature. To resolve them, Nd:YAG laser has been used – which allowed the evacuation of the blood collection mixed with viscoelastic² – or even a surgical approach using deep corneal dissection followed by irrigation with BSS, which facilitated the elimination of the blood 6 weeks after having occurred, without creating any tear in the DM³. In our case, the conservative manoeuvres to eliminate the blood deposit were not effective. A partial descemetorhexis allowed to remove it, but it was not followed by endothelial repair and the persistent corneal oedema made the ELK (DSAEK) necessary. This finally gave an excellent anatomical and visual recovery.

We present this case as an example of a rare but already described complication, of which we still ignore the cause. This is a 72-year-old patient with endothelial decompensation in her right eye, secondary to a complicated cataract surgery. No other history of systemic or ophthalmological disorders. The examination revealed corneal oedema, pseudophakia of the posterior chamber and discoria by vitreous fiber attached to the main incision. Intraocular pressure of 16 mmHg and visual acuity (VA) of 0.1. The retina could not be explored in detail.

Endothelial lamellar keratoplasty (ELK or DSAEK) was indicated and practiced. We prepared the donor cornea – 77 years old, with an endothelium of 2,470 cells/mm², preserved in Eusol for 5 days – with the ALTK system (Moria). Its central pachymetry was 597 μm and we used a 350 μm head. It was trephined at 8.5 mm and kept until its implantation in the preservation liquid. During lamellar keratectomy, hydroxypropyl methylcellulose (HPMC) was used to protect the endothelium.

In the receiver, we performed a mechanized anterior vitrectomy with decanted triamcinolone staining (Trigon). Lower iridectomy with the vitrectome itself and miosis with acetylcholine. Descemetorhexis under continuous infusion of air and implant of the graft with Busin’s spatula and traction with Busin’s forceps under saline solution. Pneumopexy for 2 hours in the supine position. Patient was sent home after checking that there was no pupillary block. In the early postoperative period, he presented a flat detachment of the graft, with whitish rough-fibrillar material on its anterior face (Figures 1 and 2), which OCT showed as deposits on a 120-μm disc (Figure 3). The recipient cornea presented without oedema. A treatment with prednisolone eye drops every 2 hours was prescribed, with progressive reduction in the following months. The evolution was a slow disappearance of the deposits, finely reticular (Figure 4), and of the interface fluid as shown by the OCT (Figure 5). VA improved from 0.15 to 1.0 (with 170° -2.5 -1) one year after surgery.

Various materials have been proposed as the cause of the deposits in the ELK interface: epithelial growth, infection, textile fibres, metallic remnants of the blade, calcium deposits, talc and others^1-3. In our case, the clinical aspect resembled more to cases that have been associated with the retention of viscoelastic with phosphate, although we only used 2.4% HPMC for the protection of the endothelium during the preparation and not during the rest of the surgery, as well as triamcinolone – momentarily and quickly washed from the anterior chamber –. In the absence of more precise tests, such as contrast sensitivity or light scattering, our patient recovered a good VA with complete satisfaction and did not require graft replacement.

A 64-year-old woman with a history of Fuchs' dystrophy, corneal decompensation and cataract (Figure 1), who was operated on her right eye for endothelial lamellar keratoplasty (ELK or DSAEK) combined with phacoemulsification and posterior chamber intraocular lens (IOL) implantation. The surgery ran without incident and we used sulphur hexafluoride (SF₆) to fix the graft. In the immediate postoperative period, a hypertensive peak of up to 41 mmHg occurred, which was normalized after 24 hours with oral acetazolamide and topical beta-blockers. In the following days, a course of accentuated ocular hypotonia (between 0 and 6 mmHg) was developed. The incisions were watertight (Seidel -), the graft was transparent, with abundant folds but correctly centered and adhered in its entirety to the recipient stroma, and the anterior chamber (AC) was very deep, completely occupied by the gas (Figure 2). Ultrasound B revealed no choroidal detachments.

Despite treatment with resting, atropine, topical and oral corticosteroids, normal intraocular pressure (IOP) was not recovered and, after reabsorption of the gas within 5-6 days, the disc began to detach (Figure 3a). Optical coherence tomography (OCT) confirmed the separation in the central area (Figure 3b) and the ultrasonic biomicroscopy (UBM) showed a de-insertion of the scleral spur in a large part of the angular circumference, mainly in its lower half (Figure 4). The air bubble was reformed (Figure 5) and the IOP normalized over the next 4 weeks. The cornea was transparent, with the graft well attached and without folds (Figure 6), the AC regained its normal depth and the IOL stable in the bag. The pupil was in hyporeflexic mydriasis and the iris acquired a moth-eaten appearance with peripheral radiated atrophy by backlighting (Figure 7). A new UBM confirmed reapplication of the ciliary body to the sclera (Figure 8).

Both air and SF₆ have been used to fix the donor disk to the recipient stroma at the end of different modalities of endothelial transplantation. Some comparative studies have shown a higher incidence of graft dislocation with air, as well as a lower endothelial cell density one year after the intervention¹. The diagnosis of a cyclodialysis is usually clinical through a gonioscopy, although studies such as UBM may be necessary. Treatment options, depending on their extension, include corticoids and cycloplegia, expandable intravitreal gases, and surgical cycloplexy or by diathermy, photocoagulation or cryotherapy².

We have not found in the literature the review of a case similar to ours. We believe that this could be due to the excessive expansion of SF₆ in the AC, which first caused hypertension due to antegrade pupillary block and later a cyclodialysis. As the angle was the zone of least resistance, the scleral spur was disinserted, which resulted in hypotonia due to increased uveoscleral drainage. This hypotonia would be responsible for the late dislocation of the lenticule and could also be due to transient inhibition of aqueous humor production, hypertensive ischemia of the ciliary body, or a hypersensitivity reaction to hypotensors³.

The UBM was definitive to demonstrate the detachment of the ciliary body as a cause of the hypotonia. Likewise, iris atrophy, pigmentary dispersion and hyporeflexic mydriasis could be due to trauma and ischemia of the iris due to excessive pressure caused by the SF₆ gas.

A lamellar interface in the cornea, be it in a transplant or refractive surgery, constitutes a niche where germs can easily be quartered and grow differently from other locations, since the usual washing and defense mechanisms do not have easy access to that space. In lamellar keratoplasties, both anterior (DALK) and posterior or endothelial keratoplasty (ELK/DSAEK), one of the infections that has been most observed in this plane is that due to yeast Candida sp^1-4. We present a case of insidious onset, months after an ELK.

This is a 76-year-old woman who presented with a very narrow anterior chamber (AC), cataract and Fuchs' dystrophy. His visual acuity (VA) was 0.15. In November 2008 we performed combined surgery of ELK and phacoemulsification in his left eye, without incidents. The immediate postoperative period was within normal, with a good evolution of transparency and VA, which in the second month was 0.4 without correction. In the third month the patient consulted for red eye, lacrimation and blurred vision of 7 days of evolution. His VA had decreased to 0.05 and the slit lamp examination showed yellowish white placoid infiltrates at the donor-recipient interface (Figure 1). We decided to perform sampling in the operating room for culture, copious irrigation of the AC and intracameral and intravitreal injection of ceftazidime, vancomycin and voriconazole. Topical treatment with broad-spectrum antibiotics and clarithromycin was continued under suspicion of P. acnes or S. epidermidis infection. There were no isolates of the donor samples or incidences with the fellow cornea that had also been transplanted.

The confocal microscopy examination revealed granular, hyperreflective deposits in the lenticule and stroma (Figure 2a, b), inflammation with intense veil (haze) at the interface (Figure 2c) and hyperreflective needle-type material, both at the interface and in the recipient stroma, with hyperintense granular structures similar to epithelial cells (Figure 2d). The presence of hyphae was not observed. In spite of the treatment, the infection progressed rapidly; in days it presented an increase in the corneal infiltrate with overlying epithelial defect, hypopyon, corneal edema and progressive reduction of vision to hand movement (Figure 3). Anterior vitrectomy was performed with extraction of the donor lenticule and samples of aqueous and vitreous humor were obtained. The culture confirmed the presence of Candida albicans.

The clinic improved with intravenous and topical voriconazole, a treatment that was gradually diminished, although the cornea remained opaque. Several weeks later she presented an increase in intraocular pressure (IOP), justified by the presence of multiple anterior synechiae. Finally, we performed penetrating keratoplasty (PK) combined with trabeculectomy and AC reconstruction. At the last visit, the cornea showed acceptable transparency (Figure 4) and IOP of 17 mmHg, but VA was limited to 0.1.

Fungal keratitis is an uncommon complication, but more frequent in lamellar keratoplasties than in PK^3-5. The appearance of infiltrates in the donor lenticule in the third month after an ELK posed a diagnostic and therapeutic challenge. The late response allowed an aggressive evolution and the final result was functionally limited. In view of the appearance of infiltrates of this type at the interface of an ELK, one should react without delay and surgically, with a block resection by PK – with adequate antimicrobial coverage –, since manipulating the graft in the AC exposes us to the intraocular spread of the infection.