We present the case of a 67-year-old man with no systemic history of interest, who reported a 5-month history of redness and pain in his left eye (LE). He had been previously diagnosed with scleritis in three centers and was receiving treatment with topical and systemic corticosteroids, topical antibiotics and nonsteroidal anti-inflammatory drugs and oral cyclophosphamide.

The systemic clinical studies, laboratory and serology, did not present alterations. The Mantoux test was negative. The patient was referred to our service for not responding to medical treatment. The corrected visual acuity (BCVA) was 1.0 decimal in both eyes. In the biomicroscopic examination, whitish scleral nodules with hyperemia were observed in the LE, compatible with necrotizing scleritis and areas of scleromalacia (Figure 1). The rest of the eye examination was normal.

A biopsy of the scleral nodules was performed, with microbiological and histopathological studies that only revealed the presence of inflammatory cells. The patient reported intense pain and was given intravenous infliximab. The pain decreased and the scleritis improved. After two months, he presented a corneal infiltrate at the ​​3 to 6 o’clock area, with a satellite lesion at 10 o’clock and an inferior scleral nodule (Figure 2). Samples were taken for culture of bacteria, fungi, mycobacteria and viruses. The agar culture of Sabouraud showed growth of the fungus Paecilomyces lilacinus. We started treatment with topical, intracorneal and oral voriconazole, 5% natamycin suspension and topical ofloxacin, and oral terbinafine. Despite this, the evolution of the corneoscleral infection was unfavorable (Figure 3) and we performed a penetrating keratoplasty of 8.75 mm in diameter (Figure 4, video 3.7.1.1). The patient improved, with complete healing of both keratitis and scleritis. At 12 months, he underwent cataract surgery and his VACC is currently 0.4 (Figure 5).

Anti-infective or therapeutic keratoplasties are a resource for infections that, like the one described, do not respond to medical treatment. Although operating "in hot" involves an increased risk of failure of a penetrating keratoplasty, and even the intraocular dissemination of the infection, this case shows that it is possible to resolve these situations with them, and even obtain a success in terms of the transparency of the graft.

This 51-year-old woman consulted our emergency department in April 1980 for eye pain, burning and decreased visual acuity (VA) of 15 days of evolution. The examination showed in both eyes (OU) a mild corneal edema, mydriasis with posterior synechiae and pigment in the endothelium. The anterior chamber (AC) was narrow, the angle very narrow and the intraocular pressure (IOP) Goldmann = 40 mmHg in OU. The crystalline lenses were sclerotic, with some pigment on the anterior capsule. In the fundus of the eye, excavation and advanced papillary pallor (grade III) was observed, more in the left eye (LE). The decimal VA was in the right eye (RE) = 0.63 (90° -1.50 +3.00), and in the LE = 0.10 (90° -0.75 +3.50).

With the diagnosis of acute glaucoma by angular closure, a sectorial iridectomy was performed and, later on, over the years, trabeculectomy and transconjunctival cyclo-diathermy were performed. She developed cataracts that were operated on in OU in 1987 by intracapsular extraction with corneal incision, without perioperative complications. Four months after cataract surgery, a whitish retro-corneal membrane was observed in the RE that compromised the upper half of the cornea (Figure 1), with IOP = 14 mmHg under treatment.

Given the suspicion of epithelization on the endothelium and AC, in January 1988 I performed a penetrating sclerokeratoplasty of 16 mm in diameter, with total iridectomy (Figure 2). The histopathological study confirmed the diagnosis, with epithelial cells growing under the Descemet’s membrane and on the anterior surface of the iris (Figure 3). Controls were continued over the years, always under treatment with timolol 0.5% 2 times/day, topical corticosteroids and vasodilators.

In 1995 a melanoma of the big toe of the left foot was diagnosed. At that time, the graft was transparent, with IOP = 18.5 mmHg under treatment. The II pair was pale, with IV grade excavation, and VA = 0.5 (10° -2.50 + 10.50), but tubular. The last visual field performed showed foveal intensity of 25 dB with preservation of a central remnant that excluded the physiological blind spot and respected the fixation area. One year later the graft was still transparent, with IOP = 16 mmHg, under treatment. A reduction of the corneal diameter in the graft was observed (Figure 4). An echo-Doppler of the central artery of the retina showed severely decreased diastolic and systolic velocities, evidence of reduced perfusion of the eyeball. The patient died a few months later.

Sclerokeratoplasties are a rarely travelled field, but sometimes can solve problems such as in this patient, that are difficult to deal with other techniques. In our experience the results of sclerokeratoplasties up to 16 mm in diameter have been variable. This was one of the best and that is why I chose it at the request of the editors to present "my best case".

This is a 50-year-old patient who consulted us for the first time in May 1985 for dry eye. She had been diagnosed with rheumatoid arthritis 12 years earlier and was treated with penicillamine and indomethacin. Four months earlier she had presented with a perforation of the cornea in her right eye (RE) and in another service she had underwent closure of the lacrimal dots and conjunctival covering, which occupied the inferior temporal sector of the cornea. In both eyes (OU) we observed a punctate epithelial keratopathy in the exploration, and in the left (LE) there was a loss of substance with epithelial groove in the temporal cornea, with thinning in the infero-nasal sector. There was no Seidel +. The visual acuity (VA) decimal was RE = 0.12; LE = 0.40, without improving with correction. The diagnostic impression included: (1) ulcerative keratitis in OU, with corneal perforation by keratolysis, partially covered with conjunctiva in RE; (2) cataract in evolution; (3) peripheral ulcerative keratitis in LE, and (4) rheumatoid arthritis with secondary Sjögren's syndrome in OU.

Treatment was started in OU with topical corticosteroids, atropine, acetylcysteine ​​and artificial tears, as well as systemic corticosteroids. Three months later (August 1985) she presented with a new perforation in the RE, and we performed a penetrating keratoplasty with extracapsular cataract extraction and intraocular lens implantation. In the early postoperative period, 3 areas of keratolysis appeared in the graft: central, at 5 o'clock and at 7 o'clock (Figure 1). The most central involved full-thickness and formed a descemetocele.

In January 1986, a perforation of the graft occurred, and a pedunculated conjunctival covering was performed (Figure 2). In April of the same year the graft was replaced, with 16 independent sutures. At the 4^th postoperative month of this second graft, a new zone of keratolysis appeared at 10 o'clock on the graft-recipient junction at the height of a suture. It was accompanied by an iris hernia and an abscess (Figure 3), which were surgically repaired. In the photo, a hyperemic eye is seen, with advancement of the vascularization of the superior limbus and a neovessel on the 8 o’clock suture, with an incipient keratolysis area. The central area of ​​the graft remained transparent. The LE continued without modification, with the lesions described in the first consultation. In February 1987 the VA was RE = 0.35 (10° -7.50 -1.75); LE = 1.00 (40° -0.75 +0.25). The patient suffered a myocardial infarction and did not return to the controls.

I chose this patient in response to the editors' request for "my worst case". This is a difficult decision since, after an experience of five decades, there have been many cases that have made us suffer or have not turned out as we wanted. This in particular shows the difficulties of keratoplasty in the presence of chronic systemic processes difficult to treat such as rheumatoid arthritis.

An 85-year-old male patient was referred to the Cornea and Ocular Surface Unit of our hospital. He had edematous keratopathy (bullous) in his left eye, paralytic mydriasis, and three-piece intraocular lens (IOL) sutured to the sclera, dislocated inferiorly (Figure 1). This had been implanted in a replacement surgery of the IOL-bag complex partially subluxated in the anterior chamber (AC) performed 3 years earlier. Visual acuity (VA) was counting fingers and intraocular pressure (IOP) of 17 mmHg. No details of the fundus could be appreciated, but one could see a myopic conus.

Given the subepithelial stromal opacity secondary to long-standing edema, the indication of endothelial keratoplasty was not considered adequate. Penetrating keratoplasty was then performed with pupillary cerclage and replacement of the subluxated IOL with another one anchored to the iris retropupillary (video 3.7.4.1). The postoperative period was uneventful and in the last revision, performed 2 years later, the graft is transparent (Figure 2), the IOP = 14 mmHg and the VA with correction is 0.2 decimal, limited by an amblyopia due to myopia magna. In the fundus, the myopic conus was confirmed, and the macula was normal looking, corroborated by OCT.

COMMENT

Edematous keratopathy secondary to cataract surgery or complicated IOL replacement is a frequent cause of corneal transplantation. The preferred procedure today to treat this is endothelial keratoplasty, but there are cases in which it is necessary to perform a penetrating keratoplasty, when there is a stromal opacity secondary to long-standing edema. In these situations, it is common to have an AC IOL replaced by another type of secondary implant. The case we present is atypical since the replaced IOL was from the posterior chamber sutured to the sclera that was dislocated. In addition, mydriasis limited surgical options and further complicated the situation.

There is no agreement on the type of optimal secondary implant in case of keratoplasty. In general, it is advisable to avoid IOLs with angular support, as they have been associated with higher rates of endothelial loss among other complications¹. The optimal position of the IOL would be in the posterior chamber. Those sutured to the sclera have not shown superior results, their learning curve is long, they require more surgical time and are associated with complications such as bleeding or incarceration of the vitreous in the sclerotomies and tilting². Likewise, IOLs sutured to the iris have a long learning curve, require a prolonged surgical time and can cause pupillary ovalization³. As for those fixed by scleral tunnel, experience is limited in the context of keratoplasty. Therefore, our first option for secondary implantation, including cases of keratoplasty, is now the IOL anchored to the iris in a retropupillary position. Gicquel et al. compared those fixed to the iris in the pre- and retropupillary positions and found that the latter were associated with a lower rate of complications and lower endothelial loss⁴, results that coincide with those of the Gonnermann group in 23 cases⁵.

Our experience in 15 cases of corneal transplantation (including two of ELK/DSAEK) and Artisan retropupillary IOL implant is very favorable. This is an implant that combines the advantages of the posterior chamber – in terms of a greater distance to the endothelium – with those of a short learning curve, minimal surgical manipulation – since it avoids sclerotomies – and short duration of the procedure. Many authors agree on the need for a healthy iris diaphragm for its implantation. The patient we reviewed presented the difficulty of paralytic mydriasis, but it allows us to show how, with an adequate technique such as pupillary cerclage, an IOL anchored to the iris retropupillary can be implanted. That is why we have chosen this in response to the editors' request to recount "my best case".

A 56-year-old woman developed Acanthamoeba keratitis after trauma to her left eye (LE). The infection occurred in a torpid manner despite treatment with topical chlorhexidine and propamidine and oral voriconazole. She had a large epithelial defect that was superinfected with Pseudomonas spp. and suffered a very rapid corneal lysis despite adequate treatment with ceftazidime and tobramycin to which this germ was sensitive (Figure 1).

We performed urgent penetrating keratoplasty of large diameter (10.5 mm donor and 9.5 mm window). The crystalline lens was at the beginning of the surgery herniated above the plane of the limbal belt, supported by the iris to which it was joined by posterior synechiae in 360°. To allow suture of the donor button, once the membrane covering the synechiae was released, the spontaneous extraction of the lens was performed, with integrity of the anterior hyaloid. An iridectomy was performed at 12 o’clock (Figure 2).

In the postoperative period, topical treatment with chlorhexidine and propamidine was maintained and oral mycophenolate was started. Three and a half months later it was necessary to perform a second keratoplasty (sclerokeratoplasty), due to recurrence of the infection in both the graft and the sclera, with cultures positive for Acanthamoeba (Figures 3-5). Two months later and despite immunosuppression with systemic corticosteroids and mycophenolate, graft rejection occurred (Figure 6). The cultures were negative.

We then carried out the third sclerokeratoplasty of 12 mm in diameter and simultaneous coating with amniotic membrane, under triple immunosuppression (corticoids – in descending pattern –, mycophenolate and tacrolimus). The graft currently survives, 18 months after surgery (video 3.7.5.1). In the last follow-up, the vision with correction (+12 sphere) is 0.15 decimal, the intraocular pressure of 18 mmHg under double topical hypotensive treatment, the graft is transparent and presents a pupillary membrane that covers the hyaloid without signs of inflammation. The patient remains aphakic for the moment (Figure 7).

COMMENT

Corneal or corneoscleral transplantation of large diameter carries a high risk by itself, especially when performed in a context of inflammation and infection not medically controlled and repeated transplants. Therefore, it constitutes one of the worst prognosis scenarios in penetrating keratoplasty¹. In this case, it was a severe corneal infection caused by Acanthamoeba, which is difficult to manage, and complicated by Pseudomonas superinfection and Acanthamoeba extension to the sclera, rarely described^2,3.

This situation requires meticulous planning of immunosuppression, since the standard topical is not enough to prevent rejections in high-risk transplants. Systemic immunosuppression is justified because allogeneic immunization occurs not only in the eye but also in the lymph nodes and spleen⁴. However, nowadays there is no evidence of high quality that supports the use of a given immunosuppression regimen⁵. Although most of the published studies employ a single drug, several authors favor immunosuppression with triple medication, similar to that of renal transplants^4,6, and it is also recommended in the reconstruction of the ocular surface with stem cells⁷.

In our case, graft rejection occurred after the 2^nd transplant despite treatment with mycophenolate, but survival of the 3^rd was achieved by combining two drugs (mycophenolate and tacrolimus). We added an amniotic membrane coating at the end of the surgery which, according to some authors, could improve the prognosis in these cases⁸. We present this patient in response to the request of the editors to report "my worst case" since, although the final result has been favorable for now, it shows the challenge that represents the combination of infection refractory to medical treatment with the need for transplants of big size to control it.

This is a 71-year-old woman with glaucoma diagnosed 10 years earlier, operated in another center twice for glaucoma in both eyes and later of cataract with IOL implant in the posterior chamber in her left eye (LE). The patient presented corneal edema with visual acuity with correction (VACC) = 0.007, with good perception and luminous localization (Figures 1b, d). A trans-scleral cyclophotocoagulation was performed with Nd:YAG laser in thermal mode and the intraocular pressure (IOP) was controlled at 16 mmHg. The VACC improved somewhat (0.04) but the edema persisted. The right eye (RE) did not perceive light but had a transparent cornea with an endothelial population of 1610 cells/mm² (Figures 1a, c). An auto-keratoplasty of the RE to the LE cornea was proposed, but the patient refused, and we decided to perform an 8 mm penetrating homograft in the LE.

The initial result was correct, with the graft well adapted, transparent and the IOP = 16 mm Hg with timolol 2 times/day (Figure 2a). The vision improved subjectively, although the VACC continued at 0.04, with a preserved Goldmann field of vision. At 4 months, she went to the emergency room due to loss of vision, with IOP = 35 mmHg. Gonio-synechiae were seen in 360° and another session of cyclophotocoagulation was performed (Figure 2b). During this period, a loop of the continuous suture was broken at 6 o'clock and the next day an ectasia of the scar appeared at that level, which was sutured immediately (Figure 2c). A month and a half later the IOP was controlled at 10 mmHg. Fibrin deposits had formed on the IOL (Figure 2d). The VA did not exceed 0.06 with correction.

Nine months after the transplant, she returned to the emergency department for loss of vision (VA = 0.03). Graft edema was observed (Figure 3a), which did not improve with intense medical treatment of possible atypical rejection. The IOP was 15 mmHg and the IOL was seen wrapped in fibrotic material near the endothelium but not touching it. In the periphery the iris contacted the endothelium in 360°. She was proposed again the auto-keratoplasty of 10 mm in diameter and this time she accepted. An explant of the IOL with its fibrotic sheath was also performed and a new IOL was implanted, sutured to the sclera after performing a vitrectomy (Figure 3b).

The postoperative evolution was correct, with a transparent graft, the new IOL well separated (chamber type IV), the IOP = 16 mmHg with 1 hypotensive eye drops and the VACC = 0.05. The endothelial population was 751 cells/mm² (Figure 3 c). However, 40 days after the operation, the IOP rose to 32 mmHg and discrete epithelial veil appeared. VACC = 0.06. A new cyclophotocoagulation was performed. The IOP was controlled again at 8-13 mmHg without treatment. Two years after auto-keratoplasty, VA = 0.08 with IOP = 14 mmHg and good Goldmann visual field (Figure 4c). However, she began to have certain alterations of the epithelium (Figure 4a) and, despite the treatment with lubricants without preservatives, 8 months later she presented with an epithelial defect requiring a therapeutic contact lens (TCL) (Figure 4b). There was a slight edema apparently caused by the epithelial defect, with no signs of immune reaction. VACC = 0.07. The IOL appeared to be advancing, with the upper haptic near the graft scar. Nine months later the situation remains stagnant, with a persistent epithelial defect that forms a pit, a somewhat thickened graft with opacities, digital hypotonia. It starts 360° peripheral vascularization. VACC = 0.04 (Figure 4d).

A new cornea homograft of 7.5 mm in diameter was decided. The initial evolution was good (Figure 5a), but the VACC remained low (0.03). At 11 months the graft began to veil, with hypotonia and loose sutures under the TCL (Figure 5b). Despite all efforts, it evolved towards the phthisis bulbi, with marked folds in the sclera (Figure 5c). Even with a relatively transparent graft, she developed choroidal detachments (Figure 5d) and the vision fell to hand movement. The patient died a few months later.

COMMENT

We present this patient in response to the editors' request to recount "our worst case". As you can see, it is not a particularly spectacular clinical course. Rather it corresponds to the experience of that struggle, sometimes deaf and a little confused in the day to day, to try to keep a few hundredths of vision in a single eye, a struggle that in this case we end up losing before the inexorable glaucoma and its other side, the hypotonia.

Among the lessons we can draw is the observation that the endothelium of an auto-keratoplasty, despite its absence of immunogenicity, is still damaged by other mechanisms, here possibly related to glaucoma or chronic inflammation. On the other hand, cyclo-destructive procedures, particularly when they are repeated, in the long run suppose a significant risk of developing hypotonia and phthisis bulbi, and in the short term of inducing an inflammatory reaction with intraocular fibrosis. It is a case of more than two decades ago, but it is not clear that today we could achieve another outcome.