Cultivated limbal epithelial transplantation and penetrating keratoplasty postchemical injury: a 14-year follow-up

Abstract

An 11-year-old girl presented to our tertiary eye care centre with a 9-month-old history of lime injury in the left eye in 1999 with vision of counting fingers close to the face. She initially underwent superficial keratectomy with amniotic membrane graft in 1999. Subsequently, cultivated limbal epithelial transplantation was performed in 2001 which improved her vision to 20/400. Following development of pannus and symblepharon in the left eye, she underwent pannus resection and conjunctival limbal autograft in 2002 and, a month later, optical penetrating keratoplasty (PK) following which her vision improved to 20/125. She was under regular follow-up, underwent exotropia correction and subsequent levator palpebrae superioris resection for ptosis and was maintaining good vision. On her last follow-up 14 years after PK in 2016, her vision in the left eye was 20/40 with lipid keratopathy and her right eye showed no signs of focal limbal stem cell deficiency.

Background

Ocular surface burn results in limbal stem cell deficiency (LSCD). Management of unilateral LSCD involves limbal stem cell transplantation (LSCT) and is then followed by penetrating keratoplasty (PK) for visual rehabilitation. Long-term outcomes up to 8 years of cultivated limbal epithelial transplantation (CLET) were reported earlier.¹ We describe here the 14-year follow-up and successful outcome of PK following CLET.

Case presentation

An 11-year-old girl presented to our tertiary eye care centre on 28 June 1999 with a history of lime injury in her left eye in September 1998. At presentation, her visual acuity in the right eye was 20/20 and her left eye was counting fingers close to the face, with intraocular pressure (IOP) of 12 and 6 mm Hg in the right and left eyes, respectively. Slit lamp examination of the right eye was within normal limits and that of the left eye showed meibomitis, traces of conjunctival congestion and 360° corneal vascularisation with scarring. The anterior chamber was deep and the pupil was hazily seen. Fundus examination of the right eye was within normal limits and the left eye could not be evaluated.

Investigations

An ultrasound B-scan of the left eye showed no apparent abnormality of the posterior segment.

Differential diagnosis

LSCD can be primary due to congenital causes like aniridia, sclerocornea, ectodermal dysplasia and multiple endocrine neoplasia or it can be secondary due to direct stem cell loss, postchemical injury, postautoimmune conditions such as ocular cicatricial pemphigoid, non-autoimmune disorders as vernal keratoconjunctivitis, atopy, postherpetic keratoconjunctivitis and iatrogenic such as postsurgery for ocular surface squamous neoplasia or pterygium.² On the basis of the clinical history and ocular features, a diagnosis of the left eye total LSCD status postchemical injury was made.

Treatment

The patient was advised to undergo superficial keratectomy with amniotic membrane graft in the left eye which was performed on 12 October 1999 following which postoperatively she was doing well. During a follow-up visit in September 2001, she was advised LSCT of the left eye. She underwent limbal biopsy of the right eye on 6 November 2001, followed by in vitro xeno-free cultivation of the cells at our stem cell laboratory, the details of which procedure were described earlier,³ and CLET of the left eye on 16 November 2001.

Outcome and follow-up

Postoperatively, she followed up regularly and was maintained on tapering doses of prednisolone acetate 1% and topical antibiotics and her vision improved to 20/400. She had vascularised pannus and symblepharon in the left eye (figure 1). She was advised to undergo pannus resection with conjunctival limbal autograft (CLAG), which was performed on 4 April 2002, following which she was doing well but vision remained 20/400 (figure 2). At this time, she was registered for optical PK for the left eye as her ocular surface had stabilised and showed no evidence of LSCD. She underwent left eye optical PK on 30 May 2002. Postoperatively, her vision was 20/400 improving to 20/125 (figure 3). She was maintained on a postoperative regimen of topical prednisolone acetate hourly with gradual tapering, lubricants and cycloplegics. On 23 June 2004, she was noted to have left eye exotropia with a vision of 20/125 improving to 20/40 with refraction. She underwent left eye lateral rectus recession 8.5 mm and medial rectus resection 5.5 mm following which she was orthotropic. When she was reviewed again in September 2004, she was noted to have ptosis in the left eye secondary to levator palpebrae superioris (LPS) disinsertion caused due to the previous symblepharon release and was advised LPS resection which she underwent on 17 November 2004. She was doing well postoperatively. At the next follow-up in 2009, her visual acuity was 20/20 in the right eye and 20/100 in the left eye. Slit lamp examination was normal in the right eye with no evidence of focal LSCD at the donor site and the left eye showed scarring with a well-apposed graft, deep anterior chamber, round pupil and clear lens. IOP was 10 and 12 mm Hg in the right and left eyes, respectively. She was advised a scleral contact lens trial for her left eye with lubricants and was asked to review when required. She then reviewed locally where appearance of scarring was noted in the graft (figure 4). On her last follow-up visit on 30 March 2016, her visual acuity in the right eye was 20/20 and the left eye was 20/60 improving to 20/40. On examination, the right eye was normal (figure 5) and the left eye showed lipid keratopathy on graft (figure 6) with a deep anterior chamber, round pupil and clear lens, with a normal fundus examination. She was advised topical lubricants and topical prednisolone acetate 1% eye drops in the left eye three times a day and then tapered monthly.

Figure 1.

Left eye status postcultivated limbal epithelial transplantation showing superior symblepharon.

Figure 2.

Left eye status postcultivated limbal epithelial transplantation and conjunctival autograft with symblepharon release.

Figure 3.

Left eye status postoptical penetrating keratoplasty showing clear graft with intact sutures.

Figure 4.

Left eye slit lamp image showing superotemporal appearance of scarring in graft.

Figure 5.

Right eye slit lamp photograph showing donor site area for limbal tissue with no evidence of focal limbal stem cell deficiency.

Figure 6.

Left eye slit lamp photograph at the 14-year follow-up showing compact graft with lipid keratopathy in superotemporal area sparing visual axis.

Discussion

In unilateral LSCD, donor tissue can be taken from the healthy eye and cultivated ex vivo (autologous CLET or Holoclar, Holostem Terapie Avanzate, Modena, Italy) or in vivo (simple limbal epithelial transplantation, SLET). Autologous CLET has been shown to produce successful outcomes in non-immune conditions like chemical injury where there is no chronic insult. The goal of such surgery is to establish an intact and stable epithelium of corneal phenotype and also possibly to replenish the source of corneal epithelial cells. A number of reports have supported such a restoration after autologous CLET.¹ Holoclar is recently approved in Europe for patients with LSCD postchemical or thermal injury.⁴ Holoclar consists of corneal epithelial cells dissociated and expanded ex vivo. In 2013, Pellegrini et al⁵ published results from 152 people treated with Holoclar for burn-related eye damage. The patients were followed for at least 5 years and showed full symptom abatement and corneal restoration in 66% of eyes and partial improvements in another 19%, with no adverse events related to the cells or their culture components. Stimulation of the endogenous limbal stem cell population, and/or recruitment of precursor cells from the bone marrow have been the proposed mechanisms of action. However, how long these cells persist in vivo and how they respond to insult or injury on a long-term basis has not been established.

In cases of severe ocular injury with chemicals, in addition to LSCD, there may also be stromal scarring and vascularisation which preclude attaining good visual outcomes after CLET or SLET, necessitating a PK. In a series by Basu et al,² the longest follow-up of successful PK post-CLET was 8.7 years. Of the 47 cases in this cohort, four cases (15.4%) failed because of recurrence of LSCD with central persistent epithelial defects. However, all these cases presented within the first 3 years of follow-up. Outcomes of CLET followed by PK in children are also not well documented in the literature. In a study by Sejpal et al,⁶ optical PK was performed in 16 children and at an average of 1-year post-CLET, the authors found that 10 grafts (62.5%) survived. The decision to plan early visual rehabilitation in children with postocular surface burn in a single-stage procedure with simultaneous CLET to prevent amblyopia or as a two-staged procedure where CLET is followed by PK is difficult. However, the study by Basu et al² noted that outcomes with a two-staged procedure are better compared with a single-staged surgery. Numerous other studies have reported keratoplasty performed postautologous CLET (table 1).^{2 7–11} It is interesting to note that all cases of graft failure due to recurrence of LSCD occurred within the first 3 years postkeratoplasty. However, how long this reserve of replenished stem cells lasts in view of challenges to epithelial integrity or an allogenic graft is a matter of speculation. In our case, it has lasted 14 years. Although this is the longest follow-up of a case of PK for chemical injury, it is tending to draw the attention of the reader to the fact that CLET has its limitations. The recurrence of pannus as well as symblepharon post-CLET early in our case indicate a breakdown of the surface necessitating salvage measures in the form of pannus resection with CLAG and surgical resection of ptosis. Undoubtedly, CLAG may have contributed to the subsequent stability of the surface as there was no episode of breakdown of the surface since then. Our recent experience with SLET intuitively corroborates the same.¹² In fact, CLET may have provided a stable milieu for the CLAG cells to function effectively. Our case gives the first report of a long follow-up of 14 years with a good outcome in terms of epithelial stability, graft clarity and survival in a case of chemical injury.

Table 1.

Summary of the outcomes of keratoplasty following autologous cultivated limbal epithelial transplantation in the published literature

Study	Number of cases	Visual outcomes	Graft rejection	Graft survival	Epithelial survival
Basu et al2	47				Recurrence of LSCD* in four cases (15.4%)
Baradaran-Rafii et al7	4	Better than 20/80 in three cases; Worse than 20/200 in one	Five episodes of endothelial rejection in three grafts	One failed with melt and three success	All showed varying degrees of conjunctivalisation as early as 6 months, with one showing complete recurrence of LSCD
Colabelli Gisoldi et al8	4	Better than 20/100 in three cases; Hand movement in one case	Not reported	Three survived for at least 10 months; one had recurrent LSCD	Stable in three cases; one case had recurrent epithelial defects
Di Iorio et al9	33	Not reported	Not reported	Clear cornea reported in 90% with a mean follow-up of 29.9 months	Stable epithelium in 90% of cases
Kawashima et al10	1	20/20 at 41 months	Not reported	Not reported	Stable epithelium for 41 months
Rama et al11	46	Permanent recovery in 21 patients ranging from 0.60 to 1.00 logMAR	Not reported	Not reported	Regeneration of epithelium sufficient to resurface donor in all cases

*Limbal stem cell deficiency.

logMAR, logarithm of the minimum angle of resolution; LSCD, limbal stem cell deficiency.

Learning points.

• Visual rehabilitation post-thermal injury can be done with penetrating keratoplasty (PK) and, with good follow-up, can give good outcomes as long as a stable ocular surface has been established.

• Two-staged intervention with initial measures to stabilise the surface followed by optical procedures in the form of PK is likely to give better outcomes.

• Optimal outcomes may mandate multiple surgeries in patients with limbal stem cell deficiency, especially in paediatric cases.