Long-Term Ocular Surface Stability in Conjunctival Limbal Autograft Donor Eyes

CLINICAL SCIENCE

Albert Y. Cheung, MD,*† Enrica Sarnicola, MD,*†‡ and Edward J. Holland, MD*†

(OSST) procedures can provide healthy limbal stem cells Purpose: fi – To investigate the incidence of limbal stem cell de ciency and conjunctiva to rehabilitate a damaged ocular surface.2 8 (LSCD) in donor eyes after conjunctival limbal autograft (CLAU). Conjunctival limbal autograft (CLAU) is a technique that ; Methods: An observational retrospective review was performed on transplants a portion (typically 3to6clockhours)ofthe all patients who underwent CLAU alone, combined keratolimbal healthy stem cells from an unaffected eye to the contralat- allograft with CLAU (“Modified Cincinnati Procedure”), or com- eral affected eye in individuals affected by unilateral LSCD.6 bined living-related conjunctival limbal allograft (lr-CLAL) with 7 $ Cultivated limbal epithelial transplantation (CLET) CLAU having 6 months of follow-up after surgery. The outcome 8 measures were best-corrected visual acuity (BCVA) and ocular and simple limbal epithelial transplantation (SLET) provide surface status. ex vivo and in vivo expansion of stem cells, respectively, to repopulate the ocular surface. In contrast to CLET and SLET, Results: The inclusion criteria were fulfilled by 45 patients. Of CLAU is ideal for rehabilitating severe unilateral LSCD these, 26 patients underwent CLAU, 18 underwent combined requiring combined conjunctival limbal transplantation and keratolimbal allograft/CLAU, and 1 underwent combined lr- for reconstruction of the ocular surface. Proponents of CLET CLAL/CLAU. Mean age at the time of surgery was 39.6 years. and SLET have raised concerns regarding the potential Mean logMAR preoperative BCVA was 20.08. There were no iatrogenic damage of CLAU to the healthy donor eye.8,9 operative complications. The mean follow-up duration after To date, there remains a paucity of data on the long- surgery was 48.3 months (range 8.3–181.5 mo). At last follow- term effects of CLAU on donor eyes. In this study, we up, all eyes maintained a stable ocular surface, and mean logMAR investigate the incidence of LSCD in donor eyes after CLAU. BCVA was 20.05.

Conclusions: With the advent of newer ocular surface transplantation methods, there has been concern that CLAU carries the METHODS theoretical risk of inducing LSCD. Our long-term clinical results A retrospective chart review was performed on all following donor eyes after CLAU demonstrate no signs of LSCD. patients with LSCD who required surgical management between May 2000 and July 2016 at the Cincinnati Eye Key Words: donor, conjunctival limbal autograft, limbal stem cell Institute. This study was conducted according to a protocol deficiency, ocular surface stem cell transplantation approved by the University of Cincinnati Institutional (Cornea 2017;36:1031–1035) Review Board (Cincinnati, OH). The protocol and methods usedalsocompliedwiththestandardssetforthbythe Declaration of Helsinki. The inclusion criteria comprised 1) imbal stem cell deficiency (LSCD) occurs when the all patients who underwent CLAU surgery alone or com- Llimbal stem cells are destroyed or become dysfunctional, bined with another OSST procedure 2) with a minimum of leading to the hallmark characteristic of corneal conjunctiv- 6-month follow-up. The parameters assessed included alization with invasion of conjunctival goblet cells onto the demographics, presenting diagnosis and clinical features, corneal surface.1 Ocular surface stem cell transplantation preoperative and postoperative Snellen best-corrected visual acuities (BCVAs) of the donor eye, intraoperative and postoperative complications of the donor eye, and ocular Received for publication March 19, 2017; revision received April 24, 2017; surface stability of the donor eye at last follow-up. If care accepted April 25, 2017. Published online ahead of print June 21, 2017. had been transferred to a local corneal specialist, the most From the *Cincinnati Eye Institute, Cincinnati, OH; †University of Cincinnati, recent records were requested. A stable ocular surface was ‡ Department of Ophthalmology, Cincinnati, OH; and Department of defined as intact corneal epithelium devoid of conjunctival- Medicine Surgery and Neuroscience, University of Siena, Siena, Italy. fl E. J. Holland has consulted for Alcon Laboratories, Allergan, Bausch & ization or in ammation. An improved ocular surface was Lomb, Kala Pharmaceuticals, Mati Pharmaceuticals, Omeros, PRN, RPS, defined as an eye with signs of partial failure with areas of Senju Pharmaceuticals, Shire, TearLab, and TearScience. The remaining healthy corneal epithelium and areas of abnormal epithe- authors have no funding or conflicts of interest to disclose. lium, late fluorescein staining, or conjunctival epithelium Reprints: Edward J. Holland, MD, Cincinnati Eye Institute, 580 South Loop fi Rd, Suite 200, Edgewood, KY 41017 (e-mail: Eholland@holprovision. over the cornea. Ocular surface failure was de ned as com). recurrent diffuse late staining, conjunctivalization, or per- Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. sistent epithelial defects despite all treatment measures.

Cornea Volume 36, Number 9, September 2017 www.corneajrnl.com | 1031

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cheung et al Cornea Volume 36, Number 9, September 2017

Statistical analysis was performed using a 2-tailed t test with included on the CLAU. The donor sites are closed with P , 0.05 considered statistically significant. 10-0 nylon sutures, and a large-diameter (16.0 or 18.0 mm) bandage contact lens (Kontur, Hercules, CA) is placed over the eye. The harvested CLAU tissue is then fixated into the Operative Procedure respective 12- and 6-o’clock positions at the limbal edge of Details of the surgical technique for CLAU6 (Fig. 1) the recipient eye with 10-0 nylon sutures and tissue glue. and combined keratolimbal allograft (KLAL) with CLAU For the Modified Cincinnati Procedure, the CLAU pro- (“Modified Cincinnati Procedure”)10 have previously been cedure is augmented by also securing KLAL segments to described, and the surgical techniques implemented for the 3- and 9-o’clock positions of the recipient eye.10 There eyes in this study do not differ significantly. Importantly, it was 1 eye that underwent combined living-related con- should be noted that approximately 2 clock hours in the junctival limbal allograft (lr-CLAL) with CLAU. Here, the limbal circumference (;6to6.25mm)atboth12-and6- CLAU procedure is augmented instead by securing 2 lr- o’clock positions with corresponding conjunctiva [dissec- CLAL segments to the 3- and 9-o’clock positions of the tion carried ;5mm(inferior)–8 mm (superior) posteriorly] recipient eye. The lr-CLAL donor tissue is harvested in are excised. Calipers are used to ensure that the extent of a manner identical to that of CLAU donor tissue but in an tissue to be harvested is always less than half the circum- eye of a living relative. Living relatives undergo ABO ference (ie. ,6 clock hours). Blunt dissection is performed blood typing, human leukocyte antigen typing, crossmatch- to a point 1 mm anteriorly beyond the limbus and periph- ing, and panel reactive antibody testing to identify the best- eral corneal vasculature to ensure that limbal stem cells are matched relative. Acceptable donor relatives had to be

FIGURE 1. Intraoperative photo- graphs demonstrating the CLAU procedure. Two clock hours in the limbal circumference (;6 to 6.25 mm) at both 12- and 6-o’clock positions are measured and marked (A). The conjunctiva is elevated with balance salt solution on a 30-guage needle (black arrow, B). Blunt dissection is performed to a point 1 mm anteriorly beyond the limbus and peripheral corneal vasculature (C). The CLAU segment is excised (D). The donor sites are closed with 10- 0 nylon sutures (E). The harvested CLAU tissue is then ﬁxated to the respective 12- and 6-o’clock positions (black arrows) at the limbal edge of the recipient eye with 10- 0 nylon sutures and tissue glue (F).

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Conjunctival Limbal Autograft Donor Eyes viable ABO blood donors and have a negative crossmatch developed. There were 2 eyes with a history of amblyopia with the recipient; the fewer the human leukocyte antigen and 1 eye with a remote history of contact lens wear for 6 mismatches between a recipient–donor pair, the better years. All other donor eyes were healthy eyes without contact the match. lens wear. There were no intraoperative complications in the donor eyes. In the early postoperative period, there was 1 eye that developed a delle at both CLAU sites, which resolved Follow-up Care quickly with conservative therapy, 2 eyes that demonstrated Postoperatively, the donor eye is administered a top- mild conjunctival thinning at a CLAU site, and 1 eye that ical corticosteroid (prednisolone 1% or difluprednate developed a small pyogenic granuloma over the conjunctival 0.05%) and fluoroquinolone (moxifloxacin 0.5% or levo- CLAU site (this resolved with topical corticosteroids). floxacin 0.5%). Once the conjunctiva has reepithelialized Before OSST, the mean BCVA for donor eyes was (typically within 1 wk), the bandage contact lens and 20/24 (91.1% $ 20/40). At last follow-up, mean BCVA for fluoroquinolone are discontinued, whereas the corticoste- donor eyes was 20/22 (93.3% $ 20/40). There was no roid is tapered over 1 month. statistically significant difference between preoperative BCVA and BCVA at last follow-up (P =0.12).Atlast follow-up, all donor eyes maintained a stable ocular surface RESULTS (100%) and clear corneas (except for the prephthisical/ During the 15-year retrospective study period, 51 phthisical eye, which has since developed band keratop- patients were referred for LSCD and underwent a CLAU athy) without any late staining, neovascularization, or procedure (alone or combined with other OSST) by a single conjunctivalization (Fig. 2). surgeon (E.J.H.). Of these, 45 patients [37 males (82%) and 8 females (18%)] met the inclusion criteria. There were 26 patients who underwent CLAU, 18 combined KLAL/CLAU, DISCUSSION and 1 combined lr-CLAL/CLAU. Mean age at the time of CLAU is a successful approach to restore the stem surgery was 39.6 years (range 4.9–79.1 yrs). Chemical/ cell population in unilateral LSCD.6,11,12 Autologous limbal thermal injury (41 eyes, 91%) was the most common cause transplantation was first described by Strampelli in 1960 of LSCD followed by iatrogenic causes (4 eyes, 11%), such and subsequently by Barraquer in 1964.13 In 1989, Kenyon as treatment of ocular surface squamous neoplasia and and Tseng6 described their CLAU technique, which repre- radiation. Optical keratoplasty (penetrating keratoplasty or sents the first true stem cell autograft technique. This deep anterior lamellar keratoplasty) was needed in 27 eyes remained the predominant autograft technique until the after OSST to further rehabilitate vision. advent of SLET and CLET.7,8 Some authors have attempted Mean follow-up after OSST was 48.3 months (range limiting the transplantation to 1 CLAU segment or using 8.3–181.5 mo). Except in 2 eyes, each CLAU procedure smaller donor segments to decrease potential iatrogenic transplanted just over 4 clock hours (2 CLAU segments) of damage to donor eyes.14,15 Kheirkhah et al16 reported limbal stem cell tissue from the donor eye to the recipient eye a successful case of using a smaller single 2 clock-hour- with LSCD. In one eye, only 1 CLAU segment was needed to long lenticule in combination with amniotic membrane treat an eye with partial LSCD affecting the visual axis. In the transplantation to treat total LSCD with 12 months other case, the donor eye was a no light perception (NLP), of follow-up. prephthisical eye with a history of multiple retinal detach- CLET and SLET are 2 rapidly emerging OSST ments and surgeries. A repeat CLAU procedure was techniques. CLET involves culturing multiple 1- to 2-mm performed 4 years after the first CLAU procedure using the biopsies from either an uninjured eye (unilateral injuries) or same eye as a donor (given its NLP nature). Despite a donor corneoscleral ring to produce layers of cells.5,7 The transplanting 8 clock hours of stem cells, the eye did not sheet of stem cells is placed on the prepared recipient corneal demonstrate conjunctivalization or other signs of LSCD with bed, and a human amniotic membrane is sutured in place over 7.3 years of follow-up although band keratopathy has these cells.5,7 Although less donor tissue is used, CLET is

FIGURE 2. Slit-lamp photograph demonstrating the superior (A) and inferior (B) CLAU donor sites more than 8 years after CLAU surgery.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cheung et al Cornea Volume 36, Number 9, September 2017 costly and unable to provide conjunctival goblet cells to the of the remaining stem cells from the surrounding untouched host surface. In SLET, a small donor limbal graft is expanded areas.26 Theoretically, this would imply that there would be an in vivo in the stem cell–deficient eye with the use of amniotic opportunity to harvest stem cells a second time from the same membrane and fibrin glue.8 In 125 cases, 76% maintained site, sparing untouched areas from any surgical trauma.26 a successful outcome with SLET with a median follow-up The concern for potential LSCD is valid as abnormal duration of 1.5 years.17 SLET is less costly than CLET and corneal epithelial healing has been observed as late as 4 to 8 similarly does not provide conjunctival tissue to the host. months after repeat insult to the donor eye in animal Advocates of CLET and SLET have raised concerns models.27–29 Therefore, a second harvest of stem cells from regarding the potential iatrogenic damage of CLAU to the the same CLAU site should be considered with extreme healthy donor eye and inadvertent induction of LSCD.8,9 caution and only in the absence of other alternatives,26 and we CLAU donor eyes typically heal quickly without any are not advocating that notion. complication. Minor adverse events are uncommon but have As part of harvesting the donor tissue, we aim to excise been described such as microperforation,18 partial conjunc- approximately 2 clock hours of limbal stem cells per CLAU tivalization limited to the keratectomy site,19,20 haze,21 segment and advocate that the total amount harvested is less pseudopterygium,22 abnormal epithelium,23 corneal depres- than 5 clock hours. In our series, there was one NLP donor sion,24 filamentary keratitis,11,25 and conjunctival scarring.25 eye from which we did remove more than 5 clock hours Development of LSCD has occurred in donor eyes with (performed in 2 separate CLAU procedures more than 4 yrs a concurrent history of contact lens wear. Jenkins et al21 apart); however, we do not routinely recommend this practice. reported that 1 of 5 donor eyes developed epitheliopathy after As this eye did not demonstrate frank signs of LSCD with autologous limbal transplantation for ocular surface disease. more than 7 years of follow-up, its stem cells may have This eye had a history of chronic contact lens wear. Tan repopulated during the interval between surgeries as sug- et al23 similarly reported abnormal epithelium in a donor eye gested by Busin et al.26 Even if the stem cells repopulate, it after CLAU for clinically unilateral contact lens-related would not be recommended to reduce the number of stem LSCD; this donor eye had a history of contact lens wear cells to less than half, as this may be too great a threshold to and exposure to thiomersal. Frank LSCD in a healthy donor overcome, potentially resulting in LSCD. eye after CLAU not wearing long-term contact lenses has not Limitations of this study include the retrospective been described. nature. Our series was compiled from a patient population CLAU is a procedure for unilateral disease. The at a tertiary care corneal referral practice (introducing referral majority of our eyes undergoing a CLAU developed LSCD bias); however, it would be difficult to study such a large secondary to chemical/thermal injury. Although there were number of these cases if not in this setting. a few eyes with LSCD from iatrogenic etiology, there were no CLAU is a very effective treatment for eyes with unilateral LSCD cases secondary to contact lens wear in unilateral LSCD. We present the largest series studying patients with exposure to contact lens bilaterally, even if the long-term outcomes of donor eyes after CLAU, with a mean fellow eye appeared clinically normal. As seen in cases follow-upof48.3months(withlongestfollow-up15yrs) described by Jenkins and Tan, this could potentially lead to after CLAU. This study supports the safety of this pro- development of LSCD if the seemingly unaffected eye was cedure for the donor eye. Although in most cases, just over used as a donor.21,23 These cases reinforce our policy of not 4 clock hours of limbal tissue are harvested, tolerating up to using long-term contact lens wearers as donors for CLAU or 4.5to5totalclockhoursfor2CLAUsegmentsis lr-CLAL. Although the limbus can appear normal on a conservative estimate. The exact number of clock hours examination, there may be significant limbal dysplasia and that can be harvested safely from a donor has not been limbal stem cell drop out. established. Based on our long-term study and excellent To study the donor sites at a cellular level, Miri et al25 outcomes of not inducing stem cell deficiency in the donor, examined eyes that underwent CLAU with $1-year follow-up we recommend harvesting at most 5 clock hours of limbal using in vivo confocal microscopy. They found that 17 donor stem cells to avoid inducing LSCD. sites had reepithelialized the donor area with conjunctival epithelium, whereas 7 other donor sites demonstrated reepi- 25 thelialization with normal corneal epithelial cells. Con- REFERENCES 26 versely, Busin et al biopsied previously harvested and 1. Dua HS, Miri A, Alomar T, et al. The role of the limbal stem cells in control sites in 12 CLAU donor eyes. One year or more after corneal epithelial maintenance. Ophthalmology. 2009;116:856–863. CLAU, they isolated limbal stem cells from both sites and 2. Holland EJ, Schwartz GS, Daya SM et al. Surgical techniques for ocular found that cells from both sites exhibited comparable surface reconstruction. In: Krachmer JH, Mannis MJ, Holland EJ, eds. biological characteristics.26 Interestingly, clonogenicity of Cornea: Surgery of the Cornea and Conjunctiva, Vol 2, 3rd ed. New York, NY: Mosby Elsevier; 2013:1727–1744. greater than 3% was obtained from both sites, representing 3. Holland EJ. Management of limbal stem cell deficiency: a historical a value considered normal for nonoperated, healthy eyes.26 perspective, past, present, and future. Cornea. 2015;34(suppl 10):S9– The limbal architecture did not always regenerate, as the S15. palisades of Vogt were noted only in a third of harvested sites. 4. Tsubota K, Satake Y, Kaido M, et al. Treatment of severe ocular-surface disorders with corneal epithelial stem-cell transplantation. N Engl J Med. They concluded that the presence of limbal stem cells in the 1999;340:1697–1703. harvested sites may result from true restoration of the entire 5. Schwab IR. Cultured corneal epithelia for ocular surface disease. Trans limbal stem cell population, rather than from simple spreading Am Ophthalmol Soc. 2000;97:891–986.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Conjunctival Limbal Autograft Donor Eyes

6. Kenyon KR, Tseng SCG. Limbal autograft transplantation for ocular 18. Morgan S, Murray A. Limbal auto-transplantation in the acute and surface disorders. Ophthalmology. 1989;96:709–722; discussion 722–723. chronic phases of severe chemical injuries. Eye (Lond). 1996;10: 7. Pellegrini G, Traverso CE, Franzi AT, et al. Long-term restoration of 349–354. damaged corneal surfaces with autologous cultivated corneal epithelium. 19. Han ES, Wee WR, Lee JH, et al. The long-term safety of donor eye for Lancet. 1997;349:990–993. 180 degrees limbal transplantation. Graefes Arch Clin Exp Ophthalmol. 8. Sangwan VS, Basu S, MacNeil S, et al. Simple limbal epithelial 2007;245:745–748. transplantation (SLET): a novel surgical technique for the treatment 20. Barros Jde N, Santos MS, Barreiro TR, et al. Cytological features of live of unilateral limbal stem cell deficiency. Br J Ophthalmol. 2012;96: limbal tissue donor eyes for autograft or allograft limbal stem cell 931–934. transplantation. Arq Bras Oftalmol. 2011;74:248–250. 9. Ramachandran C, Basu S, Sangwan VS, et al. Concise review: the 21. Jenkins C, Tuft S, Liu C, et al. Limbal transplantation in the management coming of age of stem cell treatment for corneal surface damage. Stem of chronic contact-lens-associated epitheliopathy. Eye (Lond). 1993;7: Cells Transl Med. 2014;3:1160–1168. 629–633. 10. Chan CC, Biber JM, Holland EJ. The modified Cincinnati procedure: 22. Basti S, Mathur U. Unusual intermediate-term outcome in three cases of combined conjunctival limbal autografts and keratolimbal allografts for limbal autograft transplantation. Ophthalmology. 1999;106:958–963. severe unilateral ocular surface failure. Cornea. 2012;31:1264–1272. 23. Tan DT, Ficker LA, Buckley RJ. Limbal transplantation. Ophthalmology. 11. Dua HS, Azuara-Blanco A. Autologous limbal transplantation in patients 1996;103:29–36. with unilateral corneal stem cell deficiency. Br J Ophthalmol. 2000;84: 24. Shimazaki J, Shinozaki N, Tsubota K. Transplantation of amniotic 273–278. membrane and limbal autograft for patients with recurrent pterygium 12. Basti S, Rao SK. Current status of limbal conjunctival autograft. Curr associated with symblepharon. Br J Ophthalmol. 1998;82:235–240. Opin Ophthalmol. 2000;11:224–232. 25. Miri A, Said DG, Dua HS. Donor site complications in autolimbal and 13. King JH Jr, McTigue JW. The Cornea. World Congress. Baltimore, living related allolimbal transplantation. Ophthalmology. 2011;118: MD: Butterworths; 1965:354. 1265–1271. 14. Rao SK, Rajagopal R, Sitalakshmi G, et al. Limbal autografting: 26. Busin M, Breda C, Bertolin M, et al. Corneal epithelial stem cells comparison of results in the acute and chronic phases ocular surface repopulate the donor area within 1 Year from limbus removal for limbal burns. Cornea. 1999;18:164–171. autograft. Ophthalmology. 2016;123:2481–2488. 15. Moldovan SM, Borderie V, Baudrimont M, et al. Treatment of unilateral 27. Chen JJY, Tseng SCG. Corneal epithelial wound healing in partial limbal limbal stem cell deficiency syndrome by limbal autograft. JFr deficiency. Invest Ophthalmol Vis Sci. 1990;31:1301–1314. Ophtalmol. 1999;22:302–309. 28. Kruse FE, Chen JJY, Tsai RJF, et al. Conjunctival transdifferentiation is 16. Kheirkhah A, Raju VK, Tseng SC. Minimal conjunctival limbal autograft due to incomplete removal of limbal basal epithelium. Invest Ophthalmol for total limbal stem cell deficiency. Cornea. 2008;27:730–733. Vis Sci. 1990;31:1903–1913. 17. Basu S, Sureka SP, Shanbhag SS, et al. Simple limbal epithelial 29. Huang AJW, Tseng SCG. Corneal epithelial wound healing in the transplantation: long-term clinical outcomes in 125 cases of unilateral absence of limbal epithelium. Invest Ophthalmol Vis Sci. 1991;32: chronic ocular surface burns. Ophthalmology. 2016;123:1000–1010. 96–105.