Limbal stem cell deficiency (LSCD) is a disorder of the ocular surface where corneal epithelial homeostasis is disrupted by dysfunction or an insufficient quantity of limbal stem cells (LSCs). The past few decades have ushered in significant advancements in the detection and treatment of LSCD, and recently the lack of clear diagnostic standards was addressed by the Cornea Society’s Limbal Stem Cell Working Group, which established a consensus amongst international experts on LSCD diagnosis and classification (1). Accurate diagnosis and staging is critical for selecting the treatment that is best suited for a patient’s disease severity (2). Treatments can range from medical management alone to surgery. Kim, et al. described cases of LSC dysfunction where medical therapy alone was sufficient to reverse signs of LSCD (3). Corneal epithelial removal with or without amniotic membrane transplantation, conjunctival limbal autograft, conjunctival limbal allograft, keratolimbal allograft, cultivated limbal epithelial stem cells, simple limbal epithelial transplantation and keratoprosthesis surgery have been described as surgical treatments for cases of LSCD; in unilateral cases, autologous versions of LSC transplant can be considered with harvesting from the contralateral eye (1). Mucous membrane graft can also be considered, especially in the presence of keratinization of the lid margin and palpebral conjunctiva (4). In this review, we summarize the current diagnostic paradigm for LSCD.

LSCs reside in the limbus, which is a 1-2mm transition zone where corneal and conjunctival epithelia meet (5). The limbus hosts LSCs in the basal region, in fibrovascular ridges called the palisades of Vogt (6). Based on the widely accepted XYZ hypothesis of corneal maintenance, these LSCs give rise to transient amplifying cells (TACs) that migrate centripetally along the basal layer of the epithelium. Differentiated offsprings of these TACs then undertake anterior migration and further differentiation into mature corneal epithelium (7). The limbus also plays an important role as the barrier between the avascular cornea and vascular conjunctiva (8). Deficiency of LSCs can lead to dysfunction in corneal epithelial homeostasis, resulting in persistent epithelial defects, neovascularization, ulceration, opacification, and conjunctivalization of the cornea (5).

Causes of LSCD are best divided into acquired and non-acquired. LSCD can be acquired by immune-mediated conditions including Stevens-Johnson Syndrome/toxic necrolysis (SJS/TEN) (9, 10), chronic ocular graft vs host disease (GVHD) (11), atopic/allergic ocular surface disease, and mucous membrane pemphigoid (12). Acquired, non-immune mediated causes include direct damage from radiation, contact lens wear, chemical or thermal burns, drug-induced injury such as systemic anti-neoplastic medications, limbal surgeries, infectious ocular diseases, ocular surface tumors, chronic lid margin disease, severe pterygium, and toxic exposures such as mustard gas (1, 13–19). Traumatic etiologies are more likely to cause asymmetrical LSCD (1). Non-acquired causes include genetic diseases such as congenital aniridia, xeroderma pigmentosum, and dyskeratosis congenita (20–22).

A multi-center study in India including 1331 patients identified ocular surface burns as the most common cause of unilateral LSCD (84%). For bilateral LSCD, the most frequent causes were allergic conjunctivitis (29%), SJS/TEN (23%), and congenital aniridia (9%) (23). Despite LSCD placing a significant burden on selected cornea subspecialty clinics, there are limited epidemiological data for LSCD in the U.S (24). A handful of epidemiological studies have been published to date. Goldberg, et al. found an LSCD prevalence of 4.25% over a two year period at a cornea subspeciality tertiary referral center (24). While the authors note the likely impact of referral bias and underdiagnosis prior to arrival at their tertiary center, this prevalence is significantly higher than the estimated prevalence in the U.S. reported by Orphanet rare disease (0.0001-0.0005%) (24). Cheung et al. conducted a single center study of 434 patients and found that congenital aniridia was the most common cause of LSCD accounting for 31% of cases, followed by chemical or thermal injuries accounting for 21% (25). Cheung also found that 70% of their patients had bilateral disease, and the average patient was middle-aged with no sex predominance (25). Haring, et al. studied epidemiological trends in the U.S. specifically for chemical ocular burns and found that children aged 1 to 2 years old were the highest-risk group due to accidental access to dangerous substances, and males were more likely to experience these injuries (26).

Additionally, LSC dysfunction has been described as a potential precursor to, or subtype of, LSCD. LSC dysfunction has been described previously in a series of patients in whom corneal epithelial changes were reversed by medical therapy (3). Kim, et al. proposed that LSC dysfunction can occur due to disruption of the limbal microenvironment, or niche, in which the limbal stem cells may have the potential to be rehabilitated (3). This may occur most commonly with chronic contact lens use or BAK toxicity. If left untreated, these patients can develop persistent damage to the limbal niche, and subsequent irreversible LSCD.

LSCD symptoms can vary from asymptomatic to severe. Photophobia, redness, and foreign body sensation are common (27). Other symptoms may include tearing, dryness or general ocular discomfort and pain. This wide variety of symptoms make LSCD’s presentation non-specific, which necessitates clinical evaluation, including with slit lamp biomicroscopy.

In 2019, the Cornea Society’s Limbal Stem Cell Working Group established a clear definition of LSCD and criteria for diagnosis, classification, and staging. The Working Group first divided LSCD into partial and total LSCD (1). Total LCSD was characterized as loss of all LSCs that causes full conjunctivalization of the cornea. LSCD staging based on clinical presentation was defined as the following: Stage 1 is disease where the central 5mm of the cornea maintains normal corneal epithelium; Stage 2 is disease where the central 5mm of the cornea is affected; Stage 3 is disease where the entire corneal surface is affected (1). These stages are sub-divided further based on the extent of limbal involvement: A) less than 50% limbal involvement, B) greater than 50% but less than 100% limbal involvement, and C) 100% limbal involvement ( Figure 1 ) (1).

Findings on slit lamp examination included late vortex or punctate fluorescein staining, reduced epithelial transparency, epithelial irregularities and recurrent epithelial defects, neovascularization, absence of palisades of Vogt, an opaque or scarred cornea, and an inflamed ocular surface (28). Figure 2 is a previously published descriptive grading scale adapted from Le, et al.’s review on LSCD. This grading scale is useful in determining disease severity but is not widely used clinically (28).

Slit lamp examination remains an important clinical method for detection of LSCD, especially given its wide availability. In 2020, Le, et al. outlined 105 studies where LSCD was diagnosed based on only clinical examination in 62.9% of eyes (2). A confirmatory diagnostic test was used in 25.8% of eyes, with impression cytology (IC) used in 24.2% of eyes, in vivo confocal microscopy (IVCM) used in 0.7% of eyes, and a combination IVCM and IC used in 0.9% of eyes (2). The remaining 11.3% of eyes diagnosed with LSCD had no diagnostic criteria provided in their studies (2). In patients with LSC dysfunction, slit lamp examination is also critical, with the predominant finding being a whorl-like or wave-like epitheliopathy (3). Pathologically, these patients often have a continuous sheet of single clones of late staining epithelium that follow a whorl-like pattern, which may represent a form of metaplasia. Other studies have demonstrated the pathologic presence of goblet cells on the cornea in cases of LSCD resulting from trauma or inflammation (29).

While slit lamp examination is often the initial method to detect LSCD, the use of confirmatory measures is important to review.

Recent years have seen significant advancements in the diagnosis of LSCD. Efforts to establish an international consensus on LSCD diagnostic criteria and staging have helped standardize our definition of LSCD. While slit lamp examination remains the primary method for diagnosing LSCD in most clinical settings, advanced diagnostic tools, such as IVCM, AS-OCT, impression cytology, and OCT-A, have improved our ability to confirm the diagnosis. Though the detection of conjunctival epithelium and goblet cells on impression cytology remains the gold standard for LSCD diagnosis, the use of IVCM, AS-OCT, and OCT-A are important non-invasive tools for supporting the diagnostic approach to LSCD.

Such tools are not only important for diagnosing LSCD, but also may help guide treatment. For example, three-dimensional mapping by either OCT or IVCM may help identify deep limbal lacunae and determine if there are residual limbal epithelial cells present.

Furthermore, diagnosis and staging of LSCD may be important for determining whether medical or surgical therapy is appropriate. For example, if LSCD is identified, medical therapy can immediately be initiated to optimize the ocular surface including frequent lubrication, serum tears, or therapeutic contact lenses (18). Those with confirmed LSCD may also be candidates for surgical treatments including corneal epithelial removal with or without amniotic membrane transplantation or one of various forms of limbal stem cell transplantation, depending on unilateral versus bilateral involvement as well as disease severity. The staging and diagnosis of LSCD is therefore critical to determining which treatment options may be indicated.

There is an ongoing need for improved diagnostic modalities for LSCD. Quantifiable markers such as basal cell density, sub-basal nerve density, and epithelial thickness are useful for diagnosing LSCD but often require a skilled operator to reliably obtain them. Further improvements in image resolution, field of view, and image analysis software will be important for expanding the use of ICVM and AS-OCT in the diagnosis of LSCD. Having clear diagnostic criteria for LSCD and being able to diagnose it early with accurate severity grading is critical to ensuring patients with LSCD receive appropriate treatment. The current and ongoing research on the diagnosis, grading, and management of LSCD will help achieve this goal.