Autoimmune disorders represent a wide range of conditions in which the immune system is unable to differentiate normal from foreign tissue and attacks healthy cells as a result. While many of these diseases can cause ocular pathology, Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) and mucous membrane pemphigoid (MMP)/ocular cicatricial pemphigoid (OCP) yield the most devastating destruction of the ocular surface.

Stevens-Johnson syndrome and toxic epidermal necrolysis represent a spectrum of life-threatening immune-mediated mucocutaneous reactions. SJS/TEN are quite rare, with a reported incidence ranging from 1.89 cases per million per year in all-comer populations to 1 per thousand cases per year in patients with HIV [2]. This variation in incidence reflects the contribution of several factors that increase the risk of SJS/TEN in certain groups of patients. Specifically, the HLA genotypes B*5801 and B*1502 have been associated with increased risks of reactions to specific drugs. In addition, patients who are treated with many medications, slow acetylators [3], immunocompromised, or have brain tumors treated with both radiotherapy and anticonvulsants [4] have also been shown to be more likely to get SJS/TEN. While the pathophysiology is not entirely understood, current theories attribute these disorders to drug- or infection-related hypersensitivity reactions involving cytotoxic T lymphocyte-induced apoptosis of keratinocytes. Research has also implicated a myriad of cytokines such as TNF-alpha, FasL, and granulysin contributing to the aggressive nature of cell death. Classically, drugs like allopurinol, phenytoin, carbamazepine, oxicam NSAIDs, and sulfonamides [5] and infections such as mycoplasma pneumonia and the herpes simplex virus [2] are the frequent causes of SJS/TEN. These conditions are clinically characterized by widespread skin erythema and blisters in conjunction with mucous membrane erosions. The primary distinction between SJS and TEN is in the extent of these mucocutaneous lesions with SJS affecting less than 10 % body surface area and TEN affecting more than 30 % body surface area. In 2007, Yip et al. [6] examined 117 patients with SJS/TEN over a 9-year period. They noted that 69 % percent of these patients developed acute ocular involvement. Specifically, the SJS/TEN will present with an acute cicatrizing conjunctivitis with corneal ulceration and membrane formation [2]. In the chronic phase, 46 % of patients will demonstrate severe dry eye, while other patients have limbal stem cell deficiency, symblepharon, and trichiasis. These changes lead to severe damage to the ocular surface, which can result in bilateral corneal blindness.

Mucous membrane pemphigoid is a group of rare chronic autoimmune disorders characterized by subepithelial blisters predominately affecting mucous membranes, and to a lesser degree the skin, across the body. Ocular cicatricial pemphigoid is a subtype of MMP in which the ocular surface is primarily affected, classically in a chronic cicatrizing conjunctivitis. In 1995, two epidemiologic studies conducted in Germany and France suggested an annual incidence of the disease of 0.87–1.16 per million people [7]. However, ophthalmic literature reports a much more common incidence ranging from 1 in 60,000 to 1 in 8,000 patients [8]. Typically, MMP/OCP is seen in patients in their fifth or sixth decade of life, with up to a 3-to-1 female-to-male ratio and follows a relapsing-remitting course [9]. The disease is caused by autoantibodies directed at components of the hemidesmosomes located on epithelial keratinocytes of the basement membrane zone and has been associated with various HLA alleles including HLA-DR4 and HLA-DQB1*0301 [9]. This pathophysiologic mechanism is important in the diagnosis of OCP, as a biopsy involved tissue will show a linear deposition of IgG, IgA, or C3 at the level of the basement membrane on direct immunofluorescence [10]. OCP presents as an asymmetric, though eventually bilateral, chronic progressive cicatrizing conjunctivitis. The early in the course of the disease is characterized by nonspecific conjunctival hyperemia with tearing and ocular surface epitheliopathy. As the OCP progresses, however, patients begin to demonstrate subepithelial fibrosis, fornix foreshortening, symblepharon, and ankyloblepharon [8]. Eventually, significant xerosis with limbal stem cell dysfunction, trichiasis, and ocular surface keratinization result in an opacified cornea and severe vision loss. Messmer et al. examined the outcome of 28 patients with OCP over a 12-year period. They noted that upon presentation, 38 % of patients already had a visual acuity of 20/200 or worse. Despite medical therapy, 53 % of patients progressed with further vision loss. At the end of the study only 35 % of patients demonstrated useful reading vision [11].

In 2008, Saito et al. [12] examined the corneas of patients with SJS, OCP, and alkali burns excised during the chronic phase of each disease. The authors noted that despite a clinically quiescent appearance, the cornea stroma demonstrated abnormal keratocytes and persistent inflammation with an increase in CD14-positive and CD45-positive cells. They also identified elevated concentrations of macrophage inflammatory proteins 1-alpha and 1-beta and monocyte chemoattractant protein 1 as compared to normal cornea stroma. Dudenhoefer et al. [13] studied histopathological specimens of explanted keratoprostheses in patients with autoimmune versus non-autoimmune pathologies. They reported that patients with autoimmune pathologies had a massive inflammatory cell infiltrate and stromal necrosis as compared to essentially normal corneal tissue in their noninflammatory counterparts. In addition, studies outside of the ophthalmic literature have shown that patients with autoimmune diseases have increased rates of fungal colonization of oral mucosa as well as more bacterial endotoxin present at the site of aseptically loosened joint prosthetics [14]. These studies suggest that the inflammatory nature of autoimmune-related conjunctivitides alters corneal tissue at the level of the cellular, microbiological, and immunological microenvironments and directly leads to the failure of standard treatment methodologies in these patients.

The vision loss from ocular surface dysfunction and subsequent corneal opacification in patients with SJS/TEN and MMP/OCP frequently require radical efforts for visual rehabilitation. These options range from limbal stem cell transplantation and ocular surface reconstruction to artificially produced keratoprostheses and biologic keratoprostheses. With regard to autoimmune-related cicatrizing conjunctivitis, each of these surgical options has its indications and limitations.

Ocular surface reconstruction attempts to repopulate the surface of the eye with limbal stem cells and clear corneal tissue in order to reapproximate healthy cellular functioning. Patients with bilateral limbal stem cell deficiency typically require cells harvested from either living-related donors or cadaver eyes with the use of long-term immunosuppression after the transplant. In general, this procedure can yield positive outcomes, with one study reporting success rates ranging from 33 % for cadaveric transplants to 89 % for living-related transplants at 32 months of follow-up [15]. However, prior to considering this technique, all concurrent comorbidities must be addressed. Such issues include glaucoma, eyelid abnormalities, ocular surface lubrication, and ocular surface inflammation. In 2010, Gomes et al. set out to explore these issues [16]. The authors determined that dry eye was the most important prognostic factor for graft survival. To address this, the group performed a labial mucous membrane and minor salivary gland transplantation with good outcomes. As severe sicca syndrome often accompanies autoimmune condition, this step might prove invaluable in success of these patients. In addition, ocular surface inflammation must be strictly controlled via both topical and systemic immune suppression to avoid graft rejection and failure. In spite of these options, though, in severe SJS/TEN and MMP/OCP ocular surface dysfunction is advanced enough to preclude this as a viable treatment. More recently, Biber et al. developed a technique whereby patients with limbal stem cell and conjunctival deficiency associated with SJS and OCP received a combined living-related conjunctival limbal allograft (lr-CLAL) and keratolimbal allograft (KLAL) [17]. Of these patients, 63 % received a staged keratoplasty and 17 % underwent a staged Boston type 1 keratoprosthesis and were followed over an average of 43.4 months. The authors demonstrated a successful improvement in visual acuity in 75 % of patients at their last follow-up. Specifically, of those patients who received a Boston keratoprosthesis, 90 % had improved vision with 70 % achieving a visual acuity of 20/125 or better at the last follow-up.

The Boston type 1 keratoprostheses is currently the most frequently used artificial corneal replacements today, with more than 3,500 prosthetics implanted worldwide [1]. Developed at the Massachusetts Eye and Ear Infirmary, this keratoprosthesis is a collar-button design that incorporates a donor corneal button sandwiched between a central PMMA optic and a locking backplate. These implants are intended for use in patients with a relatively well functioning tear film, blink reflex, and ocular surface. In general, the Boston type 1 keratoprosthesis is a successful and well-tolerated device. Recent studies focusing on this design have shown retention rates ranging from 80 to 85 % with patients achieving a final visual acuity of 20/200 or greater in 50–82 % [18]. However, this overall prognosis varies drastically based on a patient’s initial diagnosis which necessitated the implantation of such a prosthetic. Many studies have shown that while noninflammatory conditions like graft failures and traumas have the best overall prognosis, autoimmune diseases fare much more poorly [19]. In 2011, Sejpal et al. examined 23 eyes with limbal stem cell deficiency of various etiologies which required a Boston type 1 keratoprosthesis (28 total implants). Of this patient set, seven devices were removed—five of these patients had SJS, while another patient had OCP. The authors determined that when the patients with SJS were excluded from analysis, the remaining patients in the study had an 89 % retention rate. From this, they concluded that patients with SJS should not be considered candidates for implantation of the Boston type 1 keratoprosthesis [20].

In contrast to the type 1, the Boston type 2 keratoprosthesis was designed for patients with significant symblepharon, ocular surface dryness and keratinization, and absence of normal lid function. This device is the same as its namesake counterpart with the exception of a 2-mm-long anterior nub built to extend through the lids or a permanent tarsorrhaphy. A large retrospective study out of the Massachusetts Eye and Ear Infirmary followed 29 eyes that underwent an implantation of a Boston type 2 keratoprosthesis over an average of 3.7 years. Fifty-one percent of the patients in this study had an initial diagnosis of OCP and 41 % had SJS. The results demonstrated that vision improved to 20/200 or greater in 79 % and 20/30 or greater in 34 % of patients. In addition, this visual acuity was maintained for more than 1 year in 57 % of patients in the study. Overall, 58 % of the implanted keratoprostheses were retained without expulsion or replacement for a total of 107.9 person-years with a higher cumulative functional success rate for SJS than for OCP. Given these outcomes, the authors proposed that the Boston type 2 keratoprosthesis is a viable option for patients with severe autoimmune disease [21].

Neither the Boston type 1 or type 2 keratoprosthesis is free from complications however, and these failings are highlighted in patients with autoimmune disease. For instance, irrespective of the type of keratoprosthesis, patients with SJS/TEN and MMP/OCP require more major repairs, minor repairs, and YAG laser membranectomies than their noninflammatory counterparts. Furthermore, they demonstrate elevated risks of endophthalmitis, advance glaucoma, and retinal detachment [22]. Many efforts have been taken in recent years to reduce these risks, with comparative studies suggesting an improvement in overall prognosis over a 7-year period [23]. Specifically, a 16.0–18.0-mm, 9.8-mm base curve, Kontur bandage contact lens or a Boston scleral lens is now placed long term to control tissue necrosis and resultant aqueous leaks, glaucoma drainage implants are being placed in extra-orbital locations to improve functioning, and a long-term regimen of a third- or fourth-generation fluoroquinolone and vancomycin is employed to minimize risk of endophthalmitis in this at-risk patient group [23].