Allogeneic HCT offers the best chance of cure for several malignant hematological as well as non-malignant disorders like bone marrow failure, hemoglobinopathies or immunodeficiencies. Currently. over 30,000 allogeneic HCT are performed annually worldwide with over 18,000 in Europe in 2020 (1).

GVHD is one of the most important causes for non-relapse mortality post-transplantation. The current understanding of the pathophysiologic concepts and therapeutic targets has tremendously expanded during the last 20 years and recently been summarized in three excellent reviews (2–4). Chronic GVHD is the most common long-term complication after allogeneic HCT with an important impact on survival, morbidity, and quality of life. Traditionally, acute and chronic GVHD was differentiated depending on the time of the initial manifestation before or after 100 days post-transplant. These criteria were revised in the 2005 and 2014 National Institute of Health (NIH) Consensus Conference, introducing new criteria/definition for acute and chronic GVHD (5–7). Acute GVHD is defined as an immediate multi-organ inflammatory syndrome following HCT primarily affecting the skin, liver, and digestive tract, whereas chronic GVHD is a pleiotropic, multi-organ syndrome characterized by tissue inflammation and fibrosis that involves multiple sites including the skin, lungs, liver, gastrointestinal tract, mouth, genitalia, and eyes (5–8). Accordingly, the diagnosis of chronic GVHD requires at least one diagnostic sign of chronic GVHD or a distinctive manifestation plus a pertinent biopsy or another test (e.g., Schirmer test, evaluation by an ophthalmologist) showing or confirming chronic GVHD (Table 1).

After the first HCT in 1968 survival rates have increased in the last decades, due to human leukocyte antigen (HLA) matching, continuously improved preconditioning protocols and immunosuppressive regimen (7, 9, 10). Both, acute and chronic GVHD occur in about 30%–70% of patients after HCT depending on transplant regimens and GVHD prophylaxis strategies (7, 11). A variety of risk factors for GVHD related to donor as well as to recipients’ characteristics have been identified. The most important are the degree of histocompatibility, the source of hematopoietic progenitor cells, sex mismatch (transplantation from female donor to male recipient), the intensity of conditioning and immunosuppression, the age of donor and recipient and for chronic GVHD prior acute GVHD (2, 3, 8, 12–14).

Different criteria for the diagnosis of oGVHD have been proposed in the last decades (8). The original NIH criteria defined new onset of dry eye after HCT documented by low Schirmer test values with a mean value of both eyes <5 mm at 5 min or a new onset of keratoconjunctivitis sicca by slit-lamp examination with mean values of 6 to 10 mm at 5 min on the Schirmer test as sufficient for the diagnosis of chronic oGVHD if accompanied by distinctive manifestations in at least one other organ (6). An international consensus group proposed criteria based on Ocular Surface Disease Index (OSDI), Schirmer test score without anesthesia, corneal fluorescein staining and conjunctival injection (15). A score of 4–5 and ≥ 6 indicates probable or definite oGVHD, accordingly (15).

Acute GVHD has been reported in 40%–50% of HCT patients (16). Ocular affection in acute GVHD is quite rare and has been reported in about 7.2% after HCT (17, 18). On the other hand, occurrence of chronic oGVHD was observed in 30%–60% in the further course after HCT (19, 20), and in 60%–90% of patients with systemic GVHD (7, 10, 19, 21, 22). Lower incidences have been found in Asian studies (23–25). The mean latency of oGVHD after HCT is about 1.5 years (26). Cumulative increase of incidences over time after HCT has been reported, with a prevalence of 16% by 100 days and 35% after 2 years (21). In children, symptoms consistent with chronic oGVHD have been found at highly variating rates from 4% up to 62% (27–33). In a large prospective study, a total of 29.4% of patients with chronic oGVHD were identified using the NIH consensus criteria (34).

In the absence of overt ocular symptoms and signs during the acute disease stage, diagnosis may be delayed. Disease may partially mimic other immune-mediated inflammatory processes of the ocular surface. While no pathognomonic symptoms or clinical signs of oGVHD have been defined, certain combinations of findings are frequently present, and are provided within several recent publications (15, 67; Table 2). Key features of disease are new onset of refractory dry eye, being the most frequent manifestation (40%–70%), and secondary ocular surface damage (52). Patients suffer from diverse symptoms of the autoinflammatory reaction (particularly dry eye), including irritation, pain, burning, dryness, itchiness, blurred vision, foreign body sensation, photophobia, and redness (70, 74, 75). Visual disturbance may be the consequence from corneal higher order aberrations resulting from corneal pathology (76).

Severe ocular discomfort from dry eye, corneal epitheliopathy by means of fluorescein staining and vision loss are resulting in impaired quality of life (22). Patients with oGVHD had worse quality of life than patients without ocular involvement (77). In clinical studies, symptoms are quantified using validated QOL instruments such as Ocular Surface Disease Index (OSDI), National Eye Institute Visual Function Questionnaire (NEI-VFQ-25), and Symptom Assessment in Dry Eye (SANDE). Respective studies show that disease impact on QOL was comparable to herpetic uveitis or retinal vein occlusion (22).

By en-face evaluation, photophobia, pseudoptosis, frequent blinking or periorbital hyperpigmentation may be seen. Findings at the lid margin are common in oGVHD. Blepharitis and Meibomian gland dysfunction (50%) are probably the first signs of disease. Subsequently, atrophy, irregularity and keratinization of the eyelid margin may occur.

Conjunctival involvement mostly manifests as hyperemia (Figures 2A,B) and chemosis. Qualitative and quantitative alterations of the tear film are common, probably with serosanguineous exudation (78). In severe course, pseudo-membrane formation may be observed. Conjunctival fibrosis and subsequent scarring (Figures 2B,C) may not only result in loss of goblet cells, but also to entropion, distichiasis and trichiasis. Therefore, thorough subtarsal inspection is mandatory to determine the pathology also under the upper lid. Indeed, subtarsal fibrosis may correlate with worsening of corneal epitheliopathy. Inflammation and staining of the superior tarsal and bulbar conjunctiva with alteration of the superior limbal epithelium may be present (superior limbal keratoconjunctivitis; SLK-like appearance). The ICCGVHD grading system for conjunctival involvement in oGVHD is shown in Table 3 (15, 67).

Morphological abnormalities of the cornea involve punctate keratopathy (Figure 2A), erosions, or filamentary keratitis (Figure 2D) in the more severe cases. Further, limbal stem cell deficiency, Bowman abnormalities, stromal thinning, ulceration (Figure 2E), scarring, calcification and neovascularization may appear. Corneal perforation (Figure 2F) may be secondary to epithelial barrier dysfunction and microorganisms (herpes simplex virus or bacteria), or as sterile “melt” probably in the setting of immunosuppression (80). According to previous reports, corneal ulceration or perforation is found in about 5% of cases (69).

Further, signs of episcleritis or scleritis, secondary cataract (10%, mostly from steroids), or glaucoma (also including steroid-induced ocular hypertension) may appear (81). Within a cohort of 635 patients undergoing HCT, 7.6% had secondary posterior eye segment complications, e.g., retinal hemorrhage, cytomegalovirus retinitis, or uveitis (40, 82).

A thorough ophthalmological examination is essential in patients with (suspected) oGVHD (83). For assessing the course of disease and response to treatment, a standardized documentation of ocular findings should be performed (Table 4). Assessing ocular findings at baseline before HCT and during follow-up visits allow to early detect worsening of the ocular surface (52, 68, 71, 84). A minimal set of data as visual acuity, slit lamp findings and intraocular pressure should be collected at each visit. Further investigations should be performed as appropriate.

The Schirmer test I (without topical anesthesia) and II (with prior topical anesthesia) allows to assess the tear production during a defined time of 5 min. A folded filter paper strip is placed in the temporal third of the lower lid margin and the length of the wetting is measured (52). The Schirmer test without anesthesia is also included in the oGVHD (ICCGVHD) consensus group diagnostic criteria (67). While the Schirmer test is useful for diagnosing disease, it was removed from scoring recommendations, as values were not useful for follow-up due to poor correlation with symptom change (64). Due to its low reproducibility it has been removed in the revision of the 2005 NIH criteria and not been included in the 2014 NIH severity scoring, nor in the 2016 Japanese and Asian diagnostic criteria for dry eye disease (23, 63, 85).

Esthesiometry allows to assess the corneal sensitivity, which may be decreased due to pre-conditioning irradiation and neurotrophic keratopathy in patients with oGVHD (86–89).

Impairment of conjunctival and/or corneal epithelial integrity can be depicted with vital dye staining. Fluorescein is commonly used to evaluate the corneal staining according to the Oxford grading scheme (Figure 3) and/or the NEI grading for corneal and conjunctival staining (Figure 4) (52, 74, 92). Fluorescein dye is disclosing any disruption in superficial cell tight junctions, or defective glycocalyx of damaged epithelial cells (52). Additional dyes as Bengal rosa or lyssamine green can additionally be used in selected patients (90).

After fluorescein installation, the tear film break-up-time (TBUT) can be evaluated at the slit lamp (52, 70). A decreased TBUT indicates qualitative tear film impairment primarily due to Meibomian gland dysfunction (93).

Tear film osmolarity measurements reveal increased values in oGVHD (74, 94–96) and may be used as an additional factor in therapeutic decisions (19). The tear film osmolarity is also used in the ICCGVHD criteria (23, 95).

Meibomian gland imaging enables the assessment of Meibomian glands, which are often impaired in patients with oGVHD (52, 97–99).

In patients with keratitis, viral and/or microbial tests from corneal smears should be considered to identify viral (mainly by herpes simplex or varicella zoster virus), bacterial or fungal keratitis. The risk for infectious keratitis may be increased in patients under corticosteroid treatment.

In vivo confocal microscopy can be used as a diagnostic tool in patients with oGVHD to image epithelial cell density, epithelial dendritic cells and other inflammatory cells, subtarsal fibrosis and conjunctival changes (23, 52, 59, 99–103).

The use of anterior segment photography may be considered to document ocular findings (e.g., staining of ocular surface, conjunctival scarring/fibrosis, blepharitis). It may especially be useful for follow-up comparison of clinical course (75, 104).

Conjunctival impression cytology enables identification of epithelial cell necrosis, keratinization, goblet cells loss and also HLA-DR expression (83, 105, 106). As an alternative, Brush cytology is also a minimally invasive procedure to harvest ocular surface epithelium and inflammatory cells and to monitor pathological progress (88, 107), but interpretation might be difficult due to mechanical alteration of the harvested cells.

Tear film biomarkers (cytokines) can either directly be measured with specific antigen tests (e.g., MMP-9) (108) or (currently mainly for research purpose and not in clinical routine) by performing proteomics from tear fluid or tear-film soaked Schirmer stripes (52, 109). In eyes with oGVHD a variety of cytokines are differently expressed. Especially nucleic acid binding and cytoskeletal proteins are upregulated, while the most extensively downregulated proteins belong to an array of classes including transfer and receptor proteins, enzyme modulators, and hydrolases (109).

Tear flow cytometry is a novel approach, currently used mainly for research purpose, that allows differentiation of cells non-invasively from tear samples (51).

Histopathology may confirm the diagnosis of oGVHD. However lacrimal gland biopsies should not be performed routinely due to the increased risk of further impairment of its function. Previous investigations found mononuclear infiltration, loss of acinar lobules and fibrosis of the lacrimal gland in oGVHD (11, 110, 111). Also, conjunctival biopsies are not performed routinely but may be considered in selected patients, e.g., to rule out malignancy. In conjunctival specimen of oGVHD, lymphocyte exocytosis, vacuolization of the basal epithelium, and epithelial cell necrosis, similar to changes that are observed in other organs, have been found (11, 110, 111). Furthermore, T cells—probably driving alloreactivity in GVHD—have been found in conjunctival biopsies (112).

Ocular surface inflammation and dryness may have a relevant impact on the quality of life and activities of daily living in patients with oGVHD (22). Different validated questionnaires are used to quantify symptoms, to assess the burden of disease and to track response to treatment (11, 22). The Ocular Surface Disease Index (OSDI), consisting of 12 patient-related questions of dry eye, and the Symptom Assessment in Dry Eye (SANDE) are commonly used questionnaires to assess symptoms in these patients (22, 72, 113–115). Alternatively, or additionally, the glaucoma symptom scale (GSS) may be used (116). On the other hand, the National Eye Institute Visual Function Questionnaire (NEI-VFQ-25) allows to assess vision related quality of life (22). Saboo et al. evaluated patients with oGVHD using the NEI-VFQ-25, OSDI and SANDE questionnaires and found a relevant impact of this disease on quality of life, that is comparable to other eye diseases as for example herpetic uveitis (22).

An intensive lubrication for dry and inflamed ocular surface is essential in oGVHD (5, 83, 117). A variety of artificial tears, viscous eye drops, and viscous ointments are available and only limited data on specific preferences for oGVHD is available. In any case, preservative-free formulations should be preferred to avoid the negative impact of preservatives on the epithelium, especially if applied at high frequencies (118). Hyaluronic acid eye drops allow stabilization of the tear film and improvement of epithelial wound healing, ocular symptoms, and visual acuity (53, 117). Increasing the lubrification may also reduce the concentrations of proinflammatory cytokines on the ocular surface (5, 119). Mucolytic eye drops, i.e., topical N-acetylcysteine 5%–10%, should be considered in filamentary keratitis, which is often observed in eyes with a very dry ocular surface (5, 120).

Reducing ocular surface inflammation is a key concept in the management of oGVHD. Topical corticosteroids are effective in treating dry eye in these patients (52, 66, 75). However, due to their probable adverse effects and risks, their application over a longer time periods, or at high dosages and/or with highly potent formulations should be avoided, or regular ophthalmological checks (intervals depending on corticosteroid dosage and duration, eye pressure and lens status) be instituted. Potential risks include cataract formation, infections, ocular hypertension/glaucoma, impaired epithelialization and impaired corneal wound healing (5, 11, 66). Nevertheless, they are used commonly in oGVHD patients (5, 52, 121, 122). However, topical corticosteroids are not able to sufficiently control oGVHD in about half of the patients (7). Low-dose/−less potent topical corticosteroids or their analogs seem to be less effective in patients with oGVHD compared to dry eye patients without oGVHD (11, 123). As an anti-inflammatory treatment option, cyclosporine (CsA) eye drops are used in patients with treatment refractory dry eye disease. CsA acts as a calcineurin inhibitor and suppresses T-cell activation (11, 124), and its efficacy has also been proven in patients with oGVHD (11, 125). Hereby, it reduces ocular surface inflammation, increases conjunctival goblet cell density and tear production and improves symptoms of dry eye (5, 75, 125–129). If treatment is initiated before HCT, it probably reduces the risk for oGVHD manifestation (130). However, a reduced tolerance (burning sensation) of topical CsA may limit its use in some patients. Furthermore, tacrolimus eye drops or ointment have been studied in patients with oGVHD, probably allowing corticosteroid sparing (11, 131–133). Tacrolimus ointment may also be applied to the eyelids as an off-label treatment. Although topical non-steroidal anti-inflammatory drugs (NSAIDs) are also used in oGVHD, there is no evidence for their efficacy.

Based on several uncontrolled trials in oGVHD and in analogy to other forms of dry eye disease, autologous serum eye drops are also used in patients with oGVHD, especially in severe cases (86, 117, 134). Although the exact mechanism of action is not known, the high concentration of several growth factors combined with anti-inflammatory effects are suggested to improve healing of epithelial defects (129, 135, 136). Systemically applied cyclosporin A or mycophenolic acid might also be detectable in serum eye drops (137) and could contribute to the observed beneficial effect. Patients impaired condition to donate blood (poor venous access, severe anemia, active infection, low body weight, cardiovascular comorbidities) as well as regulatory restrictions are potential obstacles that prevent access to this therapy. Other options that have been reported are allogeneic serum eye drops (136), cord blood sera (117, 138, 139) and platelet lysate (116, 140). None of these options have become more widely available yet due to a couple of logistics and regulatory reasons.

Improving the lipid layer of the tear film with viscous eye drops and ointments, improving the Meibomian gland outflow with eyelid massage and eventually lipid sprays reduce evaporation of the tear film. The evidence for eyelid massage in oGVHD is low and the mechanical friction might even be counterproductive in oGVHD with affection of the corneal epithelium. Occlusive eye wear (52, 141) and an improvement of environmental factors as air humidity may also be helpful (117, 142).

Systemic treatment with oral muscarinic agonists as pilocarpine or cevimeline may increase tear production (117, 143, 144). As adjuvant treatment approaches, secretagogue eye drops as diquafosol and rebamipide may be used in patients with oGVHD (52, 101, 145). They stimulate secretion of aqueous and mucin and improve wound healing of the corneal surface (5, 101, 146).

Reduction of the lacrimal drainage is a further approach to improve the tear film (11). Here, collagen or silicone punctal plugs (Figure 5A) may be inserted into the lacrimal ducts, or permanent punctal occlusion by thermal cauterization may be considered (147, 148). It has been speculated that reducing the tear drainage might result in a pooling of pro-inflammatory cytokines and increase damage of the ocular surface and patient discomfort (149). Positive effects of punctal occlusion predominate in the clinical situation (86, 147).

The use of scleral lenses (Figure 5B) in patients with severe oGVHD has been shown to reduce ocular symptoms and especially ocular pain (83, 150) and improve visual acuity due to their uniform surface (11, 83, 107, 119, 150–155). These gas-permeable lenses cover most of the ocular surface, vault the cornea and limbus providing a fluid reservoir between the cornea and the lens (83). Furthermore, they protect the ocular surface from mechanical “scratching” from blinking (83). In a study by Schornack et al., most patients were still on scleral lenses after a 32-month observation period, indicating a relevant patient satisfaction (152). High costs, inadequate fitting, discomfort with blinking may be potential drawbacks (119). As an alternative to scleral lenses also soft contact lenses have been investigated in oGVHD (156), but may potentially bear a higher risk of infection (83).

Especially in eyes with severe oGVHD, epithelial defects or even corneal melting may occur, due to the very dry and inflamed ocular surface. In this situation, infectious prophylaxis with topical antibiotics should be taken into consideration (19). In patients with extended wear of contact lenses (especially soft contact lenses and topical corticosteroid treatment) topical antibiotic prophylaxis should be considered (157). Furthermore, topical antibiotic ointments or eye drops but also systemic tetracyclines (e.g., doxycycline or minocycline) may be considered in patients with blepharitis as a sign of bacterial superinfection of the eyelids (11, 52, 75, 158).

Systemic treatment of oGVHD is absolutely indicated if severe oGVHD cannot be controlled with topical treatment alone.

High dose corticosteroids (methylprednisolone 1 mg/kg) remain the mainstay of initial systemic treatment of chronic GVHD, either given alone or in combination with calcineurin inhibitors, especially in high-risk disease (159). Second line treatment is indicated in case of steroid-refractory chronic GHVD with an increasing number of treatment options (160). Up to now, there is no standard yet (161). Levels of evidence for efficacy and treatment costs vary considerably and numbers of patients reported for eye response are usually low (162). Extracorporeal photopheresis (ECP) has been reported to resolve or improve eye manifestation in 30% compared to 7% with standard therapy alone by Flowers et al. (163). Other studies could confirm these results in similar or higher magnitude. Recently, ruxolitinib (Janus kinase 1/2 inhibitor; FDA and EMA) and belumosudil (inhibitor of Rho-associated coiled-coil-containing protein kinase 2; FDA) have been approved for treatment of steroid-refractory chronic GVHD. Both have been shown to be effective in a proportion of patients with oGVHD. In the randomized open-label REACH3 trial overall response was 26% with ruxolitinib versus 10.8% with best available treatment (164). Belumosudil was studied in the phase 2 ROCKstar trial mainly in patients with advanced, steroid-refractory chronic GVHD with a remarkable overall response rate of 42% (14% complete responses, 28% partial remissions) (165). In contrast, there are no conclusive data with the third FDA-approved agent ibrutinib (inhibitor of Bruton’s tyrosine kinase) in oGVHD (166). Other agents that are frequently used are sirolimus (mTOR inhibitor), bortezomib (proteosome inhibitor), imatinib (tyrosine kinase inhibitor) and low-dose methotrexate (162). However, there are no randomized controlled trials that evaluated the effect of systemic treatment specifically on oGHVD, or that investigated superiority of one agent to another.

Several new systemic therapeutic principles are tested in preclinical studies including bromodomain inhibitors (167) and SYK inhibition by entospletinib (168).

Currently, no specific treatment strategy is available for fibrosis. Given the pathophysiology of chronic oGVHD, anti-inflammatory and anti-fibrotic treatment regiments might be beneficial. Topically, corticosteroids may have some local antifibrotic effect, but clinical relevance is unknown, and risks do not justify prolonged application. TGF-b signaling inhibition (tranilast) may be useful (169, 170). In contrast to topically applied agents, systemic DMARDs therapy is commonly recommended for severe oGVHD not properly responding to topical agents, as untoward side effects may occur. Agents such as corticosteroids and steroid sparing agents may be applied, including ciclosporin, tacrolimus, sirolimus, mycophenolate mofetil, and particularly B cell blockade with rituximab. Case reports document the value of amniotic membrane transplantation (AMT) for preventing excessive fibrosis (171).

Punctal occlusion with punctal plugs has been shown to be safe to treat severe dry eye in oGVHD (147) and is often used. In rare cases plugs are not supported or extruded repeatedly. In such situations permanent surgical occlusion is possible. Yaguchi et al. described their method of punctal cauterization with a high-temperature sterile disposable cautery device in 23 puncta from 10 oGVHD patients (148). They achieved a 100% anatomical success without recanalization after 1 year and reported no surgical complications. Several other methods for surgical punctal occlusion in other etiologies of dry eye disease have been described, including thermal cautery, diathermy, laser coagulation and punctal suturing (172–177).

Inflammation and tear deficiency in oGVHD can lead to severe corneal ulcerations (87, 178). In such situations, temporal or complete temporary tarsorrhaphies or botulinum toxin A induced protective ptosis (179) are good options to protect the cornea and gain time when systemic immunomodulatory treatment is initiated or escalated and not fully effective yet. Yeh et al. described a patient with oGVHD in whom even tarsorrhaphy and amniotic membrane transplantation (AMT) were not enough, and eventually the eye had to be eviscerated (180).

Amniotic membrane transplantation (Figure 5C) is a surgical procedure that may help to prevent or stop corneal melting by reconstructing the ocular surface and supporting the epithelialization of the cornea (181–184). Epithelial recovery and suppression of inflammation may be achieved due to the contained cytokines and growth factors, additionally the amnion membrane acts as a mechanical barrier for frictional forces (184–187). Indeed, AMT has also successfully been used in progressive corneal ulcers in oGVHD patients (171, 188–190). However, only limited data are available about its success rate up to now. In deep corneal ulcers or descemetocele with pending perforation, cyanoacrylate glue may be an option to avoid or delay more invasive corneal surgery (80, 189). Conjunctival (Gundersen) flap may be another option to cover a corneal ulcer or a fresh corneal transplant. Xu et al. described four oGVHD patients in whom they combined tectonic penetrating keratoplasty with conjunctival flaps (191). Furthermore, Pellegrini et al. reported on one patient receiving a Gundersen flap for impending perforation in their case series of 283 patients with HCT (192).

Despite intensive topical and systemic treatment and tarsorrhaphy and/or AMT, corneal perforations might still occur in severe oGVHD. In such situations, keratoplasties might be required. One option is to perform an urgent tectonic keratoplasty (Figure 5D) with the primary aim of saving the eye and gaining time to escalate the anti-inflammatory treatment. Another possibility is to perform a penetrating keratoplasty with the aim of restoring vision and globe integrity at the same time. Corneal transplant diameters from only few millimeters to large may be used for such keratoplasties depending on the individual need. Sinha et al. determined that the prevalence of corneal perforation in patients with oGVHD was 3.7% (193). Zhang et al. reported 14 corneal perforations in patients with oGVHD during an observation period of 59 years at 4 large centers (80). They all were initially glued and 8 needed penetrating keratoplasty, which had diameters of 2 to 9.5 mm. The best corrected visual acuity outcomes at last visit were 20/100 or better in 5 patients (36%), and hand motion or worse in 7 patients (50%) (80). Xu et al. reviewed 198 oGVHD patients within an observation period of 9 years and identified 9 eyes of 7 patients with corneal perforation necessitating penetrating keratoplasty (trepanation diameters of 2 to 8 mm were used). Only two eyes of two patients achieved a final best corrected visual acuity of 20/100 or better (191). Sometimes even repeat keratoplasty cannot prevent perforations and re-establish functional visual acuity, reason why we had to perform a through-lid Osteo-Odonto-Keratoprosthesis (OOKP; Figure 5E) in one patient (194) and Osteo-Keratoprosthesis (OKP) in another. The outcome was successful in both patients with a best corrected visual acuity of 20/32 or better. Liu et al. mentioned one oGVHD patient in their 10-years review on 36 patients with OOKP (195). Furthermore, Orive Bañuelos et al. also described an oGVHD patient who received a Boston keratoprosthesis Type II after several corneal perforations with repeated keratoplasties. As a further complication, probably related to the keratoprosthesis surgery, two cyclophotocoagulations had to performed. The final visual acuity was 20/20 but the visual field revealed glaucoma related damage (196). OOKP and OKP are high risk procedures that are not commonly performed but might sometimes be the last resort to restore vision in selected patients.

Chronic conjunctival inflammation and subepithelial fibrosis, are often found in oGVHD and can eventually lead to progressive conjunctival scarring with entropion and trichiasis. In combination with keratoconjunctivitis sicca these complications can be devastating for the ocular surface, reason why cicatricial entropion and trichiasis have to be treated without delay (197). Komai et al. described the cultivated oral mucosal epithelial transplantation (COMET) as a method to treat fornix shortening/symblepharon in different chronic cicatrizing conjunctival diseases (198). One of their patients suffered from oGVHD and was successfully treated with this surgical method (198). Dulz et al. described a 7-year-old boy who developed a massive bilateral cicatricial entropion with trichiasis 5 years after HCT. They performed bilateral lamellar splitting via an eyelid crease and gray line incision. Cryocoagulation of persistent trichiatic lashes was additionally performed (199). Kheirkhah et al. utilized a combined approach with mucous membrane transplantation from the lower lip covering it with AMT for their series of symblepharon, among which was also a successfully treated oGVHD patient (200).

Cataracts frequently develop in patients after HCT. This is probably a side effect of the treatments with corticosteroids or total body irradiation, and not due to GVHD directly. The long-term use of topical corticosteroids, particularly when given at higher dosages increases the risk for cataract formation. In patients with oGVHD inactivity of the ocular surface inflammation and optimal stabilization of the dry eye disease is required before surgery (201), and a good peri-operative management is critical. Bae et al. described 77 cataract surgeries in 42 patients suffering from oGVHD. Out of these patients, 19 postoperatively developed punctate keratopathy, that was being treated with artificial tears or autologous serum drops; another 7 eyes developed corneal epithelial defects, requiring non-steroidal anti-inflammatory eye drops, and another 3 eyes had cystoid macular edema (202). These findings are supported by others, additionally reporting on corneal melts and perforation after surgery (203–207). Taken together, oGVHD patients require close post-operative monitoring and prolonged anti-inflammatory treatment.