Introduction Ocular surface squamous neoplasia (OSSN) is a spectrum of disease that ranges from noninvasive intra-epithelial dysplasia of the conjunctival and cornea (CCIN), through to invasive squamous cell carcinoma (SCC). It often presents with unilateral tumours on the eyeball. The tumours may cause blindness, disfigurement and even death. In East Africa, OSSN is relatively common and aggressive, affecting younger adults and proportionally more women than in other parts of the world. The management of OSSN is challenging for various reasons. Its risk factors are not clearly understood. Studies have implicated HIV, human papilloma virus (HPV) and solar radiation however about 30% of cases are HIV-negative while some studies have implicated HPV and others found no association. The importance of vitamin A for a healthy ocular surface is known, yet its role in OSSN has not been studied. Early diagnosis relies on the clinical impression yet OSSN appears similar to other conjunctival tumours and histopathology services are generally unavailable in Africa. Surgery is the mainstay of treatment but recurrence is an issue. There is no trial evidence for the various treatments used in HIV-infected persons. This project was an integrated set of studies to improve our understanding of the epidemiology and management of OSSN in Kenya. Methods We conducted three systematic reviews on the epidemiology of OSSN in Africa, the pathophysiology of OSSN and updated a Cochrane review on the interventions for OSSN in HIV-infected individuals. Working in four eye care centres in Kenya between July 2012 and July 2015, we conducted the following six studies: (i) clinical assessment of a series of patients with conjunctival lesions to describe OSSN to determine how OSSN may differ clinically from benign lesions, (ii) evaluated vital staining with a special dye called Toluidine Blue (ToB) for making the diagnosis of OSSN, (iii) developed a diagnostic algorithm based on clinical features and vital staining, (iv) conducted a large case-control study to investigate risk factors that may contribute to the development of OSSN, (v) investigated the care-seeking journey of OSSN patients to assess referral pathway and treatment delay, and finally, (vi) conducted a randomised placebo-controlled trial of 5-Fluorouracil (5FU) chemotherapy eyedrops given after surgery to investigate if this can reduce recurrence of the lesions. Results Meta-analysis of data from cancer registries worldwide showed that Africa has the highest incidence of OSSN in the world with a peak at latitude 160 South and males and females are equally affected, unlike other continents where male disease predominates. Here the agestandardized rate in cases/year /100,000 population (95%CI) is 1.38 (1.00–3.75) and 1.18 (1.08–3.43) in males and females, respectively (p=0.853). Incidence rises with increasing exposure to direct sunlight (2–4 h, OR = 1.7, 95% CI: 1.2–2.4 and ≥5 h OR = 1.8, 95% CI: 1.1–3.1) and outdoor occupations (OR = 1.7, 95% CI: 1.1–2.6). Fixed-effect meta-analysis shows a strong association with HIV (6 studies: OR = 6.17, 95% CI: 4.83–7.89) but not cigarette smoking (2 studies: OR = 1.40, 95% CI: 0.94–2.09). HPV shows heterogeneous association (random effects meta-analysis of 7 studies: OR = 2.64, 95% CI: 1.27–5.49). The pathophysiology review concluded that limbal epithelial stem cells are the likely progenitor cells of OSSN. UV radiation probably causes DNA damage via pyrimidine dimers (involving the p53 tumour suppressor gene), photo-immunosuppression and reactivates latent HPV. HPV E6 inhibits p53 gene allowing DNA-damaged cells past the G1-S checkpoint of the cell cycle. HPV E7 inhibits the retinoblastoma (pRB) gene anti-transcription at G1 so infected cells continue replicating. HIV, photo-immunosuppression and vitamin A deficiency may impair tumour surveillance. The Cochrane review found no trials for the interventions used in OSSN in HIV-infected populations. There was one trial in Australia that found topical Mitomycin C (MMC) effective. The results from case series reviewed were difficult to compare. They reported a wide variety of combinations of surgery and adjuvant treatment used during surgery or post operatively; varying doses of adjuvant agents used; different inclusion criteria of patients and recurrences reported at varying periods after treatment. Surgery with adjuvant 5FU or MMC was often associated with recurrences of 11% to 67% about 30 months later. We enrolled 496 adults with any conjunctival lesions requiring excision and 131 controls. OSSN was themost common lesion diagnosed in 187 (38%). Patients with OSSN were slightly older (mean [SD] age, 41 [11.6] vs 38 [10.9] years; p =0.002) and tended to have lower levels of education than patients with benign lesions (p = 0.001). Females predominated (67% of OSSN vs 64% of benign lesions; p = 0.65). HIV infection was common among patients with OSSN (74%). Although some clinical signs were more frequent in OSSN, all OSSN signs were also observed in benign lesions. OSSN and benign conjunctival lesions have overlapping phenotypes and cannot always be reliably distinguished on clinical grounds. The positive predictive value of clinical appearance in identifying OSSN was 54%. Inter-observer agreement was modest (κ= 0.1-0.4). Any blue colour on vital staining with ToB 0.05% had a sensitivity of 92%, specificity of 31%, positive predictive value of 41%, and negative predictive value of 88% for OSSN. Interobserver agreement was substantial for staining (k=0.8) and moderate for overall diagnosis (OSSN or benign) (κ =0.4). Use of ToB caused mild discomfort in 88 (21%) patients; mild superficial punctate keratopathy seen in 7 (1.7%) and no histological evidence of corneal toxicity was observed. ToB had a high rate of false positives (69%). We developed a simple probability-tree clinical algorithm that shows the probability of OSSN with various combinations of clinical features. A multivariable regression model found 8 features strongly associated with OSSN; prior excision, corneal involvement, feeder vessels, dark blue ToB staining, papillary or gelatinous tumour surface, severe inflammation, antiretroviral therapy and temporal or circumlimbal tumours. Using a cut-off of any 3 of these features, the sensitivity was 89%, specificity 50%, and 65%of lesions were correctly classified. This specificity was higher than any blue ToB staining (31%) but lower than clinical photoexamination (60%). A total of 131 cases were frequency-matched to 131 controls by age, sex and eye center. Risk factors for OSSN were HIV infection without antiretroviral therapy (ART) use (OR=48.30; 95%CI 7.53-309.90) and with ART use (OR=19.02; 95%CI 6.55-55.26), longer duration of exposure to the sun in the main occupation (6.9 hrs/day vs. 4.6 hrs/day, OR=1.23; 95%CI 1.08-1.39) and a history of allergic conjunctivitis (OR=80.20; 95%CI 8.62-746.29). Wearing hats was protective (OR=0.21; 95%CI 0.07-0.63). We studied the care-seeking journey followed by 158 new OSSN patients. About half (88/158, [56%]) presented directly to the study centres while the rest were referred. Indirect presenters sought care earlier than direct presenters (median 2.0 months vs 5.5 months) and travelled a shorter distance to the first health facility (median 20km vs 30km) but had surgery later (median 12.5 months vs 5.5 months). Visits beyond the first health facility for indirect presenters markedly increased delay (median 7.3, 29.0, 37.9, and 32.0 months for 1-4 facilities, respectively). Delay was associated with number of health facilities visited (adjusted ordered OR=9.12; 95%CI 2.83-29.4, p<0.001) and being female (adjusted ordered OR=2.42; 95%CI 1.32-4.44, p=0.004). In the randomized placebo-controlled trial we randomly allocated 49 participants to 5FU and 49 to placebo. Four participants were lost to follow-up. Treatment with 5FU was associated with fewer OSSN recurrences: there were 5/47 (10·6%) recurrences in the 5FU arm and 17/47 (36·2%) in the placebo arm (odds ratio 0·21; 95%CI 0·07-0·63, p=0·01). There was little effect from adjusting for passive smoking and antiretroviral therapy imbalance (adjOR=0·23; 95%CI 0·07-0·75, p=0·02). Adverse effects were transient, mild and more frequent with 5FU: ocular discomfort (43 [88%] vs 36 [73%]), epiphora (24 [49%] vs 5 [10%]), and eyelid skin inflammation (7 [14%] vs 0). Conclusions The clinical impression alone is unreliable for distinguishing OSSN from benign lesions. Toluidine Blue (ToB) staining is less specific and predictive than clinical examination by an Ophthalmologist. However, ToB may be a useful tool for other health care workers, with less ophthalmic training who might be involved in screening patients for the disease, as when there is no staining the disease is unlikely to be malignant. An algorithm that combines clinical features and ToB staining improves the specificity to 50%, is reasonably accurate (65%) for distinguishing OSSN from non-OSSN and shows the probability of disease with various combinations of clinical features. This algorithm cannot replace histopathology. Measures to prevent and control HIV, prevent sun exposure such as wearing hats, and control allergic conjunctivitis are recommended. Referral introduces significant delay before patients receive definitive treatment for OSSN. Women were more likely to experience delay. Despite regular contact with the health system for those with known HIV infection, delays occurred. Training in recognition and referral of OSSN cases, particularly in the HIV service, might lead to shorter delays before presentation. Post-operative topical 5FU substantially reduced recurrence of OSSN, was well-tolerated, and its use recommended in this context.