Pterygium pathogenesis is mainly related to UV light exposure. However, the exact mechanisms by which it is formed have not been elucidated. Clinical advances in surgical treatment use conjunctival autografts and amniotic membranes in combination with adjuvant therapies, including mitomycin C, β-radiation, and 5-fluoroacil, to reduce recurrence. Several studies aim to unveil the molecular mechanisms underlying pterygium growth and proliferation. They demonstrate the role of different factors, such as viruses, oxidative stress, DNA methylation, apoptotic and oncogenic proteins, loss of heterozygosity, microsatellite instability, inflammatory mediators, extracellular matrix modulators, lymphangiogenesis, cell epithelial-mesenchymal transition, and alterations in cholesterol metabolism in pterygium development. Understanding the molecular basis of pterygium provides new potential therapeutic targets for its prevention and elimination. This review focuses on providing a broad overview of what is currently known regarding molecular mechanisms of pterygium pathogenesis.

DNA mutations are identified by techniques that use the knowledge of the wild‐type DNA sequence and its mutated variant. The involved analytic methods must be accurate, rapid, and sustainable, if a clinical application is pursued. High‐performance liquid chromatography under denaturing conditions is a useful technique to screen mutations. Denaturing high‐performance liquid chromatography resultant chromatograms are suitable for feature extraction analysis with multivariate methods such as principal component analysis. In this work, principal component analysis was applied to analyze the chromatograms from 3 different genes. Fragments with verified wild‐type sequence were used as reference and samples with sequence unknown were tested. A statistical characterization based on Tukey's boxplot equation of principal component scores allowed us to analyze the distance distribution between reference and sample clusters to establish a classification criterion: an outlier could represent a mutated sample, and a typical value could be a wild‐type sample. Identified outliers were further analyzed by sequencing and proved to carry a mutation. From 72 datasets with a total of 4258 injections, we successfully assessed the classification criterion, identifying mutated samples in lymphoma and breast cancer patients with ratio of prediction Gmean = [0.89, 1.00]. Compared with sequencing analysis, this procedure reduced time and costs.

No abstract