Myxoid Stromal Histophenotype Is Associated with High-Grade and Persistent Cervical Intraepithelial Neoplasia

Objectives: To evaluate the diagnostic and prognostic value of histophenotyping of the extracellular matrix of the cervical stroma at cervical intraepithelial neoplasia (CIN). Methods: Retrospective analysis of 160 biopsies and surgical preparations of the cervix in women of reproductive age included cases of CIN 1-3 and the group with confirmed persistence or lesion progression (CIN P) at repeated biopsy. The control group (n = 40) consisted of morphologically intact cervical tissue. Histophenotypes were evaluated by staining with hematoxylin, eosin, and Masson trichrome, and classified as follows: normal (dense parallel bundles of type I collagen), intermediate (disorganized and fragmented type I collagen fibers), and myxoid (amorphous weakly fibrillar matrix). The clinical, viral, and inflammatory characteristics between histophenotypes were statistically compared. Results: The distribution of histophenotypes of the extracellular matrix of the cervix varied significantly depending on the CIN degree (p < 0.001). In the control group, the normal pattern was detected in 97.5% of cases; its frequency decreased from CIN 1 (27.5%) to CIN 2 (12.5%) and was absent at CIN 3. The frequency of the myxoid pattern increased significantly in severe and persistent forms: 55% at CIN 3 and 62.5% at CIN P. Human papillomavirus 16/18 was most frequently detected in groups with intermediate (69.1%) and myxoid (27.2%) patterns. Inflammatory changes were more often accompanied by disorganized extracellular matrix; however, intermediate and myxoid types also occurred in the absence of inflammation. Conclusions: The myxoid histophenotype of the extracellular matrix is significantly associated with the high degree of dysplasia and CIN persistence. It can reflect the morphological equivalent of tumor-associated stroma remodeling. Histophenotyping of the extracellular matrix of the cervix appears to be a promising method of risk stratification and may complement existing diagnostic algorithms for CIN.