Abstract:
BACKGROUND
Uterine fibroid is the commonest gynecologic tumour in women of reproductive age but there is a disparity in research to understand the aetiologyand risk factors of the disease in Calabar. This study analyzed the clinical and histopathological characteristics of subjects and used an artificial neural network (ANN) to predict the risk factors/biomarkersof leiomyoma.
MATERIALS AND METHODS
This retrospective cross-sectional study involved women with complete data who were diagnosed with leiomyoma. Data from 104 subjects were retrieved and analyzed from January 2020 to May 2021. Ten uterine tissue blocks were retrieved and stained with haematoxylin and eosin (H&E), Weighert-vanGieson and immunohistochemical methods for progesterone receptor (PR), Ki-67 and p53. Descriptive statistics, ANOVA, and ANN model of Statistical Package for Social Sciences (SPSS)were used for analysis.
RESULTS
The 104 subjects with leiomyoma had 67(64.4%) leiomyoma uteri and 37(35.6%) degenerative changes. The nature of the sample was related to diagnosis (p=0.036). The age range was between 24-57 years. More cases occurred between 30-39 years with 58(55.8%)casesbut were not statistically significant with age (p=0.254). The nature of the sample was significant with age (p=0.008). The ANN model predicted age(100%), p53(78.2%), Ki-67(95.9%) and collagen(59.1%) as the important risk factors/biomarkers associated with leiomyoma.
CONCLUSION
Leiomyoma mostly affects women of reproductive age and is associated with loss of p53, increase Ki-67 and increase collagen deposition. The routine application of these biomarkers may be useful in understanding the predisposing factors of leiomyoma for effective diagnosis, management and prognosis
