Background: Endometrial cancer (EC) risk stratification remains limited by the fragmented and often non-integrated use of molecular and clinical biomarkers. While markers such as p53 and CA125 are used and associated with key clinicopathological features, their independent prognostic value remains inconsistent. This highlights the need for integrated, interaction-driven models to improve risk assessment. Methods: This study was conducted within a PRIN 2022 project funded by MUR. A total of 100 patients diagnosed with EC were ambispectively enrolled, including all disease stages and histological subtypes. Clinical, pathological, and molecular data were collected into a database. Patients were stratified by histological subtype (endometrioid vs non-endometrioid) and disease stage (FIGO I–II vs III–IV). Associations between histology, p53 immunophenotype, CA125 levels, and clinicopathological variables were assessed using chi-square (x2) tests, t-tests, and logistic regression models. Results: p53 immunophenotype was significantly associated with histology (p < 0.001), and age at diagnosis (p = 0.01). In detail, the probability of endometrioid histology was higher among patients older than 61 years and wildtype p53 (prob = 0.955) and lower among those older than 61 and aberrant p53 (prob = 0.21). Conversely, among patients aged less than 61, the probabilities are 0.89 and 0.5 when p53 is wild-type or aberrant, respectively. CA125 was significantly associated with disease stage (p = 0.002) and appeared to act as a modulatory factor within the broader clinical context. In fact, the logistic model suggested that CA125 may influence the relationship between clinical variables and immunohistochemical markers, particularly hormone receptor expression. Overall, these findings suggest that the prognostic role of p53 and CA125 cannot be fully captured through isolated effects but may require a multidimensional, interaction-based interpretation. Conclusions: p53 represents a robust biomarker for histological characterization in EC, while CA125 appears to play a modulatory role in the clinical context. These findings suggest that the prognostic value of individual biomarkers may be limited when considered alone and highlight the potential of interaction-driven approaches to improve risk stratification. Such interaction-driven models may support a more accurate and clinically meaningful framework for risk stratification compared to singlemarker approaches.
Role of p53 and CA125 in endometrial cancer risk stratification: an interaction-based approach
Capoluongo Ettore Domenico;Scandurra Giuseppa;Aiello Fabio;Priulla Andrea;Spadola Saveria;Scollo Paolo;Lombardo Valentina Nunziella
2026-01-01
Abstract
Background: Endometrial cancer (EC) risk stratification remains limited by the fragmented and often non-integrated use of molecular and clinical biomarkers. While markers such as p53 and CA125 are used and associated with key clinicopathological features, their independent prognostic value remains inconsistent. This highlights the need for integrated, interaction-driven models to improve risk assessment. Methods: This study was conducted within a PRIN 2022 project funded by MUR. A total of 100 patients diagnosed with EC were ambispectively enrolled, including all disease stages and histological subtypes. Clinical, pathological, and molecular data were collected into a database. Patients were stratified by histological subtype (endometrioid vs non-endometrioid) and disease stage (FIGO I–II vs III–IV). Associations between histology, p53 immunophenotype, CA125 levels, and clinicopathological variables were assessed using chi-square (x2) tests, t-tests, and logistic regression models. Results: p53 immunophenotype was significantly associated with histology (p < 0.001), and age at diagnosis (p = 0.01). In detail, the probability of endometrioid histology was higher among patients older than 61 years and wildtype p53 (prob = 0.955) and lower among those older than 61 and aberrant p53 (prob = 0.21). Conversely, among patients aged less than 61, the probabilities are 0.89 and 0.5 when p53 is wild-type or aberrant, respectively. CA125 was significantly associated with disease stage (p = 0.002) and appeared to act as a modulatory factor within the broader clinical context. In fact, the logistic model suggested that CA125 may influence the relationship between clinical variables and immunohistochemical markers, particularly hormone receptor expression. Overall, these findings suggest that the prognostic role of p53 and CA125 cannot be fully captured through isolated effects but may require a multidimensional, interaction-based interpretation. Conclusions: p53 represents a robust biomarker for histological characterization in EC, while CA125 appears to play a modulatory role in the clinical context. These findings suggest that the prognostic value of individual biomarkers may be limited when considered alone and highlight the potential of interaction-driven approaches to improve risk stratification. Such interaction-driven models may support a more accurate and clinically meaningful framework for risk stratification compared to singlemarker approaches.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


