Pelvic radiotherapy for cervical cancer affects importantly the reproducibility of cytological alterations evaluation.

ABSTRACT

BACKGROUND: to evaluate the intraobserver and interobserver reproducibility of cervical cytopathology according to previous knowledge of whether patients received radiotherapy (RT) treatment or not. METHODS: The study analyzed a sample of 95 cervix cytological slides; 24 with cytological abnormalities (CA) and presence of RT; 21 without CA and presence of RT; 25 without CA and without previous RT; 25 with CA and without previous RT. Two cytopathology (CP) evaluations of the slides were carried out. For the first CP re-evaluation, the cytotechnologist was blinded for the information of previous RT. For the second CP re-evaluation, the cytotechnologist was informed about previous RT. The results were analyzed through inter and intraobserver agreement using the unweighted and weighted kappa. RESULTS: Post radiotherapy effects were identified in 44.4% of cases that undergone previous pelvic RT. The agreement for RT status was 66.32% (unweighted K = 0.31, 95%CI 0.13; 0.49, moderate agreement). The intraobserver agreement, regarding the cytological diagnoses, regardless of radiotherapy status, was 80.32% (weighted K = 0.52, 95%CI 0.34; 0.68). In no RT group, the intraobserver agreement was 70% (weighted K = 0.47, 95%CI 0.27;0.65) and in patients that received RT, the intraobserver agreement was 84.09% (unweighted K = 0.37, 95%CI 0.01;0.74). The interobserver agreement between cytopathology result (abnormal or normal) in the group with RT, considering normal and abnormal CP diagnosis was 14.0% and 12.5%, respectively. There was no association between the cytological alterations and the median time between the end of RT and the cytological diagnosis. CONCLUSION: This study showed that RT has an important impact in CP diagnosis because the agreement, also in interobserver and intraobserver analysis, had high discrepancy in patients that received RT. Also, demonstrated that it is difficult to recognize the presence of RT in cytological slides when this information is not provided.