Should adequacy criteria in cervicovaginal cytology be modified after radiotherapy, chemotherapy, or hysterectomy?

Abstract

BACKGROUND: The general criterion of an unsatisfactory Papanicolaou (Pap) test in the 2001 version of the Bethesda system is not applicable to patients after treatment with radiotherapy, chemotherapy, or hysterectomy. The current study was performed to determine whether specimen adequacy criteria for Pap tests should be modified for these conditions. METHODS: Consecutive patients who underwent conventional Pap tests between March and August 2006 were reviewed. The original reports were done according to the 2001 Bethesda system, with cellularity criteria modified in patients with a history of radiotherapy, chemotherapy, or hysterectomy. The slides of these patients were reviewed again. The degrees of cellularity, obscuring red blood cells, and inflammation were recorded. RESULTS: The final analyses included 7033 patients for which there were complete data. The original interpretation was unsatisfactory in 4.4% of all samples. When the 1337 slides obtained from patients with a history of radiotherapy, chemotherapy, and/or hysterectomy were reviewed using the general satisfactory threshold of >8000 squamous cells/slide and <75% of the epithelium obscured, the incidence of unsatisfactory Pap tests increased from 4.3% to 13.2% (176 of 1337 slides). The odds ratios for unsatisfactory Pap tests for a history of radiotherapy, chemotherapy, and age >50 years were 2.70, 2.51, and 1.39, respectively. The majority of unsatisfactory Pap tests were because of low cellularity. The lower limits of adequate cellularity after radiotherapy or chemotherapy can be set at 2000 cells/slide, which can lower the unsatisfactory rate while at the same time resulting in no increase in the false-negative rate. Hysterectomy alone was not found to be associated with unsatisfactory Pap tests. CONCLUSIONS: In patients who received pelvic radiotherapy or chemotherapy, the incidence of low-cellularity Pap tests was unacceptably high. A lower cellularity (estimated 2000 cell/slide) could be used as a satisfactory threshold.