Stephanie Vrede

CHAPTER 2 36 Sampling method For 549 (95.8%) patients preoperative sampling method was available. Pipelle endometrial sampling was performed in 47.2%, D&C in 14.0% and hysteroscopic biopsy in 38.8% of the patients with available sampling method (Supplementary Table S3). No significant difference was found between the diagnostic sampling methods and the concordance between pre- and postoperative low- and high-grade EC (P=0.364), nor for the individual tumor grade and histological subtype (P=0.097). Median endometrial tissue surface for the preoperative sampling method pipelle was 18.6 mm2, D&C 67.8 mm2 and hysteroscopic biopsy 15.4 mm2 (P<0.001). All preoperative sampling methods (pipelle, D&C, hysteroscopic biopsy) showed higher median endometrial tissue surface in discordant low-and high-grade diagnoses, compared to concordant low- and high-grade diagnoses. Similar was shown for individual tumor grade and histological subtype diagnoses (Supplementary Figure S3). Concordance, discordance and survival outcome The DSS of the concordant and discordant diagnoses are shown in Figure 3A-C. Figure 3A shows the DSS of the patients with concordant high-grade EC, concordant low-grade EC, and clinically relevant downgraded and upgraded diagnoses (P<0.001). Patients with concordant low-grade EC had a significant superior DSS compared to patients that were downgraded (96.5% and 88.9% respectively, P=0.039) (Figure 3B). Patients with concordant high-grade EC had a significant impaired DSS compared to patients that were upgraded (71.4% and 88.6% respectively, P=0.046) (Figure 3C). DISCUSSION This study assessed whether the amount of preoperative endometrial tissue surface is related to the degree of concordance with final classification of low- and high-grade EC, and whether discordance is influenced by the diagnostic sampling method and impacts outcome. Overall, 60% showed concordant pre- and postoperative tumor grade and histological subtype and there was 88.8% concordance in pre- and postoperative classification into low- and highgrade EC, with 92.2% concordant low-grade, and 73.3% concordant high-grade EC. The median endometrial tissue surface between concordant and discordant individual tumor grade and histological subtype showed no significant difference. Interestingly, concordant diagnoses revealed a significant lower median endometrial tissue surface compared to discordant diagnoses. Furthermore, the sampling method did not influenced the degree of concordance. Finally, patients with preoperative low-grade and postoperative high-grade EC had significant improved DSS compared to patients with concordant high-grade EC.

RkJQdWJsaXNoZXIy MTk4NDMw