Preoperative hysteroscopic assessment of cervical invasion by endometrial carcinoma: a retrospective study

OBJECTIVE: The aim of this study was to evaluate the efficacy of hysteroscopy, using normal saline (NS) or carbon dioxide (CO2) as the distention medium, in assessing tumor invasion of the uterine cervix by endometrial carcinoma. METHODS: A retrospective study was conducted in 200 consecutive patients with endometrial carcinoma diagnosed from 1993 to 2000. Prior to definitive surgical treatment, hysteroscopy was performed using either NS or CO2 as the distention medium to determine whether the tumor had spread to the cervix. The uterine specimens obtained after hysterectomies were cut open for gross examination. Tumor invasion of the cervix as determined by hysteroscopy and gross examinations was compared with the pathological findings. RESULTS: Tumor invasion of the cervix was found in 41 (20.5%) cases on pathological examination. Hysteroscopy has an accuracy of 92.5% (185/200), a sensitivity of 68.3% (28/41), and a specificity of 98.7% (157/159), with a PPV of 93.3% (28/30) and a NPV of 92.4% (157/170) in determining cervical involvement. Assessment by gross inspection had 93.0% (186/200) accuracy, 68.3% (28/41) sensitivity, 99.4% (158/159) specificity, 96.6% (28/29) PPV, and 92.4% (158/171) NPV. There was no significant difference between the two assessment methods. When the results of hysteroscopy performed with different distention mediums were compared, the use of NS had a higher accuracy in determining tumor spread to the cervix (96.8% vs 88.7%, P = 0.03) and NPV (96.4% vs 88.4%, P < 0.05) than the use of CO2. CONCLUSIONS: Hysteroscopic assessment and gross examination of the uterine specimen had similar efficacy in detecting cervical invasion by endometrial carcinoma. Hysteroscopic examination using NS is more accurate than that which uses CO2.