To assess the feasibility, tolerance and diagnostic accuracy of endometrial biopsy (EB) during sonohysterography (SH) compared to EB after hysteroscopy (HSC) in endometrial disorders.
105 consecutive patients with irregular uterine bleeding were included prospectively in the calendar year 2001. SH and flexible HSC were performed in an office setting, subsequently and in a random order, by two different practitioners blind to the former experiment. SH-EB was performed using a 3.1-mm ultrasound-guided Bernard catheter in the uterine cavity still partly distended and with a gentle 20-ml syringe vacuum aspiration. The biopsy was directed on focal lesions or else randomly when no targets had previously been found. A Cornier Pipelle was used to perform EB after HSC. We investigated all patients by biopsy, independent of the endometrial thickness. HSC-EB was the gold standard.
For both methods, cervical catheterism was impossible in 20 patients, 75 of them successfully underwent both exams. Duration and tolerance were similar. SH was effective in the distinction between normal and pathologic cavities, as well as in the diagnosis of polyps. Endometrium appeared significantly thinner with HSC (1.8 mm) when compared to SH (2.9 mm, p < 0.05). Histological endometrial assessment failed in 30 cases of SH-EB and in 22 cases in HSC-EB (NS). There was a poor correlation of the histological results of both techniques. Hyperplasia has never been diagnosed by SH-EB, whereas 3 EB issued from HSC-EB brought up this diagnosis.
SH-EB with our technique did not improve the diagnostic potential of SH and severe diagnosis was missed. Histological assessment should fail less when we exclude endometrial atrophy. The diameter of the catheter and the aspiration technique must be revised and the learning curve must be considered. Our technique cannot replace EB by HSC.