The trusted source for
healthcare information and
Abstract & Commentary
Synopsis: The results of this study suggest that in women with postmenopausal bleeding, transvaginal ultrasonography in conjunction with hysterosonography can improve diagnostic accuracy, clinical decision making, and the clinician’s diagnostic certainty.
Source: Bree RL, et al. US evaluation of the uterus in patients with postmenopausal bleeding: A positive effect on diagnostic decision making. Radiology 2000;21:260-264.
The purpose of this investigation was to examine women with postmenopausal bleeding (PMB), in an effort to determine not only the accuracy of hysterosonography (HSG), but also to evaluate its role with regard to diagnostic confidence and therapeutic decision making. To accomplish these goals, Bree collaborated with physicians from two academic and one community medical center. Initially, 123 patients were enrolled in the study, but due to either a technically unsuccessful HSG (10 patients) or incomplete follow-up (15 patients), 98 women ultimately formed the basis of the study. In these women, transvaginal ultrasound in conjunction with HSG revealed the following results: normal findings in 29 patients (30%); polyps in 46 (47%); fibroids in 11 (11%); hyperplasia in eight (8%); and cancer in four (4%). The results of the ultrasound examinations suggested, therefore, that 69 of these women had pathologic conditions that caused their PMB. This was confirmed by histopathology in 65 of these patients. Bree et al claimed that with regard to the four false-positive ultrasound diagnoses, two of these patients were bleeding at the time of hysteroscopy and dilatation and curettage, and this may have caused a significant lesion to be missed during the surgical procedure. In one patient who was bleeding during the HSG, a blood clot was misinterpreted as a polyp. Of the 29 negative HSG results, 28 were in agreement with histopathologic analysis. In one patient with a small focus of endometrial carcinoma, only a large leiomyoma was detected by HSG. Statistical analysis of these results revealed a sensitivity of 98%, specificity of 88%, positive predictive value of 94%, and negative predictive value of 97%.
Further data analysis suggested that in 86 patients (88% of the cases), the results of the ultrasound examinations substantially influenced the diagnostic certainty for the referring physicians, and that in 78 patients (80% of the cases), treatment decisions were influenced by the HSG results.
Another interesting aspect of this study related to measuring endometrial thickness on transvaginal ultrasound examinations. In women with PMB, multiple reports suggest 4 mm or 5 mm should be used as the cutoff value for determing an abnormally thick endometrium. Interestingly, when Bree et al determined endometrial thickness in 55 women with technically adequate transvaginal examinations, and in whom the HSG study revealed either a polyp or leiomyoma, the mean endometrial thickness was 7.2 mm ± 4.4 mm (range, 2-18 mm). Importantly, in 22 of these women, endometrial thickness was determined to be 5 mm or less.
Comment by Faye C. Laing, MD
Until recently, the traditional work-up for a woman with PMB consisted of doing either an initial endometrial biopsy or a more invasive dilatation and curettage. If the pathology results were negative, it was assumed the bleeding was not due to a significant problem, with atrophy being the likely cause. When hysteroscopy came on the scene, it became apparent that many cases of PMB did indeed have an anatomic basis, and that many polyps and fibroids were overlooked when the work-up consisted of biopsy, or dilatation and curettage.
Since the advent of transvaginal sonography and its superior resolution compared to transvaginal images, the question has been raised as to whether this noninvasive appproach can be used initially, in lieu of more invasive techniques to evaluate women with PMB. The addition of HSG to the ultrasound armamentarium, and results such as those reported by Bree et al further strengthen this thesis. There are several compelling reasons to advocate using his approach. First, as reported by Bree et al, sonohysterography has high sensitivity and accuracy, and it can alter patient treatment in 80% of cases. Second, given a choice, most women would likely prefer an initial HSG, as opposed to a more invasive procedure. Third, a cost analysis study has shown that when ultrasound is used initally, the total work-up of the patient is less costly.
One surprising result of this investigation is that 22 of 55 women (40%) with either a submucosal fibroid or polyp had an endometrial thickness that was 5 mm or less. Most authorities suggest that when a conventional transvaginal sonogram reveals a normal appearing endometrium with a diameter of less than 5 mm thick (some use < 4 mm), atrophy associated with superficial endometrial ulceration is likely, and that no further diagnostic procedures are required. Unfortunately, in this report, Bree et al only discussed endometrial thickness and did not address whether the endometrium had an abnormal appearance. I suspect close endometrial scrutiny in at least some of these cases would reveal focal pathology.
The question of when to do a HSG remains somewhat controversial. At one extreme are sonologists such as Bree who recommend HSG for all patients with PMB. At the opposite end of the spectrum are sonologists who claim that conventional transvaginal sonography can frequently identify focal edometrial pathology, in which case the HSG can be omitted. Many recommend HSG whenever conventional transvaginal sonography reveals an abnormal endometrium, or for cases that demonstrate either diffuse thickening or indeterminate findings.