Transvaginal ultrasound (TVUS) may not be as accurate as MRI when it comes to preoperative staging of endometrial cancer, but it can be relied on as a more accessible first-line test, a study published October 19 in Clinical Radiology suggests.
Researchers led by Dr. Magnus Palmér from the University of Gothenburg found that MRI has higher specificity than TVUS performed by gynecologists for assessing deep myometrial invasion and cervical stroma invasion in low-grade endometrial cancer. However, the two modalities have similar sensitivities in these areas.
"The accuracy of preoperative endometrial cancer staging by TVUS in the present clinical setting may be regarded as sufficiently adequate as a first-line technique, with a low extra cost and without delaying surgery, when performed by the gynecologist at the preoperative visit," the team wrote.
Imaging is an important part of assessing local growth patterns in low-grade endometrial cancer tumors to predict the risk of extra-uterine malignancy. TVUS is high in accuracy when it comes to evaluating myometrial invasion, but it is more up-and-down for looking at cervical stroma invasion. Researchers suggested that this is because TVUS is operator-dependent, and the reported results are mainly from centers with highly experienced specialists.
Previous research suggests that MRI, which is not operator dependent, is at least on par with TVUS in these areas. While guidelines recommend both modalities as suitable for preoperative staging, they don't specify which one is preferable.
Palmér and colleagues wanted to test the diagnostic performance of TVUS performed by gynecologists to see if it is adequate for use by comparing results to that of MRI. They also sought to evaluate two different MRI readers, one with more experience in gynecological MRI and one with limited experience. The team wanted to form an algorithm for preoperative imaging of low-grade endometrial cancer that may be used in a variety of clinical settings.
They looked at data from 259 women with biopsy-verified endometrial cancer. Out of these, 232 had low-grade carcinoma. They were without apparent extra-uterine manifestations and planned for primary surgical treatment.
| Comparison between TVUS, MRI for preoperative endometrial cancer staging | ||
| MRI | TVUS | |
| Specificity | 0.88 | 0.68 |
| Sensitivity | 0.73 | 0.68 |
When the modalities assessed cervical stroma infiltration, MRI had the edge on specificity with 0.96 while TVUS demonstrated a specificity of 0.9. However, researchers noted no significant differences again between the two when it came to sensitivity (MRI 0.41, TVUS 0.32).
"Interestingly, MRI had a higher accuracy assessing deep myometrial invasion than TVUS regardless of the experience of the MRI reader," the study authors wrote. "Although heterogeneous results, the overall diagnostic performance of TVUS performed by 32 gynecologists with a short training session, was almost as good as the results from studies with ultrasound experts; however, not as good as MRI."
The researchers also found no significant differences between MR readers and TVUS for negative predictive values.
While Palmér and colleagues suggested that TVUS can be reliably used first in such imaging where MRI access is difficult, they added that patients should be referred for multiparametric MRIs in cases of inconclusive TVUS exams.
"The MRI can be performed at any unit by using a dedicated protocol for endometrial cancer, given the possibility to send the images to tertiary centers for a second opinion," they wrote.
