Portal de Periódicos da CAPES

Reply: New ultrasound technologies to classify deep pelvic endometriosis

2015; Wiley; Volume: 45; Issue: 3 Linguagem: Inglês

10.1002/uog.14756

1469-0705

S. Reid, Chuan Lü, Nigel Hardy, I. Casikar, Gavin E. Reid, G Cario, Danny Chou, D. Almashat, G. Condous,

Gynecological conditions and treatments

We read with interest the comments from Noventa et al. regarding the difficulty in identifying the best ‘first-line’ approach for the diagnosis of site-specific deep infiltrating endometriosis (DIE) of the posterior compartment. Indeed, the diagnostic accuracy of transvaginal sonography (TVS) and modified-TVS approaches for specific pelvic DIE locations fluctuates widely among different research groups and, in the absence of external validation and reproducibility studies, it is difficult to conclude which TVS technique is the most accurate. The systematic review and meta-analysis by Hudelist et al.1 found that TVS alone was sufficient to predict accurately the presence of bowel endometriosis, and our current gel sonovaginography (SVG) study was also able to predict accurately the presence of rectal/rectosigmoid lesions. However, the accuracy of both TVS and modified-TVS approaches varies widely with regard to the prediction of DIE located in structures such as the uterosacral ligaments (USL), vagina and rectovaginal septum (RVS). Indeed, our gel SVG study did not show a high sensitivity for vaginal, RVS or USL DIE, and we believe that these locations are much more difficult to diagnose, particularly when the lesions are subtle. In addition, our study lacked an adequate number of cases with infiltration of these areas to allow us to draw any firm conclusions. Encouragingly, in their recent saline contrast sonovaginography study, Saccardi et al.2 were able to obtain sensitivities of 95% and 81% for the prediction of DIE in the vaginal fornix and RVS, respectively. Interestingly, however, the same study could predict bowel endometriosis with a sensitivity of only 67%. Again, further studies are required to evaluate these modified-TVS approaches. It may be argued that the preoperative diagnosis of RVS DIE is unnecessary for surgical planning, as these lesions occur only rarely in isolation, generally occurring in combination with rectal and/or vaginal DIE. In our gel SVG study, 15/16 (94%) women with RVS and/or vaginal DIE had an associated rectal lesion and obliteration of the pouch of Douglas (POD). Since the TVS sliding sign and gel SVG were able to identify the vast majority (87%) of these women with POD obliteration and rectal DIE, their surgical management was not compromised by the failure of gel SVG to detect RVS DIE. The solitary case of vaginal + RVS DIE that was not associated at surgery with bowel DIE or POD obliteration was identified correctly with gel SVG. The value of the TVS sliding sign for the prediction of POD obliteration cannot be understated; we believe that this simple ultrasound technique should be used in all TVS and modified-TVS approaches which investigate pelvic DIE. Several studies have consistently confirmed the accuracy of this technique for the prediction of POD obliteration, as well as the significant association between a negative sliding sign and the presence of DIE (particularly involving the rectum/rectosigmoid)3-5. In addition, the inter- and intraobserver reproducibility of this technique among sonographers/sonologists of varying experience has been demonstrated, albeit by means of prerecorded videos of the sliding sign6. Reproducibility studies for the prediction of POD obliteration by TVS performed by the actual examiners would certainly strengthen the clinical applicability of this site-specific ultrasound finding. An additional (and not uncommon) location for endometriosis that has been essentially overlooked/excluded in the majority of previously published ultrasound studies is the peri-ureteric area. General gynecologists often choose not to perform ureterolysis for women with peri-ureteric disease, even if the disease is superficial. These women then require a second laparoscopy by an advanced laparoscopic surgeon, who is able to dissect safely the pelvic sidewall and excise the disease. Although superficial pelvic sidewall disease is undetectable with TVS, markers such as ovarian immobility or endometrioma may be helpful in identifying women who are at increased risk of requiring ureterolysis. With regard to deep infiltrative ureteric/pelvic sidewall disease, a recent study was able to demonstrate a high specificity and negative predictive value for TVS in the prediction of ureteric involvement7. Again, further studies are required to determine the accuracy of TVS for the prediction of infiltrative lateral compartment disease. With regard to the usefulness of gel SVG, as well as that of most of the other proposed modified-TVS imaging approaches, for the prediction of pelvic location of DIE, a systematic review would not be helpful at this point as there have not been sufficient studies employing these specific techniques. Reproducibility studies among different levels of sonographer and sonologist are also required to validate modified-TVS approaches for the prediction of DIE location. Our current gel SVG study was performed to provide a starting point for possible improvement in the technical aspects related to creating an acoustic window in the posterior vaginal fornix, thereby improving visualization of posterior compartment structures. We do not propose gel SVG as a ‘first-line’ imaging technique for the detection of DIE in the pelvis; however, we believe that gel SVG will be found to be an invaluable training tool for those learning how to identify both normal and abnormal structures of the posterior pelvic compartment. S. Reid*†, C. Lu‡, N. Hardy‡, I. Casikar†, G. Reid§¶**††, G. Cario¶‡‡, D. Chou¶, D. Almashat† and G. Condous†§§¶¶ †Acute Gynaecology, Early Pregnancy and Advanced Endosurgery Unit, Nepean Medical School, Nepean Hospital, University of Sydney, Penrith, NSW, Australia; ‡Department of Computer Sciences, University of Wales, Aberystwyth, UK; §St Luke's Private Hospital, Potts Point, NSW, Australia; ¶St George Private Hospital, Kogarah, NSW, Australia; **Prince of Wales Private Hospital, Randwick, NSW, Australia; ††iLiverpool Public Hospital, Liverpool, NSW, Australia; ‡‡Hurstville Private Hospital, Hurstville, NSW, Australia; §§OMNI Gynaecological Care Centre for Women's Ultrasound and Early Pregnancy, St Leonards, NSW, Australia; ¶¶Norwest Private Hospital, Bella Vista, NSW, Australia *Correspondence. (e-mail: [email protected])