Accuracy of Endovaginal Sonography for the Detection of Fallopian Tube Blockage

Size: px

Start display at page:

Download "Accuracy of Endovaginal Sonography for the Detection of Fallopian Tube Blockage"

Hester Booker
5 years ago
Views:

ccuracy of Endovaginal Sonography for the Detection of Fallopian Tube Blockage Mostafa tri, MD, Cuong N. Tran, MD, Patrice M. Bret, MD, nn E. ldis, MD, George M.

2 ccuracy of Endovaginal Sonography for the Detection of Fallopian Tube Blockage Mostafa tri, MD, Cuong N. Tran, MD, Patrice M. Bret, MD, nn E. ldis, MD, George M. Kintzen, MD The patency of 814 fallopian tubes in 414 patients was evaluated by endovaginal sonography immediately prior to hysterosalpingography. In the 659 fallopian tubes that were normal with free spillage, endovaginal sonography did not reveal any tubal or peritubal abnormality (specificity 100% ). Of the 64 fallopian tubes with definite hydrosalpinx on hysterosalpingography, only 22 were detected on endovaginal sonography (sensitivity 34% ). Four of 57 (7%) fallopian tubes with definite proximal blockage on the hysterosalpingogram showed hydrosalpinx on the same side on endovaginal sonography, indicating the association of proximal and distal tubal blockages in a small group of patients with blocked fallopian tubes. This combination can only be detected by the addition of endovaginal sonography to hysterosalpingography. Ten of 11 (91 %) hydrosalpinges in seven pa- tients who underwent endovaginal sonography immediately after hysterosalpingography were detected by ultrasonography. Only two of these had been visible on pre-hysterosalpingography endovaginal sonograms. This would indicate that the poor sensitivity of endovaginal sonography for diagnosing hydrosalpinx is at least partly due to its lack of distention. We conclude that an abnormal endovaginal sonogram is highly predictive of the presence of a blocked tube, but endovaginal sonography has a poor sensitivity for the diagnosis of a hydrosalpinx detectable by hysterosalpingography. Endovaginal sonography would be useful to detect a combination of proximal and distal blockage in a subgroup of patients with tubal blockage. KEY WORDS: Endovaginal sonography; Fallopian tube blockage; hydrosalpinx; Blockage, fallopian tube. E ndovaginal sonography has significantly improved the ability to evaluate female pelvic structures, both by better localization of disease to different organs and by better tissue charac- BBREVITIONS EVS, Endovaginal sonography; HSG, Hysterosalpingraphy Received ugust 17, 1993, from the Department of Radiology, Montreal General Hospital, and McGill University, Montreal, Que bee, Canada. Revised manuscript accepted for publication January 7, Presented at the RSN nnual Meeting. Chicago, November 29- December 4, ddress correspondence and reprint requests to Mostafa lri, MD, Division of Ultrasound, Department of Diagnostic Radiology, Montreal General Hospital, 1650 Cedar venue, 5th Floor, Mon!real, Quebec, Canada H3G 14. terization. 1 2 The role of EVS in the detection of different diseases of the uterus and ovaries has been recognized.3-5 lthough EVS plays a major role in the evaluation of the diseases of the fallopian tube, as is shown in both diagnosis and management of the patients with suspected ectopic pregnancy,6-8 its accuracy to diagnose fallopian tube blockage is not recorded. We undertook a study in a group of pa tients who had undergone an EVS examination immediately prior to HSG to determine the accuracy of EVS to detect blocked fallopian tubes. MTERILS ND METHODS The study population included 414 consecutive patients ranging in age from 21 to 44 years (mean, years) with history of primary or secondary infertility who had undergone EVS examination immediately 1994 by the merican Institute of Ultrasound in Medicine J Ultrasound Med 13: , /94/$3.50

430 FLLOPIN TUBE BLOCKGE J Ultrasound Med U429-434, 1994 prior to HSG.

3 430 FLLOPIN TUBE BLOCKGE J Ultrasound Med U , 1994 prior to HSG. EVS had been performed with knowledge of clinical history to evaluate its role in the detection of any abnormality of the fallopian tube or uterus relevant to infertility. ll EVS examinations were videotaped. Fourteen patients had previous unilateral salpingectomy leaving 814 fallopian tubes to be correlated. Of these 814 fallopian tubes, 155 were reported as revealing definite or questionable blockage on HSG. Videotapes of the examinations done on the 155 "blocked" fallopian tubes were reviewed retrospectively. In the 659 fallopian tubes with normal HSG, comparison was made with the EVS reports. Seven patients had EVS both prior to and immediately after HSG. These post-hsg endovaginal sonograms were obtained randomly to find a possible explanation for missed hydrosalpinx on pre HSG sonograms. HSG was performed with a Foley catheter that was placed in the uterine cavity, and 7 to 12 ml of Sinografin (38% iodine) (Bristol-Myers Squibb Pharmaceutical Group, Montreal, Quebec) was injected to opacify the uterine cavity and the fallopian tubes. nteropos terior and both obliques views were obtained routinely.ten-minute delayed films were taken if there was any question of tube blockage or loculated spill age. EVS was performed with an empty bladder and a transvaginal probe ranging from 5 to 7.5 MHz (Diasonics DRF400, Milpitas, C; Toshiba SL90 or SS270, Markham, Ontario; and cuson 128, Mountain View, C). Color Doppler sonography (cuson, 5 MHz probe) was used if any problem was encountered in differentiating a dilated fallopian tube from an adnexal vessel. The presence or absence of tubal blockage, manifested as a hydrosalpinx or loculated peritubal fluid, was evaluated on EVS. The criterion for the diagnosis of a blocked tube was visualization of an extraovarian, tubular, fluid-filled structure with or without the presence of longitudinal ampullary folds. On HSG, proximal blockage either at the comual or isthmic portion, presence of a hydrosalpinx, and loculated spillage were recorded. Hydrosalpinx was considered definite when it appeared as an elongated and tortuous dilation of the ampullary segment of the fallopian tube with obliteration of the ampullary folds. 9 normal fallopian tube with loculated spillage on HSG also was considered a definite indication of a blocked tube. questionable hydrosalpinx was the presence of mild ampullary dilation with preservation of the folds and some free spill. 9 lso, spillage with or without loculation was documented. Chi square analysis was used for the calculation of statistical significance. RESULTS Of the 814 tubes, 659 were normal on the HSG, and 149 were abnormal or inconclusive. The remaining six fallopian tubes were normal but had loculated spillage. mong the "abnormal or inconclusive" group, 127 were definitely abnormal and in the remaining 28 the abnormality was questionable, either owing to the technique (poor filling of one fallopian tube due to preferential flow of contrast medium to the normal opposite fallopian tube) or asymmetry of the fallopian tubes, raising the possibility of adhesions at the distal end of the tube. The 155 "abnormal" fallopian tubes on the HSG were divided into three categories: (1) those with distal tubal blockage due to blockage of the tube itself, (2) those with distal tubal blockage with normal fallopian tube but loculated spillage due to paritubal adhesion, and (3) those with proximal blockage of the tube in the comual or isthmic portion of the fallopian tube (fable 1). Seventy fallopian tubes showed hydrosalpinx with or without loculated spillage on the HSG. Sixty-four hydrosalpinges were definite and six were questionable. Of the 64 fallopian tubes with definite distal blockage, nine showed free spillage, two (22%) of which were detected on EVS. No spillage was seen in 47 cases, 16 (34%) of them were detected by EVS (Fig. 1); EVS was normal in the remaining 31 (Fig. 2). Eight fallopian tubes showed loculated spillage, which in four cases (50%) was picked up on EVS. In total, 22 of 64 of the tubes with a definite distal blockage on HSG were detected on EVS (sensitivity 34% ). None of the group with questionable hydrosalpinx was abnormal on EVS. No statistically significant difference was found when comparing the EVS sensitivity in detecting distal blockage among the free, loculated, and no-spill categories (P > 0.47). Six ''blocked tubes" with loculated spillage but no dilation of the fallopian tube were found. One of the six (17%) with loculated spillages but without a hydrosalpinx was abnormal on EVS (Fig. 3). Seventy-nine fallopian tubes had proximal blockage; 57 of these were definite blockages and 22 were questionable. Four of the 57 (7%) with definite blockage on the HSG showed hydrosalpinx on the same side on EVS (Fig. 4). EVS was normal in the remaining 53 patients. None of the fallopian tubes with questionable blockage was abnormal on EVS. Of the 659 fallopian tubes that were normal on HSG, none was abnormal on EVS (specificity = 100% ). Eleven hydrosalpinges were present in the seven patients who underwent EVS both prior to and after HSG. Two (18%) of these hydrosalpinges were

: Results of Ultrasonography Versus Hysterosalpingography salpingography fallopian tubes ULTRSOUND Normal bnormal To 659 (1003) 0 6 fallopian tubes.

4 : Results of Ultrasonography Versus Hysterosalpingography salpingography fallopian tubes ULTRSOUND Normal bnormal To 659 (1003) 0 6 fallopian tubes. /Definite hydrosalpinx ( Free spillage (9) 1. Distal tubal blockage ( 7 0)..._ )---No lッ spillage spillage (47) ( 2. Normal fallopian tubes Questionable hydrosalpinx (6) + loculated spillage (6) mitant hydrosalpinx not seen on HSG.- 3. Proximal blockage ( 79 ) --Definite (57) ---Q. uestionable (22) 8 ) 7 (78%) 2 (22%) 4 (50%) 4 (503) 31 (66%) 16 (343) 6 (1003) 0 (03) 5 (833) 1 (173) 53 (933) 4 (73)' 22 (100%) 0 (0%) 8 'T:t tllil l'1) - I» c ::t l'd... Ul... -X l'd- :+' > a- =-... O" ::r ::t l'1) 0 (/) E. C') Cll Cll "O :r - 0 6f OQ Cll 2. "ti.. ::n n l'1) I» ::t.. ::s Ul O" C') ::: -I» 0... ::s!'.!: oゥ N :r s. '< :I -... Q. I» 0 -Cll Cll I» 0.;r :I - 0 :I ' I >

432 FLLOPIN TUBE BLOCKGE J Ultrasound Med 13:429-434, 1994 B Figure 3, HSG reveals a normal left fallopian tube with ipsilateral loculated spillage (arrows). The right tube has been removed.

5 432 FLLOPIN TUBE BLOCKGE J Ultrasound Med 13: , 1994 B Figure 3, HSG reveals a normal left fallopian tube with ipsilateral loculated spillage (arrows). The right tube has been removed. B, Pre-HSG EVS shows a tubular fluid-filled structure (arrows) on the left side corresponding to the Joculated spillage. detected prior to HSG and 10 (91 % ) were found after HSG (P 0.003) (Fig. 5). DISCUSSION Normal fallopian tubes are seldom visualized on transabdominal ultrasonography. With the introduction of EVS, significant progress has been made in the approach to the diagnosis and treatment of diseases of the fallopian tube. lthough normal fallopian tubes can only be recognized in the presence of a small amount of pelvic fluid, 10 the course of the proximal portion of the tube often is demonstrated in the absence of pelvic fluid. One of the most significant contributions of EVS in the evaluation of adnexal 8 Figure 4, HSG shows a right-sided moderate hydrosala pinx and left-sided cornual blockage (c11roed arrow). B, Pre+ HSG EVS demonstrates a left hydrosalpinx (arrows), but the right hydrosalpinx was not evident. Post-HSG EVS showed bilateral hydrosalpinges. pathologic conditions is the ability to differentiate between an ovarian and an extraovarian lesion. dilated fallopian tube is best recognized by its extraovarian location, tubular appearance, and, in some cases, the presence of typical longitudinal folds in its ampullary portion (Fig. 6). The dilated tube normally is completely separate from the ovary. n extraovarian cystic adnexal mass lying in close proximity to the ovary could represent a dilated fallopian tube as well as a parovarian cyst. The elongated nature of this cystic structure and especially identification of the longitudinal folds allow distinction from a parovarian cyst. In this series, only 22 of 64 (34%) of the distally blocked tubes were diagnosed by EVS by identifying

I Ultrasound Med 13:429-434, 1994 TRI ET L 433 B Figure 5, HSG shows a mildly dilated left fallopian tube without spillage. The right tube is normal.

6 I Ultrasound Med 13: , 1994 TRI ET L 433 B Figure 5, HSG shows a mildly dilated left fallopian tube without spillage. The right tube is normal. B, The pre-hsg ultrasonogram did not show any abnormality of the fallo pian tubes. However, post-hsg EVS revealed a dilated left tube (arrows). a hydrosalpinx. Hydrosalpinges caused by distal tubal blockage are divided histopathologically into two types: thin-walled and thick-walled. t laparoscopy, the thin-walled hydrosalpinx is grossly distended by copious straw-colored fluid, which makes it appear translucent, whereas in the thick-walled type, the wall is fibrous and the lumen is smaller and contaihs little fluid. In both types, the terminal part of the tube is totally blocked and the fimbriae are obscured.11 The distinction between these two types is important clinically as the pregnancy rate after tubal microsurgery is much lower for thick-walled hydrosalpinges. The low sensitivity of EVS in detecting hydrosalpinx could possibly be due to a combination of lack of recognition of nondistended thick-walled hydrosalpinx and evacuation of its contents from the open proximal end of the tube into the uterine cavity. This theory is supported by the high sensitivity of B Figure 6 typical hydrosalpimc seen on EVS., Longitudinal. B, Transverse. Note the longitudinal folds (arrows) in the ampullary portion of the fallopian tube. EVS performed immediately after an HSG (sensitivity = 91%) in our series. lso, 7% of patients with proximal tubal blockage on HSG showed a hydrosalpinx on EVS, indicating that the tubes also were blocked distally, allowing for accumulation of fluid between the two blocked ends of the tube. We also have noticed this phenomenon in patients who undergo transcervical tuboplasty and who demonstrate distal blockage after opening of the proximal end of the fallopian tube. The routine performance of EVS prior to tuboplasty may allow the physician to be aware of this possibility, although surgical lysis of the distal adhesion can be performed after cannulation of the proximal end by transcervical tuboplasty. In this series, EVS had a specificity of 100% with no false-positive diagnoses of tubal blockage. However, the specificity for the diagnosis of hydrosalpinx was 99.8%. This was because of one false-positive diagnosis of hydrosalpinx by EVS in a patient who, although she had normal fallopian tubes, also had distal blockage due to peritubal loculated spillage

434 FLLOPIN TUBE BLOCKGE J Ultrasound Med 13:429-434, 1994 (Fig. 3).

7 434 FLLOPIN TUBE BLOCKGE J Ultrasound Med 13: , 1994 (Fig. 3). Mistaking a dilated vessel for a hydrosalpinx is a potential pitfall that can be avoided by visualizing slow venous flow on the real-time ultrasonogram or utilizing color or pulsed Doppler to detect higher velocity flows. Other potential pitfalls are a loop of bowel, which can be differentiated by the presence of peristalsis and distinct bowel wall layers, and a paraovarian cyst, which has a thin wall and is rounded or oval in shape. lso, a hydrosalpinx may appear as a complex mass if there are significant adhesions, in which case the lesion does not have the typical elongated tubular appearance and may be mistaken for an ovarian mass. In summary, although EVS is very specific when the diagnosis of hydrosalpinx is made, its sensitivity for the diagnosis of fallopian tube blockage is low. This is likely due to the type of hydrosalpinx and drainage of the blocked tube through its proximal end into the uterine cavity. EVS can detect the associated HSX in a subgroup of patients with cornual blockage. This presumably is due to associated distal blockage. REFERENCES 1. Mendelson EB, Bohm-Velez M, Joseph N, et al: Gynecologic imaging: Comparison of transabdominal and transvaginal sonography. Radiology 166:321, Leibman J, Kruse B, Mcsweeney MB: Transvaginal sonography: Comparison with transabdominal sonography in the diagnosis of pelvic masses. JR 151:89, Mendelson EB, Bohm Velez M, Joseph N, et al: Endometrial abnormalities: Evaluation with transvaginal sonography. JR 150:139, Lewit N, Thaler I, Rottem 5: Uterus: new look with transvaginal sonography. J Clin Ultrasound 18:331, Coleman BG: Transvaginal sonography of adnexal masses. Radiol Clin North m 30:677, Cacciatore B, Senterman UH, Ylostalo P: Diagnosis of ectopic pregnancy by vaginal ultrasonography in combination with a discriminatory serum 13-hCG level of 1000 IU/I (IRP). Br J Obstet Gynaecol 97:904, tri M, Bret P, Tulandi T, et al: Ectopic pregnancy: Evolution after treatment with transvaginal methotrexate. Radiology 185:749, tri M, Bret P, Tulandi T: Spontaneous resolution of ectopic pregnancy: Initial appearance and evolution at transvaginal ultrasound. Radiology 186:83, Yoder I: Disease of the fallopjan tubes. In Yoder I: Hysterosalpingography and Pelvic Ultrasound: Imaging in Infertility Gynecology. Boston, Little, Brown Company, 1988, p Timar-Tritsch IE, Rottem 5: Transvaginal ultrasonographic study of the fallopian tubes. Obstet Gynecol 70:424, Brosens I: Hydrosalpinx. In Brosens I, Gordon G: Tubal Infertility. Philadelphia, JB Lippincott, 1990

Likelihood Ratio of Sonographic Findings in Discriminating Hydrosalpinx from Other Adnexal Masses

Likelihood Ratio of Sonographic Findings in Discriminating Hydrosalpinx from Other Adnexal Masses Sonography of Adnexal Masses Women s Imaging Original Research WOMEN S IMAGING Maitray D. Patel 1 Debra L. Acord 1,2 Scott W. Young 1 Patel MD, Acord DL, Young SW Keywords: adnexal masses, hydrosalpinx,