| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 Departments of Radiology (T.A.T., H.J.A., P.T.K.)
2 Obstetrics and Gynecology (P.M., J.P.)
3 Pathology (T.P.), Helsinki University Central Hospital and Haartman Institute, Haartmaninkatu 4, FIN-00290 Helsinki, Finland.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Thirty consecutive patients hospitalized because they were clinically suspected of having PID underwent transvaginal US and T1-weighted spin-echo, T2-weighted turbo spin-echo, and inversion-recovery MR imaging at 1.5 T. All patients underwent laparoscopy after MR imaging.
RESULTS: PID was laparoscopically proved in 21 (70%) patients. The MR imaging diagnosis agreed with that obtained with laparoscopy in 20 (95%) of the 21 patients with PID. The imaging findings for PID were as follows: fluid-filled tube, pyosalpinx, tubo-ovarian abscess, or polycystic-like ovaries and free pelvic fluid. Findings at transvaginal US agreed with those at laparoscopy in 17 (81%) of the 21 patients with PID. The sensitivity of MR imaging in the diagnosis of PID was 95%, the specificity was 89%, and the overall accuracy was 93%. For transvaginal US, the corresponding values were 81%, 78%, and 80%.
CONCLUSION: MR imaging is more accurate than transvaginal US in the diagnosis of PID and provides information about the differential diagnosis of PID. MR imaging may reduce the need for diagnostic laparoscopy.
Index terms: Fallopian tubes, torsion, 853.2179 • Magnetic resonance (MR), comparative studies, • Pelvic organs, abnormalities, 85.2171, 85.2172, 85.2173, 85.2174, 85.2179 • Pelvic organs, abscess, 85.2174 • Pelvic organs, inflammation, 85.217 • Pelvic organs, MR, 85.121411, 85.121413 • Pelvic organs, US, 85.12989
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Laparoscopy has long been the standard of reference in the diagnosis of PID (4,6), but it requires general anesthesia. Laparoscopy is usually performed in patients with moderate to severe pelvic pain (6). Laparoscopy performed to diagnose PID is an invasive procedure and may lead to complications (4). Endometrial biopsy is less invasive than laparoscopy, but the results are not readily available (7,8).
Transvaginal ultrasonography (US) is a noninvasive bedside procedure that is routinely performed in patients with pelvic pain (7). Earlier studies have shown that transvaginal US performs well in the diagnosis of PID when the criteria include thickened fluid-filled tubes (8). Transvaginal US is superior to transabdominal US in the diagnosis of endometrial abnormalities, pelvic masses, and PID (9–11).
One option for the noninvasive diagnosis of PID is magnetic resonance (MR) imaging. Tubal enlargement can be easily seen on MR images and is characterized by the tortuous folding of fluid-filled structures on T2-weighted images (12).
Although computed tomography (CT) is commonly used for imaging the abdomen, MR imaging has the potential to replace a considerable portion of abdominal CT examinations (13–15). CT has been used in the diagnosis of complex tubo-ovarian abscesses (16) and has proved to be useful in the imaging of PID (17,18). CT, however, exposes patients to ionizing radiation, which is problematic among young women.
MR imaging plays a limited role in gynecologic imaging. Although MR imaging can be used in the characterization of adnexal masses (19–21), to our knowledge its role in the standard work-up of common adnexal disorders has not yet been well defined (15). In one study (22), MR imaging was less specific than transvaginal US in the assessment of pelvic masses. Only a few studies have used MR imaging for diagnosing gynecologic infections (15,23). MR imaging provides good contrast between various pelvic organs and may perform better than CT in soft-tissue imaging. The multiplanar capability of MR imaging is a particular advantage when compared with CT (15). The recent development of fast MR imaging technology has shortened the imaging time required to study the female pelvis.
To our knowledge, no prospective, comparative studies of MR imaging and laparoscopy in the diagnosis of PID have been performed. The purpose of this study was to evaluate the usefulness of MR imaging in the diagnosis of PID and to compare MR imaging findings with those from laparoscopy and transvaginal US.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Ultrasonography
All patients underwent routine bedside transvaginal US with a 6.5-MHZ (RT-X 2000; GE Medical Systems, Milwaukee, Wis) or 5–9-MHZ (HDI 3000; ATL, Bothell, Wash) transducer. US was performed in the emergency room by an experienced gynecologist, who also admitted the patient to the hospital. The presence of free pelvic fluid, a fluid-filled tubal lumen, dilated fallopian tubes, and an adnexal mass and the characterization of the adnexal mass was recorded during real-time examination by an experienced gynecologist (24). The mass was considered to be a fluid-filled tube or pyosalpinx when a tortuous, fluid-filled tubal lumen was identified. A tubo-ovarian abscess was defined as an ill-defined, fluid-containing adnexal mass. The presence of thickened walls in the mass or dilated tube was systematically recorded.
MR Imaging
Within 24 hours of undergoing sonography, all patients underwent MR imaging at the department of radiology. The radiologists (T.A.T., H.J.A.) who interpreted the MR imaging findings was blinded to the sonographic, clinical, and laboratory findings. The minor pelvic cavity was examined with a 1.5-T imager (Vision; Siemens, Erlangen, Germany). Phased-array receiver coils were used, and patients were imaged in the supine position. Sagittal T2-weighted turbo spin-echo images were obtained to cover the minor pelvic cavity with a repetition time of 4,100 msec, an echo time of 99 msec (4,100/99), an echo train length of seven, two signals acquired, and an acquisition time of 3 minutes 4 seconds. The section thickness was 5 mm, and the intersection gap was 1 mm. The same sagittal images were obtained with a turbo inversion-recovery sequence (4,900/60/150 [inversion time, msec]) with three signals acquired and an acquisition time of 4 minutes 26 seconds. T2-weighted turbo spin-echo images (5,600/132) were obtained in the axial oblique plane with an echo train length of 15, three signals acquired, an acquisition time of 5 minutes 8 seconds, 5-mm-thick sections, and a 1-mm gap. The axial oblique plane was selected so that it was perpendicular to the long axis of the uterus. The phase-encoding direction was from head to feet. The same axial oblique sections were also obtained with a turbo short-inversion-time inversion-recovery (STIR) sequence (6,500/60/150, three signals acquired, acquisition time of 4 minutes 40 seconds) and a T1-weighted spin-echo sequence (680–792/14, two signals acquired, acquisition time of 3 minutes 34 seconds). A rectangular field of view of 200 x 320 mm was used for all images. A 242 x 512 matrix was used for the sagittal T2-weighted sequence, and a 156 x 256 matrix was used for all other sequences. No antiperistaltic agent was used. The imaging time was 21 minutes, and the total examination time needed was approximately 30 minutes (including the time needed for positioning).
MR Image Interpretation
The presence of a fluid-filled tubal lumen, dilated tubes, adnexal mass, or polycystic-like ovaries and free pelvic fluid in the cul-de-sac or the minor pelvic cavity was systematically studied independently by two radiologists (T.A.T., H.J.A.) without knowledge of the laparoscopic or laboratory findings. Discrepancies in interpretation were resolved by consensus. Interobserver reproducibility was calculated with Cohen 
statistics. values were 0.78 for a fluid-filled tube and 0.90 for free pelvic fluid. There was no disagreement between the readers about the presence of adnexal masses or polycystic-like ovaries.
MR images were also evaluated with regard to the morphologic appearance and signal intensity characteristics of the mass (21,22,25). Urine was used as a reference for high signal intensity on T2-weighted images and low signal intensity on T1-weighted images. Fat was used as a reference for high signal intensity on T1-weighted images. A fluid-filled tube (Fig 1) was diagnosed when a tortuous, fluid-filled tubal lumen was identified; it had low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. A pyosalpinx (Fig 2) was diagnosed when the tubal lumen was dilated and the walls were thickened; it had the same signal intensity characteristics as those of a fluid-filled tube. Hydrosalpinx had a similar appearance to that of pyosalpinx, but it had thin walls (21). Criteria for tubo-ovarian abscess included an ill-defined adnexal mass with thick, irregular walls containing fluid (Fig 3). An abscess usually had low signal intensity on T1-weighted images and high signal intensity on T2-weighted images, but the signal intensity varied from low signal intensity on T1-weighted images to isointensity or even high signal intensity on T2-weighted images. In addition, the abscess was heterogeneous on some T2-weighted images (23,26). The presence of a fluid-filled tubal lumen, if it could be distinguished from the adnexal mass, was noted. The diagnosis of polycystic-like ovaries was based on the presence of multiple, small (2–10-mm) follicles scattered within increased ovarian stroma in an enlarged ovary (8,24,27,28). PID was diagnosed at MR imaging if any of the following were detected: a tubo-ovarian abscess (23,26), a pyosalpinx (15,21), a fluid-filled tube (15) (major findings), or polycystic-like ovaries (8,24,28) with free pelvic fluid (8,29) (minor findings).
|
|
|
|
|
|
|
|
Laparoscopy
Laparoscopy was performed immediately after MR imaging (4,32). The minimum criteria for the visual diagnosis of acute PID were as follows: pronounced hyperemia of the tubal surface, edema of the tubal wall, and sticky exudate on the tubal surface and from the fimbriated ends when patent (4). In some cases, PID can be limited to endometritis only. The presence or absence of endometritis without salpingitis or frank tubo-ovarian abscess formation, however, was not considered in this study. Peritoneal lavage was performed. Surgical specimens were obtained from the adnexal mass in 18 patients, fixed in formalin, and embedded in paraffin. Three- to four-micrometer-thick slices were cut and routinely stained with hematoxylin-eosin or van Gieson. Cytologic peritoneal lavage samples were stained with May-Grünwald-Giemsa.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
Of the nine patients without PID (Table 2), three had tubal torsion (Fig 4), one had a simple cyst, one had a dermoid cyst, one had endometrioma (Fig 5), and two had free pelvic fluid on MR images. One patient had no signs of a gynecologic disorder on MR images.
|
|
|
|
|
|
|
Transvaginal US findings were consistent with the diagnosis of PID in 17 (81%) of the 21 patients with laparoscopically proved PID and two of the nine patients without PID. Of the latter two patients, one had an endometrioma and the other had tubal torsion. Both patients were believed to have tubo-ovarian abscesses at transvaginal US.
Thus, when comparing MR imaging with laparoscopy in the diagnosis of PID, MR imaging had a sensitivity of 95%, a specificity of 89%, and an accuracy of 93%. For transvaginal US, the corresponding figures were 81%, 78%, and 80%.
Comparison of MR Imaging and Transvaginal US
Three cases of PID were missed with transvaginal US. In two patients (patients 2 and 16), abscesses were misinterpreted as an ovarian tumor. In one patient (patient 7), a pyosalpinx was missed. Among patients without PID, an endometrioma and a torsion were misinterpreted as abscesses (patients 19 and 29). One case of PID (patient 18) was missed with MR imaging, and there was one false-positive diagnosis of PID with MR imaging (patient 15).
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In the present study, MR imaging was sensitive and specific in the diagnosis of PID when laparoscopy was used as a standard of reference. We emphasize that MR imaging was useful not only for establishing the diagnosis of PID but also for detecting other processes accounting for the patients' clinical presentation.
The combination of T1-weighted, T2-weighted, and STIR images allowed us to further characterize the detected lesions. We used both sagittal and axial oblique sections, with the axial sections perpendicular to the plane of the uterus to standardize the orientation of the images. This is important because straight axial or coronal planes are not optimal for imaging the adnexa (19). We found the combination of axial oblique and sagittal images most useful in the differential diagnosis of fluid-filled bowel loops, fluid-filled tubes, and free pelvic fluid. It may be difficult to differentiate fluid-filled fallopian tubes from other cystic structures with transvaginal US. Transvaginal US is not very accurate in the demonstration of small quantities of fluid in the minor pelvic cavity (34). Fluid is easy to detect with MR imaging because subcutaneous tissue or pelvic organs do not affect the signal.
Visualization of normal fallopian tubes with MR imaging seems difficult (15), but fluid-filled fallopian tubes have high signal intensity on T2-weighted and STIR images. Fluid-filled bowel loops may sometimes mimic fluid-filled tubal lumina on MR images, but, because the motility of the bowel usually generates motion artifacts, differentiation is not difficult. Furthermore, bowel extends outside the minor pelvic cavity. In acute cases when dilated fallopian tubes are detected, it is extremely important to be able to differentiate tubal torsion from a pyosalpinx. A tubal torsion and a hematosalpinx may have a similar appearance to that of a fluid-filled tube on T2-weighted and STIR images; on T1-weighted images, however, a fluid-filled tube has low signal intensity. Layering is common with hemorrhagic lesions, and tubal torsion may have a comma-shaped appearance. Three (10%) of our 30 patients had a tubal torsion, and all had high signal intensity within the lesion on T1- and T2-weighted images and layering on T2-weighted images. A pyosalpinx may have a similar appearance to that of a hydrosalpinx, but a hydrosalpinx usually has thinner walls than a pyosalpinx (21) and may be less tortuous. An abscess usually has low signal intensity on T1-weighted images and high signal intensity on T2-weighted images, but there can be a large variation in the signal intensity on T1-weighted images and heterogeneity on T2-weighted images. Thick, irregular walls are typical of abscesses. Layering has been described in hemorrhagic lesions (31) and can be considered a criterion in the diagnosis of hemorrhagic or infectious lesions (22). Although patients with large masses that have irregular, thick walls must undergo laparoscopy and be properly evaluated because of the possibility of malignancy, MR imaging enables us to characterize adnexal masses. Polycystic-like ovaries are diagnosed if multiple, small follicles are scattered within increased ovarian stroma in an enlarged ovary; however, multiple follicles can be seen in a normal ovary as well.
In the present study, MR imaging was more accurate than transvaginal US in the diagnosis of PID. Transvaginal US was performed in the emergency room by the gynecologist who admitted the patient to the hospital. In clinical practice, treatment decisions are usually based on this information. We compared unenhanced MR imaging with real-time transvaginal US. Intravenous and oral contrast media have been useful (21,35,36), but we wanted to keep the MR examination as noninvasive, fast, and cost-effective as possible.
The study population was highly selected—all patients were hospitalized because they were clinically suspected of having PID. Larger studies among unselected women with pelvic pain are needed to test the overall utility of MR imaging. More studies are also needed to determine the effect of gynecologic MR imaging on the treatment decisions and net cost (37). Many physicians argue that transvaginal US is far more cost-effective than MR imaging. However, MR imaging may still be cost-effective if proper therapy is provided early in the disease course and if explorative laparoscopy or laparotomy can be avoided on the basis of the information obtained with MR imaging. Fast and ultrafast techniques can shorten the imaging time, and this may further lower the cost of MR imaging (38).
In conclusion, MR imaging was accurate in the diagnosis of PID and in the detection of other processes accounting for the patients' clinical presentation. MR imaging may reduce the need for diagnostic laparoscopy.
|
Footnotes |
|---|
Address reprint requests to T.A.T.
From the 1997 RSNA scientific assembly.
Abbreviations: PID = pelvic inflammatory disease STIR = short inversion time inversion recovery
Author contributions: Guarantors of integrity of entire study, T.A.T., H.J.A., J.P.; study concepts and design, T.A.T., H.J.A., T.P.; definition of intellectual content, H.J.A., J.P.; literature research, T.A.T.; clinical studies, T.A.T., P.T.K., P.M., J.P.; experimental studies, T.A.T., P.T.K.; data acquisition, T.A.T., P.T.K., P.M., J.P.; data analysis, T.A.T., H.J.A., P.T.K., P.M.; statistical analysis, T.A.T., H.J.A., P.T.K.; manuscript preparation, T.A.T., H.J.A., P.T.K., P.M., T.P., J.P.; manuscript editing, T.A.T., H.J.A., J.P.; manuscript review, H.J.A., J.P.
Received December 5, 1997;
revision requested February 24, 1998; revision received May 28, 1998;
accepted July 20, 1998.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. A. Telischak, B. M. Yeh, B. N. Joe, A. C. Westphalen, L. Poder, and F. V. Coakley MRI of Adnexal Masses in Pregnancy Am. J. Roentgenol., August 1, 2008; 191(2): 364 - 370. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Balogun Pelvic infections Imaging, December 1, 2007; 19(4): 369 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Kitamura, S. M. Ascher, C. Cooper, S. J. Allison, R. C. Jha, P. A. Flick, and J. B. Spies Imaging Manifestations of Complications Associated with Uterine Artery Embolization RadioGraphics, October 1, 2005; 25(suppl_1): S119 - S132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Oto, R. D. Ernst, R. Shah, M. Koroglu, G. Chaljub, A. F. Gei, N. Zacharias, and G. Saade Right-Lower-Quadrant Pain and Suspected Appendicitis in Pregnant Women: Evaluation with MR Imaging--Initial Experience Radiology, February 1, 2005; 234(2): 445 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Imaoka, A. Wada, M. Matsuo, M. Yoshida, H. Kitagaki, and K. Sugimura MR Imaging of Disorders Associated with Female Infertility: Use in Diagnosis, Treatment, and Management RadioGraphics, November 1, 2003; 23(6): 1401 - 1421. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M Balogun Pelvic infections Imaging, June 1, 2003; 15(2): 72 - 78. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Molander, P. Finne, J. Sjoberg, J. Sellors, and J. Paavonen Observer Agreement With Laparoscopic Diagnosis of Pelvic Inflammatory Disease Using Photographs Obstet. Gynecol., May 1, 2003; 101(5): 875 - 880. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nagayama, Y. Watanabe, A. Okumura, Y. Amoh, S. Nakashita, and Y. Dodo Fast MR Imaging in Obstetrics RadioGraphics, May 1, 2002; 22(3): 563 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D C Ross Outpatient antibiotics for pelvic inflammatory disease BMJ, February 3, 2001; 322(7281): 251 - 252. [Full Text]
|
|
|
|
|
|
M. Dohke, Y. Watanabe, A. Okumura, Y. Amoh, T. Hayashi, T. Yoshizako, M. Yasui, S. Nakashita, J. Nakanishi, and Y. Dodo Comprehensive MR Imaging of Acute Gynecologic Diseases RadioGraphics, November 1, 2000; 20(6): 1551 - 1566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kinnunen, P. Molander, A. Laurila, I. Rantala, R. Morrison, M. Lehtinen, R. Karttunen, A. Tiitinen, J. Paavonen, and H.-M. Surcel Chlamydia trachomatis reactive T lymphocytes from upper genital tract tissue specimens Hum. Reprod., July 1, 2000; 15(7): 1484 - 1489. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Guerriero, S. Ajossa, M.P. Lai, V. Mais, A.M. Paoletti, and G.B. Melis Transvaginal ultrasonography associated with colour Doppler energy in the diagnosis of hydrosalpinx Hum. Reprod., July 1, 2000; 15(7): 1568 - 1572. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I Simms and J M Stephenson Pelvic inflammatory disease epidemiology: what do we know and what do we need to know? Sex. Transm. Inf., April 1, 2000; 76(2): 80 - 87. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
