Diagnosing Uterine Cavity Abnormalities

HSG has a high sensitivity but a low specificity for the diagnosis of uterine cavity abnormalities.⁶ HSG and diagnostic hysteroscopy performed on 336 infertile women showed that HSG had a sensitivity of 98% but a specificity of only 35% due to difficulties distinguishing between polyps and myomas. False-negative results were mild intrauterine adhesions of doubtful clinical significance.⁷ Thus, HSG fulfills the requirements as a good first-line screening test for revealing abnormalities of the uterine cavity, although any abnormalities found will likely need further evaluation to make a definitive diagnosis. Sonohysterography or diagnostic hysteroscopy can distinguish between polyps and submucous myomas, which appear similar on HSG (Fig. 29-2). A uterine septum and a bicornuate uterus cannot be differentiated on an HSG (Fig. 29-3). Evaluation of the external fundal contour by laparoscopy, magnetic resonance imaging (MRI), or three-dimensional ultrasonography is required to make a definitive diagnosis. The arcuate uterus has a mild convex fundal margin and is a normal variant. Extrinsic compression from an intramural fundal myoma may give a similar appearance and is easily recognized on routine transvaginal ultrasonography.

Figure 29-2 A round intrauterine filling defect indicates either a polyp or a leiomyoma (black arrows). This study also shows bilateral hydrosalpinges. There is some preservation of the ampullary mucosal folds within the left tube (white arrows), and the end of the right tube is occluded (arrowheads).

Figure 29-3 A uterine septum cannot be differentiated from bicornuate uterus by HSG.

Other conditions visualized on HSG are adhesions, changes caused by diethylstilbestrol (DES), and adenomyosis. Adhesions appear as irregular filling defects and may be very mild or completely obliterate the cavity (Fig. 29-4). DES was used from the 1940s up to 1971 as prophylaxis for spontaneous abortions. The classic appearance associated with in utero DES exposure is a hypoplastic T-shaped cavity. (Fig. 29-5). Adenomyosis can occasionally be diagnosed by HSG as a cavity with shaggy borders (Fig. 29-6). The sensitivity for detecting this condition is unknown because it is uncommon in younger women and is definitely diagnosed histologically after hysterectomy.

Figure 29-4 Extensive intrauterine adhesions associated with Asherman’s syndrome. Only some irregular areas of contrast are seen above the tip of the cannula.

Figure 29-5 Hypoplastic and T-shaped uterine cavity associated with in utero exposure to diethylstilbestrol.

Figure 29-6 Severe adenomyosis will sometimes appear as contrast filling numerous intramural diverticuli. This condition can be confused with intravasation.

Diagnosing Tubal Abnormalities

An evidence-based study found HSG to be “a valid and accurate diagnostic test to be applied in a general population of subfertile couples to assess tubal patency but an unreliable test for diagnosing tubal occlusion.” Tubal blockage on HSG is not confirmed by laparoscopy in up to 62% of patients, but if HSG suggests patent tubes, tubal blockage is highly unlikely. Laparoscopy is needed to confirm or exclude tubal occlusion on HSG.⁸ It should be noted that laparoscopy is not the perfect gold standard; 2% of patients with bilateral tubal occlusion subsequently conceived spontaneously.⁹ One study noted that 60% of patients with proximal tubal occlusion on HSG were patent on repeat HSG 1 month later.¹⁰

Surprisingly, hydrosalpinges may be both overdiagnosed and underdiagnosed by HSG. They may be only mildly dilated with preservation of mucosal folds or massively dilated with complete loss of the normal intratubal architecture (Fig. 29-7). HSG can also diagnose salpingitis isthmica nodosa. This condition, which predisposes to tubal occlusion and ectopic pregnancy, is similar to adenomyosis of the uterus in that there are diverticuli from the mucosa into the muscularis (Fig. 29-8). HSG is also not an ideal test for diagnosing pelvic adhesions because it detects them in only half of the cases in which they are present.² Adhesions are usually diagnosed on HSG by the presence of loculated spill of contrast medium (Fig. 29-9).

Figure 29-7 A very large hydrosalpinx. It has no potential for repair and should be excised laparoscopically because it can reduce the pregnancy rate resulting from in vitro fertilization by half by compromising endometrial receptivity. In addition, the fluid within the tube is directly embryotoxic.

Figure 29-8 Salpingitis isthmica nodosa is diagnosed by HSG when diverticuli from the mucosa are seen extending into the muscularis in the cornual region of the uterus. The arrows point to these outpouchings.

Figure 29-9 Pelvic adhesions can be diagnosed when contrast media is seen to be loculated behind peritubal adhesions after spill from a patent tube, as seen on the left. The right tube is normal with contrast spilling freely outlining loops of bowel.

INDICATIONS

HSG is part of the basic infertility workup and is normally required in every patient.¹¹ It is still the first-line technique for excluding anatomic defects in the uterine cavity and documenting tubal patency in infertility patients. A 1994 survey of board-certified reproductive endocrinologists in the United States reported that 96% obtained an HSG as part of the basic infertility workup.¹² In addition to its role in assessing the uterus and tubes, HSG may have a therapeutic effect.

Hysterosalpingography has also been the mainstay for diagnosing uterine cavity abnormalities in women with recurrent pregnancy loss. Such abnormalities have been reported in 15% to 27% of these patients. The most common anomaly is the septate uterus; surgical correction improves the live birth rate from 3% to 20% to 70% to 90%.¹³ Because tubal status is not an issue with these patients, HSG may be replaced with sonohysterography.

Other indications for HSG include pretubal anastomosis to determine the length of the isthmic tubal segments and exclude uterine cavity defects. It may also be used for postoperative evaluation of the uterine cavity after hysteroscopic myomectomy, polypectomy, adhesiolysis or septoplasty or to document tubal patency after tubal anastomosis, neosalpingostomy, or ectopic pregnancy. It can aid in the delineation of known or suspected uterine anomalies, such as in DES progeny, as well as confirming proper IUD placement or tubal occlusion after ligation.

Active Pelvic Infection

Pelvic inflammatory disease (PID) is an uncommon but potentially serious complication of HSG. Active cervicitis, endometritis, or salpingitis are absolute contraindications to performing an HSG. Cervical cultures for gonorrhea and chlamydia, a white blood cell count, and erythrocyte sedimentation rate should be obtained to rule out an active infection in patients with adnexal tenderness. They may also be considered in patients with a history of prior PID.

Severe Iodine Allergy

Patients with a history of severe iodine allergy may elect to forgo HSG in favor of diagnostic laparoscopy and hysteroscopy or sonohysterosalpingography. However, the risk of a severe reaction with HSG is small because the volume of contrast (most of which escapes vaginally) is small and it is not injected intravascularly. If a decision is made to proceed with HSG in patients with known iodine allergy, a nonionic water-soluble media such as Hexabrix (ioxaglate), Isovue 370 (iopamidol), or Omnipaque (iohexol) should be used because the incidence of iodine allergy is lower with them.² Patients with severe iodine allergy should be premedicated with prednisone 50mg administered 13 hours before HSG and diphenhydramine 50mg 1 hour before. In a study in 563 patients with a history of anaphylactic reactions to iodine contrast treated with this protocol, there were fewer than 10 reactions, none life threatening.¹⁴

Bleeding

HSG should not be performed while the patient is menstruating because blood clots may give the false impression of an intrauterine filling defect or may occlude the uterotubal ostia. Bleeding also increases the risk for intravasation of contrast and may theoretically predispose to endometriosis by the retrograde flushing of endometrial fragments. Further, performing the procedure in a patient with unscheduled bleeding opens the possibilities that she may have a pregnancy complication or endometrial carcinoma.

Endometrial Carcinoma

It is felt to be prudent to avoid performing HSG in known or suspected cases of endometrial carcinoma for fear of disseminating tumor cells and worsening the prognosis for survival. Studies on using HSG for the diagnosis and follow-up of endometrial carcinoma noted that injecting contrast media under low pressure is unlikely to disseminate tumor cells and that it is doubtful that shed cells have the potential for independent metastasis.¹⁵ The 5-year survival in endometrial carcinoma patients evaluated with HSG was not significantly different in those who had positive washings after the procedure versus those who did not. Venous or lymphatic extravasation also failed to influence survival rates.¹⁶

Pregnancy

HSG is scheduled immediately after menses to prevent performing the procedure in the presence of an unrecognized pregnancy due to the concerns of disrupting the pregnancy as well as radiation exposure. Before the advent of pregnancy testing and ultrasonography, HSG was performed specifically to diagnose pregnancy, with no untoward effects.² HSG during inadvertent pregnancy is uncommonly reported. In a recent case report the calculated dose to the embryo was 3.7 mGy. The authors state that the teratogenic risk with less than 20 mGy during the first trimester does not justify pregnancy termination.¹⁷ The possibility of spontaneous abortion should be discussed with the patient.

Who Should Perform the HSG?

It is always preferable for the gynecologist to perform the study. Patients are comforted by having their physician present during this stressful test. Performing the procedure provides additional information not available from the static films, such as the appearance on initial filling, how rapidly the contrast traversed the tubes, how much contrast was required, and how much pressure was used. The patient can also be turned to her side to better visualize the uterus or tubes or to differentiate a true filling defect from an air bubble. The films should always be obtained for review whenever the study is performed by someone else because radiologists’ reports vary greatly depending on their training and experience. Also, the gynecologist has the advantage of being able to correlate the radiographic findings with the actual pathology at surgery. Adhering to the following protocol will reduce patient risk and discomfort and provide optimal images.

Timing

The procedure should be performed during the window between the end of menses and before ovulation. This will reduce the risk of flushing an embryo out of the uterus or tube, disrupting an implanted embryo, or exposing it to radiation. The endometrium is thin during the early follicular phase, which reduces false positives for filling defects or proximal tubal occlusion due to thick mucosal folds. Waiting for the completion of menses reduces the risk of intravasation, filling defects from clots, and the theoretical risk of inducing endometriosis by retrograde flushing of endometrial fragments through the tubes.

Analgesia

Unfortunately, most patients experience uterine cramping during the HSG, but the entire procedure usually only lasts about 3 minutes and the discomfort resolves rapidly at the end of the procedure. A prospective, double-blind, randomized, controlled trial comparing 1 g of acetominophen to placebo taken 30 minutes before HSG found no significant difference in mean pain scores during or 24 hours after the procedure.¹⁸ Several studies noted that nonsteroidal anti-inflammatory drugs (NSAIDs) did provide significant analgesia for HSG.^19–21

Two randomized, placebo-controlled studies found no difference with intrauterine anesthetics delivered before the HSG. Patients in both studies also received an NSAID, which may have decreased pain enough to mask any beneficial effect of the anesthetic.^22,23 There are no references on the use of paracervical blockade for HSG.

Oil-soluble versus Water-soluble Contrast Media

The water-soluble media and oil-soluble media camps have been arguing the relative merits of their favored contrast for decades; as yet there is no clear winner. Water-soluble media offers better image quality because the higher density oil-soluble media tends to obscure fine details in the uterus and tubal mucosal folds. Also, because water-soluble media dissipates quickly, there is no need for delayed films, whereas 1- to 24-hour delayed films are necessary with oil-soluble media. Oil-soluble media also carries increased risks for oil embolism and granuloma formation. Proponents of oil-soluble media counter that water-soluble media causes increased pain on peritoneal spill, although it is mild and very transient. Water-soluble media with lower osmolality may cause less irritation and discomfort.² The procedure should be performed with the smallest volume of contrast possible to provide the needed information. The biggest argument in favor of oil-soluble media is that it has higher postprocedure pregnancy rates. Most still advocate water-soluble media because it provides better uterine and ampullary mucosal detail and has no serious secondary effects.^3,6

Types of Cannulas

Rigid metal cannulas have been the standard device for performing HSG because they are inexpensive, reusable, and readily available. A balloon catheter can be employed in rare cases when cervical stenosis or an inadequate cervical seal prevent completion of the study with the rigid cannula. In addition to the cost of the disposable balloon catheter, it is more difficult to manipulate the uterus compared to a rigid cannula with a tenaculum. Also, the balloon obscures the lower uterine segment. One study reported that the balloon catheter caused significantly greater pain after the procedure.²⁴ However, a randomized study found that HSG performed with a balloon catheter required less contrast and fluoroscopy time, produced less patient discomfort, and was easier for senior residents to perform. It has the additional advantage of allowing the clinician to perform immediate selective salpingography and transcervical tubal catheterization for proximal tubal occlusion without the need to replace the cannula or reschedule the patient.²⁵