Whatever the cause of tubal damage, recognizing tubal factor infertility is essential in the diagnostic workup of the infertile couple. HSG has been the standard initial test for assessing the uterine cavity and fallopian tubes since 1925 [12], although laparoscopic chromopertubation remains the gold standard for diagnosing tubal disease. HSG is a radiographic imaging procedure to image the uterine cavity and demonstrate tubal patency by injecting radiographic contrast media through the cervix. The technique is easy to learn and perform, is relatively low cost, has an acceptable radiation exposure, and has few complications. Indications and contraindications for HSG are listed in ◘ Tables 23.2 and 23.3. Risks, benefits, and alternatives of the procedure are outlined in ◘ Table 23.4. The most significant additional benefit of HSG is an enhanced post-procedure pregnancy rate. Alternative methods for evaluating tubal condition (such as sonohysterosalpingography or “hystero-contrast-sonography,” transvaginal hydrolaparoscopy, falloposcopy, and salpingoscopy) are used far less often.

HSG should be considered a screening test and should have a high sensitivity, so as not to miss the opportunity to treat an abnormality, but a low false-positive rate to prevent unnecessary additional testing and treatments. The accuracy of a HSG is highly dependent on technique and interpretation. The technical quality of the HSG is important to limit misinterpretations (i.e., eliminating air bubbles that may be confused with a polyp or myoma or using inadequate contrast volume or injection pressure to demonstrate tubal patency).

HSG has a high sensitivity but a low specificity for the diagnosis of uterine cavity abnormalities [13]. 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 [14]. Thus, HSG fulfills the requirements as a good first-line screening test for revealing abnormalities of the uterine cavity, though any abnormalities found will likely need further evaluation to make a definitive diagnosis. A uterine septum and a bicornuate uterus cannot be differentiated on a HSG. Evaluation of the external fundal contour by laparoscopy, MRI, or 3D ultrasonography is required to make a definitive diagnosis. Other conditions visualized on HSG are adhesions, diethylstilbestrol (DES) changes, and adenomyosis.

HSG appears to be a highly valid and accurate test for assessing tubal patency in subfertile couples; however, its reliability for diagnosing tubal occlusion is questionable. Tubal blockage on HSG is not confirmed by laparoscopy in up to 62% of patients. Laparoscopy is needed to confirm or exclude tubal occlusion on HSG [15]. Even laparoscopy is imperfect, as 2% of patients with bilateral tubal occlusion subsequently conceived spontaneously [16]. One study noted that 60% of patients with proximal tubal occlusion (PTO ) on HSG showed patency on repeat HSG 1 month later [17].

Surprisingly, hydrosalpinges may also be poorly diagnosed by HSG. When detected, they may be only mildly dilated with preservation of mucosal folds or massively dilated with complete loss of the normal intratubal architecture (◘ Fig. 23.1). HSG can also diagnose salpingitis isthmica nodosa , represented by diverticuli from the mucosa into the muscularis (◘ Fig. 23.2). HSG is also not an ideal test for diagnosing pelvic adhesions because it detects them in only one-half of the cases in which they are present [12]. Adhesions are usually diagnosed on HSG by the presence of loculated spill of contrast.

If possible, the gynecologist should perform the study; patients are comforted by having their physician present during this stressful test. At a minimum, the films should be obtained for review as the radiologists’ reports vary greatly depending on training and experience. The procedure should be performed during the window between the end of menses but before ovulation. Although most patients experience uterine cramping during the HSG, the duration of the procedure is short and the discomfort generally resolves rapidly. Several studies have noted that NSAIDs do provide significant analgesia for HSG [18–20]. Proponents of water-soluble media (WSM) vs. oil-soluble media (OSM) have been arguing the relative merits of their favored contrast for decades and as yet there is no clear winner. WSM offers better image quality as the higher density OSM tends to obscure fine details in the uterus and tubal mucosal folds. Also, since WSM dissipates quickly, there is no need for delayed films, whereas 1–24 h delayed films are necessary with OSM. OSM also carries increased risks for oil embolism and granuloma formation. OSM has been claimed to have higher post-procedure pregnancy rates, although this finding has been questioned in the current Cochrane analysis [21].

The HSG cannula is attached to a 20-cc syringe filled with contrast media, which is flushed to expel air and prevent air bubble artifacts. The acorn is advanced so that it is about 1 cm from the end of the cannula. The cannula tip should not go beyond the internal os. Using a lubricated, open-sided bivalve speculum, the cervix is cleansed with an antiseptic solution, a single-tooth tenaculum is applied to the anterior lip, and the cannula tip is inserted into the cervical canal. Gentle upward pressure on the cannula while pulling downward on the tenaculum will seal the cervix and straighten the uterine axis.

Contrast media is then injected slowly. If a filling defect is suspected to be an air bubble, the patient can be rotated to her side (defect side down). A polyp or myoma will remain stationary, whereas a bubble will rise to the elevated side. Only 5–10 mL is usually required to complete the study. Patients should be observed for several minutes afterwards for bleeding and signs of vasovagal or allergic reactions.

While bilateral PTO is usually indicative of anatomical pathology, unilateral PTO is frequently transient due to spasm of the uterotubal ostium, plugging by mucus, debris, or air bubbles. Unilateral PTO is found in 10–24%, but 16–80% are patent on repeat HSG or laparoscopy with chromotubation [22]. Increasing the hydrostatic pressure and rotating the patient may establish patency during HSG showing PTO. The use of antispasmodic agents such as glucagon to prevent PTO has also been recommended, but this practice is supported only by limited anecdotal reports in the literature.

Tubal disease may be present at multiple sites; however, tubal damage or blockages are often categorized as proximal, mid-tubal, or distal. The success of surgical management will depend on the location and degree of the damage in addition to nonsurgical factors, such as patient age and ovarian reserve. In general, patients with extensive tubal damage will be best served by IVF, whereas those without widespread disease may be good surgical candidates.

Minimally invasive surgical techniques are most appropriate for tubal surgery, but this approach requires some advanced training as well as meticulous dissection and hemostasis to prevent adhesions or further damage to the tubes.

PTO is most often caused by salpingitis isthmica nodosa, chronic pelvic infection, intratubal endometriosis, mucus plugging, or anatomic malformations [23]. Persistent PTO can be treated by therapies such as selective salpingography, tubal cannulation under fluoroscopy, hysteroscopic cannulation, or tubal resection and reanastomosis with a high degree of success.

Selective salpingography is performed by injecting contrast media through a transcervical catheter positioned at the tubal ostia hysteroscopically or under fluoroscopic guidance (◘ Fig. 23.3). Patency is confirmed by laparoscopy or fluoroscopy, respectively. Often, the hydrostatic pressure will relieve an obstruction. If unsuccessful, this is immediately followed with the introduction of a smaller catheter with an atraumatic guidewire through the selective salpingography catheter. The inner catheter and guidewire are advanced through the tubal ostia into the proximal isthmus. Chromotubation is then performed though the catheter to demonstrate patency through the fimbria. Although there is a high success rate of 75–95% in terms of achieving patency, reocclusion rates average 30% [23–25]. Despite similar tubal patency rates between hysteroscopic and fluoroscopic tubal cannulation, pregnancy rates are significantly higher with the former technique, 48.9% vs. 15.6% [23]. Tubal perforation occurs 2–10% of the time but is innocuous. Although microsurgical resection and reanastomosis can be used when tubal cannulation fails or when PTO is due to salpingitis isthmica nodosa, IVF is the preferred treatment.

Tubal sterilization is the most common contraceptive used by women worldwide [26]. Although it should be considered a permanent birth control option, 5–20% of women experience regret [27], and 1–2% of women seek reversal of sterilization [28]. In a large prospective, multicenter cohort study of over 11,000 women, regret appeared most strongly associated with younger age at the time of sterilization [27]. Compared with IVF, tubal anastomosis offers patients desiring sterilization reversal the advantage of a one-time, minimally invasive treatment with high success rates. In addition, some patients prefer the ability to attempt conception each month as well as to conceive more than once as well as the avoidance of the inconvenience and risks of IVF. The disadvantages are the possibility of bleeding, infection, inadvertent injury to other organs, and anesthetic complications inherent with any surgery. There is also a higher risk for ectopic pregnancy following tubal reanastomosis. Other factors involved in the decision are the IVF program’s success rates, surgeon’s ability, presence of other infertility factors, cost, and patient preference. A retrospective cohort study reported significantly higher cumulative pregnancy rates for tubal anastomosis compared to IVF for women less than 37 years of age, but there was no significant difference in women aged 37 years or older [29]. Furthermore, the average cost per delivery for tubal anastomosis was almost half that of IVF.

The procedure is performed by first mobilizing and opening the occluded tubal ends. A stitch is then placed in the mesosalpinx beneath the tubal ends to align them and relieve tension on the anastomosis. The anastomosis is accomplished with interrupted sutures through the tubal muscularis (◘ Fig. 23.4). The serosa is also repaired with interrupted stitches. Transcervical chromotubation is used to confirm tubal patency. A HSG is recommended if the patient has not conceived within six cycles postoperatively. Tubal anastomosis has usually been performed utilizing an operating microscope by laparotomy with an overnight hospitalization, but outpatient minilaparotomy achieves the same success rate with less cost and discomfort.

Cumulative pregnancy rates after tubal reanastomosis in women under age 40 range from 70% to over 90% [28, 30–32]. Even women ages 40–45 years have good success rates of 13–70%, with subsequent ectopic risks of 2–10% [28, 30, 32–35]. Most of the pregnancies occur within the first year of surgery. High success rates were also reported with laparoscopic tubal anastomosis, though few possess the skills to perform this [28, 31]. Robotic surgery is a technique that facilitates laparoscopic tubal reanastomosis. Two small studies comparing tubal anastomosis by robotic-assisted laparoscopy vs. laparotomy or minilaparotomy found no difference in pregnancy rates, but robotic surgery took significantly longer and cost more, even compared with laparotomy with overnight hospitalization, though recovery was quicker after robotic surgery [36, 37].

The most frequent site for tubal infertility is the distal tube, which manifests as hydrosalpinges and fimbrial phimosis (agglutination of fimbria leading to a narrowed tubal opening). As described earlier, these most often follow PID but can also be preceded by peritonitis from any cause, trauma from previous surgery, or endometriosis. The decision to attempt repair of distal tubal disease is often made at the time of surgery, and categorizing patients as either favorable or poor prognosis may be helpful in guiding management. Tubal reconstruction should not be considered in patients with both proximal and distal occlusion or for women with severe disease, and patients should be counseled preoperatively about the possibility of neosalpingostomy/fimbrioplasty vs. salpingectomy depending on the surgical findings. According to American Fertility Society classifications, a favorable-prognosis patient generally has minimal adnexal adhesions, tubal dilation <3 cm, preservation of normal tubal walls, and endosalpinx [38]. In contrast, poor-prognosis patients may have dense tubal adhesions; significantly dilated tubes; thickened, fibrotic tubal walls; and damaged mucosa.

For favorable-prognosis patients, a neosalpingostomy or fimbrioplasty can be attempted to open up a hydrosalpinx or narrowed tubal opening (◘ Fig. 23.5). This is accomplished laparoscopically, with the release of adhesions and incision at the distal end of the occluded or narrowed tube. The mucosa and fimbria are everted and attached to the serosa using fine suture or electrosurgery. Transcervical chromopertubation is used to confirm patency at the completion of the procedure. Depending on the severity of the disease, pregnancy rates can range from 58% to 77% in favorable-prognosis patients, with ectopic rates of 2–8% [39]. As expected, for poor-prognosis patients, pregnancy rates fall to 0–22%, with ectopic rates of 0–17% [39]. It should be noted that although patency can often be achieved surgically in both favorable- and poor-prognosis patients, the irreversible damage caused by pelvic infection to the endosalpinx may account for compromised tubal function after surgery.

The harmful effect of hydrosalpinx on IVF outcomes has been well documented [40–43]. This observation may be explained by toxic effects of the hydrosalpinx fluid on the embryo, flushing of the embryo from the endometrium by hydrosalpinx fluid, or impaired endometrial receptivity. A meta-analysis of over 5500 women showed that the implantation and delivery rate per transfer is halved and the miscarriage rate is increased in women with untreated hydrosalpinx undergoing IVF [44]. Several prospective randomized trials have demonstrated that hydrosalpinges treated with salpingectomy prior to IVF result in restoration of comparable pregnancy rates to controls [45–47]. This finding has been replicated even in patients with unilateral salpingectomy for unilateral hydrosalpinx [48]. A large hydrosalpinx visualized by ultrasound should be removed prior to IVF, as these appear to be associated with the poorest outcomes [44, 45, 49]. An area of greater controversy is whether less pronounced hydrosalpinges (such as those identified by HSG or on laparoscopy) should also be removed in this circumstance.

The gold standard for treatment of hydrosalpinx prior to IVF is laparoscopic salpingectomy; however, other techniques such as proximal tubal ligation or occlusion, ultrasound-guided drainage, or salpingostomy have been suggested. Salpingectomy prior to IVF should be performed laparoscopically, taking care to remain close to the tube to preserve the ovarian blood supply and maintain ovarian reserve. PTO has been demonstrated to be an effective alternative to salpingectomy in cases where salpingectomy is technically difficult or surgery is contraindicated [47, 50]. Pregnancy rates following tubal occlusion appear comparable to treatment with salpingectomy [51]. Most recently, several small case series have demonstrated the effectiveness of using tubal inserts to hysteroscopically occlude hydrosalpinges [52, 53]; however, persistent coils within the endometrial cavity may theoretically limit the success of subsequent ART cycles.

Women with a history of tubal adhesions are at an increased risk of ectopic pregnancy . Ectopic pregnancy is defined by the abnormal implantation of an embryo outside of the endometrial cavity. These pregnancies represent approximately 1.55–2% of pregnancies and most frequently affect the fallopian tube (>90%), although they can uncommonly involve sites such as the abdomen, ovary, cervix, or cesarean scar [54]. Further, within the tube itself, ectopic pregnancies have a predilection for the ampullary portion of the tube where fertilization occurs (70%); an additional 10% occur in the isthmus, 10% in the fimbria, and 2% in the uterine cornu or interstitium [55]. Although rates of ectopic pregnancy in the USA appear to have peaked and plateaued in the 1990s [56], the true incidence is difficult to estimate because these pregnancies are increasingly managed as outpatients and therefore may not be included in hospital databases. It does appear that enhanced awareness and improved detection methods have resulted in more favorable outcomes. Nevertheless, ectopic pregnancies remain the leading cause of maternal deaths in the first trimester [57] and must be recognized and managed promptly. Failure to do so can result in fallopian tube rupture, intraperitoneal hemorrhage, shock, and even death.

Abnormal implantations are thought to result mainly from inflammation or blockage within the tubal lumen. As with infertility, most tubal damage that precedes ectopic pregnancy is caused by infection with N. gonorrhoeae or C. trachomatis. As many as half of women with ectopic pregnancies will have no identifiable risk factors [58]. Known risk factors for ectopic pregnancy include infection, prior ectopic pregnancy, and prior tubal surgery. Although IUD use does not increase the overall risk of ectopic pregnancy, a positive pregnancy test in an IUD user warrants suspicion for ectopic, as the location of the gestation is most likely extrauterine [58]. A complete list of maternal risk factors is shown in ◘ Table 23.5. It has also been theorized that chromosomally abnormal embryos may have a higher rate of inappropriate implantation. However, more recent studies with larger patient numbers are not consistent with earlier case reports showing high percentages of abnormal karyotypes from ectopic pregnancies [59, 60].