AJR 2000; 175:1173-1176
© American Roentgen Ray Society
Improvement of Hysterosalpingographic Accuracy in the Diagnosis of Peritubal Adhesions
Anna Lia Valentini1,
Ludovico Muzii2,
Riccardo Marana2,
Giovan Fiore Catalano2,
Vincenzo Summaria1,
Fabrizio Felici1,
Alfonso Rossetti2 and
Carmelo Destito3
1
Department of Radiology, Università Cattolica
del Sacro Cuore di Roma, Policlinico A. Gemelli, Largo A. Gemelli 8, 00168
Roma, Italy.
2
Department of Obstetrics and Gynecology,
Università Cattolica del Sacro Cuore di Roma,
00168 Roma, Italy.
3
Department of Surgery, Università Cattolica
del Sacro Cuore di Roma, 00168 Roma, Italy.
Received September 8, 1999;
accepted after revision March 23, 2000.
Address correspondence to A. L. Valentini.
Abstract
OBJECTIVE. Our purpose was to compare hysterosalpingography with
laparoscopy in the diagnosis of peritubal adhesions and to verify whether a
combination of radiographic signs improves hysterosalpingographic
accuracy.
SUBJECTS AND METHODS. Thirty candidates for laparoscopy underwent
hysterosalpingography before surgery. Two radiologists evaluated the presence
or absence and types of radiographic signs of peritubal adhesions (convoluted
tubes, vertical tubes, loculation of contrast medium in peritoneum, halo
effect, and fixed laterodeviation of the uterus) using two different criteria
for normality or abnormality: no sign means a normal result, one or more signs
mean an abnormal result (first criterion); no sign or one sign means a normal
result, two or more signs mean an abnormal result (second criterion).
Interpretation discrepancies were resolved by consensus. Peritubal and
periovarian adhesions were evaluated by a single operating surgeon during
laparoscopy (recorded on S-VHS videotape) and by a different surgeon reviewing
the videotape. The radiographic results obtained using the two criteria in
radiologically patent as well as in distally nonpatent tubes were compared
with corresponding laparoscopic results by 2 x 2 tables and were
statistically analyzed (kappa statistics).
RESULTS. The first criterion displayed poor diagnostic accuracy. The
correlation with laparoscopy was not statistically significant in either
radiologically patent or distally nonpatent tubes. The second criterion
greatly improved the agreement with laparoscopy, but only in patent tubes
(
= 0.7789; p < 0.001).
CONCLUSION. Hysterosalpingographic accuracy in peritubal adhesion
diagnosis can be improved in patent tubes by taking into account more than one
of the reported radiographic signs.
Introduction
Peritubal adhesions are a significant cause of infertility in women,
altering the normal anatomic relationship between ovarian fimbriae and ovary
and interfering or preventing the normal capture and transport of the ovum
[1]. In peritubal adhesion
diagnosis, hysterosalpingography is considered less accurate than laparoscopy,
which shows a direct view of the pelvic abnormality. High rates of
false-positive and false-negative results are found on hysterosalpingography
compared with laparoscopy [2].
In a retrospective study comparing these two techniques, Karasick and Goldfarb
[1] identified the following
signs as radiographic findings corresponding to peritubal disease in 75% of
their laparoscopic control subjects: convoluted tubes, vertical tubes,
loculation of contrast medium in peritoneum, halo effect (double-contour
appearance of the tubal wall), and ampullary dilatation in patent tubes.
Karasick and Goldfarb did not include the fixed laterodeviation of the uterus
in the list of the radiographic signs of peritubal adhesions. Nevertheless,
they described the laterodeviation of the uterus as a possible cause of
stretched or convoluted tubes in peritubal adhesive disease. The fixed
laterodeviation of the uterus has been described as a radiographic finding of
peritubal adhesions in another retrospective study
[3].
In this present study the diagnostic accuracy of hysterosalpingography in
enabling prediction of peritubal adhesions identified at laparoscopy was
evaluated. The aim was to verify whether hysterosalpingographic accuracy might
be improved by taking into account a combination of selected radiographic
signs of peritubal adhesions.
Subjects and Methods
This study was approved by the institutional review board and followed the
tenets of the Declaration of Helsinki
[4]. Informed consent was
obtained from patients.
From January 1996 to December 1998, 30 consecutive patients (age range,
26-42 years; mean, 32 years) in whom laparoscopy had been planned for
infertility of more than 1 year underwent hysterosalpingographic procedure
before surgery. Twenty-two patients had both tubes studied; in the remaining
eight patients, only one tube was analyzed either because of a surgical
removal of the other tube (four patients) or because the other tube was
inadequately visualized during hysterosalpingography (four patients). In
total, 52 tubes were compared.
Hysterosalpingography and laparoscopy were performed in a blind manner and
analyzed by the same team of radiologists and surgeons with expertise in
hysterosalpingography and laparoscopy, respectively. Laparoscopy was performed
under general anesthesia in the proliferative phase of the cycle and was
recorded on S-VHS videotape. A 10-mm laparoscope was introduced through the
umbilicus, and two or three 5-mm accessory trocars for ancillary instruments
were introduced suprapubically. The chromopertubation test was always done.
Laparoscopic results were evaluated by a single operating surgeon, who
considered the presence of peritubal or periovarian adhesions abnormal
[5] independently from the
severity. A different surgeon repeated this evaluation by reviewing the S-VHS
videotape. Hysterosalpingography was performed in the same phase of the cycle
3-6 months before laparoscopy. For the intrauterine injection, a nonionic
water-soluble contrast medium (Isovist 240 [iotrolan]; Schering, Berlin,
Germany) was introduced by cervical catheter or vacuum cervix adaptor. Two
supine and two oblique images were obtained. The prone position was used when
the fallopian tubes were not completely visualized in the supine position.
Hysterosalpingographic findings were evaluated by two radiologists
considering the following reported radiographic findings of peritubal
adhesions: convoluted tubes, vertical tubes, loculation of contrast medium in
peritoneum, halo effect, and fixed laterodeviation of the uterus
(Fig. 1). The fixed
laterodeviation of the uterus was assessed by gripping, pulling, and then
releasing the traction of the hysteroinjection system. Both a fixed
laterodeviated uterus as well as a normal nonmedian uterus initially follow
the traction direction; however, during the releasing phase, the fixed
laterodeviated uterus tends to keep the original deviated position, and the
normal nonmedian uterus keeps the pelvis midline. The ampullary dilatation
described by Karasick and Goldfarb
[1] was not included in this
list. Hysterosalpingograms were assessed as showing normal or abnormal
findings using two different criteria: no sign means a normal result, one or
more signs mean an abnormal result (first criterion); no sign or one sign
means a normal result, two or more signs mean an abnormal result (second
criterion). Discrepancies in interpretation of radiographic signs were
resolved by consensus. Hysterosalpingographic results obtained using the two
criteria in patent as well as in distally nonpatent tubes were then compared
with corresponding laparoscopic results by 2 x 2 tables and were
statistically analyzed (kappa statistics).
View larger version (98K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —In 34-year-old infertile woman, hysterosalpingogram shows
radiographic signs of peritubal adhesions. Left tube lies vertically in
ampullary segment (small black arrow). Contrast medium spills into
peritoneal cavity from fimbrial end (black arrowhead) and falls to
show normal inhomogeneous pattern. Note loculation of contrast medium
(White arrowhead). Uterus is deviated on left side. Right tube is
convoluted in ampullary segment (white arrows). Also note
double-contour appearance of right tubal wall (halo effect) (large black
arrows).
|
|
Results
At hysterosalpingography, 28 tubes were patent and 24 were distally
obstructed. At laparoscopy, 32 tubes were patent and 20 tubes were distally
obstructed. Discrepancies in tubal patency were found in four tubes in which a
distal ampullary diverticulum, identified at laparoscopy, was misdiagnosed at
hysterosalpingography as a distal occlusion with no dilatation of the
ampullary tract [6].
At laparoscopy, 31 of 52 tubes showed peritubal or periovarian adhesions,
and 21 tubes did not present any pelvic adhesive disease. Both the operating
surgeon and the reviewing surgeon always agreed with either the presence or
absence of adhesions. At hysterosalpingography, no sign was found in eight
tubes (three patent tubes in which peritubal adhesions were not found at
laparoscopy and five distally nonpatent tubes in which adhesions at
laparoscopy were found in four cases); only one sign was identified in 23
tubes (13 patent tubes in which adhesions were found at laparoscopy in only
one case and 10 nonpatent tubes with adhesions at corresponding laparoscopy in
eight cases); two or more associated signs were seen in the remaining 21 tubes
(12 patent tubes in which adhesions were diagnosed at laparoscopy in 10 cases
and nine non-patent tubes with adhesions at laparoscopy in eight cases).
Hysterosalpingographic results formulated on the basis of the two adopted
criteria described earlier in this article are plotted against those of
laparoscopy in Tables 1 and
2, which refer to
radiologically patent as well as distally nonpatent tubes and in which
corresponding statistical analysis (kappa statistics) of results has also been
done. Diagnostic sensitivity, specificity, and accuracy derived from
corresponding analyses are plotted as histograms in Figures
2 and
3.
View this table:
[in this window]
[in a new window]
|
TABLE 2 Hysterosalpingography (HSG) and Laparoscopy of Peritubal Adhesions in 24
Radiologically Distally Nonpatent Tubes
|
|
View larger version (44K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2. —Bar chart shows peritubal adhesions revealed by
hysterosalpingography compared with laparoscopy using both adopted criteria in
patent tubes. Histogram of sensitivity (Sens), specificity (Spec,)
false-negative (FN), false-positive (FP), positive predictive value (PPV),
negative predictive value (NPV), and overall accuracy (Acc) rates shows high
diagnostic precision of hysterosalpingography when second criterion is
applied. Light gray bars represent first criterion; dark gray bars represent
second criterion.
|
|
View larger version (39K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3. —Bar chart shows peritubal adhesions revealed by
hysterosalpingography versus laparoscopy using both adopted criteria in
distally nonpatent tubes. Histogram of sensitivity (Sens), specificity (Spec),
false-negative (FN), false-positive (FP), positive predictive value (PPV),
negative predictive value (NPV), and overall accuracy (Acc) rates suggests
that hysterosalpingography does not perform as well in patent tubes in
predicting peritubal adhesions, whichever criterion is applied. Light gray
bars represent first criterion; dark gray bars represent second criterion.
|
|
The first criterion (no sign is a normal radiographic result) generally
displayed a low diagnostic accuracy (Figs.
2 and
3 and Tables
1 and
2). In radiologically patent as
well as distally nonpatent tubes, the agreement with laparoscopy was quite
low, with an unacceptable rate of false-positive results. The kappa value was
0.144 and 0.045, respectively. The p value was equally not
significant.
The second criterion (no sign or one sign is a normal radiographic result)
improved the overall accuracy of hysterosalpingography in patent tubes (89.2%)
(Fig. 2 and
Table 1), reducing
false-positive results from 82.3% to 11.7% with an acceptable rate of
false-negative diagnoses (9%). Kappa value improved by a factor of 5.4; the
p value was 0.001 (Fig.
2 and Table 1). The
more frequently associated signs were loculation of contrast medium,
convoluted tube, and laterodeviation of the uterus. Vertical tubes and
laterodeviation of the uterus always corresponded to peritubal and periovarian
adhesions at laparoscopy. Figure
4A,4B
shows a case of agreement between laparoscopy and hysterosalpingography. The
patent left tube shows more than one radiographic sign of peritubal adhesions.
The corresponding laparoscopic findings show an adhesive disease involving the
ampullary tract of the left tube and the peritoneum. In distally nonpatent
tubes, hysterosalpingography did not perform as well as in patent tubes in
predicting peritubal adhesions at laparoscopy using either criterion adopted
in this study (Fig. 3 and
Table 2). Kappa statistics
never reached statistical significance.
View larger version (81K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —30-year-old infertile woman. Hysterosalpingogram shows deviation of
uterus to left side with convoluted ampullary segment of left tube (small
arrow) and loculation of contrast medium in peritoneum (large
arrows). Right tube is convoluted with normal spillage pattern
(arrowheads) in peritoneum, which appeared normal on laparoscopy.
|
|
Discussion
Surgery and pelvic inflammatory disease can cause peritubal adhesions. In
the literature, the incidence of peritubal adhesions after salpingitis, at
second-look laparoscopy, ranges from 38.5% to 54.6%
[7]. The relationship between
pelvic adhesions and female infertility is known
[1,
2,
5,
7,
8] and seems to be connected to
peritubal or periovarian involvement interfering with the ovum cycle
[1]. However, the direct
relationship between the severity of pelvic adhesions and reproductive outcome
after surgery has been discussed
[9,10,11,12,13].
Laparoscopy is considered the best technique because of its direct view of
pelvic abnormality and the possibility of a one-session treatment
(adhesiolysis). Hysterosalpingography is usually included in the female
infertility workup and often performed before surgery for a preliminary
evaluation of the female genital tract (uterus, fallopian tube lumen, and
peritoneum). An accurate radiographic map could be very useful in planning
treatment.
From this study, the statistical analysis performed on the results of
radiologically patent and distally nonpatent tubes shows that the first
criterion is not recommended in the radiographic diagnosis of peritubal
adhesions (one radiographic sign is not enough for a correct diagnosis). In
both radiologically patent as well as distally nonpatent tubes, the number of
false-positive results was unacceptable (Figs.
2 and
3), the overall diagnostic
accuracy was low, and the correlation with laparoscopy was not significant
(Tables 1 and
2). The better diagnostic
accuracy of hysterosalpingography in patent tubes using the second proposed
method might depend on the number of radiographic signs. In distally
obstructed tubes, hysterosalpingography cannot give information about
peritubal findings, thus limiting its usefulness in diagnosing peritubal
adhesions. In the study of Karasick and Goldfarb
[1], the researchers observed
the tendency for fallopian tubes with peritubal adhesive disease to have more
radiographic findings at laparoscopy, but these researchers were unable to
determine a statistically significant correlation between
hysterosalpingography and laparoscopy when more signs were associated. In our
opinion, this failure might depend on a particular sign that they included in
the list of radiographic findings of peritubal adhesions, which is the
ampullary tract dilatation in nonoccluded tubes. Intraluminal lesions, and not
necessarily a peritubal adhesive disease, can determine ampullary dilatation
with partial occlusion of the abdominal ostium
[14]. Even if intraluminal
lesions and tubal occlusions may be associated with peritubal or pelvic
adhesions [7,
8,
12], the ampullary dilatation
in non-occluded tubes cannot be considered, in our opinion, as an exclusive
sign of peritubal adhesive disease. That is why this sign was not considered
in our study, which shows what Karasick and Goldfarb were unable to determine.
From this present study, it is easy to understand that the complete
visualization of fallopian tubes, particularly if tubal transit is documented
(with more radiographic signs available), is very useful in the
hysterosalpingographic diagnosis because of peritubal adhesions. Even if
laparoscopy remains the gold standard in this diagnosis because of its direct
view of pelvic disease, hysterosalpingography, a noninvasive procedure not
burdened by general anesthesia, can be considered an effective technique in
the investigation of peritubal factors of infertility, provided that the
appropriate diagnostic criteria are used. Additional information alerting the
clinician to the possibility of a peritubal adhesive disease could be given if
the proposed second diagnostic criterion is used in patent tubes. Laparoscopy
is usually performed after hysterosalpingography in cases of either abnormal
or normal radiographic result. The common policy during an infertility workup
is to perform laparoscopy shortly after hysterosalpingography when findings
are abnormal and to wait approximately 6 months
[15] or more
[16] when
hysterosalpingography reveals normal findings because hysterosalpingography
might induce pregnancies. The reported rate of live births after
hysterosalpingography is 20.4%, 19.4%, and 21.8% depending on the contrast
medium used (water-soluble, oil-soluble, or a combination of both,
respectively) [17,
18]. From this study, the
radiographic diagnosis of peritubal adhesions in normal patent tubes is
credible. Therefore, the 6-month interval between hysterosalpingography and
laparoscopy might be shortened
[2] when adhesive peritubal
involvement is radiologically suspected. In these cases, the preliminary
radiographic diagnosis can also simplify the laparoscopic procedure by
obviating diagnostic laparoscopy before surgery. In our opinion and according
to the literature [2,
19,20,21],
because of its high diagnostic precision in predicting peritubal adhesions at
laparoscopy in patent tubes, hysterosalpingography should always be performed
before surgery.
Acknowledgments
We thank Benedetto Falsini (Department of Ophthalmology,
Università Cattolica del Sacro Cuore di Roma)
for statistical advice.
References
-
Karasick S, Goldfarb AF. Peritubal adhesions in infertile women:
diagnosis with hysterosalpingography. AJR
1989;152:777
-779[Abstract/Free Full Text]
-
Henig I, Prough SG, Cheatwood M, De Long E. Hysterosalpingography
laparoscopy and hysteroscopy in infertility: a comparative study. J
Reprod Med 1991;36:573
-575[Medline]
-
Valentini AL, Danza FM, De Vivo D, Colavita N, Vincenzoni M.
Aderenze peritubariche e isterosalpingografia. Radiol
Med 1985;71:326
-328
-
Declaration of Helsinki: recommendations guiding medical doctors in
biomedical research involving human subjects. Med J
Aust 1976;1:206
-207[Medline]
-
American Fertility Society-Birmingham Alabama. The American
Fertility Society classification of adnexal adhesions, distal tubal occlusion,
tubal occlusion secondary to tubal ligation, tubal pregnancies, Mullerian
anomalies and intrauterine adhesions. Fertil Steril
1988;49:944
-955[Medline]
-
Muzii L, Marana R, Mancuso S. Distal fallopian tube occlusion:
false diagnosis with hysterosalpingography in cases of tubal diverticula.
Radiology
1996;199:469
-471[Abstract/Free Full Text]
-
Gerber B, Krause A. A study of second-look laparoscopy after acute
salpingitis. Arch Gynecol Obstet
1996;258:193
-200[Medline]
-
Bowman MC, Cooke ID. Comparison of fallopian tube intraluminal
pathology as assessed by salpingoscopy with pelvic adhesions.
Fertil Steril
1994;61:464
-469[Medline]
-
Gomel V. Salpingostomy by microsurgery. Fertil
Steril 1978;29:380
-385[Medline]
-
Boer-Meisel ME, Te Velde ER, Habbema JDF, Kardaum JWPF. Predicting
the pregnancy outcome in patients treated for hydrosalpinx: a prospective
study. Fertil Steril
1986;45:23
-29[Medline]
-
Dubuisson JB, Chapron C, Morice F, Aubriot FX, Foulot H, Bouquet de
Jolinière J. Laparoscopic salpingoscopy:
fertility results according to the tubal mucosal appearance. Human
Reprod 1994;9:334
-339[Abstract/Free Full Text]
-
Marana R, Rizzi M, Muzii L, Catalano GF, Caruana P, Mancuso S.
Correlation between the American Fertility Society classification of adnexal
adhesions and distal tubal occlusion, salpingoscopy and reproductive outcome
in tubal surgery. Fertil Steril
1995;64:924
-929[Medline]
-
Marana R, Catalano GF, Caruana P, Malaguti F, Mancuso S. The
prognostic role of salpingoscopy in laparoscopic tubal surgery.
Human Reprod
1999;14:2991
-2995[Abstract/Free Full Text]
-
Vincenzoni M, Valentini AL. Le ostruzioni tubariche. In: Vincenzoni
M, Valentini AL, eds. L'Isterosalpingografia, 1st ed.
Milano: Bracco Italia, 1986;74
-86
-
Swart P, Mol BWJ, Van der Veen F, Van der Beurden M, Redekop WK,
Bossuyt PMM. The accuracy of hysterosalpingography in the diagnosis of tubal
pathology: a meta-analysis. Fertil Steril
1995;64:486
-491[Medline]
-
Mol WJ, Collins JA, Burrows EA, van der Veen F, Bossuyt PMM.
Comparison of hysterosalpingography and laparoscopy in predicting fertility
outcome. Human Reprod
1999;14:1237
-1242[Abstract/Free Full Text]
-
Duff DE, Fried AM, Wilson EA, Ibrahim AA, Wahby O. Comparative
evaluation of laparoscopy and hysterosalpingography in infertile patients.
Obstet Gynecol
1978;51:29
-32[Abstract/Free Full Text]
-
Spring DB, Barkan HE, Pruyn SC. Potential therapeutic effects of
contrast materials in hysterosalpingography: a prospective randomized clinical
trial. Radiology
2000;214:53
-57[Abstract/Free Full Text]
-
Hutchins CJ. Laparoscopy and hysterosalpingography in the
assessment of tubal patency. Obstet Gynecol
1976;49:325
-327
-
Rice JP, London SN, Olive DL. Revaluation of hysterosalpingography
in infertility investigation. Obstet Gynecol
1986;67:18
-721
-
La Sala CG, Sacchetti F, Degl'Incerti-Tocci F, Dessanti L, Torelli
MG. Complementary use of hysterosalpingography, hysteroscopy and laparoscopy
in 100 infertile patients: results and comparison of their diagnostic
accuracy. Acta Eur Fertil
1987;18:369
-374[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
