Human Reproduction vol 11 no. 10 pp.2130-2133, 1996
Comparison of in-vitro falloposcopy with tubal histology
in the diagnosis of Fallopian tube pathology
S.Rimbach1, D.Wallwiener, C.Barth, D.Heberling
and G.Bastert
Department of Obstetrics and Gynecology, Vosstrasse 9, University
of Heidelberg, D-69115 Heidelberg, Germany
'To whom correspondence should be addressed
In order to assess the diagnostic quality of falloposcopy in
relation to pathomorphology, a consecutive series of 30
Fallopian tubes obtained from surgical salpingectomy cases
were prospectively examined by in-vitro falloposcopy and
histology. Falloposcopy was performed using an over-thewire catheterization system and a 0.5 mm falloposcope
with 3000 pixels. Assessment of the specimens included the
description of lumen geometry, intraluminal changes and
status of the mucosal surface. Falloposcopy classified 14
tubes as normal and 16 pathological. Histology resulted in
17 normal versus 13 pathological tubes. Pathologies
included lumen obstructions and dilatations, intraluminal
synechiae and mucosal damage. Sensitivity and specificity
of falloposcopy were calculated to be 0.85 and 0.71; positive
and negative predictive values were 0.69 and 0.86. It was
concluded that falloposcopic findings indeed reflect and
successfully differentiate normal and pathological conditions allowing adequate and reproducible image interpretation. However, variations of the diagnostic accuracy with
the type of pathology and the tubal segment have to be
taken into account before clinical consequences are drawn
from a falloposcopic investigation.
Key words: falloposcopy/infertility/reproductive surgery/tubal
endoscopy
Introduction
First attempts to visualize intraluminal pathology in tubal
sterility date back to the year 1970, when Mohri et al. (1970)
described the use of a flexible glass fibre endoscope for
intratubal observations. More than a decade later, the method
started to gain clinical importance, mainly with the work of
Henry-Suchet et al. (1981, 1985), Cornier (1982) and Brosens
et al. (1987), who used tuboscopy to predict reproductive
outcome and the chances of tubal reconstructive surgery.
Whereas they approached the tube by its distal end, limiting
the tuboscopic evaluation to fimbria and ampulla, a transcervical access was described in 1990 by Kerin et al (1990),
allowing visualization over the entire tubal length including
intramural and isthmic segments. This technique has since
been referred to as falloposcopy.
Already, a number of reports exist in the literature
2130
establishing clinical scores for falloposcopic findings (Kerin
et al., 1992; Lower et al., 1992; Dunphy and Pattinson,
1994; Rimbach et al, 1994). However, so far, no study has
systematically evaluated falloposcopic images in comparison
to histopathology. This verification, however, seems of utmost
importance to ensure reliable and reproducible image interpretation.
Falloposcopy is on the threshold of becoming an important
infertility investigation technique, but its ability to describe
intraluminal findings adequately, especially to differentiate
accurately normal and pathological tubal conditions, has yet
to be proved.
The aim of the present study was therefore to assess the
diagnostic quality of falloposcopy in relation to histomorphology.
Material and methods
Specimens
A consecutive series of 30 Fallopian tubes were prospectively
examined by falloposcopy and histology (Table I). The tubes were
obtained from hysterectomy cases (n = 11), unilateral salpingectomies
(n = 6) and proximal tubal microsurgery for sterilization reversal
(n = 8) or proximal occlusion (n = 5).
Falloposcopy procedure
Falloposcopy was performed in vitro immediately after removal of
the organ using an over-the-wire catheterization system and a 0.5
mm falloposcope with 3000 pixels (Conceptus Inc., San Carlos,
CA, USA) under conunuous fluid irrigation with Ringer's solution.
Falloposcopic findings were instantly recorded on standardized
documentation sheets.
Histology
The procedure lasted <5 min overall time per tube and the specimen
was then immediately fixed in formalin before histological preparation.
After paraffin embedding, routine histological serial cuts from the
different tuba! segments (intramural, isthmic, ampullary and fimbrial)
as available were stained with haematoxylin and eosin for bght
microscopy The histologist was blinded to the result of the falloposcopy
Table L Sources of specimens for in-vitro falloposcopy
Hysterectomy + salpingectomy
11
Unilateral salpingectomy
Microsurgery for stenlizaiion reversal
Microsurgery for proximal occlusion
Total
5
30
O European Society for Human Reproduction and Embryology
Falloposcopy versus tubal histology
Table H. Morphological assessment cntena and parameters as applied
during falloposcopy and for histology
Lumen geometry
regular
obstructed
obliterated
dilated
debns/plug
polyp
synechiae (thin/solid)
septation
intact
atrophic
disturbed
intact
reduced
Intraluminal findings
Status of the mucosal surface
Fold architecture
Table HI. Diagnoses of normal versus pathological tubes according to the
results of falloposcopy and histology
Falloposcopy
Histology
Normal
Pathological
14
17
16
13
False negative
False positive
Figure 1. Falloposcopic picture of a normal ampulla.
•Endometnosis with moderate dilatation (n = 1), inflammatory epithelial
proliferations (n = 1)
"Mucosal atrophy (n = 5, four isthmic, one ampullary), one case including
thin synechiae
Analysis
Assessment of the specimens used the same descriptive morphological
cntena for both methods, falloposcopy and histology the lumen
geometry was recorded as regular, obstructed, obliterated or dilated;
the presence of intraluminal changes and the status of the mucosa
were recorded (Table II)
Biometncal analysis included the determination of sensitivity and
specifity of falloposcopy as well as calculation of negative and
positive predictive values
Results
Falloposcopic findings
As Table EH shows, 14 of 30 tubes were found by falloposcopy
to have a non-dilated, patent lumen, intraluminal findings
absent, giving the visual impression of an intact mucosal layer
characterized by a velvet-like, smooth surface and, depending
on the tuba! segment, floating primary and secondary mucosal
folds. Those tubes were described as normal (Figure 1).
The other 16 tubes appeared pathological Two were found
to have a distally dilated lumen, one appeared proximally
obstructed and two were completely obliterated.
Intraluminal findings in four tubes showed filmy synechiae,
and in one tube, solid synechiae crossing the lumen (Figure
2). In 14 tubes the mucosa was described as flattened along
with reduced fold structures in five tubes (Figure 3).
Thus, among the 30 tubes evaluated, 14 were considered
normal and 16 pathological by falloposcopy.
Histological diagnosis
Histology revealed 17 normal and 13 pathological tubes
(Table HI)
The pathological specimens included seven sactosalpmges,
Figure 2. Falloposcopic picture of solid intraluminal synechiae
three cases of tuba! endometnosis, two cases of tuba! fibrosis
and one case of salpingitis isthmica nodosa.
All 17 normal cases were characterized by patent lumina
with regular diameter, absence of intraluminal findings, predominantly cylindrical epithelium and the presence of well
vascularized pnmary and secondary fold structures distributed
according to the anatomical segments of the tube.
The pathological diagnoses were associated with lumen
dilatation in seven cases (five distal, two proximal), obstruction
in two cases (proximal) and obliteration in one case (proximal),
intraluminal synechiae in seven cases (including four cases of
thin non-vascularized formations and three cases of solid
vasculanzed pseudofollicular septations), flattening of the
mucosa in eight cases (among those seven in isthmus and
ampulla, one in isthmus), disturbed mucosal structure in
2131
S.Rimbacta el at
were concerned in detail, lumen patency was correctly diagnosed in all cases with only one case of subtotal fibrotic
obstruction incorrectly-diagnosed as obliteration. In contrast,
dilation of the lumen was correctly described only in two
cases but missed in five cases. Intraluminal falloposcopic
findings were confirmed by histology in three of four cases of
synechiae, but only one of three cases of solid septation
apparent in the histological specimens was correctly described
by falloposcopy. In 14 falloposcopic diagnoses of flattened
mucosa, only eight could be verified by histology. No case
of mucosal atrophy described in histology was missed by
falloposcopy, whereas four mucosal changes by inflammation
and endometnosis were not detected in falloposcopy. Five
cases of reduced fold architecture in the ampullary part
were confirmed histologically, but two cases of histologically
diagnosed fold destruction along with septation were missed
by falloposcopy
Figure 3. Falloposcopic picture of partial mucosal atrophy in a
sactosalpinx
Table IV. Test quality parameter and predictive values of falloposcopy in
comparison to histology as a reference
Test parameter
P value
Sensitivity (%)
Specificity (%)
Positive predictive value
Negative predictive value
0.85
071
0 69
0 86
four cases (two luminal endometriosis and two inflammatory
epithelial proliferative changes) and reduced fold structures in
seven cases.
In addition, histology revealed a number of pathologies not
interfering with the lumen side of the examined tubes. Among
these, there was one case of tubal wall endometnosis, one
case of subserosal endosalpingiosis, four cases of thickened
muscular layer and one case of fibrotic changes within the
wall not extending towards the mucosa. In one case, Walthardcell nodules were found in the subserosal layer.
Comparison of falloposcopic and histological data
The falloposcopic diagnosis of normal was therefore correct
in 12 tubes and false in two, missing in one case endometriotic infiltration of the isthmic mucosa along with
moderate dilation, in one other inflammatory epithelial proliferations within an otherwise normal mucosa.
The diagnosis of pathological was correct in 11 tubes and
false in five. In all five cases the mucosa was described
atrophic, four times in the isthmic and one time in the
ampullary region. In one case, thin intraluminal synechiae
were additionally seen not being confirmed by histology. As
a result, sensitivity and specifity of the procedure were calculated as 0.85 and 0.71 respectively; positive and negative
predictive values were 0 69 and 0.86 (Table IV).
As far as the different morphological evaluation criteria
2132
Discussion
The results of the present study suggest that falloposcopy
allows differentiation of normal and pathological tubal conditions in reasonable agreement with histology. Nevertheless,
inherent limitations of the method as well as typical traps for
possible misinterpretation of falloposcopic images became
equally apparent
A careful interpretation of the results given above and
discussed below certainly needs to take into account the
in-vitro setting of the study. Whereas ln-vivo falloposcopy
allows successful intraluminal visualization only in a variable
percentage of the cases, the m-vitro examination usually
results in technically sufficient quality. The organ can easily
be manipulated and white-out, vision-disturbing light reflexes
etc. can be eliminated.
On the other hand, artefacts are potentially caused by the
in-vitro situation such as tubal dilatation. Therefore the results
of m-vitro falloposcopy may not reflect in-vivo images in
all cases.
Histologically confirmed endotubal pathologies were
detected by falloposcopy with a high sensitivity of 85%, a
result corresponding to earlier data by Hershlag et al. (1991),
whose observations were restricted to distal salpingoscopy but
who also found a good relationship to histological findings for
severe pathologies.
The positive predictive value of 69% in our study, however,
indicates that the diagnosis 'pathological' can only be considered true in about two thirds of cases.
Whereas the diagnosis 'normal' is correct in 86% of cases
(negative predictive value), pathologies were falsely described
in morphologically normal tubes in almost 30% of the cases.
For example, it proved most difficult to discriminate atrophic
from normal mucosa in the proximal tube. Folds, as the most
reliable parameter for interpretation, are physiologically flat
and rare in this segment of the tube, and the mucosa gives an
impression of atrophy per se with its even surface in a tunnelshaped geometry.
In contrast, the diagnoses of severe lesions such as
obstructive fibrosis or intraluminal synechiae were usually
Falloposcopy versus tuba] histology
accurate, both in the intramural and isthmic tubal segments.
In particular, the reproducible diagnosis of synechiae may be
of utmost clinical importance, since Dunphy and Greene
(1995) have shown a significant association between intraluminal adhesions and the occurrence of ectopic pregnancy.
Tubal wall pathologies, well known for their prognostic
impact (Fortier and Haney, 1985; Wiedemann et al., 1987),
may be associated with intraluminal lesions and then accessible
for falloposcopic diagnosis. If strictly localized within the wall,
as observed in cases of subacute and chronic inflammation,
endometriosis and also fibrosis, they are, however, missed by
falloposcopy for obvious reasons.
Another clinically most important limitation to falloposcopic
investigation became apparent during this study. Septations in
high degree sactosalpinges, also known for their important
clinical impact (De Bruyne et al, 1989), were easily misinterpreted, considering that one of the compartments is only
the tubal lumen surrounded by flattened epithelium.
Depending on the clinical situation, the diagnostic assessment of both proximal as well as distal tubal disease may
therefore require complementary methods such as laparoscopic
and salpingoscopic inspection for complete information.
In conclusion, the present study shows that falloposcopic
findings indeed reflect and successfully differentiate normal
and pathological conditions of the Fallopian tube. The use
of defined pathomorphological criteria allows adequate and
reproducible image interpretation. The accuracy of the interpretation and therefore the diagnostic quality of the method
varies, however, with the type of pathology and the tubal
segment. Both have to be taken into account before clinical
decisions are made from a falloposcopic investigation.
Lower, A , Magurness, S , Djahanbakhch, O and Grudzinskas, J (1992)
Transcervical tuba] endoscopy (falloposcopy) - a clinically useful tool?
GynaecoL Endosc, 1, 155—158.
Mohri, T, Mohn, C and Yamadon, F. (1970) Tubaloscope flexible glass
fibre endoscope for intratubal observations Endoscopy, 2, 226
Rimbach, S., Wallwiener, D. and Bastert, G (1994) Tuben-Endoslcopiemirumal invasive Diagnostik bei gestorter Eileiterfunktion Minimal Invasive
Medrnn, 5, 9-12
Wiedemann, R, Scheidel, P., Wiesinger, H and Hepp, H (1987) Die Pathologie
des proximalen Tuhenverschlusses - morphologische Auswertungen
Gebunsh. FrauenheiUc, 47, 96-100
Received on October 26, 1996, accepted on July 31, 1996
References
Brosens, I., Boeckx, W., Delamn, Ph et aL (1987) Salpmgoscopy: a new
preoperative diagnostic tool in tubal infertility? Br J Obstet Gynaecol,
94, 722-728
Cornier, E (1982) La fibroscopie en gynecologie la fibrohysteroscopie et la
fibrotuboscopie Nouv. Prtsse Med, 11, 2841-2843
De Bruyne, E, Puttemans, P, Boeckx, W and Brosens, I. (1989) The clinical
value of salpmgoscopy in tuba) infertility FertiL StenL, 51, 339-340
Dunphy, B and Greene, C (1995) Falloposcopic carinulauon, oviductal
appearances and prediction of treatment independent intrautenne pregnancy
Hum. Reprod, 10, 3313-3316
Dunphy, B and Pattinson, H A (1994) Office falloposcopy, a tertiary level
assessment for planning the management of infertile women AusL N. Z. J
ObstcL GynecoL, 34, 189-190.
Fortier, K J and Haney, A F (1985) The pathologic spectrum of uterotubal
junction obstruction. Obstei GynecoL, 65, 93-98
Henry-Suchet, J , Tesquier, L and Loffredo, V (1981) Endoscopies tubaires
premiers resultats Gynecologie, 23, 293-295
Henry-Suchet, J , Loffredo, V, Tesquier, L and Pez, J P (1985) Endoscopy
of the tube (=tuboscopy) its prognostic value for tuboplasties Acta Eur
FemL, 16, 139-145.
Hershlag, A., Seifer, D , Carcangiu, M et aL (1991) Salpmgoscopy light
microscopic and electron microscopic correlations Obstet Gynecol, 77,
399-^*05
Kenn, J., Daykhovsky, L., Segalowitz, J et aL (1990) Falloposcopy a
microendoscopic technique for visual exploration of the human Fallopian
tube from the uterotubal ostium to the fimbna using a transvaginal approach
FertiL StenL, 54, 390-^MX)
Kenn, J , Williams, D , Roman, G et aL (1992) Falloposcopic classification
and treatment of Fallopian tube lumen disease FertiL Stenl, 57, 731-741
2133