Clin. Radiol. (1974) 25, 535-542 DIVERTICULOSIS
OF THE
FALLOPIAN
TUBES
G. FREAKLEY, W. J. NORMAN,* J. T. ENNIS** and E. R. DAVIES
Radiology Departments, University of the West Indies, United Bristol Hospitals Diverticulosis of the Fallopian tubes is uncommon, but is nine times as common in Negro women of Jamaican origin as in Caucasian women living in Bristol. 58 examples were encountered in 918 hysterosalpingograms done in Kingston, Jamaica (426) and in Bristol (492) from 1968 to 1972. In 41 cases (72~) diverticulosis was bilateral. There are several possible predisposing clauses, of which infection, including tuberculosis, is the most important. The radiological appearances are distinctive, but, with the possible exception of tuberculosis, do not give any indication of the cause. Infertility was the commonest clinical association of diverticulosis. Among infertile patients, hydrosalpinx and tubal obstruction were slightly commoner in the presence of diverticulosis, and the significance of this is discussed. Ectopic pregnancy is the most important associated condition and diverticulosis predisposes to its occurrence. Hysterosalpingography is indicated following ectopic pregnancy to show whether there is diverticulosis of the contralateral tube, as this may influence prognosis and management. Two kinds of diverticula of the Fallopian tubes are described: (a) Solitary. These are exceedingly rare and very little is known about them (Troell, 1970). (b) Multiple. These are much more common, though still relatively rare. The pathological features of multiple diverticula are well known (Chiari, 1887; Schenken and Burns, 1943). Macroscopically, the tube is sometimes normal, but usually it has focal fusiform swellings involving its entire circumference and covered by a smooth serosal surface. Occasionally more than one nodule is visible, giving the tube a beaded appearance. The nodules are up to 2.5 cm. in diameter, though some are so small that they can only be detected by microscopy (Schenkem and Burns, 1943). Other causes of nodularity of the tubes, e.g. endometriosis, neoplasms and inflammatory lesions are rarer, but can give rise to similar macroscopic appearances. Microscopically the characteristic features of diverticulosis are innumerable collections of irregular alveolar spaces in the muscular coat of the tube (Fig. 1). These spaces are lined by epithelium which is continuous with, and usually indistinguishable from, that of the tubal mucosa (Persaud, 1970). Usually there are several foci of epithelial inclusions, but in the
*Present Address - Royal Salop Infirmary, Shrewsbury. **Present Address - Mater Mesericordiae Hospital, Dublin.
extreme case the whole wall is converted into a honeycomb of spaces. Generally these spaces are rounded, but when they are numerous they become angular and elongated, with their long axes parallel to that of the tube. Adjacent fibrosis or leucocytic infiltration is the rule, and there may even be intramural abscess formation. Exceptionally there is no evidence of associated inflammatory change. Hypertrophy of adjacent muscle is frequently observed, and sometimes the lumen of the tube is occluded (Schenken and Burns, 1943). These pathological findings are based on careful studies, which include contrast injections of postmortem and operative specimens. By comparison there has been little radiological interest in tubal diverticulosis (Mills et al., 1968). The present study was undertaken to compare the incidence of the condition in Negro (Kingston, Jamaica) and Caucasian (Bristol) women undergoing hysterosalpingography and to evaluate the clinical and radiological significance of the findings. PATIENTS AND M E T H O D 918 hysterosalpingograms done during 1968-72 were reviewed. 426 of these were done on Jamaican Negroes at the University Hospital of the West Indies. 492 were done at Bristol General Hospital, 471 of them on Caucasians and 21 on Negroes of Jamaican origin. The indications for the investigation are sum535
536
CLINICAL RADIOLOGY
Oiverticulo . , : 1 ~
-~
""'/ tube (~Zollopion
\
5 1B
FIG. 1A and 1B Transverse section of Fallopian tube with diverticulosis. The diverticula penetrate the muscular coats of the tube and their lining epithelium is similar to that of the tube itself.
D I V E R T I C U L O S I S OF THE F A L L O P I A N TUBES
537
FIo. 2 Both tubes are patent. There are many diverticula in the proximal parts of b o t h tubes.
FIo. 3 The fight tube is patent. There are diverticula in its proximal third. The left tube is occluded. No diverticula are shown in its proximal half.
FIG. 4 The shape of the uterus is areuate. Both tubes are slightly occluded. Diverticula are bilateral.
FIO. 5 The left tube is beaded and several isolated diverticula arise from it, the so-called "peas in a pod". There is a terminal hydrosalpinx. The right tube is slightly beaded also. The histological diagnosis was tuberculosis.
marised in Table 1. The patients were examined in mid-cycle, usually as outpatients. No premeditation or other preparation was given to Bristol patients but 10 rag. hyoscine-N-butyl (Buscopan) was given to Kingston patients 20-60 rains, before the examination. In both centres the examinations were done jointly by a gynaecologist and a radiologist. Up to 20 ml. of water-soluble contrast medium (Diaginol-viscous) were injected under fluoroscopic control until satisfactory filling was achieved. The radiographs were taken at the end of the injection and 15 mins. later. RADIOLOGY Diverticula were demonstrated in a total of
58 patients, that is in 46 (10.5 ~) of the Kingston, Jamaica, patients, 6 (29 ~) of the Negro Bristol patients, and 6 (1.2~) of the Caucasian Bristol patients (Table 2). The condition was bilateral in 41 cases (72~) (Fig. 2), that is in 33 (75~) of the Kingston patients and in 8 (66~) of the Bristol patients. Usually the diverticula were seen on screening as minute quantities of contrast medium adjacent to the lumen of the tube during the injection, as reported in other accounts (Mills et al., 1968) and on the subsequent films these pools of contrast are shown to be in continuity with the lumen of the tube. The diverticula were up to 2 mm. in diameter and usually they were clustered together over
CLINICALRADIOLOGY
538
TABLE2 INCIDENCEOFDIVERTICULOSIS
TABLE 1 Indications for Hysterosatpingography Primary infertility 80 Recurrent abortion 5 Post-operative assessment (following salpingoplastymyomectomy, etc.) 5 Dysfunctional bleeding 4"5 Previous ectopic pregnancy 2.4 Miscellaneous 3"1 (pelvic mass, cysts, location of IUD, etc.)
Ethnic Group of Patients
Total No. of Patients
No. with Diverticulosis
Negro : Kingston Bristol Total
426 21 447
46 (10.5 ~) 6 (29 ~) 52 (11.6 ~)
Caucasian: Bristol
471
6 (1-2~)
918
58 (6.4 ~)
Total 1-2 cm. (Fig. 3). Less often a longer segment was involved. The diverticula were confined to the proximal region of the tube in 38 cases (66 ~ ) and were slightly more distal in 6 cases (100~). In a further 13 ( 2 2 ~ ) there was extensive involvement of both these regions (Fig. 4). In one patient the diverticula were arranged singly along the tube (Fig. 5), an appearance likened to 'peas in a pod' (Persaud, 1970). The main distribution patterns are shown in Fig. 6. A S S O C I A T E D C O N D I T I O N S (Table 3)
Ectopic pregnancy was known to have occurred in 16 patients, 8 of whom had diverticulosis. In 6 of
these, the ectopic pregnancy had led to salpingectomy before the hysterosalpingography, and diverticula were shown in the remnant of the resected tube (Fig. 7). In the remaining 2 cases the ectopic pregnancy occurred three years after the hysterosalpingogram, and in each the pregnancy was shown histologically to be within the area of diverticulosis (Fig. 8). Tubal obstruction was present in 206 patients, i.e. in 190 (22 7oo)of those without diverticulosis and in 16 ( 2 8 ~ ) of those with diverticulosis (Figs. 4 and 9). Hydrosalpinx was present in 145 patients, i.e. in 134 ( 1 7 ~ ) of those without diverticulosis and in
Distribution of diverticula
Cases
6 ;.%
~Cases ~
e
Fio. 6 Distribution of diverticula in 58 cases.
DIVERTICULOSIS OF THE FALLOPIAN TUBES TABLE 4
TABLE 3 INCIDENCE OFASSOCIATED CONDHIONS
INCIDENCE OF T U B A L OBSTRUCTION AND H Y D R O S A P I N X IN BRISTOL AND K I N G S T O N
Diverticulosis Total Present Absent (58 cases) (860 cases) Ectopic pregnancy 8 (14%) Tubal obstruction 16(28%) ' Hydrosalpinx 11 (28~) Tuberculosis l -Endometriosis 1
8 (0.99/00) 190(22%) 134(22~) --
539
Bristol Patients
918 cases 16 206 145 1 1
11 (19 ~ ) of those with diverticulosis (Figs. 4 and 7). Tubal obstruction and hydrosalpinx were slightly commoner among the Jamaican patients (Table 4).
Tuberculosis of the Fallopian tubes was shown histologically in one patient. Endometriosis was shown histologically in one patient. DISCUSSION Several theories have been put forward to explain diverticulosis of the Fallopian tubes. The theory that they are congenital has found little support because they have not been shown in children (Persuad, 1970) and are most common in the fourth and fifth decades of life. Inflammatory lesions are often found in association with diverticulosis (Schenken and Burns, 1943; Benjamin and Beaver, 1951). However, they are not always present, and even when present may not coincide with the site of diverticulosis (Benjamin and Beaver, 1951). In the present series, hydrosalpinx and tubal obstruction were c o m m o n but were only slightly commoner in those with diverticulosis than in the remainder. In fact, the majority of patients with diverticulosis did not have radiological evidence of these sequelae of pelvic infection. Thus it is improbable that infection is the sole cause of diverticulosis. Indeed it has been suggested that diverticulosis predisposes to infection, which is a secondary event (Benjamin and Beaver, 1951). In the present series tuberculosis was the only specific infection shown, and that in only one instance. Tubal endometriosis is another rare condition said to be an occasional cause of diverticulosis (Wrork and Broders, 1942). Only one example was proven in this series but its true incidence may be under-estimated because of the lack of histological information in the vast majority of cases. Finall~ ulceration of the mucosal surface
Total Number Number with tubal obstruction Number with hydrosalpinx
492
Khlgston Patients
(20 %)
426 106 (25 %)
56 (11.3 %)
89 (21 ~)
100
has been put forward as a means of giving the epithelium access to the muscle coat (Schenken and Burns, 1943) and clearly infection or even endometriosis could act in this way. The absence of muscularis mucosae in the Fallopian tube may make the muscle coats more accessible after the epithelium has been breached and, once this invasion has occurred, the epithelium proliferates slowly, possibly under the influence of the normal hormonal cycle (Schenken and Burns, 1943). The significance of muscle hypertrophy is unclear, but as it is not present in all cases it may also be a secondary phenomenon, possibly influenced by the hormonal cycle that controls tubal peristalsis (Mai and Coutinlio, 1970). It is very likely that many factors influence the development of tubal diverticula, and it is virtually never possible to predict the dominant cause from the radiographic appearances. The exception is that a series of isolated
Fio 7 Hysterosalpingogram after left salpingectomy for ectopic pregnancy. There are diverticula of the remnant of the left tube. There is mild hydrosalpinx of the right tube with numerous diverticula as well.
540
CLINICAL RADIOLOGY
Chorionic villus
FIG. 8A and 8B Transverse section of a tubal pregnancy showing the close relationship between the decidua and the diverticula
DIVERTICULOSIS
OF T H E F A L L O P I A N
diverticula along the tube (Fig. 5) (Troell, 1970) should raise the possibility of tuberculosis. In the present series the incidence of diverticulosis among the Negro patients was nine times as great as among the Caucasian patients irrespective of the geographic location of the former. A similar preponderance was found in other series, though the ratios varied from 3:1 (Schenken and Burns, 1943) to 9:1 (Benjamin and Beaver, 1952). There is no certain explanation for this preponderance, but the increased incidence of pelvic sepsis among Negro women may be a contributory factor. Radiological evidence of pelvic sepsis was greater among Jamaicans in the present series, though not to the same degree as diverticulosis, again suggesting that infection is not the only factor involved. The most important clinical association of diverticulosis is ectopic pregnancy, which still represents a grave clinical problem. It is the cause of 11.6 ~ of all admissions to the gynaecology ward of the University Hospital of the West Indies (Stuart and Skinner, 1967), and accounts for 6 ~ to 1 0 ~ of maternal mortality in the United States (Hallatt, 1968). Diverticulosis was present in as many as 49 ~ of one series of 100 Fallopian tubes removed at operation for ectopic pregnancy (Persaud, 1970). The pregnancy had occurred in the area of diverticulosis in 98 ~ of these, and was demonstrated within the diverticulum in 4 ~ . The association between the two conditions in this series was equally striking, and in the only two
TUBES
541
cases in which histology was available, the cyesis was at the site of diverticulosis (Fig. 8). Thus the demonstration of diverticulosis has considerable importance in management and prognosis. Because it is often bilateral, it is justifiable to recommend hysterosalpingography following ectopic pregnancy, particularly if there is histological evidence of diverticulosis in the removed tube. This is especially true in countries such as Jamaica, where there is a high incidence of ectopic pregnancy (Sterling and Castor, 1968) and of diverticulosis. Finally, the relatively localised nature of the disease in many cases may make it amenable to resection with reimplantation of the distal portion of the tube (Mills et al., 1968). Infertility was the major indication for investigation in the present series in both those with and those without diverticulosis, but it is difficult to establish a causal relationship between infertility and diverticulosis. The associated lesions were not common enough to cause infertility in all of those who did not have ectopic pregnancy, neither was diverticulosis common enough to cause infertility in all those without hydrosalpinx or obstruction. The causes of infertility are complex and often there is no demonstrable radiographic abnormality. Further, the incidence of tubal diverticulosis in fertile women who have not had ectopic pregnancy is not known widely, though in one series of Jamaican women it was as high at 11 ~ (Persaud, 1970). No direct comparison is therefore possible. CONCLUSION Diverticulosis of the Fallopian tubes is far commoner among Jamaican Negro women than Caucasian women living in Bristol The radiological appearances are distinctive, and rarely they raise the possibility of tuberculosis. There is an important association with both ectopic pregnancy and infertility. Hysterosalpingography may yield information that is important for management and prognosis in patients who have had ectopic pregnancy. Acknowledgements.--Weare grateful to Mr. E. J. Turnbull for the line drawings, and to Mr. J. Hancock for the illustrations. Dr J. S. Comes kindly reviewed the histological appearances for us.
FIG. 9 The right tube is patent. There are diverticula in its proximal third, The left tube is occluded at the isthmus.
REFERENCES BENJAMIN, C. L. and BEAVER,D. C. (1951). Pathogenesis of salpingitis isthmica nodosa. American Journal of Clinical Pathology, 21, 212-222. CrtIARI, 4. (1887). Zur pathologischen Anatomie des Eleiter-
542
C L I N I C A L RADIOLOGY
catarrhs Zeitschrift fur Herlk. 8, 457, cited by Benjamin and Beaver, 195l. HALLATT, J. F. (1968). Ectopic pregnancy in perspective. Postgraduate Medichze, 44, 100-103. MAIA, H. S. and COUTINLIO, E. (1970). Peristalsis and Antiperistalsis of the Human Fallopian Tube during the Menstrual Cycle. Biology of Reproduction, 2, 035-314. MILLS, J., UPADHAYAY,P. P. and CARSTAmS,L. S. (1968). Diverticulosis of the Fallopian tubes - Salpingitis Isthimica Nodosa. Clinical Radiology, 19, 233-235. P~RSAUD0V. (1970). Etiology of Tubal Ectopic Pregnancy. American Journal of Obstetrics and Gynaecology, 36, 256-263. P~IILLIe, E. and HUBER. H. (1940). Sterilitate der Alterden
Fron, Zeutralblatt fur Gynakologie. 67, 49. SCHENKEN,J. R. and BURNS,E. L. (1943). A Study and Classi-
fication of Nodular Lesions of the Fallopian Tubes. American Journal of Obstetrics and Gynaecology, 45, 624-636. STIRLING,G. A. and CASTOR,U. S. (1962). Tubal Pregnancy in Jamaicans. West Indian Medical Journal, 11, 45-47. STEWART,D. B. and SKINNER,S. M. (1967). Tubal Ectopic Pregnancy. Manitoba Medical Review, 47, 552-555. TROELL, S. (1970). Diverticula of the Walls of the Fallopian Tube. Acta Obstetrica et Gynaecologica Scandinavica, 49, 17-20. WRORK, D. H. and BRODERS,A. C. (1942). Adenomyosis of the Fallopian Tubes. American Journal of Obstetrics and Gynaecology, 44, 412-432.
The Indices for V o l u m e X X V which n o r m a l l y appear in the October 1974 issue will n o w be included in the J a n u a r y 1975 issue, V o l u m e X X V I No. 1. The Publishers a n d Editor apologise for the inconvenience caused.