Diagnostic Radiology
Hysterographic Evidence of Polypoid Filling Defects in the Uterine Cavity 1 Premysl Slezak, M.D., and K. G. Tlllinger, M.D. According to earlier studies, radiographic evidence of polypoid filling defects in the uterine cavity has a physiological background if hysterography is carried out during the secretory phase of the menstrual cycle, but if performed during the proliferative phase, the defects indicate endometrial hyperplasia. This view was not confirmed in the present investigation. Of 8,000 hysterograms reviewed, 200 (2.5 %) showed filling defects. Of these, 162 hysterograms were from cases in which the coexisting histological appearance of the endometrium was available; 78% of the histological examinations showed no abnormalities. There was no difference between the secretory and proliferative phases in this respect. and no correlation between radiographic evidence of polypoid filling defects and abnormal uterine bleeding. INDEX TERMS: phy
•
Uterine Neoplasms, diagnosis. Uterus. abnormalities • Uterus, radiogra-
Radiology 115:79-83, April 1975
•
AN IRREGULAR OUTLINE of the 1-\ onstrated by hysterography
(8) described the radiographic appearance of the endometrium. The concensus of opinion appears to be that the hysterogram reflects endometrial hyperplasia as round or oval filling defects. Beclere (4) expressed the view that 90 % of the "images lacunaires " represent polyps, carcinoma or myoma, while 10 % are due to endometrial hyperplasia. According to Kjellberg (21), Erbsloh (17) and Schultze and Erbsloh (36), endometrial hy-
uterine cavity dem(HSG) may indicate endometrial lesions such as hyperplasia, polyps or endometritis which can cause abnormal bleeding. The importance of HSG in the diagnosis of bleeding problems was first emphasized by French investigators (3-7, 10, 13, 14, 17,23-25,30-37,40). In Scandinavia, Kjellberg (21), Bakke (2), Norman (26) and Bergman and Wehlin
b
Fig. 1. A through D. Hysterogram showing polypoid filling defects. The histological findings associated with these defects were as follows: (A) Normal endometrium on the 25th day of the menstrual cycle; (B) the 19th day; (C) 15th day; and (D) late in the proliferative phase. 1 From the Department of Diagnostic Radiology (P. S.)} the University School of Medicine. Linkoping, Sweden, and the Women's Clinic SOdersjukhuset (K. G. T., Head), Stockholm, Sweden. Accepted for publication in October 1974. shan
79
PREMYSL SLEZAK AND K. G. TILLINGER
80
Table I:
• •••
April 1975
Size and Distribution of Filling Defects No. of Hysterograms
Group
41 94
I II III
Table II:
Size in mm
10
32
Histological Findings Coexisting with Those in the
162 Hysterograms and Their Distribution in the Three SizeGroups Histological
Fig. 2. Hysterogram showing the size of the polypoid filling defects included in Group I.
perplasia is suggested if the endometrial folds formed in the proliferative phase of the menstrual cycle are unduly thick. MATERIAL AND METHODS
In a review of 8,000 hysterograms, 200 (2.5 %) showed polypoid filling defects (Fig. 1). Excluded were 33 hysterograms because the case records afforded incomplete or no information about the histological appearance of the endometrium at the time of HSG. The remaining 167 hysterograms, derived from 153 patients, form the basis of this investigation. The patients ranged from 17 to 55 years of age (average 32 ± 7 years). The filling defects were measured directly on the hysterograms, and divided into three groups according to size (TABLE I) (Figs. 2, 3 and 4). The histological diagnosis in Figure 2 was cystic glandular hyperplasia. Figure 3, A demonstrates histologically confirmed mild endometritis. The histological findings in Figure 3, B were of a normal endometrium typical of the 18th day of the menstrual cycle. There was no evidence of polyp fragments. HSG carried out a few days before curettage disclosed a bean-sized polyp in the isthmus. The histological findings for Figure 4, A were of the normal endometrium on the 25th day of the menstrual cycle. There was no evidence of polyp fragments. In the patient in Figure 4, B, a submucosal fundal myoma and a small polyp in the isthmus were identified at hysteroscopy. The histological finding was normal endometrium at about the 22d day of the menstrual cycle. Eighteen days after hysteroscopy, the myoma was enucleated. The hospital records of the 153 patients were reviewed with special reference to gynecological diagnosis, bleeding pattern, findings at hysterectomy and hysteroscopy, and phase of the menstrual cycle at the time of hysteroscopy, hysterectomy and curettage. The conditions for which HSG Was carried out and their distribution among the three size groups are as follows (patients with several diagnoses are included).
Group I
Group II
Group III
Appearan~e of (39 HSGs) (92 HSGs) (31 HSGs) Total HSGs Endometrium - - - - - -- _ . _ - - - - - - - 126 (78%) 24 Normal 30 72 36 (22%) 20 7 Abnormal 9
Group I Primary infertility Secondary infertility Menometrorrhagia Uterine myoma Dysmenorrhea Ovarian cystoma Chronic non-specific salpingitis Habitual abortion Cause of complaint unspecified
No. of Hysterograms 17
5 9 4 1 1 1 1 1
Group II Primary infertility Secondary infertility Menometrorrhagia Uterine myoma Habitual abortion Dysmenorrhea Ovarian endometriosis Non-specific endometritis Chronic salpingitis Cause of complaint unspecified
47 10 15 7 3 8 3 2
1
6
Group III Primary infertility Secondary infertility Uterine myoma Menometrorrhagia Dysmenorrhea Ovarian endometritis Ovarian thecoma
14 9 6
3 3 1 1
Endometrial tissue was obtained for histological examination by fractional curettage, diagnostic curettage and hysterectomy. Of the 167 hysterograms studied, histological findings were available for 162. The endometrial specimen was obtained in the same phase of the menstrual cycle as the hysterograph in 120 cases (74 %); 116 (97 %) were obtained within seven days after HSG. In 42 specimens (26 %), the cycle phases did not correspond. Endometrial thickness was measured in 84 specimens.
81
POL YPOID FILLING DEFECTS IN THE UTERINE CAVITY
Vol. 115
Diagnostic Radiology
Fig. 3. A. Hysterogram demonstrating polypoid filling defects in the whole uterine cavity corresponding in size to those included in Group II. B. Polypoid filling defects in the lower part of the uterine cavity corresponding in size to those included in Group II.
RESULTS
The phase of the menstrual cycle could be determined for 140 of the 167 hysterograms. In 90 specimens (70 %), the endometrium was in the secretory phase at the time of HSG. TABLE II shows the histological findings for the 162 hysterograms analyzed. The abnormal histological findings and their distribution among the three size groups were as follows:
Histologi cal Findings Group I:
Glandular and/ or cystic glandular hyperpla sia Prol ife rative phase and superficial necrot ic area s Protra cted proliferative phase
Group II :
Group II I:
Glandular and/ or cystic glandular hyperpl asia Premature shedding Non-tub erculou s endometriti s Polyps of the body of the uteru s Cystic glandular hyperpla sia Irregular sec retio n Exce ssive endom etr ial edema Polyps of the body of Ihe ute rus
No. of iHysterograms
Table III : Type of Abnormal Uterine Bleed ing Assoc iated with the Find ings in 167 Hysterograms and Th eir Distribution in the Thre e Size Groups Type of Uterin e Blee ding
Group I Group II Group III (41 HSGs) (94 HSGs) (32 HSGs)
- - - - - - - -20 21
Norm al Abnormal
Table IV:
Total HSGs 102 (61%) 65 (39%)
20 12
Average Thickn ess (Il) of the Endometrium in the Three Size Groups
Group I
Group III
3,015 ± 1,213
3,424 ± 567
4 1 4 7
1 2 10 1
1 1
4
Almost four fifths of the histological examinations of the endometrium showed no abnormality. There was no difference between the three groups with respect to the proportion of abnormal-to-normal findings. More than 50 % of the abnormal findings were due to endocrine disturbances. T ABLE III shows the type of uterine bleeding associated with the findings in the 167 hysterograms and their distribution among the three size groups. Abnormal uterine bleeding resembling menometror-
a Fig. 4. A. Hysterogram showing polypoid filling defects in the whole uterine cavity corresponding in size to those included in Group 111. B. Polypoid filling defect in the fundus of the uterine cavity correspond ing in size to those included in Group III.
82
PREMYSL SLEZAK AND K. G. TILLINGER
rhaqla, i.e., copious and prolonged menstrual flow, was associated with the findings in 39 % of the hysterograms. The relatively high incidence of abnormal uterine bleeding is not surprising as this condition was the indication for HSG in many of the cases in this series. Abnormal uterine bleeding was most frequently associated with small filling defects (Group I) in the uterine cavity. As already mentioned, the thickness of the endome.trium at the time of HSG was measured in 84 endometrial specimens (TABLE IV). There was no noteworthy difference between the three size groups with respect to the thickness of the endometrium. ENDOMETRIAL POLYPS
In view of the low incidence of histologically confirmed polyps of the corpus uteri (about 8 %) in the cases in this series in which HSG was carried out on 167 occasions, one is inclined to question the value of histological examination in the diagnosis of endometrial polyps, particularly as HSG often discloses changes suggesting the presence of polyps in the absence of histological evidence. In order to throw some light on this point, this investigation was extended to 28 cases with hysteroscopically verified endometrial polyps; 24 patients had been examined by HSG prior to hysteroscopy. In no case did the interval between the two examinations exceed seven days. In all cases, the endometrium was histologically examined. In 25, the polyps were removed under hysteroscopic control and were subsequently examined separately. Of the 28 cases, the clinical diagnosis was histologically confirmed in 18. In 23 of the 24 patients examined by HSG prior to hysteroscopy, there was radiographic evidence of polypoid filling defects whose size and location corresponded to the polypoid lesions identified at hysteroscopy. Close correlation was demonstrated between the hysterographic and hysteroscopic findings but these did not correspond to the histological findings. (Ten in vitro experiments in which mucosal secretion was collected at HSG and transferred to a glass tube containing opaque medium, have shown that mucosal secretion may cause filling defects as late as 180 minutes after the beginning of the experiment.) DISCUSSION
Histological examination disclosed the normal endometrium to be associated with four fifths of hysterograms displaying polypoid filling defects in the uterine cavity. This does not agree with the observations of French investigators (3, 4, 10, 32). Malignant changes were not observed in the present investigation, nor any lesions thought to be precancerous in character such as "adenomatous hyperplasia." According to the French studies, there is a relationship between filling defects in the uterine cavity and menstrual disturbances.
April 1975
The findings in this investigation do not support this view as only 1 patient in 3 had a history of metrorrhagia or menometrorrhagia. Abnormal changes in the endometrium were found in only 1 in 5 cases. The difference between these results and the findings of earlier investigators probably lies in the meaning of the term "endometrial hyperplasia" as it has not yet been clearly defined. Many investigators have pointed out that if the endometrium is in the secretory phase of the menstrual cycle at the time of HSG, radiographic evidence of an irregular outline of the uterine cavity may have a physiological background (17, 21, 24, 36). On the other hand, Schultze and Erbsloh (36) feel that if the endometrium is in the proliferative phase at the time of HSG, the radiographic appearance of polypoid filling defects suggests endometrial hyperplasia. The observations made in the present investigation do not support this view: there was no difference between the secretory and proliferative phase with respect to the proportion of endometrial lesions. The lack of correlation between the histological and hysterographic and hysteroscopic findings of endometrial polyps in this series of cases may reside in the difficulty involved in the histological diagnosis of an endometrial polyp. A polyp may be missed at a curettage seemingly performed lege artis. Word (38) found 7 intact polyps in 178 uteri which were removed following curettage. In a subsequent investigation (39), he identified 38 polyps in 512 uteri removed after curettage. Englund et al. (16) observed polyps at hysteroscopy performed immediately after curettage in 33 of 124 cases. Burnett (11) found polyps in 121 of 1,289 uteri removed by hysterectomy. He reported that the incidence of "polyps missed at recent curettage [was] 47 %, [that of] failure to remove all polyps at recent curettage [being] 5 % ... " There may be no difference between the endometrium of a polyp and the endometrium in general. Novak (27) expressed the view that "the term polyp is a clinical rather than a pathological one .... " The histological diagnosis of a polyp is easy, provided the entire polyp is removed at curettage. Peterson and Novak (29) expressed the view that so-called functional uterine bleeding in reality is due to a polyp that has been so fragmented during curettage that it is not histologically recognized as such. The results of earlier studies (27, 29) and the findings in this investigation substantiate the view that histological examination is not the most reliable method for confirming the presence of endometrial polyps. As already mentioned, there was no correlation between the histological findings and hysterographic or hysteroscopic findings. However, this does not belittle the value of tissue study as it is the only available method for determining the nature of a polyp. The diagnosis of a polyp in cases of abnormal uterine bleeding is of limited value. In Burnett's (11) large series of abnormal bleeding cases, polyps accounted for only
Vol. 115
POLYPOID FILLING DEFECTS IN THE UTERINE CAVITY
3.2%. A polyp per se does not necessarily give rise to bleeding. Malignant degeneration of a polyp of the body of the uterus may occasionally occur (27, 29, 35). According to Peterson and Novak (29), the incidence of malignant degeneration is 0.36 %. On the other hand, the association of endometrial polyps with carcinoma of the reproductive organs is frequently encountered. According to the literature, the association of endometrial polyps and endometrial carcinoma in post-menopausal women ranges from 4.6% (22) to 37.6% (18); that of endometrial polyps and genital carcinoma as a whole is 57.9% (19).In a 12-year follow-up study of 1,482 women with endometrial polyps, Armenia (1) found that the risk of carcinoma developing in the body of the uterus is nine times greater in these cases. According to Peterson and Novak (29), the incidence of the association of endometrial polyps and endometrial carcinoma is 10%. They stressed however, that this applied primarily to post-menopausal women. The existence of a relationship between endometrial hyperplasia and endometrial carcinoma has been observed by earlier investigators (9, 12, 15,20,28,35). The incidence of endometrial hyperplasia progressing to carcinoma reportedly ranges from 0.31 % to 25% (28). Novak (28) wrote: " ... glandular hyperplasia of the endometrium is a common precursor of carcinoma but carcinoma occurs only in a small proportion of women with endometrial hyperplasia.... " Premysl Slezak, M.D. Department of Diagnostic Radiology Link6ping University School of Medicine 581 85 Link6ping, Sweden
REFERENCES 1. Armenia CS: Sequential relationship between endometrial polyps and carcinoma of the endometrium. Obstet Gynecol 30: 524-529, Oct 1967 2. Bakke SN: Uterosalpingography. Bergen. 1927 3. Beclere Diagnostic differentiel et etiologique des hemmorragies uterlnes fonctionelles. Semin Hop Paris 42: 1152-1160, 14 Apr 1966 4. Beclere C: L'hystero-salpingographie dans Ie diagnostic des hemorragies uterines fonctionelles. Gynec Prat 16: 17-42, 1965 5. Beclere C: Les hsmorraqtes uterines avant et apres la menopause. Paris, Masson et Cie, 1936 6. Beclere C: l.'hysteroqraphls dans Ie diagnostic des lesions lntra-uterlnes et des rnetrorrhagles fonctionelles. Bull Soc Obstet Gynecol 22:815-822, Dec 1933 7. Belbeoch P: Forme tumorale d'une hyperplasie de I'endomefre. Bull Fed Gynec Obstet Franc 17: 196, Apr-May, 1965 8. Bergman P, Wehlin L: The hysterographic appearance of cystic glandular hyperplasia. Acta RadioI50:255-260. Sep 1958 9. Bettinger HF: Hyperplasia and carcinoma of the endometrium. Am J Obstet GynecoI109:194-197. 15 Jan 1971 10. Brocq P, Moulonguet-Doleris P. Maricot R. et al: Atlas d'hysterographie. Paris, Masson & Cie, 1953 11. BurnettJE Jr: Hysteroscopy-controlled curettage for endometrial polyps. Obstet GynecoI24:621-625, Oct 1964
C:"
Diagnostic Radiology
83
12. Chamlian DL, Taylor HB: Endometrial hyperplasia in young women. Obstet GynecoI36:659-666, Nov 1970 13. Contreras-Sanches J: Hiperplasia endometrial y su correlation clfnica; presentaci6n de 58 casos. Ginecol Obstet Mex 18: 343-353, Jul-Aug 1963 14. Dalsace J. Garcia-Calderon J: Gynecologic Radiography. NeW York, HoeberDiv., Harper and Row, 1959 15. Dcmotorl J, Rechnitz K, Agoston J: Ober hyperplastisches Endometrium bei Korpuskarzinom. Zbl Gynaek 85:1362-1365. 21 Sep 1963 16. Englund S, Ingelman-Sundberg A, Westin B: Hysteroscopy in diagnosis and treatment of uterine bleeding. Gynaecologia (Basel) 143:217-222. Mar 1957 17. Erbsloh J: Das R6ntgenbild der Metrorrhagien und seine Differentialdiagnosen. Zbl Gynaek 79:1047-1056, 6 Jul 1957 18. Huber H: Ober die Polypen der Zervix- und Korpusschleimhaul. Zbl Gynaek 72:833-849, 1950 19. Huber H: Zur Klinik der Schleimhautpolypen des Uterus. Geburtsch Frauenh 11:675-690. Aug 1951 20. Hunter RG, Henry GW: Cervicohysterosalpingography. Preliminary report on a new cannula and technic. Fertil Steril 6:68-79, Jan-Feb 1955 21. Kjellberg SR: Hystero-salpingo-pelvigraphie. Acta Radiol (SuppI43): 1942 22. Lau H, Stoll P: Das Adenom des Corpus uteri. Deutsch Med Wschr87:1005-1012, 11 May 1962 23. Lemaitre G. Du Bois R, Empereur-Buisson R: Les etats hyperplasiques de l'endornetre. J Radiol Electr 45:603-607, Oct 1964 24. Menees T, Miller JD: Recent advances in hysterography. Am J Obstet Gynecol 30:540-595. Oct 1935 25. Mikaelian S, Rochet Y, Tommasi M, et al: Un cas excepune tionnel d'hyperplasie monstrueuse de l'endornetre associee alutelnle et une endornetriose pelvienne chez une jeune fille de 18 ans, Bull Fed Gynec Obstet Franc 18:268-269, Jun-Aug 1966 26. Norman 0: Hysterography in cancer of the corpus of the uterus. Acta Radiol (SuppI79): 1-156. 1950 27. Novak ER, Woodruff JD: Novak's Gynecologic and Obstetric Pathology. Philadelphia, Saunders, 6th ed, 1967 28. Novak ER: Relationship of endometrial hyperplasia and adenocarcinoma of the uterine fundus. JAMA 154:217-220, 16 Jan 1954 29. Peterson WF, Novak ER: Endometrial polyps. Obstet Gynecol 8:40-49, Jul 1956 30. ProustR, B€lclere C: De i'lnteret de I'examen radiologique apres injection de Lipiodol dans Ie diagnostic des rnetrorraqles. Gynecol Obstet (Paris) 16:408-413, Nov 1927 31. Robins SA, Shapiro AD: Uterotubography. [In] Golden's Diagnostic Roentgenology. Baltimore, Williams & Wilkins, 1956 Vol II, chap 9 32. Rochet Y, Mikaelian S: Valeur de l'hysterographie dans la diagnostic des rnetrorrhaqies post-mencpauslques et plus partlouIierement du cancer de l'endometre, Lyon Chir 66:321-324, SepOct 1970. 33. Ross L, Bensusen CI: Giant polypoid cystic endometrial hyperp/asia. Am J Obstet Gynec 95: 1164-1165. Aug 1966 34. Rozin S: Uterosalpingography in Gynecology. Springfield, III., Thomas, 1965 35. Schroder R: Endometrial hyperplasia in relation to genital function. Am J Obstet Gynecol 68:295-309, Jul 1954 36. Schultze GKF. Erbsl6h J: Gynakologische R6ntgendiagnostik. Stuttgart, Ferdinand Enke Verlag, 1954 37. Siegler AM: Hysterosalpingography. New York, Hoeber Med. Div.• Harper & Row, 1967 38. Word B: Pitfalls of uterine curettage. South Med J 47:3847, Jan 1954 39. Word B, Gravlee LC, Wideman GL: The fallacy of simple uterine curettage. Obstet Gynecol 12:642-648, Dec 1958 40. Zuppinger A: Rontgendiagnostik in der Gynakoloqie. [In] Lehrbuch der R6ntgendiagnostik. Schinz HR, Baensch WE, ed. Stuttgart, Georg Thieme Verlag, 1952
a
a
Hysterographic Evidence of Polypoid Filling Defects in the Uterine Cavity 1 Premysl Slezak, M.D., and K. G. Tlllinger, M.D. According to earlier studies, radiographic evidence of polypoid filling defects in the uterine cavity has a physiological background if hysterography is carried out during the secretory phase of the menstrual cycle, but if performed during the proliferative phase, the defects indicate endometrial hyperplasia. This view was not confirmed in the present investigation. Of 8,000 hysterograms reviewed, 200 (2.5 %) showed filling defects. Of these, 162 hysterograms were from cases in which the coexisting histological appearance of the endometrium was available; 78% of the histological examinations showed no abnormalities. There was no difference between the secretory and proliferative phases in this respect. and no correlation between radiographic evidence of polypoid filling defects and abnormal uterine bleeding. INDEX TERMS: phy
•
Uterine Neoplasms, diagnosis. Uterus. abnormalities • Uterus, radiogra-
Radiology 115:79-83, April 1975
•
AN IRREGULAR OUTLINE of the 1-\ onstrated by hysterography
(8) described the radiographic appearance of the endometrium. The concensus of opinion appears to be that the hysterogram reflects endometrial hyperplasia as round or oval filling defects. Beclere (4) expressed the view that 90 % of the "images lacunaires " represent polyps, carcinoma or myoma, while 10 % are due to endometrial hyperplasia. According to Kjellberg (21), Erbsloh (17) and Schultze and Erbsloh (36), endometrial hy-
uterine cavity dem(HSG) may indicate endometrial lesions such as hyperplasia, polyps or endometritis which can cause abnormal bleeding. The importance of HSG in the diagnosis of bleeding problems was first emphasized by French investigators (3-7, 10, 13, 14, 17,23-25,30-37,40). In Scandinavia, Kjellberg (21), Bakke (2), Norman (26) and Bergman and Wehlin
b
Fig. 1. A through D. Hysterogram showing polypoid filling defects. The histological findings associated with these defects were as follows: (A) Normal endometrium on the 25th day of the menstrual cycle; (B) the 19th day; (C) 15th day; and (D) late in the proliferative phase. 1 From the Department of Diagnostic Radiology (P. S.)} the University School of Medicine. Linkoping, Sweden, and the Women's Clinic SOdersjukhuset (K. G. T., Head), Stockholm, Sweden. Accepted for publication in October 1974. shan
79
PREMYSL SLEZAK AND K. G. TILLINGER
80
Table I:
• •••
April 1975
Size and Distribution of Filling Defects No. of Hysterograms
Group
41 94
I II III
Table II:
Size in mm
10
32
Histological Findings Coexisting with Those in the
162 Hysterograms and Their Distribution in the Three SizeGroups Histological
Fig. 2. Hysterogram showing the size of the polypoid filling defects included in Group I.
perplasia is suggested if the endometrial folds formed in the proliferative phase of the menstrual cycle are unduly thick. MATERIAL AND METHODS
In a review of 8,000 hysterograms, 200 (2.5 %) showed polypoid filling defects (Fig. 1). Excluded were 33 hysterograms because the case records afforded incomplete or no information about the histological appearance of the endometrium at the time of HSG. The remaining 167 hysterograms, derived from 153 patients, form the basis of this investigation. The patients ranged from 17 to 55 years of age (average 32 ± 7 years). The filling defects were measured directly on the hysterograms, and divided into three groups according to size (TABLE I) (Figs. 2, 3 and 4). The histological diagnosis in Figure 2 was cystic glandular hyperplasia. Figure 3, A demonstrates histologically confirmed mild endometritis. The histological findings in Figure 3, B were of a normal endometrium typical of the 18th day of the menstrual cycle. There was no evidence of polyp fragments. HSG carried out a few days before curettage disclosed a bean-sized polyp in the isthmus. The histological findings for Figure 4, A were of the normal endometrium on the 25th day of the menstrual cycle. There was no evidence of polyp fragments. In the patient in Figure 4, B, a submucosal fundal myoma and a small polyp in the isthmus were identified at hysteroscopy. The histological finding was normal endometrium at about the 22d day of the menstrual cycle. Eighteen days after hysteroscopy, the myoma was enucleated. The hospital records of the 153 patients were reviewed with special reference to gynecological diagnosis, bleeding pattern, findings at hysterectomy and hysteroscopy, and phase of the menstrual cycle at the time of hysteroscopy, hysterectomy and curettage. The conditions for which HSG Was carried out and their distribution among the three size groups are as follows (patients with several diagnoses are included).
Group I
Group II
Group III
Appearan~e of (39 HSGs) (92 HSGs) (31 HSGs) Total HSGs Endometrium - - - - - -- _ . _ - - - - - - - 126 (78%) 24 Normal 30 72 36 (22%) 20 7 Abnormal 9
Group I Primary infertility Secondary infertility Menometrorrhagia Uterine myoma Dysmenorrhea Ovarian cystoma Chronic non-specific salpingitis Habitual abortion Cause of complaint unspecified
No. of Hysterograms 17
5 9 4 1 1 1 1 1
Group II Primary infertility Secondary infertility Menometrorrhagia Uterine myoma Habitual abortion Dysmenorrhea Ovarian endometriosis Non-specific endometritis Chronic salpingitis Cause of complaint unspecified
47 10 15 7 3 8 3 2
1
6
Group III Primary infertility Secondary infertility Uterine myoma Menometrorrhagia Dysmenorrhea Ovarian endometritis Ovarian thecoma
14 9 6
3 3 1 1
Endometrial tissue was obtained for histological examination by fractional curettage, diagnostic curettage and hysterectomy. Of the 167 hysterograms studied, histological findings were available for 162. The endometrial specimen was obtained in the same phase of the menstrual cycle as the hysterograph in 120 cases (74 %); 116 (97 %) were obtained within seven days after HSG. In 42 specimens (26 %), the cycle phases did not correspond. Endometrial thickness was measured in 84 specimens.
81
POL YPOID FILLING DEFECTS IN THE UTERINE CAVITY
Vol. 115
Diagnostic Radiology
Fig. 3. A. Hysterogram demonstrating polypoid filling defects in the whole uterine cavity corresponding in size to those included in Group II. B. Polypoid filling defects in the lower part of the uterine cavity corresponding in size to those included in Group II.
RESULTS
The phase of the menstrual cycle could be determined for 140 of the 167 hysterograms. In 90 specimens (70 %), the endometrium was in the secretory phase at the time of HSG. TABLE II shows the histological findings for the 162 hysterograms analyzed. The abnormal histological findings and their distribution among the three size groups were as follows:
Histologi cal Findings Group I:
Glandular and/ or cystic glandular hyperpla sia Prol ife rative phase and superficial necrot ic area s Protra cted proliferative phase
Group II :
Group II I:
Glandular and/ or cystic glandular hyperpl asia Premature shedding Non-tub erculou s endometriti s Polyps of the body of the uteru s Cystic glandular hyperpla sia Irregular sec retio n Exce ssive endom etr ial edema Polyps of the body of Ihe ute rus
No. of iHysterograms
Table III : Type of Abnormal Uterine Bleed ing Assoc iated with the Find ings in 167 Hysterograms and Th eir Distribution in the Thre e Size Groups Type of Uterin e Blee ding
Group I Group II Group III (41 HSGs) (94 HSGs) (32 HSGs)
- - - - - - - -20 21
Norm al Abnormal
Table IV:
Total HSGs 102 (61%) 65 (39%)
20 12
Average Thickn ess (Il) of the Endometrium in the Three Size Groups
Group I
Group III
3,015 ± 1,213
3,424 ± 567
4 1 4 7
1 2 10 1
1 1
4
Almost four fifths of the histological examinations of the endometrium showed no abnormality. There was no difference between the three groups with respect to the proportion of abnormal-to-normal findings. More than 50 % of the abnormal findings were due to endocrine disturbances. T ABLE III shows the type of uterine bleeding associated with the findings in the 167 hysterograms and their distribution among the three size groups. Abnormal uterine bleeding resembling menometror-
a Fig. 4. A. Hysterogram showing polypoid filling defects in the whole uterine cavity corresponding in size to those included in Group 111. B. Polypoid filling defect in the fundus of the uterine cavity correspond ing in size to those included in Group III.
82
PREMYSL SLEZAK AND K. G. TILLINGER
rhaqla, i.e., copious and prolonged menstrual flow, was associated with the findings in 39 % of the hysterograms. The relatively high incidence of abnormal uterine bleeding is not surprising as this condition was the indication for HSG in many of the cases in this series. Abnormal uterine bleeding was most frequently associated with small filling defects (Group I) in the uterine cavity. As already mentioned, the thickness of the endome.trium at the time of HSG was measured in 84 endometrial specimens (TABLE IV). There was no noteworthy difference between the three size groups with respect to the thickness of the endometrium. ENDOMETRIAL POLYPS
In view of the low incidence of histologically confirmed polyps of the corpus uteri (about 8 %) in the cases in this series in which HSG was carried out on 167 occasions, one is inclined to question the value of histological examination in the diagnosis of endometrial polyps, particularly as HSG often discloses changes suggesting the presence of polyps in the absence of histological evidence. In order to throw some light on this point, this investigation was extended to 28 cases with hysteroscopically verified endometrial polyps; 24 patients had been examined by HSG prior to hysteroscopy. In no case did the interval between the two examinations exceed seven days. In all cases, the endometrium was histologically examined. In 25, the polyps were removed under hysteroscopic control and were subsequently examined separately. Of the 28 cases, the clinical diagnosis was histologically confirmed in 18. In 23 of the 24 patients examined by HSG prior to hysteroscopy, there was radiographic evidence of polypoid filling defects whose size and location corresponded to the polypoid lesions identified at hysteroscopy. Close correlation was demonstrated between the hysterographic and hysteroscopic findings but these did not correspond to the histological findings. (Ten in vitro experiments in which mucosal secretion was collected at HSG and transferred to a glass tube containing opaque medium, have shown that mucosal secretion may cause filling defects as late as 180 minutes after the beginning of the experiment.) DISCUSSION
Histological examination disclosed the normal endometrium to be associated with four fifths of hysterograms displaying polypoid filling defects in the uterine cavity. This does not agree with the observations of French investigators (3, 4, 10, 32). Malignant changes were not observed in the present investigation, nor any lesions thought to be precancerous in character such as "adenomatous hyperplasia." According to the French studies, there is a relationship between filling defects in the uterine cavity and menstrual disturbances.
April 1975
The findings in this investigation do not support this view as only 1 patient in 3 had a history of metrorrhagia or menometrorrhagia. Abnormal changes in the endometrium were found in only 1 in 5 cases. The difference between these results and the findings of earlier investigators probably lies in the meaning of the term "endometrial hyperplasia" as it has not yet been clearly defined. Many investigators have pointed out that if the endometrium is in the secretory phase of the menstrual cycle at the time of HSG, radiographic evidence of an irregular outline of the uterine cavity may have a physiological background (17, 21, 24, 36). On the other hand, Schultze and Erbsloh (36) feel that if the endometrium is in the proliferative phase at the time of HSG, the radiographic appearance of polypoid filling defects suggests endometrial hyperplasia. The observations made in the present investigation do not support this view: there was no difference between the secretory and proliferative phase with respect to the proportion of endometrial lesions. The lack of correlation between the histological and hysterographic and hysteroscopic findings of endometrial polyps in this series of cases may reside in the difficulty involved in the histological diagnosis of an endometrial polyp. A polyp may be missed at a curettage seemingly performed lege artis. Word (38) found 7 intact polyps in 178 uteri which were removed following curettage. In a subsequent investigation (39), he identified 38 polyps in 512 uteri removed after curettage. Englund et al. (16) observed polyps at hysteroscopy performed immediately after curettage in 33 of 124 cases. Burnett (11) found polyps in 121 of 1,289 uteri removed by hysterectomy. He reported that the incidence of "polyps missed at recent curettage [was] 47 %, [that of] failure to remove all polyps at recent curettage [being] 5 % ... " There may be no difference between the endometrium of a polyp and the endometrium in general. Novak (27) expressed the view that "the term polyp is a clinical rather than a pathological one .... " The histological diagnosis of a polyp is easy, provided the entire polyp is removed at curettage. Peterson and Novak (29) expressed the view that so-called functional uterine bleeding in reality is due to a polyp that has been so fragmented during curettage that it is not histologically recognized as such. The results of earlier studies (27, 29) and the findings in this investigation substantiate the view that histological examination is not the most reliable method for confirming the presence of endometrial polyps. As already mentioned, there was no correlation between the histological findings and hysterographic or hysteroscopic findings. However, this does not belittle the value of tissue study as it is the only available method for determining the nature of a polyp. The diagnosis of a polyp in cases of abnormal uterine bleeding is of limited value. In Burnett's (11) large series of abnormal bleeding cases, polyps accounted for only
Vol. 115
POLYPOID FILLING DEFECTS IN THE UTERINE CAVITY
3.2%. A polyp per se does not necessarily give rise to bleeding. Malignant degeneration of a polyp of the body of the uterus may occasionally occur (27, 29, 35). According to Peterson and Novak (29), the incidence of malignant degeneration is 0.36 %. On the other hand, the association of endometrial polyps with carcinoma of the reproductive organs is frequently encountered. According to the literature, the association of endometrial polyps and endometrial carcinoma in post-menopausal women ranges from 4.6% (22) to 37.6% (18); that of endometrial polyps and genital carcinoma as a whole is 57.9% (19).In a 12-year follow-up study of 1,482 women with endometrial polyps, Armenia (1) found that the risk of carcinoma developing in the body of the uterus is nine times greater in these cases. According to Peterson and Novak (29), the incidence of the association of endometrial polyps and endometrial carcinoma is 10%. They stressed however, that this applied primarily to post-menopausal women. The existence of a relationship between endometrial hyperplasia and endometrial carcinoma has been observed by earlier investigators (9, 12, 15,20,28,35). The incidence of endometrial hyperplasia progressing to carcinoma reportedly ranges from 0.31 % to 25% (28). Novak (28) wrote: " ... glandular hyperplasia of the endometrium is a common precursor of carcinoma but carcinoma occurs only in a small proportion of women with endometrial hyperplasia.... " Premysl Slezak, M.D. Department of Diagnostic Radiology Link6ping University School of Medicine 581 85 Link6ping, Sweden
REFERENCES 1. Armenia CS: Sequential relationship between endometrial polyps and carcinoma of the endometrium. Obstet Gynecol 30: 524-529, Oct 1967 2. Bakke SN: Uterosalpingography. Bergen. 1927 3. Beclere Diagnostic differentiel et etiologique des hemmorragies uterlnes fonctionelles. Semin Hop Paris 42: 1152-1160, 14 Apr 1966 4. Beclere C: L'hystero-salpingographie dans Ie diagnostic des hemorragies uterines fonctionelles. Gynec Prat 16: 17-42, 1965 5. Beclere C: Les hsmorraqtes uterines avant et apres la menopause. Paris, Masson et Cie, 1936 6. Beclere C: l.'hysteroqraphls dans Ie diagnostic des lesions lntra-uterlnes et des rnetrorrhagles fonctionelles. Bull Soc Obstet Gynecol 22:815-822, Dec 1933 7. Belbeoch P: Forme tumorale d'une hyperplasie de I'endomefre. Bull Fed Gynec Obstet Franc 17: 196, Apr-May, 1965 8. Bergman P, Wehlin L: The hysterographic appearance of cystic glandular hyperplasia. Acta RadioI50:255-260. Sep 1958 9. Bettinger HF: Hyperplasia and carcinoma of the endometrium. Am J Obstet GynecoI109:194-197. 15 Jan 1971 10. Brocq P, Moulonguet-Doleris P. Maricot R. et al: Atlas d'hysterographie. Paris, Masson & Cie, 1953 11. BurnettJE Jr: Hysteroscopy-controlled curettage for endometrial polyps. Obstet GynecoI24:621-625, Oct 1964
C:"
Diagnostic Radiology
83
12. Chamlian DL, Taylor HB: Endometrial hyperplasia in young women. Obstet GynecoI36:659-666, Nov 1970 13. Contreras-Sanches J: Hiperplasia endometrial y su correlation clfnica; presentaci6n de 58 casos. Ginecol Obstet Mex 18: 343-353, Jul-Aug 1963 14. Dalsace J. Garcia-Calderon J: Gynecologic Radiography. NeW York, HoeberDiv., Harper and Row, 1959 15. Dcmotorl J, Rechnitz K, Agoston J: Ober hyperplastisches Endometrium bei Korpuskarzinom. Zbl Gynaek 85:1362-1365. 21 Sep 1963 16. Englund S, Ingelman-Sundberg A, Westin B: Hysteroscopy in diagnosis and treatment of uterine bleeding. Gynaecologia (Basel) 143:217-222. Mar 1957 17. Erbsloh J: Das R6ntgenbild der Metrorrhagien und seine Differentialdiagnosen. Zbl Gynaek 79:1047-1056, 6 Jul 1957 18. Huber H: Ober die Polypen der Zervix- und Korpusschleimhaul. Zbl Gynaek 72:833-849, 1950 19. Huber H: Zur Klinik der Schleimhautpolypen des Uterus. Geburtsch Frauenh 11:675-690. Aug 1951 20. Hunter RG, Henry GW: Cervicohysterosalpingography. Preliminary report on a new cannula and technic. Fertil Steril 6:68-79, Jan-Feb 1955 21. Kjellberg SR: Hystero-salpingo-pelvigraphie. Acta Radiol (SuppI43): 1942 22. Lau H, Stoll P: Das Adenom des Corpus uteri. Deutsch Med Wschr87:1005-1012, 11 May 1962 23. Lemaitre G. Du Bois R, Empereur-Buisson R: Les etats hyperplasiques de l'endornetre. J Radiol Electr 45:603-607, Oct 1964 24. Menees T, Miller JD: Recent advances in hysterography. Am J Obstet Gynecol 30:540-595. Oct 1935 25. Mikaelian S, Rochet Y, Tommasi M, et al: Un cas excepune tionnel d'hyperplasie monstrueuse de l'endornetre associee alutelnle et une endornetriose pelvienne chez une jeune fille de 18 ans, Bull Fed Gynec Obstet Franc 18:268-269, Jun-Aug 1966 26. Norman 0: Hysterography in cancer of the corpus of the uterus. Acta Radiol (SuppI79): 1-156. 1950 27. Novak ER, Woodruff JD: Novak's Gynecologic and Obstetric Pathology. Philadelphia, Saunders, 6th ed, 1967 28. Novak ER: Relationship of endometrial hyperplasia and adenocarcinoma of the uterine fundus. JAMA 154:217-220, 16 Jan 1954 29. Peterson WF, Novak ER: Endometrial polyps. Obstet Gynecol 8:40-49, Jul 1956 30. ProustR, B€lclere C: De i'lnteret de I'examen radiologique apres injection de Lipiodol dans Ie diagnostic des rnetrorraqles. Gynecol Obstet (Paris) 16:408-413, Nov 1927 31. Robins SA, Shapiro AD: Uterotubography. [In] Golden's Diagnostic Roentgenology. Baltimore, Williams & Wilkins, 1956 Vol II, chap 9 32. Rochet Y, Mikaelian S: Valeur de l'hysterographie dans la diagnostic des rnetrorrhaqies post-mencpauslques et plus partlouIierement du cancer de l'endometre, Lyon Chir 66:321-324, SepOct 1970. 33. Ross L, Bensusen CI: Giant polypoid cystic endometrial hyperp/asia. Am J Obstet Gynec 95: 1164-1165. Aug 1966 34. Rozin S: Uterosalpingography in Gynecology. Springfield, III., Thomas, 1965 35. Schroder R: Endometrial hyperplasia in relation to genital function. Am J Obstet Gynecol 68:295-309, Jul 1954 36. Schultze GKF. Erbsl6h J: Gynakologische R6ntgendiagnostik. Stuttgart, Ferdinand Enke Verlag, 1954 37. Siegler AM: Hysterosalpingography. New York, Hoeber Med. Div.• Harper & Row, 1967 38. Word B: Pitfalls of uterine curettage. South Med J 47:3847, Jan 1954 39. Word B, Gravlee LC, Wideman GL: The fallacy of simple uterine curettage. Obstet Gynecol 12:642-648, Dec 1958 40. Zuppinger A: Rontgendiagnostik in der Gynakoloqie. [In] Lehrbuch der R6ntgendiagnostik. Schinz HR, Baensch WE, ed. Stuttgart, Georg Thieme Verlag, 1952
a
a
