Endometrial carcinoma in patients with bilateral oophorectomy or irradiation castration F.
J.
HOFMEISTER,
B.
F.
VONDRAK,
Milwaukee,
M.D.,
F.A.C.S.,
F.A.C.O.G.
M.D.
Wisconsin
Seventeen instances of endometrial carcinoma when ovarian function has been eliminated by surgery or irradiation are #resented as evidence of the hazards of the retained uterus. This increased hazard is due to complacency and inadequate methods of systematic, routine evaluation. The importance of dilatation and curettage, the value of the cndometrial biopsy, and the necessity for a more aggressive attitude toward detecting endometrial lesions is stressed as being essential to a complete gynecologic examination. When bilateral oophorectomy is performed, total hysterectomy should always be fierformed.
of Gynecology and Surgery of the Lutheran Hospital of Milwaukee, Inc. (formerly Milwaukee Hospital), one will quickly realize that the mode of therapy selected for these patients has varied with the concepts prevalent during the respective time periods. The authors’ scheme of immediate patient consideration will be outlined at the end of this paper. Controversy regarding the incidence of endometrial malignancy in previously castrated women still persists.
P R o P E R M E D I c A t responsibility dictates that the physician suspicously examine and cautiously treat the entire patient despite the chief complaint, the present illness, and an often misleading past medical and surgical history. False .assumptions have led and will lead to missed diagnoses, inappropriate management, and fatal consequences. The patient who retains her uterus after overian function has been destroyed by surgery or irradiation presents to her physician the challenge of exacting observation, diagnosis, and accurate care. To emphasize the axiom of mandatory, thorough investigation, a series of patients demonstrating uterine disease reflecting pitfalls hardly conceivable in this day of advanced computerized technology is presented, It will become quite obvious that this presentation is handicapped in many respects by a lack of laboratory facts not requested in past years but presently deemed essential before therapeutic management is instituted. Likewise, since this is a review of past medical records of the Departments From the Department Gynecology, Lutheran Milwaukee.
of Obstetrics Hosjrital of
Presented before the Chicago Gynecological Society on Ian.
Endometrial carcinoma after bilateral oophorectomy Research into the medical service records of Lutheran Hospital of Milwaukee, Inc., revealed 9 cases of uterine carcinoma following castration (Table I) . Eight were confirmed at the time of hysterectomy for the management of the malignancy. The ninth was diagnosed malignant mixed mesodermal tumor of the uterus by dilatation and curettage, and was treated by radium insertion and external cobalt irradiation. Thus, in the ninth case, although the patient claimed to have had postmenopausal symptoms after a reported bilateral oophorectomy, a laparotomy was not performed to confirm the absence of ovarian tissue. Three patients in the series were treated
and
17, 1969.
1099
1100
Hofmeister
and Vondrak Amer.
Table I. Nine patients carcinoma
following Surgical
Patient L. Y.*
I,. S.”
E. S.
with uterine castration
and pathologic history
Left salpingo-oophorectomy for salpingooophoritis Right salpingo-oophorectomy for salpingooophoritis Right onphorectomy dermoid Left oophorectomy dermoid
22
1937
44
1959
23
1934
26
1937
18
1934
for for
Bilateral oophorectomy endometriosis
for
A. W.
Bilateral cysts
for 29
1930
R. L.*
Bilateral oophorectomy for dysfunctional bleeding
42
1937
E. B.
Bilateral oophorectomy unknown lesions
37
1940
Subtotal hysterectomy and a bilateral salpingooophorectomy for unknown lesions
50
1944
Bilateral oophorectomy for cysts (serous cystadenocarcinoma of ovary)
53
1943
58
1936
44
1919
B. F.
G. L.
E. B.
“Premarin,
oophorectomy
for
Laparotomy and bilateral salpingo-oophorectomy Radium insertion and x-ray 0.625
or 1.25 m~. daily.
by surgical castration elsewhere and the histologic diagnoses of these ovaries were not obtainable. By history, all of the patients demonstrated amenorrhea following their initial pelvic procedure for a benign condition Five described sensations of estrogen loss. Of the 9 patients, 2 were known to be using conjugated estrogens-equine* daily during a period of 2 years prior to the diagnosis of malignancy, One had been taking this medication for 5 years. The dosage varied from 0.625 to 1.25 mg. once daily. Data concerning the 9 patients are presented in Table II. Two patients, L. Y. and L. S., had an ovary removed as a single procedure at intervals of 22 years and 3 *Premarin,
Ayerst
Laboratories,
New
York,
New
York.
August J. Obstet.
1, 1970 Gynec.
years, respectively. B. F. had a fundectomy, designated as a subtotal hysterectomy, along with bilateral salpingo-oophorectomy. Of the 9 patients, 6 were symptomatic with a bloody discharge present during periods of as short as 2 months (L. S.) to as long as 2 years (G. L.) , In only 4 of these patients was a dilatation and currettage or endometrial sampling done. Three were asymptomatic. In one (R. L.), the malignancy was discovered when the patient was treated for prolapse and stress incontinence by vaginal hysterectomy and vaginal reconstruction The preliminary dilatation and curettage was positive. In the second (E. B.) an enlarged uterus treated for fibroids led to the discovery. This was found coincidental to surgery in the removed specimen. In the third (L. Y.) , the presence of an enlarging uterus with a stenosed cervix prompted her gynecologist to initiate a dilatation and curettage. The malignancy was discovered and treatment initiated. The age when malignancy was discovered varied from 47 to 69 years. The average age of discovery of the uterine malignancy was 56.4 years in our series. A time lapse of 5 to 33 years after castration was noted, with an average time interval of 16.6 years. Eliminating the patient operated on for ovarian carcinoma from our statistical analysis, one would arrive at 55.7 years as the average age of diagnosis and a time interval of 14.7 years for the adjusted series. These values are of significance after considering Hertig’s and Sommersy4 conclusions, and Cianfrani’sl series with an average age of 54.3 years and a time lapse of 10.6 years. Henriksen and MurriettaZ~ 3 documented 2 such cases with a time lapse of 21 and 27 years after castration, without an accompanying presence of endometrial hyperplasia. In 195 1, Randall, Mirick, and Wieben” analyzed a series of patients with fundal malignancy and found that, prior to the diagnosis of the endometrial carcinoma, 20 of their patients had been sterilized artificially and 4 by surgical castration. They failed to mention the time span between the initial and later confirming surgical procedure, and they did not indicate the type of
Volume Number
107 7
Endometrial
Table II. Data concerning
9 patients
I Patient
with
uterine
carcinoma
I Present
symptoms
Diagnosis
L. Y.
Low-grade adenocarcinoma of endometrium; Pap. neg.; no Novak; D & C pos.
L.
Bloody
In
Is.
discharge
2 mo.
situ adenocarcinoma; Pap. atyp.; Novak
following
Age
Year
50
1965
6
Total abdominal terectomy
hys-
54
1965
28
Total abdominal terectomy
hys-
47
1963
29
Radium, cobalt, total abdominal hysterectomy
I
1 Therapy
I
E. S.
Vaginal
bleeding
10 mo.
Adenocarcinoma; Pap.; no Novak
A. W.
Vaginal
bleeding
3 mo.
Malignant mixed mesodermal tumor; Pap. atyp.; D & C pos.
62
1963
33
Radium,
cobalt
R. L.
Uterine der
prolapse, stress
blad-
Invasive adenocarcinoma; Pap. neg.; D & C pos.
62
1957
20
Vaginal
hysterectomy
E. B.
Enlarged
Adenocarcinoma; Pap.; no Novak
50
1953
13
Total abdominal terectomy
B. F.
Bloody mo.
discharge
G. L.
Vaginal
E. B.
Vaginal
uterus
no
1101
castration
pas.
no
after castration
1 Time lapse I (yr.)
I
Cervical stenosis (unable to probe), uterine enlargement
carcinoma
hys-
3 to 4
Adenocarcinoma no Pap.; no Novak
55
1959
5
Removal of lower uterine segment and cervix
bleeding
2 yr.
Adenocarcinoma; Pap.; no Novak
no
59
1949
6
Subtotal abdominal terectomy
hys-
bleeding
8 mo.
Invasive adenocarcinoma; no Pap.; no Novak
69
1947
11
Subtotal abdominal terectomy
hys-
ovarian lesions present in the ovariectomized women. Cianfranil cited 8 instances of postcastrated uterine carcinoma in 1955, and quoted an additional 3 patients apparently reported by G. Van Smith. These patients developed lesions 15 years after previous bilateral oophorectomy. In Cianfrani’s series, 2 of the patients had had total oophorectomy for diagnosed ovarian adenocarcinoma 5 and 6 years before the discovery of the endometrial lesion. Only one of the latter patients had an endometrial sampling before the definitive ovarian procedure. Referring to the total series, he arrived at an average age of discovery of carcinoma at 54.3 years, and stated that the average time lapse between oophorectomy and the diagnosis of uterine malignancy was 10.6 years. He included a concise extraction of current literature regarding the ovarian stimulus as an initiating, persistent, or incidental etiologic factor in the inception of endometrial
adenocarcinoma and concluded that one could not explain the presence of uterine malignancy on the basis of altered ovarian activity. Endometrial irradiation
carcinoma castration
after
Coincidentally, 8 instances of endometrial malignancy following pelvic irradiation were recorded (Table III). All of the patients gave a history of amenorrhea after the initial procedure, but only 3 experienced vasomotor sensations. Two patients had had a preirradiation dilatation and curettage with negative endometrial samplings. The average age of diagnosis of uterine malignancy in this series was 56.5 years and the time lapse was 16.1 years. This almost parallels the statistics of the operative group. Of these, 2 were asymptomatic with negative Papanicolaou smears and were detected by endometrial biopsy. Symptoms were present over a period
1102
Hofmeister
and
Vondrak
Table III. Adenocarcinoma Age (yr.)
Previous
76
55
Amer.
following
pelvic
therapy
Year
Present
Radium insertion for menorrhagia
1930
Asymptomatic
X-ray for rhagia*
1938
menor-
August J. Obstet.
1, 1970 Gynec.
irradiation
symptoms
Asymptomatic
Pathologic
findings
Adenocarcinoma; Novak and neg.
Pap.
Adenocarcinoma; Novak and neg.
Pap.
Year
Time lapse (yr.)
1964
34
Vaginal hysterectomy; bilateral salpingooophorectomy
1964
26
Vaginal hysterectomy; bilateral salpingooophorectomy
Therapy
X-ray (6 times) for premenstrual pain
1944 to 1946
Vaginal
discharge,
Adenocarrinoma
1952
6
Total abdominal hysterectomy; bilateral salpingo-oophorectomy
58
X-ray, purpose unknown
1942
Vaginal discharge, 2 mo.
Adenocarcinoma
1951
9
Total abdominal hysterectomy; bilateral salpingo-oophorectomy
52
Radium insertion for menorrhagia*
1934
Vaginal bleeding, 2 episodes
Adenocarcinoma, invasive
1949
15
Total abdominal hysterectomy; bilateral saipingo-oophorectomy
58
Radium insertion, purpose unknown
1940
Vaginal 1 yr.
spotting,
Adenocarcinoma
1948
8
Total abdominal hysterectomy; bilateral salpingo-oophorectomy
74
Radium insertion for menorrhagia
1922
Vaginal 1 wk.
discharge,
.4denocarcinoma
1947
25
38
X-ray (3
1939
Vaginal
bleeding
Adenocarcinoma
1945
6
41
‘Dilatation
to ovaries times)
and
curettages--negative
1 yr.
Radium
insertion
Total abdominal hysterectomy; bilateral salpingo-oophorectomy
tissue.
which varied from one week to one year. Two patients were asymptomatic and the adenocarcinoma was detected by endometrial biopsy. Papanicolaou smears were negative. Comment
The authors were stimulated to search the hospital records for these unique instances of endometrial carcinoma, after being aware of 6 of these instances which occurred within the past 4 years. If these 17 instances of endometrial carcinoma were presented to emphasize the fact that they occurred during a 10 year period when there were 26,168 gynecologic admissions, while 533 dilatations and curettages were performed to diagnose 393 endometrial carcinomas, the effort would be without value.
No attempt has been made to abolish the ovary and its functional activity as an inciting factor in the etiology of malignant uterine disease. Rather, one cannot eliminate the possibility of further pelvic disease such as endometrial adenocarcinoma simply on the basis of a previous total oophorectomy. Admittedly, it is rare to uncover such a patient; however, one need only scan the surgical records of any major hospital where, for example, breast carcinoma is managed, to find instances of bilateral salpingo-oophorectomy without hysterectomy to be convinced that the potential exists. This hazard also exists where treatment for extensive endometriosis is accomplished by bilateral oophorectomy. An additional hazard is the generous sized, retained cervical stump, in effect, a fundectomy when so-called techni-
Volume Number
107 7
caI difficulties make total abdominal hysterectomy impossible. It is interesting to note that when analyzing the additional surgical procedures performed as part of the management of breast carcinoma following radical mastectomiesat Lutheran Hospital of Milwaukee over a 10 year period ending in 1965, one discovers that 42 premenopausal patients were treated. Of these 42, 27 were castrated without regard for the uterus, and 23 of the 27 were managed on the general surgical service. The remainder were treated by gynecologists who did not in all instances remove the uterus when bilateral oophorectomy was performed. Admittedly, as of this time, there was not an instance of uterine malignancy detected in the follow-up of this series.An interesting fact in this group of 27 patients, where the uterus was permitted to remain with disregard for future malignancy, is that a prophylactic appendectomy was performed in 18 of the patients. Certainly, the technicians were displaying some degree of optimism and, in our opinion, an error of direction. The existence of these unique instances directs attention to valuable facts. They are distressing evidence of the failure, even today, to completely evaluate patients. They emphasize the necessity for exacting, more specific, and more aggressivemethods of approach for the detection of endometrial carcinoma. Finally, they call attention to the necessity of performing exacting and adequate surgical therapy when indicated. In a survey? conducted in 1964, one of the authors determined that, in spite of persistent efforts of medical educators, not all gynecologists did routine Papanicolaou smears,Efforts5 to impressail physicians with the value of this technical aid when doing a complete gynecologic examination had apparently fallen on some deaf ears. This same condition existed in 1969. In studies87 g associatedwith the ‘Papanicolaou smear and endometrial biopsy, the author determined that the Papanicolaou smear was only of assistance in detecting endometriai carcinoma in 20 per cent of the patients treated.
Endometrial
carcinoma
after castratlon
1103
More important, however, it was of value only in those patients who were symptomatic and in whom pathology was suspected.This fact was in evidence also in the statistics associatedwith the 17 patients evaluated in this present study. Dilatation and curettage was not performed preliminary to definitive surgery on most of these present patients. This, surprisingly enough, confirms the findings? which indicated that not all gynecologists, surgeons, and general practitioners did dilatation and curettage or suggested it when abnormal or postmenopausal bleeding existed. Endometrial biopsy, which in the author’s seriesof 14,655’ resulted in a detection rate of 0.7 per cent, a rate comparable to the detection rate of the Papanicolaou smear in cervical carcinoma, was rarely used. Of the malignancies in this present ‘series, 5 were asymptomatic. Two of these were detected by preoperative dilatation and curettages. Three were found in the specimens after surgery. In a serie& 7 of 102 patients with endometrial carcinomas treated by the author, over 20 per cent of the endometrial malignancies were asymptomatic and unsuspected when detected by routine endometrial biopsy or immediate preoperative dilatation and currettage. In a subsequentpresentation9 reporting an additional 3,011, making an over-all total of 17,666 endometrial biopsies, another 22 patients with endometrial carcinoma were detected. Thus, during an 18 year period, 124 women with adenocarcinomas of the endometrium were detected, 21 per cent asymptomatic and unsuspected. Only 3 of the 17 patients had a history of exogenous estrogen administration; yet, it is absolutely necessary, as evidenced by the number of patients who bleed abnormally when estrogen is administered, that expert, accurate investigation be insisted upon. Repeated evaluation is necessary. The size of the uterus must be determined. The contents of the uterus must be examined and the patency of the cervix must be assured. It, therefore, becomes essential that when
1104
Hofmeister
and
August
Vondrak
Ame,.
abnormal bleeding exists, with or without ovaries present, endometrial biopsy or dilatation and currettage is mandatory. It is important to note that the endometrial biopsy is only of value if positive. If negative, hospital dilatation and curettage must be done. From the patients presented, it is evident that lack of ovarian function does not assure security from endometrial malignancy. This was recognized and reported in 1941, by Jones and Brewer.*O Ultraconservatism or excessive technical complications still result in the excuse for permitting the uterus to remain when bilateral oophorectomy is performed. A recent instance of this has just occurred in one of the large teaching institutions. Extensive endometriosis or pelvic inflammatory conditions might fit these excuses. The teen-ager, the victim of massive ovarian dermoids, could conceivably present a problem of decision. Should the uterus be retained, if, in this rare instance, both ovaries are removed? Is the psychological support offered by cyclic bleeding due to cyclic hormonal therapy necessary for the normal physical and psychologica1 development of this individual? Should this useless uterus remain as a future hazard to the patient? It is our opinion that the uterus should be removed; the confusion of abnormal bleeding and the hazard of future
J. Ohstet.
1, 1970 Gynec.
malignancy eliminated. The patient should then be supported by hormonal therapy and conscientious psychological counseling by her gynecologist. The ultimate impact of the retained uterus can include problems associated with the cervix. Cianfrani,l in his review of 130 cases in which the uterus was retained where bilateral oophorectomy was performed, discovered 19 patients with carcinoma, an incidence of 14.5 per cent. Eleven were cervica1 carcinomas and 8 were endometrial carcinomas. This represents the general incidence of cervical carcinoma and endometrial carcinoma in female patients. Thus, this fact, admittedly reduced by yearly routine Papanicolaou smears, compounds the hazard of the retained uterus. Finally, the complacency of ignorance engendered in the patient by the false security of oophorectomy and the unjustified complacency on the part of the physicians as evidenced by the author’s survey, adds greater problems and can result in inoperability. When conservative surgery is done and ovarian tissue remains, preliminary diIatation and curettage should always be performed. When bilateral oophorectomy is performed, there is never justification for retaining the uterus.
REFERENCES
1. 2. 3. 4. 5.
6.
Cianfrani, R.: AMER. J. OBSTET. GYNEC. 69: 64, 1955. Henriksen, E., and Murrietta, T.: West. J. Surg. 58: 331, 1950. Henriksen, E.: Obstet. Gynec. 15: 663, 1960. Hertig, A. T., and Sommers, S. C.: Cancer 2: 946, 1949. Hofmeister, F. J., Savage, G. W., and Wolfe, C. W.: AMER. J. OBSTET. GYNEC. 77: 1245, 1959. Hofmeister, F. J., and Barbo, D. M.: Obstet. Gynec. 23: 386, 1964.
Discussion DR. review it may
A. F. LASH, Chicago, Illinois. From of the literature and personal experience be stated that 17 patients with adenocarci-
a
7.
8.
9.
10. 11.
Hofmeister, F. J., and Barbo, D. M.: Wisconsin Med. J. 63: 358, 1964. Hofmeister, F. J., Vondrak, B. F., and Barbo, D. M.: AMER. J. OBSTET. GYNEC. 42: 91, 1966. Hofmeister, F. J.: Proceedings of Fifth World Congress of Gynaecology and Obstetrics, 1967, p. 683. Jones, H. O., and Brewer, J. L.: AMER. J. OBSTET. GYNEC. 42: 207, 1941. Randall, J. H., Mirick, D. F., and Wieben, E. E.: AMER. J. OBSTET. GYNEC. 61: 596, 1951.
noma of the endometrium after ablation of the ovaries without hysterectomy and after irradiation castration is a large series. It might be of interest to know over what period of time these
Volume 107 Number 7
cases occurred and how many endometrial adenocarcinomas were present during these years. Cianfrani,l in an analysis of 130 patients after ovarian ablation without hysterectomy, found 8 instances of endometrial adenocarcinoma and 11 instances of other uterine cancers. Norris and Taylor5 present clinical and pathologic findings in 17 patients with sarcoma of the uterus and a history of pelvic irradiation. The interval between the pelvic irradiation and development of the sarcoma varied from 7 to 40 years. The controversial question of the etiologic role of estrogens in adenocarcinoma of the endometrium has not been settled. Katayama and Jones3 believe that there arc chromosomal aberrations in endometrial carcinoma of a marked degree. Forni and Miles? studied chromosomal aberrations in atypical cervical smears. Kirkland’ analyzed the chromosomes in dysplasia, in carcinoma in situ, and in invasive carcinoma, and found 3 distinct patterns. The degree of abnormality of the patterns may be interpreted as an index of malignancy. Wynder, Escher, and Mantel6 reported an interview study of 11’2 endometrial cancer patients and 200 comparison patients. They deduced from the high incidence of obese, tall women with and without the previously described metabolic and menstrual disturbances that there is a possibility of a pituitary hormone dysfunction, most likely of the growth hormone. From the evidence presented, one may deduce that irradiation per se may be considered car-
Endometriai
carcinoma
after
castration
1105
cinogenic with its proved effect on the chromosomes whether carcinoma or sarcoma follows. In those women who have not been irradiated, the omnipresence of viruses and their carcinogenic effect on the chromosomes of cells may well be the more likely primary etiologic factor, rather than hormones. During pregnancy, when the woman is subjected to the greatest amount of hormonal influences, the rate of growth of cervical carcinoma has not been shown to be greater than during the nonpregnant state. The secondary role of hormones is uncertain, although the progestins have some influence on endometrial adenocarcinoma and its pulmonary metastases in about 25 per cent of the patients, while estrogens inlluence prostatic carcinoma growth. Therefore, we agree with the authors in their admonitions to the clinicians in their conclusions about adequate diagnostic procedures and complete surgery. (IMy under extenuating circumstances should irradiation be used instead Of surgery for benign uterine bleeding. REFERENCES 1.
2. 3. 4. 5. 6.
Cianfrani, T.: AMER. J. OBSTET. GYNEC. 69: 64, 1955. Forni, A., and Mi!es, C. P.: Acta Cytol. 10: 200, 1966. Katayama, K. P., and Jones, H. W.: AMER. J. OBSTET.GYNEC.Q~:~~~, 1967. Kirkland, J. A.: Acta Cytol. 10: 80, 1966. Norris, H. J., and Taylor, H. B.: Obstet. Gynec. 26: 689, 1965. Wynder, E. L., Escher, G. C., and Mantel, N.: Cancer 19: 489, 1966.