- Email: [email protected]
Prognosis of ovarian cancer developing in the residual ovary
GYNECOLOGIC ONCOLOGY 43, 164-166 (19%)
Prognosis of Ovarian Cancer Developing in the Residual Ovary BRUCE Division
of Gynecologic
A. FINE, M.D., ROBERTO YAZIGI, M.D. ,l
AND
RICK RISER, M.S.
Oncology, Department of Obstetrics & Gynecology, and Academic Computing Services, University Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 752359032
of Texas
Received March 8, 1991
Elective oophorectomy at the time of hysterectomy for benign disease in women during their fifth decade is an important issue for both gynecologist and patient. It has been suggested that cancer developing in the residual ovary has a worse prognosis than the national average (L. McGowan, O&et. Gynecol. 69, 386, 1987). In an effort to corroborate such finding, 36 women with epithelial ovarian cancer developing in the residual ovary after prior hysterectomy were compared to a group of 121 patients with epithelial ovarian cancer and no previous surgery. Analysis was made of age, stage at diagnosis, feasibility of cytoreductive surgery, and survival in both groups. Only age distribution was found to be significantly different between the two groups of patients (P < 0.001). Neither FIGG staging or quality of cytoreductive surgery showed a statistically significant difference between both groups. At 3 years, 41% of the subjects with cancer in the residual ovary were alive, compared to 42% in the group without previous hysterectomy. The corresponding figures for Syear survival are 34 and 27%, respectively (P = 0.939). On the basis of our findings we conclude that the overall prognosis for patients with ovarian cancer developing in the residual ovary does not appear to be any worse than that reported for ovarian cancer in general. Q 1991 Aeademk Press, Inc.
cancer developing in the residual ovary is approximately one-half the national survival rate for this disease [l]. If this holds true, it could bolster the argument even further in favor of performing prophylactic oophorectomy. The reasons contributing to the poor outcome of the patients in his study were not analyzed. Factors known to be associated with the outcome in patients with ovarian malignancies are stage at the time of diagnosis, histologic grade of the tumor, age of the patient, size of residual disease after primary surgery, and response to chemotherapy. Of these, histologic grade and response to chemotherapy would not be expected to be different in patients with cancer developing in the residual ovary. The purpose of our study is to analyze those other factors which could hypothetically contribute to a decreased survival in women who have a history of a hysterectomy prior to the diagnosis of ovarian cancer, namely, age, stage at the time of diagnosis, and the feasibility of adequate reductive tumor surgery. MATERIAL
AND METHODS
The records of 36 women with epithelial ovarian cancer developing in the residual ovary after prior hysterectomy, seen at Parkland Memorial Hospital and St. Paul Medical Elective oophorectomy at the time of hysterectomy for Center between 1980 and 1989, were reviewed and combenign disease in women during their fifth decade conpared to a group of 121 patients diagnosed with epithelial tinues to be an issue debated by gynecologists and a conovarian cancer with no previous pelvic surgery. cern frequently voiced by patients. The benefits derived The reasons for prior hysterectomy were obtained for from ovarian preservation, with all of its physiological 25 of the 36 patients. The most common were uterine aspects and personal preferences, must be balanced fibroids (7), uterine prolapse (6), abnormal vaginal bleedagainst the advantages of preventing ovarian cancer and ing (5)) cervical intraepithelial neoplasia (3)) postpartum a variety of other pathologic conditions developing in the hemorrhage (1)) endometriosis (1)) torted adnexa (1)) and residual ovary. McGowan has published a study in which pelvic inflammatory disease (1). In the remaining 11 cases he concluded that the survival of women with ovarian the reason was not recorded. Abdominal hysterectomy without oophorectomy was performed in 22 cases, ab’ To whom reprint requests should be addressed at Division of Gynecologic Oncology, Department of Obstetrics & Gynecology, 5323 dominal hysterectomy and unilateral salpingo-oophorectomy in 6 cases, and vaginal hysterectomy in 8 cases. Harry Hines Blvd., Dallas, TX 752359032. INTRODUCTION
164 0090-8258/91 $1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
CANCER IN THE RESIDUAL
165
OVARY
TABLE 1 Histology
Type
Previous hysterectomy (“ro)
Serous Undifferentiated Mutinous Endometrioid Clear cell Mixed mesodermal Brenner malignant
22 7 2 3 2
Total
36
(61) (19) (5.5) (8) (5.5)
-
No previous hysterectomy (%I 77 16 8 13
(64) (13) (7)
(11)
5 (4)
1 (1) 1 (1)
[““]
,
,
,
,
;,
121 a
The diagnosis of ovarian cancer was established in all casesby exploratory laparotomy and patients were staged according to FIG0 classification. Surgical staging consisted of peritoneal washings, diaphragmatic biopsies, peritoneal biopsies, infracolic omentectomy, and pelvic and aortic lymph node dissections. Maximal reductive tumor surgery was attempted in all patients, and optimal cytoreduction was defined as residual tumor masses less than 2 cm in diameter. All histologic sections from the original diagnostic material were reviewed by the Department of Pathology. The histologic subtypes are shown in Table 1. Comparison was made between the group of 121 patients with no previous hysterectomy and the group of 36 patients with previous hysterectomy and ovarian conservation, in regard to age at the time of diagnosis, initial stage, pathology, cytoreduction, and survival.
0
12
36 24 t.4CNWS FC5T-DlAGlWSiS
48
60
FIG. 1. Survival comparison of patients with ovarian cancer in the previous hysterectomy (...) and no previous hysterectomy (-) groups.
the two groups were compared using the log-rank test. Results of these two-sided statistical tests were considered to be significant when P < 0.05. RESULTS
Of the 121 women without a previous hysterectomy, the age at diagnosis ranged from 21 to 79 years with a median of 53. For the group with previous hysterectomy it ranged from 42 to 82 years, with a median of 58. The age distribution was found to be significantly different (P < 0.001) between both groups. Age at the time of hysterectomy ranged from 26 to 55 years with a median age of 38. The time interval from hysterectomy to the diagnosis of ovarian cancer ranged from 7 to 52 years, with a median of 19. Statistical Analysis Initial FIG0 staging at the time of diagnosis is depicted For statistical analysis, comparison of age at diagnosis in Table 2. We presumed that stage would be the most was made using the Wilcoxon rank sum test and com- likely factor to vary between the two groups of patients; parisons of initial stage and cytoreduction were made however, we found no significant difference in distribution using Fisher’s exact test. Survival was estimated using the across stages (P = 0.095). A simpler comparison of the Kaplan-Meier technique, and the survival estimates of proportion of stage I and II cases in each group was also not significant (P = 0.178). The quality of cytoreductive surgery was assessedand compared in both groups, including all patients with exTABLE 2 Stage at Diagnosis traovarian spread (stages II and above). In the group without previous hysterectomy, 60 of 102 patients (59%) No previous Previous were optimally cytoreduced compared to 17 of 32 (53%) hysterectomy hysterectomy of those with cancer developing in the residual ovary. The Stage (%) (%I difference again is not statistically significant (P = 0.682). 19 (16) I 4 (11) The estimated survival distributions between both 13 (11) II 1 (3) groups (including all stages) were not significantly differ66 (55) 28 (78) III ent. At 3 years, 41% of the subjects with cancer in the 23 (19) IV 3 (8) residual ovary were alive, compared to 42% of the group 121 36 Total without previous hysterectomy. The corresponding figures
166
FINE. YAZIGI,
for 5-year survival are 34% and 27, respectively. Comparison of survival curves showed no significant difference (P = 0.939) (Fig. 1). DISCUSSION Elective oophorectomy at the time of hysterectomy for benign disease in women over age 40 is a frequent and important consideration, requiring analysis by both patient and physician before surgery. The advantages of the procedure include prevention of ovarian neoplasms, alleviation of symptoms attributable to ovarian function, and elimination of the risk of reoperation for adnexal pathology. Disadvantages include the need for hormonal replacement therapy in order to treat vasomotor symptoms and prevent osteoporosis and cardiovascular diseasesand the potential for diminished self-image and psychosexual disfunction which may follow surgical castration
PI.
AND RISSER
come, was significantly different in the groups with and without previous hysterectomy. Age was significantly different in both groups; however, it did not affect the final outcome. Consequently, no significant difference in the 5-year survival rate was found, which was 27% in the no previous hysterectomy group and 34% in the group developing cancer in the residual ovary. Theoretically, the most plausible explanation for a possible prognostic difference between both groups would have been the finding of a higher proportion of advanced stage cases in the prior hysterectomy group. Experience indicates that a residual ovary is frequently firmly adherent to other pelvic organs and peritoneal structures. Moreover, occasionally they may even be found placed in a retroperitoneal location. It would be reasonable then to speculate that stage I cases would only rarely be found in tumors developing in the residual ovary. We were unable however to prove this theory. Previous studies have shown a similar incidence of advanced stage of disease at the time of diagnosis of ovarian cancer (60 to 70% with stage III or IV) regardless of a history of prior hysterectomy [4]. The decreased survival associated with more advanced disease states has been well documented [5]. The decision of whether to perform elective oophorectomy at the time of hysterectomy for benign disease in women over age 40 should continue to be established on an individual basis, taking into consideration age, menstrual status, individual risk factors, family history, contraindications to replacement hormonal therapy, and patient’s preference. However, on the basis of our findings, we conclude that the overall prognosis for patients with ovarian cancer developing in the residual ovary does not appear to be any worse than that reported for ovarian cancer patients in general.
The prevention of ovarian cancer constitutes the leading argument favoring elective oophorectomy. In light of the fact that survival of patients with ovarian cancer has not changed significantly over the past decades [3], despite new chemotherapeutic agents, the focus in patient management needs to be prevention. Although it has been virtually impossible to establish statistically any meaningful bearing of prophylactic oophorectomy on the incidence of ovarian cancer, for the individual patient it remains a convincing means to prevent its development. The finding reported by McGowan that patients with carcinoma of the ovary developing in the residual ovary had a 21% survival rate, quoted as being one-half of the national survival rate, is quite worrisome. This information was based on survival data obtained for 38 patients collected from 29 area hospitals, and although details REFERENCES regarding stage distribution are lacking, he reports that 51% of the patients did not have proper evaluation of I. McGowan, L. Ovarian cancer after hysterectomy, Okret. Gynecol. 69, 3866-3868 (1987). the extent of the disease. We found it necessary however 2. American College of Obstetricians and Gynecologists. Prophylactic to re-explore the issue by rationalizing the known progoophorectomy, ACOG Technical Bulletin III, 1987. nostic factors that could be responsible for this poor 3. Barber, H.R.K. Ovarian cancer, Cancer 36, 149-184 (1986). outcome. 4. Gibbs, E.K. Suggested prophylaxis for ovarian cancer, Am. J. Obstet. In the present study we found that neither stage at Gynecol. 111, 756-765 (1971). diagnosis nor adequacy of cytoreductive surgery, which 5. Boring, C. C., Squires, T. S., and Tong, T. Cancer statistics 1991, Cancer 41, 19-51 (1991). are the most influential prognostic factors related to out-
Prognosis of Ovarian Cancer Developing in the Residual Ovary BRUCE Division
of Gynecologic
A. FINE, M.D., ROBERTO YAZIGI, M.D. ,l
AND
RICK RISER, M.S.
Oncology, Department of Obstetrics & Gynecology, and Academic Computing Services, University Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 752359032
of Texas
Received March 8, 1991
Elective oophorectomy at the time of hysterectomy for benign disease in women during their fifth decade is an important issue for both gynecologist and patient. It has been suggested that cancer developing in the residual ovary has a worse prognosis than the national average (L. McGowan, O&et. Gynecol. 69, 386, 1987). In an effort to corroborate such finding, 36 women with epithelial ovarian cancer developing in the residual ovary after prior hysterectomy were compared to a group of 121 patients with epithelial ovarian cancer and no previous surgery. Analysis was made of age, stage at diagnosis, feasibility of cytoreductive surgery, and survival in both groups. Only age distribution was found to be significantly different between the two groups of patients (P < 0.001). Neither FIGG staging or quality of cytoreductive surgery showed a statistically significant difference between both groups. At 3 years, 41% of the subjects with cancer in the residual ovary were alive, compared to 42% in the group without previous hysterectomy. The corresponding figures for Syear survival are 34 and 27%, respectively (P = 0.939). On the basis of our findings we conclude that the overall prognosis for patients with ovarian cancer developing in the residual ovary does not appear to be any worse than that reported for ovarian cancer in general. Q 1991 Aeademk Press, Inc.
cancer developing in the residual ovary is approximately one-half the national survival rate for this disease [l]. If this holds true, it could bolster the argument even further in favor of performing prophylactic oophorectomy. The reasons contributing to the poor outcome of the patients in his study were not analyzed. Factors known to be associated with the outcome in patients with ovarian malignancies are stage at the time of diagnosis, histologic grade of the tumor, age of the patient, size of residual disease after primary surgery, and response to chemotherapy. Of these, histologic grade and response to chemotherapy would not be expected to be different in patients with cancer developing in the residual ovary. The purpose of our study is to analyze those other factors which could hypothetically contribute to a decreased survival in women who have a history of a hysterectomy prior to the diagnosis of ovarian cancer, namely, age, stage at the time of diagnosis, and the feasibility of adequate reductive tumor surgery. MATERIAL
AND METHODS
The records of 36 women with epithelial ovarian cancer developing in the residual ovary after prior hysterectomy, seen at Parkland Memorial Hospital and St. Paul Medical Elective oophorectomy at the time of hysterectomy for Center between 1980 and 1989, were reviewed and combenign disease in women during their fifth decade conpared to a group of 121 patients diagnosed with epithelial tinues to be an issue debated by gynecologists and a conovarian cancer with no previous pelvic surgery. cern frequently voiced by patients. The benefits derived The reasons for prior hysterectomy were obtained for from ovarian preservation, with all of its physiological 25 of the 36 patients. The most common were uterine aspects and personal preferences, must be balanced fibroids (7), uterine prolapse (6), abnormal vaginal bleedagainst the advantages of preventing ovarian cancer and ing (5)) cervical intraepithelial neoplasia (3)) postpartum a variety of other pathologic conditions developing in the hemorrhage (1)) endometriosis (1)) torted adnexa (1)) and residual ovary. McGowan has published a study in which pelvic inflammatory disease (1). In the remaining 11 cases he concluded that the survival of women with ovarian the reason was not recorded. Abdominal hysterectomy without oophorectomy was performed in 22 cases, ab’ To whom reprint requests should be addressed at Division of Gynecologic Oncology, Department of Obstetrics & Gynecology, 5323 dominal hysterectomy and unilateral salpingo-oophorectomy in 6 cases, and vaginal hysterectomy in 8 cases. Harry Hines Blvd., Dallas, TX 752359032. INTRODUCTION
164 0090-8258/91 $1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
CANCER IN THE RESIDUAL
165
OVARY
TABLE 1 Histology
Type
Previous hysterectomy (“ro)
Serous Undifferentiated Mutinous Endometrioid Clear cell Mixed mesodermal Brenner malignant
22 7 2 3 2
Total
36
(61) (19) (5.5) (8) (5.5)
-
No previous hysterectomy (%I 77 16 8 13
(64) (13) (7)
(11)
5 (4)
1 (1) 1 (1)
[““]
,
,
,
,
;,
121 a
The diagnosis of ovarian cancer was established in all casesby exploratory laparotomy and patients were staged according to FIG0 classification. Surgical staging consisted of peritoneal washings, diaphragmatic biopsies, peritoneal biopsies, infracolic omentectomy, and pelvic and aortic lymph node dissections. Maximal reductive tumor surgery was attempted in all patients, and optimal cytoreduction was defined as residual tumor masses less than 2 cm in diameter. All histologic sections from the original diagnostic material were reviewed by the Department of Pathology. The histologic subtypes are shown in Table 1. Comparison was made between the group of 121 patients with no previous hysterectomy and the group of 36 patients with previous hysterectomy and ovarian conservation, in regard to age at the time of diagnosis, initial stage, pathology, cytoreduction, and survival.
0
12
36 24 t.4CNWS FC5T-DlAGlWSiS
48
60
FIG. 1. Survival comparison of patients with ovarian cancer in the previous hysterectomy (...) and no previous hysterectomy (-) groups.
the two groups were compared using the log-rank test. Results of these two-sided statistical tests were considered to be significant when P < 0.05. RESULTS
Of the 121 women without a previous hysterectomy, the age at diagnosis ranged from 21 to 79 years with a median of 53. For the group with previous hysterectomy it ranged from 42 to 82 years, with a median of 58. The age distribution was found to be significantly different (P < 0.001) between both groups. Age at the time of hysterectomy ranged from 26 to 55 years with a median age of 38. The time interval from hysterectomy to the diagnosis of ovarian cancer ranged from 7 to 52 years, with a median of 19. Statistical Analysis Initial FIG0 staging at the time of diagnosis is depicted For statistical analysis, comparison of age at diagnosis in Table 2. We presumed that stage would be the most was made using the Wilcoxon rank sum test and com- likely factor to vary between the two groups of patients; parisons of initial stage and cytoreduction were made however, we found no significant difference in distribution using Fisher’s exact test. Survival was estimated using the across stages (P = 0.095). A simpler comparison of the Kaplan-Meier technique, and the survival estimates of proportion of stage I and II cases in each group was also not significant (P = 0.178). The quality of cytoreductive surgery was assessedand compared in both groups, including all patients with exTABLE 2 Stage at Diagnosis traovarian spread (stages II and above). In the group without previous hysterectomy, 60 of 102 patients (59%) No previous Previous were optimally cytoreduced compared to 17 of 32 (53%) hysterectomy hysterectomy of those with cancer developing in the residual ovary. The Stage (%) (%I difference again is not statistically significant (P = 0.682). 19 (16) I 4 (11) The estimated survival distributions between both 13 (11) II 1 (3) groups (including all stages) were not significantly differ66 (55) 28 (78) III ent. At 3 years, 41% of the subjects with cancer in the 23 (19) IV 3 (8) residual ovary were alive, compared to 42% of the group 121 36 Total without previous hysterectomy. The corresponding figures
166
FINE. YAZIGI,
for 5-year survival are 34% and 27, respectively. Comparison of survival curves showed no significant difference (P = 0.939) (Fig. 1). DISCUSSION Elective oophorectomy at the time of hysterectomy for benign disease in women over age 40 is a frequent and important consideration, requiring analysis by both patient and physician before surgery. The advantages of the procedure include prevention of ovarian neoplasms, alleviation of symptoms attributable to ovarian function, and elimination of the risk of reoperation for adnexal pathology. Disadvantages include the need for hormonal replacement therapy in order to treat vasomotor symptoms and prevent osteoporosis and cardiovascular diseasesand the potential for diminished self-image and psychosexual disfunction which may follow surgical castration
PI.
AND RISSER
come, was significantly different in the groups with and without previous hysterectomy. Age was significantly different in both groups; however, it did not affect the final outcome. Consequently, no significant difference in the 5-year survival rate was found, which was 27% in the no previous hysterectomy group and 34% in the group developing cancer in the residual ovary. Theoretically, the most plausible explanation for a possible prognostic difference between both groups would have been the finding of a higher proportion of advanced stage cases in the prior hysterectomy group. Experience indicates that a residual ovary is frequently firmly adherent to other pelvic organs and peritoneal structures. Moreover, occasionally they may even be found placed in a retroperitoneal location. It would be reasonable then to speculate that stage I cases would only rarely be found in tumors developing in the residual ovary. We were unable however to prove this theory. Previous studies have shown a similar incidence of advanced stage of disease at the time of diagnosis of ovarian cancer (60 to 70% with stage III or IV) regardless of a history of prior hysterectomy [4]. The decreased survival associated with more advanced disease states has been well documented [5]. The decision of whether to perform elective oophorectomy at the time of hysterectomy for benign disease in women over age 40 should continue to be established on an individual basis, taking into consideration age, menstrual status, individual risk factors, family history, contraindications to replacement hormonal therapy, and patient’s preference. However, on the basis of our findings, we conclude that the overall prognosis for patients with ovarian cancer developing in the residual ovary does not appear to be any worse than that reported for ovarian cancer patients in general.
The prevention of ovarian cancer constitutes the leading argument favoring elective oophorectomy. In light of the fact that survival of patients with ovarian cancer has not changed significantly over the past decades [3], despite new chemotherapeutic agents, the focus in patient management needs to be prevention. Although it has been virtually impossible to establish statistically any meaningful bearing of prophylactic oophorectomy on the incidence of ovarian cancer, for the individual patient it remains a convincing means to prevent its development. The finding reported by McGowan that patients with carcinoma of the ovary developing in the residual ovary had a 21% survival rate, quoted as being one-half of the national survival rate, is quite worrisome. This information was based on survival data obtained for 38 patients collected from 29 area hospitals, and although details REFERENCES regarding stage distribution are lacking, he reports that 51% of the patients did not have proper evaluation of I. McGowan, L. Ovarian cancer after hysterectomy, Okret. Gynecol. 69, 3866-3868 (1987). the extent of the disease. We found it necessary however 2. American College of Obstetricians and Gynecologists. Prophylactic to re-explore the issue by rationalizing the known progoophorectomy, ACOG Technical Bulletin III, 1987. nostic factors that could be responsible for this poor 3. Barber, H.R.K. Ovarian cancer, Cancer 36, 149-184 (1986). outcome. 4. Gibbs, E.K. Suggested prophylaxis for ovarian cancer, Am. J. Obstet. In the present study we found that neither stage at Gynecol. 111, 756-765 (1971). diagnosis nor adequacy of cytoreductive surgery, which 5. Boring, C. C., Squires, T. S., and Tong, T. Cancer statistics 1991, Cancer 41, 19-51 (1991). are the most influential prognostic factors related to out-