Incidence of ovarian metastasis from colorectal cancer (CRC) varies from 4% to 30.8% in various clinical studies.1
Between 6% and 14% of women dying with CRC are found at autopsy to have ovarian metastases.2 Less than 5% of the patients with CRC are reported to develop metachronous ovarian metastases within 2 years from primary resection.3
In case of direct invasion of the contiguous ovary from the CRC (pT4) or isolated macroscopic metastases, oophorectomy can be performed with a curative aim and the radical resection of ovaries was described to improve the overall survival.3,4
The objectives of removing macroscopically normal ovaries in a woman with primary CRC are to abolish the risk of primary ovarian cancer, to improve CRC cure rate by removing microscopic synchronous metastases, and to prevent development of metachronous ovarian metastases.
Routine oophorectomy in CRC had been suggested by Schenk and Sitzenfrey5 in 1907; however, there is no prospective randomized trial with high statistical power on the role of prophylactic oophorectomy during surgery for CRC; consequently, there is no consensus about oophorectomy during surgery for CRC.
The aims of this study were to find out the true incidence of synchronous and metachronous ovarian metastases in our patients with CRC and to define the clinicopathologic feature of ovarian involvement to clear the role of prophylactic oophorectomy.
The data from 180 women who were operated on for CRC since 1990 to 2004 in the Cancer Institute of Tehran University of Medical Science were retrospectively retrieved and analyzed.
All women who underwent radical surgery were included, regardless of the type of operation, and patients who were considered inoperable and had not undergone colectomy were excluded.
Patients who presented with direct CRC invasion of the ovary and who had a history of previous oophorectomy were also excluded.
The Cancer Research Center of the Cancer Institute of Iran approved this project on March 2005.
Clinical data were retrieved from medical records, and clinical features were recorded including age, menstruation status, presenting symptoms, preoperative ultrasound and computed tomographic findings, preoperative carcinoembryonic antigen levels, primary tumor location, tumor differentiation, additional site of tumor involvement, and ovary morphology including size and surface abnormality.
The reason for oophorectomy were retrieved from the surgeon's note. It was classified as prophylactic, ovarian abnormality, and undetermined.
Data of at least 3 years of follow-up were retrieved from our outpatient clinics' records.
The features of cases with ovarian involvement (synchronous and metachronous) were compared with those of cases with normal ovaries in the pathologic study.
Statistical analysis was done using the Statistical Package for Social Sciences (SPSS version 11.5). The study of the relationships between variables was performed using χ 2 test, and a significance level of 0.05 was applied.
Of the 180 women with CRC treated in our center (mean age, 47.5 years; range, 17-86 years), 60 women underwent oophorectomy along with large bowel resection according to the surgeon's preference. The reasons for oophorectomy are the following: prophylactic (n = 22, 36.6%), abnormal morphology (n = 35, 58.3%), and undetermined (n = 3, 5%).
No complication directly related to the oophorectomy was reported. Five patients had histologically proven ovarian involvement. Thus, the incidence of synchronous ovarian metastases was 2.7% (5 of 180) in this study. Eight patients had primary ovarian tumor including 2 hemorrhagic cysts, 1 dermoid cyst, 1 fibroma, and 4 serous cystadenoma. The remaining 47 patients had normal ovaries in the pathologic evaluation. There was no relation between abnormal morphology and metastatic involvement of the ovary in this study (P = 0.239).
Of the 120 women who underwent only colectomy, 8 developed metastasis to the ovary during the follow-up period. The mean time to presentation of metachronous ovarian metastasis was 14 months (range, 10-24 months).
The clinical features of all the patients with ovarian metastasis (synchronous and metachronous) are presented in Table 1.
The features of these patients were compared with patients who had nonmetastatic ovary in the pathologic study.
The mean age in the ovarian involvement group was 45 years, and most of them (n = 9, 69.23%) were premenopausal, whereas the mean age in the nonmetastatic group was 50.9, and half (n = 27, 50%) of them were premenopausal.
The most common presenting symptom in the ovarian involvement group was mass. However, in the nonmetastatic group, it was rectorrhagia.
Oophorectomy was mostly performed in rectal and rectosigmoid cancers. Only 3 patients in the nonmetastatic group and 1 patient in the ovarian involvement group had right and transverse colon cancers.
Most tumors in both groups were moderately to poorly differentiated. There was no correlation between ovarian metastases and tumor location and differentiation (P = 0.067 and P = 0.545 respectively). Preoperative carcinoembryonic antigen was elevated in all patients with ovarian involvement, but we could not find the level of this marker in most of the others.
In the ovarian involvement group, 9 patients (69.23%) had transmural extension of their colorectal tumor. Patients with ovarian metastasis had a higher stage of tumor compared with those without ovarian metastasis.
The survival of patients with ovarian metastasis was very poor. The survival of patients with metachronous metastases seems longer (mean 20 months) than that of those with synchronous metastases (mean, 10 months).
The true incidence of ovarian metastasis from CRC is unclear because the reported incidence varies from 0% to 30%, depending on autopsy data or clinical series.2,6-8 It varies from 5% to 31% in autopsy data and 0% to 8.6% in clinical series. Some of these clinical series are mentioned in Table 2. The mean incidence seems to be 3.5%.
In this study, the true incidence of synchronous and metachronous ovarian metastases was very low (2.7% and 6.6%, respectively).
Isolated ovarian metastases from primary CRC occurred in 3.3% of women undergoing colorectal resection. The mechanism of ovarian metastasis is uncertain.
Hematogenous spread is likely to be the main mechanism, although lymphatic dissemination has also been proposed.9
Transcoelomic spread was supported by synchronized involvement of the peritoneum in many cases of ovarian metastasis.10 However, deep location of metastatic tumor in the ovary makes transcoelomic spread unlikely.11
The high incidences of peritoneal disease (5 of 13), transmural tumor extension (9 of 13), and lymphatic disease (8 of 13) in our patients suggest that lymphatic pathway and direct peritoneal dissemination may serve as important mechanisms of ovarian involvement in CRC.
The influence of age and menstrual status on ovarian involvement is unclear. Several studies12-14 have demonstrated a higher relative frequency of ovarian metastasis in premenopausal women; some other studies15,16 have refuted the association between young age and ovarian metastases; they have reported that most women with ovarian metastases were postmenopausal.
In this study, the median age at diagnosis of ovarian metastasis was 41 years, and it seems that premenopausal women are relatively more commonly affected, but there was no correlation between menopausal status and ovarian involvement (P = 0.629).
It has been shown10,12,17-20 that differentiation of the primary colorectal tumor is not a risk factor for the development of ovarian metastasis. In this study, we got the same result (P = 0.545).
In addition, like other studies,10,12,13,18-21 there was no correlation between size and anatomic site of primary CRC and ovarian metastasis (P = 0.067). In fact, spread to the ovaries occurs from all parts of the large bowel in proportion to the natural frequency of the primary tumor.
In this study, oophorectomy was performed most commonly in women with cancer of the rectum and rectosigmoid, and this could be explained by a surgeon's trend to do prophylactic oophorectomy in rectal and sigmoid cancers.
Most CRCs with ovarian metastasis are Dukes B or more advanced.13,17,18,20,22
In this study, in the women with ovarian metastasis, the primary tumor extended through the bowel wall in 69.23% of the patients, and nodal disease was present in 61.53%. Peritoneal seeding and liver metastasis were presented in 38.46% and 15.38%, respectively.
It have been demonstrated11,18,23,24 that in one third to one half of ovarian metastasis from CRC, gross inspection of the ovaries may not reveal the presence of implants, the abnormality would be missed at operation, and only histologic examination may confirm the diagnosis.
In our study, oophorectomy was done because of morphologic abnormality of the ovaries in 35 cases, but only 8 primary ovarian tumors and 5 metastatic ovarian tumors were found.
Prophylactic bilateral oophorectomy was performed in 22 women in our series, and no one had an abnormality in the histologic examination. Several studies have been published about the role of prophylactic oophorectomy during CRC surgery.9,23-25
Although sample size and the statistical power of our study were small, we do not support prophylactic oophorectomy in our patients because we found the true incidence of synchronous and metachronous ovarian metastases to be very low.
It has been demonstrated that even if bilateral prophylactic oophorectomy had been performed, long-term survival is not affected24,26 and macroscopic metastatic disease to the ovaries is a poor prognostic factor in CRC.16
In this study, survival of patients with ovarian metastasis was disappointing. The survival of patients with metachronous metastasis seems longer than that of those with synchronous metastasis, but because of the very small sample size, it is not possible to compare the outcome of the 2 groups.
Two other studies that compared survivals in synchronous and metachronous ovarian metastases have contradictory conclusions.21,27
No patient with ovarian metastases in our study survived 5 years. Long-term survival has been reported relatively rarely in the literature. Miller et al16 reported one 5-year survivor (4.3%), whereas MacKeigam and Ferguson18 described 1 patient who was alive 10 years after diagnosis. Moreover, Wright et al28 reported 1 patient who was alive 8.9 years after diagnosis.
Blamey et al29 reported a 5-year cancer-specific survival rate of 50% in 25 patients with colon cancer metastasis to the ovary treated by curative resection.
Because of the poor prognosis of synchronous or metachronous ovarian metastases, several authors12,16 have advocated palliative surgical management only or especially when extensive intra-abdominal disease or distant metastasis is present. However, some others13,30 suggested a more aggressive surgical resection.
In conclusion, the true incidence of ovarian metastases in CRC is very low, the primary CRC is commonly advanced, and other distant metastases are frequent. Although oophorectomy reduces the chance of secondary laparotomy for resection of metachronous metastatic disease in a small percentage of patients, routine prophylactic bilateral oophorectomy is not justified.
1. Pitt J, Dawson PM. Oophorectomy
in women with colorectal cancer
. Eur J Surg Oncol
2. Abrams HL, Spiro R, Goldstein N. Metastases in carcinoma; analysisof 1000 autopsied cases. Cancer
3. Huang PP, Weber TK. Long-term survival in patients with ovarian metastases
from colorectal carcinoma. Ann Surg Oncol
4. Rayson D, Bouttell E, Whiston F, et al. Outcome after ovarian/adnexal metastasectomy in metastatic colorectal carcinoma. J Surg Oncol
5. Schenk F, Sitzenfrey A. Gleichzeitiges Karzinom des Magens, der ovarien und des uterus, mit besonderer berucksichtigung ihrer operativen Behandlung und der hitoloischen. Z Geburtshilfe Gynakol
6. Buirge RE. Carcinoma of the large intestine: review of four hundredand sixteen autopsy records. Arch Surg
7. Koves I, Vamosi-Nagy I, Besnyak I. Ovarian metastases
of colorectal tumours. Eur J Surg Oncol
. 1993;19(suppl 1):633-635.
8. Fujiwara K, Ohishi Y, Koike H, et al. Clinical implications of metastases to the ovary. Gynecol Oncol
9. Young-Fadok TM, Wolff BG, Nivatvongs S, et al. Prophylactic oophorectomy
in colorectal carcinoma. Dis Colon Rectum
10. Perdomo JA, Hizuta A, Iwagaki H, et al. Ovarian metastases
in patients with colorectal carcinoma. Acta Med Okayama
11. Graffner HOL, Alm POA, Oscarson JEA. Prophylactic oophorectomyin colorectal carcinoma. Am J Surg
12. Herrera-Ornelas L, Mittelman A. Results of synchronous surgical removal of primary colorectal adenocarcinoma and ovarian metastases
13. Morrow M, Enker WE. Late ovarian metastases
in carcinoma of the colon and rectum. Arch Surg
14. Taylor AE, Nicolson VMC, Cunningham D. Ovarian metastases
from primary gastrointestinal malignancies: the Royal Marsden Hospital experience and implications for adjuvant treatment. Br J Cancer
15. Tentes A, Markakidis S, Mirelis C, et al. Oophorectomy
during surgery for colorectal carcinoma. Tech Coloproctol
16. Miller BE, Pittman B, Wan JY, et al. Colon cancer with metastasis to the ovary at time of initial diagnosis. Gynecol Oncol
17. Herrera-Ornelas L, Natarajan N, Tsukada Y, et al. Adenocarcinomaof the colon masquerading as a primary ovarian neoplasia: an analysisof ten cases. Dis Colon Rectum
18. MacKeigan JM, Ferguson JA. Prophylactic oophorectomy
andcolorectal cancer in premenopausal patients. Dis Colon Rectum
19. Blamey SL, McDermott FT, Pihl E, et al. Resected ovarian recurrence from colorectal adenocarcinoma: a study of 13 cases. Dis Colon Rectum
20. Herrera-Ornelas L, Ledesma EJ, Natarajan N, et al. Metachronous ovarian metastases
from adenocarcinoma of the colon and rectum. Surg Gynecol Obstet
21. Tunca JC, Starling JR, Hafez GR, et al. Colon carcinoma metastatic to the ovary. J Surg Oncol
22. Mazur MT, Hsueh S, Gersall DJ. Metastases to the female genital tract; analysis of 325 cases. Cancer
23. Cutait R, Lesser ML, Enker WE. Prophylactic oophorectomy
in surgery for large-bowel cancer. Dis Colon Rectum
24. Sielezneff I, Salle E, Antoine K, et al. Simultaneous bilateral oophorectomy
does not improve prognosis of postmenopausal women undergoing colorectal resection for cancer. Dis Colon Rectum
25. Ballantyne GH, Reigel MM, Wolfe BG, et al. Oophorectomy
and colon cancer: impact on survival. Ann Surg
26. Shofield A, Pitt J, Biring G, et al. Oophorectomy
in primary colorectal cancer
. Ann R Coll Surg Engl
27. O'Brien PH, Newton BB, Metcalf JS, et al. Oophorectomy
in women with carcinoma of the colon and rectum. Surg Gynecol Obstet
28. Wright JD, Powell MA, Mutch DG, et al. Synchronous ovarian metastases
at the time of laparotomy for colon cancer. Gynecol Oncol
29. Blamey S, McDermott F, Pihl E, et al. Ovarian involvement in adenocarcinoma of the colon and rectum. Surg Gynecol Obstet
30. Webb MJ, Decker DG, Mussey E. Cancer metastatic to the ovary: factor influencing survival. Obstet Gynecol
31. Burt CA. Prophylactic oophorectomy
with resection of large bowel for cancer. Am J Surg
32. Stearns MW, Deddish MR. Five years results of abdominopelvic lymph node dissection for carcinoma of rectum. Dis Colon Rectum
33. Deddish MR. Surgical procedures for carcinoma of the colon and rectum, with five-years end results following abdominopelvic dissection of lymph nodes. Am J Surg
34. Barr SS, Valiete MA, Bacon HE. Rational for bilateral oophorectomy
concomitant with resection for carcinoma of the rectum and colon. Dis Colon Rectum
35. Sherman LF, Tenner RJ, Chadbourn WA. Prophylactic oophorectomy
with carcinoma of the rectum and colon. Dis Colon Rectum
. 1964;7: 517-520.
36. Harcourt KF, Dennis DL. Laparotomy for "ovarian tumor" in unsuspected carcinoma of the colon. Cancer
37. Antoinades K, Spector B, Hecksher RH. Prophylactic oophorectomyin conjunction with large-bowel resection for cancer: report of two cases. Dis Colon Rectum
38. Gemos K, Rizzotti L, Tsardis P, et al. Indications for prophylactic ovariectomy in patients with colorectal carcinoma. Minerva Chir
. 1995; 50:89-92.
39. Erroi F, Scarpa M, Angriman I, et al. Ovarian metastasis from colorectal cancer
: prognostic value of radical oophorectomy
. J Surg Oncol