Utility of peritoneal lavage cytology during laparoscopic salpingo-oophorectomy: a retrospective analysis
R Crawford, Cambridge University Hospitals NHS Foundation Trust, Hills Road, Cambridge CB2 0QQ, UK. Email email@example.com
Please cite this paper as: Haldar K, Giamougiannis P, Crawford R. Utility of peritoneal lavage cytology during laparoscopic salpingo-oophorectomy: a retrospective analysis. BJOG 2011;118:28–33.
Objectives To assess the significance of peritoneal washing cytology at the time of laparoscopic salpingo-oophorectomy.
Design Retrospective study.
Setting Cambridge University Hospital.
Population Four hundred and nine women who underwent laparoscopic salpingo-oophorectomy by the gynaecology oncology team between 2004 and December 2009 were included. One hundred and thirteen women had risk-reducing salpingo-oophorectomy, 103 women had salpingo-oophorectomy as part of breast cancer management, 59 had simple ovarian cysts, 111 had complex ovarian cysts and 23 had the procedure done for other reasons.
Methods Histology and cytology results were reviewed and all hospital records were checked for subsequent malignancy. Sensitivity and specificity of peritoneal washing cytology was calculated.
Main outcome measures Malignant peritoneal cytology in the absence of cancer on histopathology.
Results Eleven of the 409 women in our study had occult malignancy on histopathological examination and three of them had positive peritoneal washings. One woman had positive washings from metastatic breast cancer. Thirteen women developed different malignancies subsequently but none had primary peritoneal or ovarian cancer within a median follow-up interval of 34 months.
Conclusions Peritoneal lavage cytology did not pick up any additional malignancy in the study population. Based on the evidence presented, we suggest that peritoneal washing cytology during laparoscopic salpingo-oophorectomy is of limited value and should not be practised routinely.
Peritoneal lavage cytology is an established technique to detect malignant cells in gynaecological cancers, that is endometrial, ovarian and fallopian tube carcinoma and is a component of the staging process.1,2 This procedure has a high sensitivity for detecting cancer cells and it carries minimal risk to the patient; surgical oncologists perform this procedure on entry into the abdomen almost reflexively. A positive or a suspicious wash should lead to a re-evaluation of the peritoneal cavity, fallopian tubes and endometrium.3 Although there are several protocols recommending that cytological washings of the pelvis should be obtained when performing bilateral salpingo-oophorectomy as part of risk-reducing surgery4,5 their role in detecting early tubal or ovarian cancer remains controversial.4 The primary purpose of performing peritoneal washings in these situations is to assist in the detection of occult primary peritoneal carcinoma.6 The presence of malignant cells in peritoneal washings could theoretically also come from a tiny focus of cancer that was missed during routine histological examination. A positive peritoneal cytology in the presence of malignancy in the fallopian tubes or ovaries in the setting of risk-reducing surgery or even limited assessment of ovarian pathology does not obviate the need for a proper staging procedure.
Using the characteristics of reactive mesothelial cells and familiarity with the cytological features differentiating them from malignant cells, the cytologist can avoid the potential diagnostic pitfalls associated with reactive mesothelial cells suggesting malignancy.3 Additionally, with a lavage technique, there is a risk of a sample containing too few or no cells at all leading to a potentially false-negative result. Even though recommendations for the use of peritoneal lavage have been made4,5 the necessity of the procedure and the value of the information obtained in the management or follow up of patients are still not clearly defined.3
Laparoscopic bilateral salpingo-oophorectomy is a procedure commonly performed for a variety of indications, especially when the suspicion of malignancy is low. The main indications include: (1) risk-reducing surgery in women at high risk for ovarian carcinoma (particularly with known mutations in the BRCA1 and BRCA2 genes7 or those with a significant family history of ovarian or breast cancer; (2) hormonal manipulation of oestrogen-dependent breast cancer in premenopausal women, as an alternative to gonadotrophin-releasing hormone analogues;8 and (3) and (4) for simple or complex ovarian cysts, where the risk of malignancy, as determined clinically as well as by imaging and tumour marker levels, is low.
The advantages of laparoscopy include day-case surgery and a rapid recovery, compared with open procedures.9 Laparoscopic salpingo-oophorectomy is therefore considered as an intermediate-risk surgical procedure with high patient acceptance.9 If occult cancer is detected on histopathology, a staging operation follows. This includes peritoneal washing cytology to detect the presence of extra-ovarian disease as part of the standard staging process.
Prophylactic salpingo-oophorectomy in particular is an accepted treatment strategy for women with an inherited predisposition to ovarian carcinoma.7 About 10% of epithelial ovarian cancers are thought to be related to a germline mutation.10BRCA1 and BRCA2 are the most commonly affected genes, accounting for approximately 90% of the mutations in hereditary ovarian cancer.10 A woman with a BRCA1 mutation has a 40–65% risk of developing ovarian carcinoma in her lifetime. The lifetime risk of ovarian cancer for a woman with a BRCA2 mutation is less, approximately 15–25%. Prophylactic salpingo-oophorectomy in these women not only reduces their risk of developing breast and gynaecological cancers but may lead to the detection of asymptomatic, early-stage ovarian carcinoma, which typically carries a better outcome.4
The practice at the gynaecological oncology unit in Addenbrooke’s Hospital is to perform peritoneal washings in women undergoing laparoscopic salpingo-oophorectomy for any indication. The objective of this study is to test the hypothesis that peritoneal washing cytology is not a necessary adjunct to surgery for the above indications.
We performed a retrospective audit of practice after the protocol was reviewed by the institutional audit board. No patient-identifiable data were reported and no direct interaction with the women was necessary so ethical approval was not necessary for this project. The primary aim of this project was to audit the practice of performing peritoneal washings during laparoscopic salpingo-oophorectomy.
We used the Addenbrooke’s Hospital gynaecology oncology services database to identify all women who underwent laparoscopic salpingo-oophorectomy between January 2004 and November 2009. Every surgical procedure carried out by our team is prospectively registered into this database after surgery. After identifying all women in our study population we performed a retrospective review of hospital records and collected demographic, clinical, surgical, pathological and follow-up data.
Data regarding BRCA status and high-risk family history in the risk-reducing group was cross-checked with data from our genetic clinic database.
All laparoscopic procedures in our unit were performed by the open technique using Hasson approach. Peritoneal washings were performed before any major manipulation of the pelvic organs and samples were collected from pelvic peritoneum only.
All surgical specimens were submitted for thorough histological and cytological examination and serial sectioning of the ovaries and fallopian tubes was carried out for women with BRCA mutations.
During the study period 409 women underwent laparoscopic salpingo-oophorectomy. Of these, 395 women had peritoneal washings performed for cytological evaluation at laparoscopy. Table 1 shows the indications for performing salpingo-oophorectomy, the mean age and age range for women in each group, the histological and cytological findings, serum CA125 levels where measured and range of abnormal results as well as subsequent cancers diagnosed at follow up. Retrospective review of hospital records revealed that the range of follow up was 0–71 months, with a median of 34 months.
Table 1. Breakdown of women undergoing laparoscopic salpingo-oophorectomy for different indications, their age group, cytological and histological findings, serum CA125 levels where available and subsequent malignancies if any
|Risk reducing (n = 113)||52/35–78||2||0||111||2||108||3||20||3/38–45||7|
|Breast cancer (n = 103)||45/32–56||1||0||102||2||96||5||3||1/40||3|
|Simple cyst <5 cm (n = 22)||61/50–75||0||0||22||0||22||0||18||4/50–389||1|
|Simple cyst >5 cm (n = 37)||61/49–84||2||1||34||0||36||1||30||7/36–103||2|
|Complex cyst, 1 feature (n = 54)||58/40–82||1||3||50||0||53||1||46||8/56–750||0|
|Complex cyst, >1 feature (n = 57)||61/42–86||5||1||51||0||54||3||48||8/43–126||0|
|Other (n = 23)||47/32–64||0||0||23||0||22||1||11||10/39–436||0|
One woman out of 103 (1%) who underwent salpingo-oophorectomy as part of the treatment of hormone-sensitive breast cancer was found to have serous papillary cancer of ovary and positive peritoneal washings. Another woman had positive peritoneal washing from metastatic breast cancer and was found to have hepatic metastasis at laparoscopy. The only woman diagnosed with primary ovarian malignancy had stage III cancer with positive peritoneal washings in subsequent staging laparotomy. Three women in this group developed recurrence of breast cancer subsequently but none had any other malignancy.
None with a simple ovarian cyst <5-cm diameter was found to have any malignancy or positive cytology. One of them developed Paget’s disease of breast and myeloma later in life. Two women with >5-cm simple cysts were found to have metastases from non-Hodgkin’s lymphoma, one had normal and the other had marginally raised serum CA125 levels. Both women had negative washings. Another lady with a borderline tumour had high serum CA125 levels and negative washings and subsequent staging surgery revealed no residual disease.
One hundred and eleven women had complex ovarian cysts, 54 of the cysts had only one abnormal feature on imaging and 57 had more than one abnormal feature. Of the complex cysts with a single abnormal feature, one (1.9%) turned out to be malignant and three (5.6%) showed borderline disease. None had positive cytology. Of the three, one woman with borderline ovarian tumour had a CA125 level of 56. The two other women with borderline cysts and the patient with a malignant ovarian tumour, all had normal pre-operative CA125 levels. All had subsequent staging surgeries. Of women who had cysts with more than one abnormal imaging characteristic, five (8.8%) had malignant histology and one (1.8%) had a borderline tumour. Two were metastases from previous breast cancer. The three (5.2%) primary ovarian malignancies were a malignant mixed müllerian tumour of the ovary, a sex cord stromal tumour and a papillary cancer of the thyroid in a mature cystic teratoma. The two women with breast cancer metastases in the ovaries had raised serum CA125 levels whereas the other three malignancies and the borderline tumour were associated with normal serum CA125 levels. None of these six women had positive washings. Of the 111 women with complex cysts none developed any cancer during follow up. Twenty-three women had salpingo-oophorectomy for other reasons. None in this group had malignant histology, positive cytology, or developed cancer subsequently. Also, in women with ovarian cysts of complex morphology on ultrasonography, we noted borderline and malignant histology in 3.6 and 5.4% only. Ninety-three percent of cysts with one abnormal feature and 89% with more than one abnormal feature were benign. The four women in this group who had primary ovarian malignancy were found to have stage I disease with negative washings in staging surgery that followed. Three of these four women had a Risk of Malignancy Index <50 and for the fourth patient it was 168.
Table 2 gives the breakdown for women who had risk-reducing salpingo-oophorectomy for either BRCA mutations or for significant family history of breast or ovarian cancer or both. All patients had a consultation with both a genetic counsellor and the senior author and thorough counselling before having risk-reducing surgery.
Table 2. Outcome of women undergoing risk-reducing salpingo-oophorectomy (RRSO) in our study
|BRCA 1 (n = 27)||46/35–77||2||0||25||2||25||0||0||1/39||4|
|BRCA 2 (n = 17)||48/37–59||0||0||17||0||17||0||0||1/45||1|
|Family history (n = 69)||54/35–78||0||0||69||0||66||3||18||1/38||2|
Two women of the 113 (1.8%) who had risk-reducing salpingo-oophorectomy had histological evidence of primary peritoneal cancer and positive peritoneal cytology. Both these women were BRCA1 mutation carriers and had subsequent staging surgery followed by combination chemotherapy for stage III disease.
Among the 27 women who had the BRCA1 mutation, two (7.4%) developed breast cancer, one had malignant mixed müllerian tumour of the uterus and another developed endometrial cancer during follow up. One (5.9%) BRCA2 mutation carrier developed breast cancer and two women with high-risk family history developed breast cancer and cutaneous melanoma subsequently.
Three hundred and eighty-one procedures were performed by laparoscopy. There were complications in three patients (0.73%). One patient had a laparotomy because bilateral tubo-ovarian abscesses with extensive adhesions were noted on laparoscopy. Another patient had a perforation of the sigmoid diverticulum during adhesiolysis and had an open repair followed by bilateral salpingo-oophorectomy. One small bowel perforation was diagnosed intra-operatively and was repaired laparoscopically.
For the purpose of discussion we divided women having laparoscopic bilateral salpingo-oophorectomy into four groups based on the indication, as mentioned above. The objective of our study was to assess if peritoneal washing cytology performed at laparoscopic salpingo-oophorectomy helped to identify occult peritoneal cancer or small foci of malignancy on the ovarian surface that would otherwise go undetected. In our entire study population we found no such case of missed malignancy during subsequent follow up. We found peritoneal washing cytology to have 33% sensitivity and 100% specificity in diagnosing concomitant malignancy. In light of these findings we looked at available literature to question further the significance and necessity of carrying out peritoneal washing cytology at laparoscopic salpingo-oophorectomy.
Risk-reducing surgery in women with BRCA mutations or those with a significant family history of breast/ovarian cancer can reduce their risk of developing ovarian or fallopian tube cancer by 94–99%. Salpingo-oophorectomy was shown to be highly protective against development of breast cancer and BRCA-related gynaecological cancers with a hazard ratio of 0.21 (95% CI 0.07–0.62). Occult cancer was detected in 4.5% of our patients with BRCA mutations, which correlated well with the quoted incidence 2–10% in other studies.7,11 Serial sectioning of the tissue specimen helps in diagnosing such unsuspected malignancies.12 Peritoneal washing cytology is potentially another way of identifying malignant cells shed from the ovarian surface, fallopian tube epithelium or the peritoneum that might otherwise be missed on histopathology. This has, however, not been proven conclusively in any study.
Available guidelines suggest that peritoneal washing cytology might be of benefit.4,5 A previous study6 of 35 women who underwent risk-reducing salpingo-oophorectomy recommended peritoneal washing cytology. Twenty-eight women in the study were diagnosed to have a BRCA mutation, three were negative for such mutations and the BRCA carrier status of the remaining four women was unknown. Peritoneal lavage cytology detected three cases of carcinoma. Histology revealed multiple foci of poorly differentiated serous cancer of the ovary in one and a single focus of adenocarcinoma in situ in another woman. In the third case, cytological examination of the peritoneal lavage specimen revealed occasional clusters of malignant cells with features of serous carcinoma. Histological examination of the ovaries and fallopian tubes, however, revealed no malignancy despite extensive additional sectioning. A second laparotomy followed but no malignancy was documented on the hysterectomy specimen. No further data about staging is mentioned in this regard. Peritoneal washing cytology during the second laparotomy interestingly failed to reveal malignant cells. The woman, on the basis of the initial cytology result received chemotherapy and remained well 10 months after operation. There was no mention of whether immunocytochemistry was used to confirm the presence of malignancy or the possible site of origin of the malignant cells in the cytological specimen. The use of chemotherapy in this situation, especially when the source of primary cancer, if any, remains unknown, is debatable.
Ivazzo et al.13 reported a case of a postmenopausal woman who had an exploratory laparotomy for a benign serous cystadenoma with normal tumour markers but showed positive peritoneal cytology. Examination of the uterus revealed complex atypical hyperplasia of the endometrium. The peritoneal fluid cytology was positive for malignancy. A second laparotomy was performed and omentectomy and pelvic and para-aortic lymph node sampling was carried out. Further sectioning of the ovary and fallopian tubes was performed but no histological evidence of cancer was noted. Postoperative imaging of the thorax and abdomen was within normal limits, as were endoscopic studies of stomach and colon. The patient was treated with combined chemotherapy based on the initial cytology result and remained free of any disease 12 months later. A definite proof of malignancy was lacking in this case too and chemotherapy was administered in the absence of a histological diagnosis. There needs to be a detailed discussion with the patient in this scenario, regarding the risk and benefit of chemotherapy, given the absence of histological diagnosis of a primary malignancy and the significant toxicity of chemotherapeutic medications.
To our knowledge there is no instance where a positive peritoneal cytology has helped in definite diagnosis of occult cancer that would have otherwise been missed. Table 3 lists all the studies in which peritoneal washings were performed during risk-reducing salpingo-oophorectomy. Two of the studies were not aimed at assessing the value of peritoneal fluid washing at salpingo-oophorectomy but focused primarily on detecting fallopian tube dysplasia and mesothelial cell atypia, respectively, in these high-risk women.
Table 3. Summary of studies that looked at peritoneal cytology at risk-reducing salpingo-oophorectomy
|Colgan et al., 20026||Risk-reducing surgery||35||35||2||3||Not available||NA|
|Leunen et al., 200614||Risk-reducing surgery||51||28||0||0||None||25 months|
|Eitan et al., 20063||Risk-reducing surgery||130||117||0||0||6||20 months|
|Present study, 2010||Risk-reducing surgery||113||110||2||2||7||34 months|
In these four studies, 290 out of 329 women who underwent risk-reducing salpingo-oophorectomy had peritoneal washings. A total of four (1.2%) malignancies were noted on histology among all patients in these four studies and five women had positive peritoneal washings. As for the other woman, as discussed before, no definitive malignancy was diagnosed even after re-laparotomy. Eitan et al.3 noted a higher rate of mesothelial atypia (24%) among BRCA mutation carriers whereas Leunen et al.14 showed a higher incidence of endosalpingiosis (15%) in cytological washings in women with BRCA mutations. None of the reports were powered to detect a significant association of these abnormal findings with BRCA mutation status or clinical outcome. In our study we found endosalpingiosis in peritoneal washings in only two (4.5%) of the BRCA mutation carriers and none in BRCA-negative women. The clinical significance of endosalpingiosis or mesothelial atypia is uncertain and there is no evidence from these studies that their presence is associated with a risk of developing peritoneal carcinomatosis after salpingo-oophorectomy. In our study, one woman was noted to have mild to moderate cytological atypia in the fimbrial end of the fallopian tube. Peritoneal washings were negative in her case and she did not develop any malignancy in 15 months of follow up. Laparoscopy was the preferred route of surgery in 44 (86%), 112 (96%) and 407 (99.5%) of the women in the last three studies (Table 3), respectively. Peritoneal washing did not pick up any additional case of cancer that was subsequently demonstrated on histopathological examination and did not add information that altered management or follow up of these patients. Occult cancer was noted in only four (1.2%) of the women in these four studies combined. Hence, peritoneal washing cytology does not seem to be of any value when performed routinely during risk-reducing salpingo-oophorectomy.
We were unable to identify any study that evaluated the significance of peritoneal washing cytology during salpingo-oophorectomy, laparoscopic or open, performed for other indications where risk of malignancy is low. In our study, 296 women underwent salpingo-oophorectomy for other indications and 285 women among them had peritoneal washings performed. Five (1.7%) cases of ovarian cancer, four (1.4%) ovarian metastases from other primary cancers and five (1.7%) borderline ovarian tumours were noted. Of the two women with positive peritoneal washings one had a primary ovarian cancer whereas the other women had extensive abdominal metastasis from breast cancer. Peritoneal washing cytology did not detect any occult cancer that was otherwise missed at histopathology. One ovarian cancer (1%) was detected in women having bilateral salpingo-oophorectomy for breast cancer treatment and one (1.7%) borderline tumour was found in women with simple ovarian cysts. On the other hand, four (3.6%) primary ovarian cancers and four (3.6%) borderline tumours were noted in women with complex ovarian cysts. The risk of malignancy index was <250 in all of these women. From our results it appears unlikely that peritoneal washing is of any benefit in women undergoing bilateral salpingo-oophorectomy for breast cancer or simple ovarian cysts. If complex features are noted in the ovarian cysts then the risk of malignancy or borderline tumours is higher and peritoneal washings may help in the staging process.
This practice of carrying out peritoneal washings during laparoscopic salpingo-oophorectomy is not uniform among institutions, even among individuals from the same institution. There is no strong evidence in favour of, or against, collecting peritoneal washings during such procedures. We have hence included salpingo-oophorectomy performed for different indications in our study. We conclude that peritoneal washings are not useful in diagnosing peritoneal cancer or occult ovarian cancers that might be missed at histology. However, it would aid in the staging process when the risk of malignancy is higher and where there is a risk of cyst rupture during the procedure hence differentiating a natural from an iatrogenic 1C tumour. Studies by Sharifi et al.15 and Mathew et al.16 also do not support the use of routine peritoneal washing cytology during gynaecological surgery for apparently benign indications and suggested that there is potential for significant cost savings.
The process of collection of peritoneal washings at laparoscopy is simple and does not take up more than a few minutes of operating time. It does not have the potential of causing any direct harm to the patient. However, valuable resources are used for collection and processing of the specimens. We estimated the cost of collecting and processing each peritoneal washing specimen in our institution at approximately £120–£140, a potential saving if washings are not collected routinely at laparoscopic salpingo-oophorectomy. This includes cost of surgical equipment (about £75) used to collect the washings and the cost of processing the specimen (approximately £45–£65 depending on the sample). Additionally, apart from the cost saving, this also saves the valuable time of the specialist consultants as well as of the para-medical staff involved in the process.
Disclosure of interest
This is to declare that KH, PG and RC have no financial and personal relationships with other people or organisations that could have inappropriately influenced this work.
Contribution to authorship
KH was instrumental in planning the study, data collecting and writing the paper. PG helped with data collection and in writing the paper. RC was involved in planning the study, finalising the study protocol and writing the paper.
Details of ethics approval
No patient-identifiable data were reported and no direct interaction with women was necessary so ethical approval was not necessary for this project. This was confirmed by the Research & Development Department of the Cambridge University Hospital Foundation trust.
This study resulted from an audit and there was no funding issue associated with this paper.
We thank Mr Peter Baldwin, Mr John Latimer, Mr Mahmood Shafi, Mr Ahmed Ahmed and Dr Mercedes Jimenez-Linan for their valuable support in the study.