INTRODUCTION
Ovarian cancer is one of the most common malignant tumors in the female reproductive system, and its incidence ranks second only to cervical cancer and endometrial cancer. However, it has the highest mortality rate, and 47% of all deaths from cancer of the female genital tract occur in women with ovarian cancer.[ 1 ] Five-year survival rates of patients diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stage I epithelial ovarian cancer are 90%, stage II 65%, stage III 34%, and stage IV 15%.[ 2 ] Therefore, early diagnosis is vital for reducing mortality. However, 71% of the patients were diagnosed at advanced stages (FIGO III and IV) due to the lack of typical clinical presentations and effective means of early diagnosis.[ 3 ]
Before 2008, serum CA125 was the only Food and Drug Administration-approved tumor marker of ovarian cancer.[ 4 ] About 50% of ovarian cancer patients have a high level of serum CA125 at early stages,[ 5 ] leading to low sensitivity and a high false-negative rate in the early asymptomatic phase. Besides, an increase of serum CA125 level can also be found in 1%–2% healthy individuals, 6% patients with benign diseases, and 28% patients with non-gynecological malignant diseases,[ 6 ] which can result in a high false-positive rate. HE4 was approved to be used in the detection of the recurrence of ovarian cancer in 2008.[ 7 ] Studies have shown that it performs better than serum CA125 in sensitivity and specificity in the early stages of epithelial ovarian cancer,[ 8 , 9 ] but its ability to detect ovarian cancer in the early stage is still far from satisfactory.[ 10 ] Transvaginal ultrasound is a widely accepted noninvasive technique which performs better than transrectal and peritoneal ultrasound.[ 11 ] However, the early detection of ovarian cancer by transvaginal ultrasound requires the presence of detectable masses that often present as precursors to low-grade serous carcinoma, endometrioid carcinoma, clear cell carcinoma, or mucinous carcinoma.[ 12 ] Thus as for high-grade serous cancers (HGSCs), the use of transvaginal ultrasound for screening early lesions is ineffective because a large proportion of which develop from small lesions of the fallopian tubes.[ 13–15 ] Several large clinical studies among asymptomatic average-risk women have indicated that the screening method combining CA125 and transvaginal ultrasound didn’t help in improving the prognoses of the ovarian cancer patients.[ 16 , 17 ] So finding an effective diagnostic tool is pretty imperative.
In 1992, Luzzatto[ 18 ] first reported the presence of malignant cells in endometrial smears from a patient with early bilateral ovarian cancer. Since then, a series of case reports have claimed that malignant cells were detected in endometrial smears from ovarian cancer patients[ 19–21 ] and fallopian tube cancer patients,[ 22–25 ] which provided a possible novel way of ovarian cancer, fallopian tube cancer, and primary peritoneal cancer diagnosis.
To evaluate the feasibility of diagnosing ovarian cancer, fallopian tube cancer, and primary peritoneal cancer through endometrial cytology , we conducted a systematic review and meta-analysis to estimate the positive rates of malignant cells in endometrial cytology samples from patients with ovarian cancer, fallopian tube cancer, and primary peritoneal cancer.
MATERIALS AND METHODS
Search strategy and selection criteria
PubMed, EMBASE, Medline, and Cochrane Central Register of Controlled Trails were queried from inception to November 12, 2020, for studies estimating positive rates of malignant cells in endometrial cytology samples from patients with ovarian cancer, fallopian tube cancer, and primary peritoneal cancer [Supplementary Tables 2 -4 ]. We excluded studies of which the populations had accepted any treatment before the exam, studies of which the positive rates of patients without endometrial lesions were not accessible, or the records without the original data, case reports, case series, reviews, and letters. Included publications were restricted to English articles. All studies were reviewed by two co-authors, separately, and discrepancies were settled by a discussion with the third reviewer.
STable 1: Search strategy of Cochrane Central Register of Controlled Trails (from inception to November 12, 2020)
STable 2: Search strategy of EMBASE (from inception to November 12, 2020)
STable 3: Search strategy of Medline (from inception to November 12, 2020)
STable 4: Search strategy of PubMed (from inception to November 12, 2020)
Quality assessment and data abstraction
For each selected article, a standard data extraction form was used to extract the general characteristics (first author, year of publication, country, research institution, period, study design, patient selection, sample size) and details of participants (age, tumor site, sampling methods, staining methods) and outcomes (the number of positive malignant cell cases, including by subgroup when available).
Definitions and outcomes
Positive endometrial cytology was defined by studies as the presence of malignant cells in endometrial specimens, regardless of the number of malignant cells. The outcome parameter was the pooled positive rate of malignant cells in endometrial cytology specimens of patients with ovarian cancer, fallopian tube cancer, and primary peritoneal cancer.
Statistical methods
We calculated the pooled positive rate and 95% confidence interval (95% CI) of malignant cells in endometrial cytology specimens of patients with ovarian cancer, fallopian tube cancer, and primary peritoneal cancer. Meta-analyses of proportions were performed on logit-transformed data. Furthermore, predefined subgroup analyses based on different sampling methods were conducted. I 2 statistic was used for evaluating the statistical heterogeneity of the included studies, I 2 < 25%, 25%–50%, 50%–70%, >70% were considered low, moderate, high, and considerable heterogeneity, respectively. If the results showed high or considerable heterogeneity, we would use the random effects model to pool the data; otherwise, the fixed effects model was used. Sensitivity analyses were used to assess the stability of the results. As only seven eligible articles were included and our sensitivity analyses showed great stability for our results, we didn’t perform our preplanned meta-regression to look for sources of heterogeneity. The above statistical analyses were all carried out through R 4.0.3.
Risk of bias assessment in individual studies
Modified Methodological Index for Non-Randomized Studies consisting of six items was used to assess the quality of the included studies. Two items regarding the follow-up information were deleted because they were not applicable. Finally, we evaluated each study in terms of aim, patient selection, data collection, endpoints, endpoint assessment and loss to follow-up less than 5%. Each item was assessed as “high risk”, “unclear risk”, or “low risk” if data were “reported but not adequate”, “not reported”, or “reported and adequate”. The assessment of the risk of bias was performed through Review Manager 5.4.
Ethics approval: Not applicable.
RESULTS
Among the 9082 studies screened from the databases initially, 7159 studies were left after removing the duplicates. After checking the titles and abstracts, we excluded 7077 studies that didn’t meet the inclusion criteria. The full texts of 13 records were not available through document delivery. Of the remaining 69 full-text articles, two were excluded for not written in English and one was excluded because of the lack of detailed data. Fifty-seven case reports, case series, and two abstracts were excluded. Finally, seven retrospective studies[ 26–32 ] including 975 patients were identified [Figure 1 ].
Figure 1: Flow diagram of study selection (PRISMA 2009 Flow Diagram)
The characteristics of the selected studies were shown in Table 1 . All of the studies were single-center studies from Japan, but it turned out that they were performed by different teams from different research centers. The years of publication varied from 1997 to 2020. The gold standards of diagnoses (positive control) of the researches were all histopathology. Two studies had a population of fewer than 100 patients.[ 26 , 31 ] The positive rates of malignant cells in endometrial cytology specimens of ovarian cancer, fallopian tube cancer, and primary peritoneal cancer patients ranged from 11.7% to 50%.
Table 1: Characteristics of the included studies
Risk of bias assessment in individual studies
All of the included studies showed a low risk of bias in terms of “aim”, “data selection”, “endpoints”, “endpoint assessment”, and “loss to follow up less than 5%”. Five studies[ 27 , 29–32 ] didn’t report if the patients were selected continuously or not, and thus, they were considered “unclear risk” [Figure 2 ].
Figure 2: (a) The risk of bias of each included study based on the review authors’ judgments; (b) Summary of each risk of bias item presented as percentages of all included studies
Pooled positive rate
Statistical heterogeneity among the included studies was considerable (I 2 = 89%, P < 0.01). We used random effects model to pool the positive rates. Pooled positive rate of malignant cells in endometrial cytology specimens of ovarian cancer, fallopian tube cancer, and primary peritoneal cancer patients was 23% (95% CI: 16%–34%) [Figure 3 ].
Figure 3: Forest plot of pooled positive rate of endometrial cytology of the included studies
Subgroup analysis
The result of subgroup analysis based on different sampling methods was shown in Figure 4 . There were four studies including 605 patients in the group of aspiration smears, the pooled positive rate of which was 33% (95% CI: 25%–42%, I 2 = 80%, P < 0.01). The group of brushes included three studies, the pooled positive rate was 13% (95% CI: 10%–17%, I 2 = 0, P = 0.45). The test of difference between subgroups showed a significant difference (P < 0.001). The specific sampling tools were listed in Table 1 , two of the studies didn’t report the staining methods, and[ 28 , 30 ] all of the other studies used Papanicolaou stain.
Figure 4: Forest plot of pooled positive rate of endometrial cytology between the different sampling methods’ subgroups
Sensitivity analyses
Sensitivity analyses showed that after eliminating the researches one by one, the pooled positive rates ranged from 21% to 26%, the I 2 varied from 83.2% to 90.5%, which indicated that our results were stable [Supplementary Figure 1 ].
SFigure 1: Forest plot of sensitivity analyses
DISCUSSION
Our research showed that the pooled incidence of malignant cells in endometrial cytology samples from ovarian cancer, fallopian tube cancer, and primary peritoneal cancer patients was 23% (95% CI: 16%–34%). Patients using the aspiration smears sampling method had a positive rate of 33% (95% CI: 25%–42%), which was higher than that of the brushes group (13%, 95% CI: 10%–17%). According to the result of the sensitivity analyses, our results were reckoned to have good stability.
Based on the subgroup analysis, we consider the different sampling methods as the major source of heterogeneity. We found that the positive rate was higher in the group of aspiration. However, the previous study done by our group indicated that brushes were more likely to obtain more adequate tissues than aspiration smears in patients with endometrial cancer.[ 33 ]
Although we didn’t perform further analyses due to the paucity of detailed data, there were some other clinic-pathologic characteristics as potential sources of heterogeneity. In 2002, Singer et al .[ 34 ] proposed a secondary grading system for ovarian serous cancers, which divided ovarian serous cancers into low-grade serous cancers and HGSCs. At present, multiple studies had bolstered the notion that many or most extra-uterine HGSCs may stem from the fallopian tubes.[ 35–38 ] Two of the included studies[ 27 , 32 ] showed that HGSCs had higher positive rates than other pathological types. This finding seems reasonable since the study of Bijron et al .[ 39 ] has shown that the intraluminal shedding of tumor cells in the fallopian tubes from serous carcinoma cases is common. What’s more, several studies suggested that advanced patients of stages III or IV had higher positive rates,[ 27 , 29 , 32 ] and high positive rates might be related to the presence of the ascites.[ 26 , 29 , 32 ]
Except for endometrial cytology , malignant cells have been proven to be detectable in ascites cytology and cervicovaginal cytology. Our literature review indicated that the positive rates of malignant cells in ascites cytology of ovarian cancer, fallopian tube cancer, and primary peritoneal cancer patients ranged from 18.1% to 85%, and 6.1% to 25% in cervicovaginal cytology [Table 2 ].[ 26–32 , 40 ]
Table 2: Review of literatures reporting the positive rates of malignant cells in ascites specimens and cervicovaginal specimens
We hold that with a positive rate of only 23%, although endometrial cytology is not an ideal tool for the diagnosis of ovarian cancer, fallopian tube cancer, and primary peritoneal cancer, it is a potentially convenient, painless, and easy-to-implement adjunct to other tools. A higher positive rate may be achieved if we apply it to specific groups, such as the high-risk population of HGSCs or patients with ascites, and more experiments should be carried out to combine endometrial cytology with other tools to see if there would be an increase in the detection rate .
Strengths and limitations
For all we know, our study was the first systematic review and meta-analysis to estimate the positive rates of malignant cells in endometrial cytology samples from ovarian cancer, fallopian tube cancer, and primary peritoneal cancer patients. Our conclusions were based on a rigorous process of article retrieval, article screening, literature evaluation, data extraction, and data processing.
However, there were still some inevitable limitations. Firstly, due to the constraint of the language and literature resources, we only included English studies, which may lead to a language bias. Secondly, we didn’t conduct subgroup analyses based on tumor sites, pathological types, the presence of ascites and the FIGO stages on account of the lack of detailed data. Thirdly, owing to the small number of related studies, this study had a limited sample size. Fourthly, all of the included studies were from Japan, which lead to a geographical constraint on the applicability of the results. Finally, baseline characteristics were not clearly stated in some studies, which may affect the results.
Financial support and sponsorship
This work was supported by the Clinical Research Award of the First Affiliated Hospital of Xi’an Jiaotong University, China (XJTU1AF-2018-017, XJTU1AF-CRF-2019-002), the Natural Science Basic Research Program of Shaanxi (2018JM7073, 2017ZDJC-11), the Key Research and Development Program of Shaanxi (2017ZDXM-SF-068, 2019QYPY-138), the Innovation Capability Support Program of Shaanxi (2017XT-026, 2018XT-002), and the Medical Research Project of Xi’an Social Development Guidance Plan (2017117SF/YX011-3). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Conflicts of interest
There are no conflicts of interest.
REFERENCES
1. Berek JS, Kehoe ST, Kumar L, Friedlander M. Cancer of the ovary, fallopian tube, and peritoneum. Int J Gynaecol Obstet 2018;143 Suppl 2: 59-78
2. Jessmon P, Boulanger T, Zhou W, Patwardhan P. Epidemiology and treatment patterns of epithelial ovarian cancer. Expert Rev Anticancer Ther 2017;17:427-37
3. Nagle CM, Francis JE, Nelson AE, Zorbas H, Luxford K, de Fazio A, et al. Reducing time to diagnosis does not improve outcomes for women with symptomatic ovarian cancer: A report from the Australian Ovarian Cancer Study Group. J Clin Oncol 2011;29:2253-8
4. Moore RG, McMeekin DS, Brown AK, DiSilvestro P, Miller MC, Allard WJ, et al. A novel multiple marker bioassay utilizing HE4 and CA125 for the prediction of ovarian cancer in patients with a pelvic mass. Gynecol Oncol 2009;112:40-6
5. Frietze KM, Roden RB, Lee JH, Shi Y, Peabody DS, Chackerian B. Identification of anti-CA125 antibody responses in ovarian cancer patients by a novel deep sequence-coupled biopanning platform. Cancer Immunol Res 2016;4:157-64
6. Bast RC Jr, Klug TL, St John E, Jenison E, Niloff JM, Lazarus H, et al. A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 1983;309:883-7
7. Yuan C, Li R, Yan S, Kong B. Prognostic value of HE4 in patients with ovarian cancer. Clin Chem Lab Med 2018;56:1026-34
8. Wang J, Gao J, Yao H, Wu Z, Wang M, Qi J. Diagnostic accuracy of serum HE4, CA125 and ROMA in patients with ovarian cancer: A meta-analysis. Tumour Biol 2014;35:6127-38
9. Karlsen MA, Høgdall EV, Christensen IJ, Borgfeldt C, Kalapotharakos G, Zdrazilova-Dubska L, et al. A novel diagnostic index combining HE4, CA125 and age may improve triage of women with suspected ovarian cancer-An international multicenter study in women with an ovarian mass. Gynecol Oncol 2015;138:640-6
10. Huang J, Chen J, Huang Q. Diagnostic value of HE4 in ovarian cancer: A meta-analysis. Eur J Obstet Gynecol Reprod Biol 2018;231:35-42
11. Campbell S, Bhan V, Royston P, Whitehead MI, Collins WP. Transabdominal ultrasound screening for early ovarian cancer. BMJ 1989;299:1363-7
12. Horiuchi A, Itoh K, Shimizu M, Nakai I, Yamazaki T, Kimura K, et al. Toward understanding the natural history of ovarian carcinoma development: A clinicopathological approach. Gynecol Oncol 2003;88:309-17
13. Piek JM, Verheijen RH, Kenemans P, Massuger LF, Bulten H, van Diest PJ. BRCA1/2-related ovarian cancers are of tubal origin: A hypothesis. Gynecol Oncol 2003;90:491
14. Sowamber R, Nelson O, Dodds L, DeCastro V, Paudel I, Milea A, et al. Integrative transcriptome analyses of the human fallopian tube: Fimbria and ampulla-site of origin of serous carcinoma of the ovary. Cancers (Basel) 2020;12:1090
15. Lee Y, Miron A, Drapkin R, Nucci MR, Medeiros F, Saleemuddin A, et al. A candidate precursor to serous carcinoma that originates in the distal fallopian tube. J Pathol 2007;211:26-35
16. Buys SS, Partridge E, Black A, Johnson CC, Lamerato L, Isaacs C, et al. Effect of screening on ovarian cancer mortality: The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Randomized Controlled Trial. JAMA 2011;305:2295-303
17. Henderson JT, Webber EM, Sawaya GF. Screening for ovarian cancer: Updated evidence report and systematic review for the US preventive services task force. JAMA 2018;319:595-606
18. Luzzatto R, Rodrigues IC, Recktenvald M, Brucker N. Diagnosis of occult ovarian carcinoma in endometrial smears. Acta Cytol 1992;36:653-4
19. Nwanodi O, Choi C, Khulpateea N. Cervicovaginal cytology and diagnosis of ovarian or peritoneal cancer: Case report and literature review. Arch Gynecol Obstet 2008;277:171-4
20. Shintaku M, Taniguchi H, Yamamoto Y, Kono F, Sumitomo M. Detection of tumor cells of serous tubal intraepithelial carcinoma (STIC) in cervical smears and rapid development of the ovarian involvement: A case report. Diagn Cytopathol 2018;46:945-9
21. Tepeoğlu M, Ozen O, Ayhan A. Ovarian serous borderline tumor detected by conventional papanicolaou smear: A case report. Acta Cytol 2013;57:96-9
22. Luzzatto R, Sisson G, Luzzatto L, Recktenvald M, Brucker N. Psammoma bodies and cells from
in situ fallopian tube carcinoma in endometrial smears: A case report. Acta Cytol 1996;40:295-8
23. Minato H, Shimizu M, Hirokawa M, Fujiwara K, Kohno I, Manabe T. Adenocarcinoma
in situ of the fallopian tube. A case report. Acta Cytol 1998;42:1455-7
24. Ural UM, Balik G, Tekin YB, Sehitoglu I, Bedir R, Sahin FK. Primary fallopian tube carcinoma diagnosed preoperatively by cervical smear. Ann Saudi Med 2014;34:444-6
25. Maeda D, Takazawa Y, Ota S, Takeuchi Y, Seta A, Nakagawa S, et al. Bilateral microscopic adenocarcinoma of the fallopian tubes detected by an endometrial cytologic smear. Int J Gynecol Pathol 2010;29:273-7
26. Takeshima N, Hirai Y, Yamauchi K, Hasumi K. Clinical usefulness of endometrial aspiration cytology and CA-125 in the detection of fallopian tube carcinoma. Acta Cytol 1997;41:1445-50
27. Saji H, Kurose K, Sugiura K, Miyagi E, Onose R, Kato H, et al. Endometrial aspiration cytology for diagnosis of peritoneal lesions in extrauterine malignancies. Acta Cytol 2007;51:533-40
28. Otsuka I, Kameda S, Hoshi K. Early detection of ovarian and fallopian tube cancer by examination of cytological samples from the endometrial cavity. Br J Cancer 2013;109:603-9
29. Suzuki M, Suzuki T, Matsuura M, Iwasaki M, Tanaka R, Ito E, et al. Prediction of histologic type and lymph node metastasis for advanced ovarian cancer on uterine cervical and
endometrial cytology . Acta Cytol 2010;54:575-81
30. Ushigusa T, Yoshida H, Kuno I, Kojima N, Ishikawa M, Kato T. Characteristics and prognostic significance of incidentally detected cancer cells in uterine specimens of patients with pelvic high-grade serous carcinoma. Cytopathology 2020;31:122-9
31. Hirasawa T, Yasuda M, Muramatsu T, Itoh H, Shinozuka T, Makino T, et al. Cytologic study of ascites and the endometrium in ovarian carcinoma. Clinical significance. Acta Cytol 1997;41:1451-5
32. Jobo T, Arai M, Iwaya H, Kato Y, Ohno E, Kuramoto H. Usefulness of endometrial aspiration cytology for the preoperative diagnosis of ovarian carcinoma. Acta Cytol 1999;43:104-9
33. Du J, Li Y, Lv S, Wang Q, Sun C, Dong X, et al. Endometrial sampling devices for early diagnosis of endometrial lesions. J Cancer Res Clin Oncol 2016;142:2515-22
34. Singer G, Kurman RJ, Chang HW, Cho SK, Shih Ie M. Diverse tumorigenic pathways in ovarian serous carcinoma. Am J Pathol 2002;160:1223-8
35. Crum CP, McKeon FD, Xian W. The oviduct and ovarian cancer: Causality, clinical implications, and “targeted prevention”. Clin Obstet Gynecol 2012;55:24-35
36. Gao FF, Bhargava R, Yang H, Li Z, Zhao C. Clinicopathologic study of serous tubal intraepithelial carcinoma with invasive carcinoma: Is serous tubal intraepithelial carcinoma a reliable feature for determining the organ of origin?. Hum Pathol 2013;44:1534-43
37. Seidman JD, Zhao P, Yemelyanova A. “Primary peritoneal”high-grade serous carcinoma is very likely metastatic from serous tubal intraepithelial carcinoma: Assessing the new paradigm of ovarian and pelvic serous carcinogenesis and its implications for screening for ovarian cancer. Gynecol Oncol 2011;120:470-3
38. Kurman RJ, Shih Ie M. The origin and pathogenesis of epithelial ovarian cancer: A proposed unifying theory. Am J Surg Pathol 2010;34:433-43
39. Bijron JG, Seldenrijk CA, Zweemer RP, Lange JG, Verheijen RH, van Diest PJ. Fallopian tube intraluminal tumor spread from noninvasive precursor lesions: A novel metastatic route in early pelvic carcinogenesis. Am J Surg Pathol 2013;37:1123-30
40. Sasagawa M, Nishino K, Honma S, Kodama S, Takahashi T. Origin of adenocarcinoma cells observed on cervical cytology. Acta Cytol 2003;47:410-4
