Williams, Melony G. MPH1,2,3; Bandera, Elisa V. MD, PhD1,2,3; Demissie, Kitaw MD, PhD, MPH2,3; Rodríguez-Rodríguez, Lorna MD, PhD4
The occurrence of synchronous primary ovarian and endometrial cancers is relatively uncommon, but it is important in clinical practice because a diagnosis of synchronous malignancies would affect clinical management and prognosis.1
Currently there is no consensus on the definition of synchronous cancer. Studies attempting to estimate its occurrence have used various definitions, as summarized by Coli et al,2 ranging from simultaneous detection to detection of both tumors within 5 years. Ulbright and Roth3 were the first to derive a set of histologic criteria, later revised by Scully et al4 to distinguish between synchronous ovarian and endometrial cancers and those that are metastatic. However, these criteria have yet to be validated independently. In a case series report with a small number of patients, synchronous ovarian and endometrial cancers occurred in about 3–10% of patients with ovarian cancer.5 However, that report did not specify the time interval between detection of both tumors. Overall, patients with synchronous ovarian and endometrial cancers were more likely to have ovarian tumors with endometrioid histology and to exhibit better survival than those with metastatic tumors.1,5–8 Yet, most of these studies are hospital-based with small samples (mostly fewer than 100 synchronous cancers), limiting the generalizability of their findings. Few population-based studies have examined the occurrence and survival experience of synchronous primary ovarian and endometrial cancers.6,9
The aim of this study was to estimate the occurrence of synchronous ovarian and endometrial cancers and to compare the survival experience of women with these synchronous cancers with that of women with single ovarian cancer using data from the Surveillance, Epidemiology, and End Results (SEER) Program.
MATERIALS AND METHODS
Data from the National Cancer Institute's SEER Program were used for this study, including cancer diagnoses from 1973 to 2005.10 The SEER Program collects and compiles information on cancer cases representing about 26% of the U.S. population. For this study, cases reported from the San Francisco-Oakland, Connecticut, Metropolitan Detroit, Hawaii, Iowa, New Mexico, Seattle (Puget Sound), Utah, Metropolitan Atlanta, Alaska, San Jose-Monterey, Los Angeles, Rural Georgia, Greater California, Kentucky, Louisiana, and New Jersey registries were used.10 Recorded information includes cancer stage and grade at diagnosis for each tumor site, year of diagnosis, initial cancer treatment, vital status, and race/ethnicity. Rigorous quality-control procedures are in place to assure the quality of data being collected.
All women diagnosed with primary ovarian cancer and women with synchronous primary ovarian and endometrial cancers between January 1, 1973, and December 31, 2005, were included in our analyses (n=85,062). Women with single primary ovarian cancer were identified using the International Classification of Diseases for Oncology, Third Edition, code C56.9. For main analyses, we used the SEER definition of synchronous cancers, as cancers of the endometrium (C54.1) and ovary (C56.9) diagnosed within 2 months of each other.11 Women were excluded from the analysis if they were diagnosed with other primary cancers (eg, breast, colon, or cervical), were diagnosed with borderline and nonepithelial ovarian and endometrial cancers (n=19,205), had synchronous ovarian and endometrial cancers diagnosed more than 2 months apart (n=164), were younger than 18 years of age at diagnosis (n=57), or had inconsistencies in histologic classification (n=37). Because our study compared the survival of women with single ovarian cancer with that of women with synchronous cancers, we also excluded 56 women who had two primary epithelial ovarian cancers (eg, tumor on each ovary); among them, there were two synchronous ovarian and endometrial cancer cases. We found identical results when we repeated main analyses including them (data not shown). Two registries (Alaska and Rural Georgia) were excluded because they did not contribute any synchronous cancer cases at their sites.
Demographic information included age at diagnosis (years), race (white, African American, other, and unknown), and marital status (married, unmarried, unknown). Information on vital status (alive compared with dead) as well as ovarian and endometrial tumor characteristics, including histology, grade, stage, receipt of radiation, and surgery, was available. Histologic cell types were categorized for ovarian tumors as endometrioid, all other adenocarcinoma and adenocarcinomas not otherwise specified, papillary (including papillary, papillary serous cystadenocarcinoma, and serous cystadenocarcinoma), clear cell, cystadenocarcinoma, and mucinous cell (including mucinous cystadenocarcinoma and mucinous adenocarcinoma). Endometrial tumor histologies included endometrioid; all other adenocarcinoma, adenocarcinomas not otherwise specified, and adenocarcinoma with squamous metaplasia; clear cell; papillary and papillary serous; mucinous cell (including mucinous and mucin producing); and adenosquamous.
Using SEER grade criteria,11 tumors were classified as grade 1, 2, 3, 4, or unknown. SEER classifies tumors by stage as in situ (noninvasive, confined to epithelial tissue), localized (cancer confined to the ovary/endometrium), regional (cancer extends beyond the ovary/endometrium into surrounding tissues and organs or regional lymph nodes), distant (cancer has spread to sites remote from the ovary/endometrium), or unknown. Neither the American Joint Committee on Cancer nor the tumor-nodes-metastasis classification system was used because their staging information was available only for those women diagnosed from 1988 to 2003. Information on chemotherapy treatment was not available. Central pathology review or tumor banking is not included in the SEER Program. Initial treatment modalities include 1) radiation only, 2) surgery only, 3) surgery and radiation, and 4) type of treatment unknown. Information for women without any type of radiation or surgery also was captured.
χ2 and t-tests were used to evaluate differences in patient and tumor characteristics between single ovarian cancer cases and synchronous cases in age at diagnosis, race, ovarian tumor stage, ovarian tumor grade, and ovarian tumor histology. Receipt of surgical and radiation therapy also was compared between the two groups. Because our main interest was to evaluate the difference in survival between single and synchronous cases, Kaplan-Meier estimates of survival probability were calculated for deaths from ovarian cancer for single compared with synchronous cases and survival curves were compared with the log-rank and Wilcoxon tests. We additionally evaluated deaths from all causes. The Cox proportional hazards model was used to compare survival from synchronous cancer and single ovarian cancer, taking into account age at diagnosis, stage, grade, histology, race, treatment, year of diagnosis, and registry. A similar Cox proportional hazards analysis evaluated risk of ovarian cancer death stratified by race (white, African American, or other) and stage (local, regional, and distant). Our analysis focused on ovarian cancer outcomes because endometrial cancer survival rates tend to be favorable.12 Overall, ovarian cancer survival rates are low because the disease tends to be diagnosed at late stages, whereas endometrial cancer tends to be diagnosed at an earlier stage, resulting in better survival. Five-year relative survival rates for ovarian cancer and endometrial cancer are 53.8% and 84.7%, respectively.12,13
All analyses were performed using the SAS system 9.1 (SAS Institute Inc., Cary, NC).
This study received an exemption from the institutional review boards at Robert Wood Johnson Medical School and University of Medicine and Dentistry of New Jersey.
A total of 55,348 single ovarian cancer cases and 1,355 synchronous ovarian and endometrial cancer cases were identified when using the SEER definition of synchronous cancers (diagnoses within 2 months of each other) and excluding registries that did not have synchronous cases. However, because various definitions regarding the time interval between the detection of the first and second tumors have been used in the published literature, we also estimated the occurrence of synchronous ovarian and endometrial cancers according to these different time frames.
As shown in Table 1, using the least stringent criteria, synchronous cases accounted for approximately 2.7% of the 56,986 primary epithelial ovarian cancers. In the majority of synchronous cases (89.1%) both tumors were detected within 2 months of each other. Patients with synchronous cancers detected within 1 month of each other had similar tumor characteristics to those with synchronous cancers detected within 2 months of each other (data not shown). In contrast, patients with synchronous ovarian and endometrial cancers detected within 2 months displayed different ovarian tumor characteristics from those with synchronous cancers detected 3 or more months apart. Compared with synchronous cancers diagnosed within 2 months, patients with synchronous cancers diagnosed 3 or more months apart were older, more often had ovarian tumors with papillary and adenocarcinoma histologies, less often had ovarian tumors with endometrioid histology, had a higher prevalence of distant ovarian and endometrial tumors, had a higher frequency of ovarian tumors with unknown grade, and were more likely to have both surgery and radiation treatment (data not shown). For the remaining analyses, we used the SEER definition of synchronous cases as patients with endometrial and ovarian tumors detected within 2 months of each other (n=1,355).
Tables 2 and 3 show that women with synchronous cancers were more likely to be diagnosed at an early stage, more likely to exhibit lower ovarian tumor grade, and ovarian tumors were mostly from endometrioid histology. Women with synchronous cancers were also more likely to have endometrial tumors in early stages and lower grades and of endometrioid and adenocarcinoma histology.
Frequency of initial treatment for single compared with synchronous tumors is displayed in Table 4. Patients with single ovarian tumors were less likely than those with synchronous tumors to receive surgery followed by radiation therapy. The frequency of not receiving any treatment was higher for patients with single ovarian cancers than for those with synchronous cancers.
Differences in survival between women with synchronous and single ovarian cancers were estimated by the Kaplan-Meier method. Women with synchronous cancers had better survival than those with single ovarian cancer (log-rank P<.001, Wilcoxon P<.001). The 5-year survival probability for synchronous cancers is approximately 79% compared with 42% for those with single primary ovarian cancers. All-cause 5-year survival rates reflect similar results (results not shown).
Age-adjusted, age-and-stage-adjusted, and multivariable hazard ratios with 95% confidence intervals (CIs) for ovarian cancer deaths comparing synchronous cancers with single primary ovarian cancers are shown in Table 5. Age-adjusted hazard ratios showed a 67% reduced risk of death from ovarian cancer for synchronous cancers compared with single ovarian cancers (hazard ratio 0.33, 95% CI 0.29–0.37). However, after additionally adjusting for stage, the risk reduction for synchronous cancers was reduced to 53% (hazard ratio 0.47, 95% CI 0.42–0.53). The risk reduction in the multivariable model was further reduced after adjusting for additional prognostic and treatment characteristics, year of diagnosis, and registry (hazard ratio 0.75, 95% CI 0.66–0.85). Our stratified analysis by race revealed similar magnitude of effects. The survival advantage associated with having synchronous ovarian and endometrial cancers was not statistically significant because of small numbers of African Americans and women of other races.
Because synchronous ovarian and endometrial cancers more often are diagnosed at an early stage than single ovarian primary cancers, we estimated survival after stratifying by stage (Table 5). We found that women with synchronous cancers experienced much better survival than those with single ovarian cancers in each stage strata. The survival benefit associated with having a synchronous cancer persisted even when analyses were restricted to distant-stage cases (hazard ratio 0.70, 95% CI 0.60–0.81).
To evaluate any potential effect of changes in the classification system or treatment during the study period, we performed multivariable Cox-proportional hazards models by year of diagnosis for those diagnosed between 1990 and 1999 and between 2000 and 2005. Both timeframes reflected similar results as the overall study population results (hazard ratio 0.75, 95% CI 0.63–0.89 and hazard ratio 0.74, 95% CI 0.60–0.89, respectively). Only 14 synchronous cases were reported from 1973 to 1989, and, therefore, we did not conduct stratified analyses in that time period.
Our large population-based study found that synchronous ovarian and endometrial cancers, regardless of time between detection of both tumors, represented less than 3% of all ovarian cancer cases. Similarly, van Niekerk et al,14 using the Netherlands Cancer Registry, found that uterine or endometrial cancers detected within the same year of ovarian cancer diagnosis accounted for 2.6% of the 5,366 patients diagnosed with ovarian cancer. In contrast, other studies have estimated that approximately 10% of ovarian cancer patients had coexisting ovarian and endometrial cancers.5 The discrepancy in the estimate of these synchronous cancers might be a result of differences in study population in the different studies. Most studies reporting higher frequencies of synchronous cancers were hospital-based with fewer than 100 patients.1,3,5,7,8,15 Moreover, prior studies have used various criteria to distinguish between synchronous primary cancers and a single primary cancer with metastasis.
Our study provided further evidence that synchronous ovarian and endometrial cancers exhibit favorable survival outcomes as compared with single primary ovarian cancers. This is explained mostly by favorable characteristics associated with synchronous tumors, including younger age at diagnosis, earlier stage of disease, and better grade of disease. Eifel et al15 found that patients with synchronous ovarian and endometrial tumors of endometrioid histology had better prognoses compared with those with synchronous tumors of papillary, clear cell, mucinous, or mixed cell type. In our study, synchronous patients more often had ovarian tumors with endometrioid histology, which tend to have better outcomes, with a 5-year survival rate of about 71%.13 In addition, among women with synchronous cancers, 70% had endometrial tumors diagnosed at a localized stage and 75% were grade 1 or 2.
Better survival among patients with synchronous cancers might be due to a higher likelihood of having the disease detected at an earlier stage compared with those with exclusively primary ovarian cancers. In fact, at least 75% of endometrial cancers are detected in earlier, treatable stages because women notice and report irregular bleeding.16 Thus, synchronous cases may have a lead-time advantage over single ovarian cases because endometrial tumors more often present gynecologic-related symptoms sooner, prompting clinical examination and treatment.
We also found that women with synchronous cancers had better survival independent of stage. In other words, women with distant disease still had better survival if they had a synchronous diagnosis, rather than a single primary ovarian cancer, even after adjusting for other prognostic factors. This finding, together with the Ramus et al report,17 suggests that synchronous ovarian and endometrial cancers may have different genetic profiles than single primary ovarian cancers and that the two entities may represent two distinct biological diseases.
Survival rates have been shown to be lower for single ovarian13 and single endometrial cancers12 among African Americans compared with whites. Our study found that the survival advantage associated with having synchronous cancers was less evident among African-American women. However, owing to limited statistical power, we cannot provide a definite answer to this issue because only 43 synchronous cancers occurred among African Americans.
We found that, among women who received both surgery and radiation as initial treatment, those with synchronous cancers were more likely to receive both types of treatment than were women with single ovarian cancers, possibly related to their presentation at earlier stage.
The overall survival advantage of having synchronous cancers suggests that survival also may be affected by other related factors not captured in our study, such as genetic susceptibility, or other treatment differences, such as chemotherapy. Other factors that were not included are family history, income, education, and health insurance, as well as recurrences or comorbidities that may affect survival. Future studies should address these issues because information on these factors is not available in the SEER database.
In conclusion, although synchronous cancers are rare (probably even less common than previously reported), their identification and optimal clinical management remain important issues in the battle to reduce ovarian cancer mortality. A national registry documenting these cases, ideally with central pathology review and specimen collection, would help in achieving this goal. The possibility of synchronous ovarian and endometrial cancers being a distinct biologic entity separate from single ovarian cancer is intriguing and requires further study.
1. Chiang YC, Chen CA, Huang CY, Hsieh CY, Cheng WF. Synchronous primary cancers of the endometrium and ovary. Int J Gynecol Cancer 2008;18:159–64.
2. Coli A, Dell'Isola C, Bigotti G. Synchronous tumors: a proposal for standardization. Tumori 2002;88:433–4.
3. Ulbright TM, Roth LM. Metastatic and independent cancers of the endometrium and ovary: a clinicopathologic study of 34 cases. Hum Pathol 1985;16:28–34.
4. Scully R, Young R, Clement P. Tumors of the ovary, maldeveloped gonads, fallopian tube, and broad ligament. Atlas of Tumor Pathology. Washington (DC): Armed Forces Institute of Pathology; 1998. Tables 5-1–5-3.
5. Zaino R, Whitney C, Brady MF, DeGeest K, Burger RA, Buller RE. Simultaneously detected endometrial and ovarian carcinomas-a prospective clinicopathologic study of 74 cases: a gynecologic oncology group study. Gynecol Oncol 2001;83:355–62.
6. Sheu BC, Lin HH, Chen CK, Chao KH, Shun CT, Huang SC. Synchronous primary carcinomas of the endometrium and ovary. Int J Gynaecol Obstet 1995;51:141–6.
7. Soliman PT, Slomovitz BM, Broaddus RR, Sun CC, Oh JC, Eifel PJ, et al. Synchronous primary cancers of the endometrium and ovary: a single institution review of 84 cases. Gynecol Oncol 2004;94:456–62.
8. Kondi-Pafiti A, Grapsa D, Liapis A, Papadias K, Kairi-Vassilatou E, Hasiakos D. Synchronous ovarian and endometrial carcinoma: a strong link to endometriosis? Eur J Gynaecol Oncol 2008;29:256–9.
9. Howe HL, Weinstein R, Hotes J, Kohler B, Roffers SD, Goodman MT. Multiple primary cancers of the ovary in the United States, 1992–1997. Cancer 2003;97:2660–75.
10. Surveillance, Epidemiology, and End Results (SEER) Program. Limited-Use Data (1973–2005). DCCPS, Surveillance Research Program, Cancer Statistics Branch. Bethesda (MD): National Cancer Institute; 2008.
11. Johnson CH. SEER program coding and staging manual 2004, revision 1. Bethesda (MD): National Cancer Institute, NIH Pub. No. 04-5581;2004.
12. Kosary CL. Cancer of the corpus uteri. In: Ries LAG YJ, Keel GE, Eisner MP, Lin YD, Horner M-J, editors. SEER survival monograph: cancer survival among adults. US SEER Program, 1988–2001, patient and tumor characteristics. Bethesda (MD): National Cancer Institute, NIH Pub. No. 07-6215;2007. p. 123–32.
13. Kosary CL. Cancer of the ovary. In: Ries LAG YJ, Keel GE, Eisner MP, Lin YD, Horner M-J, editors. SEER survival monograph: cancer survival among adults. US SEER Program, 1988–2001, patient and tumor characteristics. Bethesda (MD): National Cancer Institute, NIH Pub. No. 07-6215; 2007. p. 133–44.
14. van Niekerk CC, Vooijs GP, Bulten J, van Dijck JA, Verbeek AL. Increased risk of concurrent primary malignancies in patients diagnosed with a primary malignant epithelial ovarian tumor. Mod Pathol 2007;20:384–8.
15. Eifel P, Hendrickson M, Ross J, Ballon S, Martinez A, Kempson R. Simultaneous presentation of carcinoma involving the ovary and the uterine corpus. Cancer 1982;50:163–70.
16. Rahaman J, Cohen CJ. Endometrial cancer. In: Kufe DW, Bast RC Jr, Hait WN, Hong WK, Pollock RE, Weichselbaum RR, et al, editors. Cancer medicine. 7th ed. Hamilton (ON): BC Decker; 2006. p. 1522–36.
17. Ramus SJ, Elmasry K, Luo Z, Gammerman A, Lu K, Ayhan A, et al. Predicting clinical outcome in patients diagnosed with synchronous ovarian and endometrial cancer. Clin Cancer Res 2008;14:5840–8.
© 2009 by The American College of Obstetricians and Gynecologists.