Objectives
To explore the differences between women with endometiosis associated ovarian cancer and typical epithelial ovarian cancer.
Study Design
The medical charts of total 226 patients with epithelial ovarian cancer treated at Peking Union Medical College Hospital between March 2011 and March 2012 were reviewed. Histology evaluation determined endometiosis associated ovarian cancer (n = 17) or nonendometiosis associated ovarian cancer (n = 209).
Results
Compared with nonendometiosis associated ovarian cancer, patients with endometiosis associated ovarian cancer were proved: (1) to be younger and more likely to be premenopausal at diagnosis of epithelial ovarian cancer ( P = .03 and .005, respectively); (2) to have lower preoperative serum level of Ca125 (mean: 122.9 vs 1377.5 U/mL, P < .001) and more likely to display normal Ca125 level ( P < .001); (3) to be identified at the earlier stage (stage I, P < .001); (4) to have completely different distribution of histological subtypes (significant overrepresentation of clear cell and endometrioid carcinoma).
Conclusion
As such, patients with endometiosis associated ovarian cancer differ from nonendometiosis associated ovarian cancer in many of their critical clinical and biologic characteristics.
Endometriosis is the common gynecologic disorder characterized by ectopic growth of endometrial glands and stroma. Although endometriosis is benign, its destructive, invasive, and metastatic nature mimics presentations of malignant tumors. It is especially notable that endometriosis progresses to biologic malignant tumor, often in ovaries. As early as 1925, the father of endometriosis, Sampson had described the association between endometriosis and ovarian carcinoma. Scott further defined the endometriosis-associated ovarian carcinoma (EAOC). In previous studies about the relationship between these 2 disorders, the epithelial ovarian cancers (EOC) were mostly analyzed as a single entity. However, the studies in recent years revealed that the variable histology subtypes of ovarian cancer, eg, low-grade serous (LGS), high-grade serous (HGS), clear cell, endometrioid, and mucinous subtypes, have significant heterogeneity in both clinical characteristics and outcomes. Several reports demonstrated that the prevalence of associated endometriosis significantly differed among the variable histology subtypes of ovarian cancers, and furthermore suggested that concurrent endometriosis might be the good prognostic factor for specific subtypes of EOC. However, there were still divergent, and even contrary, results reported. We therefore, retrospectively, analyzed the cases of EOC received primary cytoreductive surgery (CRS) and adjuvant chemotherapy treatment at Peking Union Medical College Hospital, a tertiary teaching hospital of China, and aimed to explore the association between EOC and endometriosis based on our data.
See Journal Club, page 415
Materials and Methods
We reviewed the medical charts of patients with EOC who were primarily treated and received complete surgical staging or CRS at the Division of Gynecological Oncology of the Department of Obstetrics and Gynecology, in Peking Union Medical College Hospital, Peking Union Medical College, China. This study included 226 patients who were admitted between March 2011 and March 2012, and obtained the University Institutional Review Board approval. All patients with tumors at stage I and II have received completed staging surgery; and women with advanced cancer (stage III and IV) have undergone optimal CRS, except for those with unresectable tumors who received suboptimal CRS. They all received adjuvant chemotherapy after primary surgery.
In this study, “endometriosis-associated ovarian carcinoma (EAOC)” was defined as follows: (1) presence of ovarian cancer and endometriosis identified histologically in the same ovary; (2) presence of endometriosis in 1 ovary and that of ovarian cancer in the contralateral ovary; or (3) the presence of ovarian cancer and extraovarian pelvic endometriosis (eg, peritoneal endometriosis). Endometriosis was identified when the tissue resembling endometrial stroma surrounding epithelial glands presented in ovaries or peritoneum.
According to the pathologic criteria listed above, we identified 17 of the 226 patients as having endometriosis (EAOC group). The remaining 209 patients had no pathologic evidence of endometriosis (non-EAOC group).
For statistical analysis, International Federation of Gynecology and Obstetrics (FIGO) stage categories were classified into early stage (FIGO stages I subjects) and late stage (FIGO stages II to IV). We also analyzed the age and menopausal status at diagnosis of ovarian cancer, the serum level of preoperative cancer antigen 125 (Ca125), and the distribution of variable histologic subtype of ovarian cancer, respectively, in 2 groups. Based on a dualistic model of carcinogenesis and Kurman and Shih’s classification, we divided EOC of our cases into 2 groups (type I and type II). Type I tumors comprise LGS cancer (LGSC), endometrioid, clear cell, mucinous, and transitional cell carcinomas. Type II tumors comprise HGS cancer (HGSC), malignant mixed mesodermal tumors (MMMT), and undifferentiated carcinomas.
Statistical analysis for clinicopathologic variables between 2 groups was performed with χ 2 or Fisher exact tests as indicated. All P values reported are 2 tailed and P value of .05 or less was considered to be statistically significant.
Results
In this entire study population, the mean age of the 226 patients was 52.2 ± 12.2 years (range, 18–79 years). The clinical and pathologic characteristics collected are listed in Table 1 . The FIGO stage distribution was as follows: FIGO I (21.2%, n = 48), FIGO II (8.4%, n = 19), FIGO III (63.8%, n = 144), and IV (6.6%, n = 15). The distribution of histologic subtypes was sequentially as follows: HGSC (59.7%, n = 135), clear cell (9.7%, n = 22), LGSC (8.0%, n = 18), endometrioid (7.5%, n = 17), mucinous (7.5%, n = 17), transitional cell (3.6%, n = 8), undifferentiated subtype (3.1%, n = 7), and MMMT (0.9%, n = 2).
| Characteristics | Number | Percentage |
|---|---|---|
| Total | 226 | |
| Age, y | ||
| Mean ± SD | 52.2 ± 12.2 | |
| Range | 18-79 | |
| FIGO stage | ||
| I | 48 | 21.2 |
| II | 19 | 8.4 |
| III | 144 | 63.8 |
| IV | 15 | 6.6 |
| Histology | ||
| LGSC | 18 | 8.0 |
| HGSC | 135 | 59.7 |
| Endometrioid | 17 | 7.5 |
| Clear cell | 22 | 9.7 |
| Transitional cell | 8 | 3.6 |
According to the criteria mentioned above, 17 (7.5%) were EAOC, whereas 209 (92.5%) patients were non-EAOC. As seen in Table 2 , the patients with EAOC were about 6 years younger at diagnosis of cancer compared with those with non-EAOC (mean age at diagnosis, 46.1 vs 52.8 years, P = .03). The former were more likely at premenopausal status than the latter (76.5% vs 40.2%, P = .005). All 226 patients’ serum Ca125 were measured preoperatively. Our results showed that serum level of preoperative Ca125 in EAOC group was significantly lower and more likely in normal range, when compared with non-EAOC goup (mean level of Ca125, 122.9 U/L vs 1377.5 U/L, P < .001; with normal Ca125, 47.1% vs 10.0%, P < .001).
| Variable | EAOC, % | Non-EAOC, % | P values |
|---|---|---|---|
| No. | 17 | 209 | |
| Age, y | |||
| Mean ± SD | 46.1 ± 10.1 | 52.8 ± 12.2 | .030 |
| Range | 33-66 | 18-79 | |
| Premenopause at the time of diagnosis | 13 (76.5%) | 84 (40.2%) | .005 |
| Ca125 before surgery, U/L | |||
| Mean ± SD | 122.9 ± 179.5 | 1377.5 ± 2496.1 | < .001 |
| Range | 6–573.4 | 3.3–24,183 | |
| With normal Ca125 | 8 (47.1%) | 21 (10.0%) | < .001 |
| FIGO stage, n (%) | |||
| I | 15 (88.2) | 33 (15.8) | |
| II | 2 (11.8) | 17 (8.1) | |
| III | 0 (0) | 144 (68.9) | |
| IV | 0 (0) | 15 (7.2) | |
| Stage comparison | |||
| Early stage (I) | 15 (88.2%) | 33 (15.8%) | < .001 |
| Late stage (II-IV) | 2 (11.8%) | 176 (84.2%) | |
| Histology, n (%) | |||
| LGSC | 1 (5.9) | 17 (8.1) | .599 |
| HGSC | 2 (11.7) | 133 (63.7) | < .001 |
| Clear cell | 8 (47.1) | 14 (6.7) | < .001 |
| Endometrioid | 6 (35.3) | 11 (5.3) | .001 |
| Mucinous | 0 (0) | 17 (8.1) | .461 |
| Transitional cell | 0 (0) | 8 (3.8) | .890 |
| Others | 0 (0) | 9 (4.3) | .819 |
| Type I or II | |||
| Type I (n = 82) | 15 (88.2%) | 67 (32.1%) | < .001 |
| Type II (n = 144) | 2 (11.8%) | 142 (67.9%) |
The FIGO stage distribution was different between 2 groups ( P < .001), 88.2% of subjects in EAOC group were diagnosed at early stage (stage I), in contrast to 15.8% of subjects in non-EAOC group.
The histology distribution in 2 groups was totally different from each other ( P < .001). The relative distribution of various histology subtypes in EAOC group was clear cell (47.1%), endometrioid (35.3%), HGSC (11.7%), and LGSC (5.9%). In contrast, the histology distribution in non-EAOC group was HGSC (63.7%), LGSC (8.1%), mucinous (8.1%), clear cell (6.7%), endometrioid (5.3%), undifferentiated and MMMT (4.3%), and transitional cell (3.8%).
When these 226 cases were reclassified according to Kurman and Shih’s classification (as described in Materials and Methods ), 82 were type I and 144 were type II tumors. In EAOC group, 88.2% of patients had type I tumors, compared with 32.1% in non-EAOC group ( P < .001). In another word, type I of EOC showed higher possibility of coexisting with endometriosis, when compared with type II tumor (18.3% vs 1.4%, P < .001).
As mentioned before, the entire study cohort was stratified into early and late stage. We specifically analyzed the features of early stage tumors with or without endometriosis (n = 15 and 33, respectively, as shown in Table 3 ). There was no significant difference between the 2 groups with regard to age at present ( P = .772), preoperative level of Ca125 ( P = .314), or the patients with normal Ca125 ( P = .367). The distribution of histology subtypes in 2 groups was completely divergent from each other ( P < .001). In decreasing order of prevalence, the histology subtypes in EAOC were clear cell (46.6%), endometrioid (40%), HGSC (6.7%), and LGSC (6.7%); whereas, which distribution in non-EAOC group was mucinous (36.3%), LGSC (21.2%), clear cell (18.2%), HGSC (9.1%), endometrioid (9.1%), and transitional cell subtype (6.1%). However, the prevalence of individual histology subtype between 2 groups showed no statistically significant difference, except for mucinous and endometrioid subtype.
| Variable | EAOC, % | Non-EAOC, % | P values |
|---|---|---|---|
| No. of early stage (stage I) | 15 | 33 | |
| Age, y | |||
| Mean ± SD | 46.9 ± 10.6 | 45.7 ± 16.1 | .772 |
| Range | 33-66 | 18-71 | |
| Ca125 before surgery, U/L | |||
| Mean ± SD | 133.8 ± 189.1 | 277.4 ± 529.0 | .314 |
| Range | 6-573.4 | 8.4-2517.85 | |
| With normal Ca125, n (%) | 8 (53.3) | 13 (39.4) | .367 |
| Histology, n (%) | < .001 | ||
| LGSC | 1 (6.7) | 7 (21.2) | .406 |
| HGSC | 1 (6.7) | 3 (9.1) | .99 |
| Clear cell | 7 (46.6) | 6 (18.2) | .088 |
| Endometrioid | 6 (40) | 3 (9.1) | .032 |
| Mucinous | 0 (0) | 12 (36.3) | .019 |
| Transitional cell | 0 (0) | 2 (6.1) | .846 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
