Endometrial stromal sarcomas are rare uterine malignancies. Cancers involving the uterine corpus are the fourth most common cancer in women. Uterine sarcomas make up 2–6% of the 40,100 cases of uterine malignancies projected to occur in the United States in 2008. Endometrial stromal sarcomas will account for less than 10% of all uterine sarcomas encountered during that year.1,2
Endometrial stromal sarcoma histologically resemble stromal cells found during the proliferative phase of the menstrual cycle. The tumor has a characteristic pattern of infiltration and a network of branching small vessels resembling spiral arteries.3 They have traditionally been classified as either low-grade endometrial stromal sarcoma or high-grade endometrial stromal sarcoma based on nuclear size, degree of atypia, and mitotic activity.4 Because of the abundance of neoplastic cells demonstrating little evidence of endometrial stromal differentiation, the most recent World Health Organization classification suggested that high-grade endometrial stromal sarcoma may be better classified as undifferentiated uterine sarcomas.5,6
The rarity of endometrial stromal sarcoma, the historical aggregation of various types of uterine sarcomas in prior reports, as well as the relatively recent distinction between undifferentiated uterine sarcomas from low-grade endometrial stromal sarcoma has left a paucity of data to formulate clinical decisions.
Since the publication of Gynecologic Oncology Group-33,7 the prognostic importance of extrauterine spread, particularly lymph node metastasis, has resulted in the recommendation for intraoperative assessment of pelvic and aortic lymph nodes in patients with endometrial carcinomas. Although the precise stratification of patients based on risk of lymph node metastasis is debated, this approach has gained acceptance among United States gynecologic oncologists (Hernandez E. ACOG Practice Bulletin #65: management of endometrial cancer [letter]. Obstet Gynecol 2006;107:952). Lymphatic involvement by endometrial stromal sarcoma tumors is well established, as evidenced by a prior pathologic designation of endolymphatic stromal myosis.8 The role of surgical staging for the treatment of low-grade endometrial stromal sarcoma remains unclear. Because endometrial stromal sarcoma is difficult to diagnose on a curettage specimen alone, it is often diagnosed postoperatively on hysterectomy specimens from patients initially thought to have benign leiomyoma or stromal nodules. The lack of information regarding the propensity of low-grade endometrial stromal sarcoma to metastasize to lymph nodes limits prognostic and therapeutic discussions.
Endometrial stromal sarcoma have a high proportion of estrogen and progesterone receptors. This has led some authors to conclude that recurrences are more frequent when ovaries are conserved,9,10 whereas others have suggested that if normal in appearance, conservation of ovaries during hysterectomy does not lead to a poorer outcome.11–13 Because low-grade endometrial stromal sarcoma commonly occur in premenopausal women, the need for bilateral salpingo-oophorectomy (BSO) requires further investigation.
This population-based study reiterates the important clinical distinction between high-grade endometrial stromal sarcoma or undifferentiated sarcoma and low-grade endometrial stromal sarcoma. The objectives of this analysis were to report among low-grade endometrial stromal sarcoma the incidence of lymph node metastasis and analyze the effect of lymph node dissection, lymph node metastasis, and ovarian preservation on survival.
MATERIALS AND METHODS
Demographic, clinicopathologic, and survival information from women diagnosed with endometrial stromal sarcoma (International Classification of Diseases code 8930) from January 1, 1988, to December 31, 2005, were extracted from the Surveillance, Epidemiology and End Results (SEER) database of the U.S. National Cancer Institute. The SEER program collects information at multiple regional sites about all incident cancer cases, including patient demographics, tumor site, stage at diagnosis, surgical procedure performed, date of diagnosis, radiotherapy performed, and annual follow-up (SEER web site: seer.cancer.gov/about/).
All patients with endometrial stromal sarcoma who were actively followed were extracted in tabular format in the “Case Listing” option of the SEER*Stat software and were analyzed using SPSS 15.0 (SPSS Inc., Chicago, IL). Exclusion criteria were patients with unknown age, presence of a more than one malignancy, nonuterine primary site of tumor, or diagnosis by autopsy or death certificate. Data recorded included age at diagnosis, race, grade, stage, surgery, lymph node involvement, date of diagnosis, radiation treatment, and overall survival.
Grade when reported was divided into low-grade endometrial stromal sarcoma, which included well- and moderately differentiated, and high-grade endometrial stromal sarcoma, which included poorly or undifferentiated/anaplastic. Race was divided into white, blacks, and others. For multivariable analysis age was divided into age groups younger than 44 years, 45–54 years, and older than 55 years to correspond to premenopausal, perimenopausal, and postmenopausal status. Stage was determined according to International Federation of Gynecology and Obstetrics (FIGO) criteria without taking into account lymph node status and then divided into disease confined to the uterus (FIGO stage I and II) and extrauterine disease (FIGO stage III and IV). Surgical procedure data were derived from site-specific surgery codes and divided into hysterectomy with BSO and hysterectomy without BSO. In some patients, ovarian status was not addressed, and these patients were labeled not specified. Lymph node dissection performed, lymph node metastasis, and extent of lymphadenectomy were derived from pathology codes.
Comparisons of characteristics between low-grade and high-grade endometrial stromal sarcoma used χ2, Pearson coefficient, or Fisher exact test and parametric t tests. Survival probability data were computed using Kaplan-Meier method and differences between groups were calculated using the log rank test. Cox proportional hazards model in a forced entry was first used to screen for variables likely to affect survival. Subsequently, a forward step-wise method (conditional logistic regression) was used to determine the independent effect of lymph node dissection performed, lymph node involvement, ovarian preservation, and radiotherapy on survival, after controlling for other potential confounding variables such as age of diagnosis, race, and stage. All P values reported are two-tailed, and a P value of less than or equal to 0.05 was considered to be statistically significant. This study was exempt from the Institutional Review Board of Wayne State University.
From 1988 to 2005, 975 patients were diagnosed with endometrial stromal sarcoma of the uterus; 127 patients had more than one malignancy and were excluded from analysis. Of the remaining 848 patients with endometrial stromal sarcoma, 384 patients had low-grade endometrial stromal sarcoma, 320 patients had high-grade endometrial stromal sarcoma, and in 144 patients grade was not specified. Comparison between demographic and clinical features of the 704 patients with low-grade endometrial stromal sarcoma and high-grade endometrial stromal sarcoma are shown in Table 1.
The mean age of patients was 55.2 years (range 17–96 years). Low-grade endometrial stromal sarcoma patients were younger than those with high-grade endometrial stromal sarcoma. The mean age of patients with low-grade endometrial stromal sarcoma compared with high-grade endometrial stromal sarcoma patients was 48.5 years and 60.5 years, respectively. (P<.001) The majority of patients in both low-grade endometrial stromal sarcoma and high-grade endometrial stromal sarcoma were white (74.5%). However blacks were more likely to have high-grade endometrial stromal sarcoma than low-grade endometrial stromal sarcoma (P=.014) and subsequently a shorter survival.
A much higher percentage of patients with high-grade endometrial stromal sarcoma (52.5%) compared with low-grade endometrial stromal sarcoma (25%) had advanced stage (FIGO III or IV) disease (P<.001) (Fig. 1). The survival of patients with disease extending beyond the uterus was worse than those with disease confined to the uterus. This was significant in both high-grade and low-grade lesions. Of the 356 patients with stage-defined low-grade endometrial stromal sarcoma, 267 (75%) had stage I or II (uterus/cervix confined) and 89(25%) had stage III or IV (extrauterine disease), with corresponding 5-year survival rates of 96% and 81%, respectively (P<.001) (Fig. 2A).
Either systematic lymphadenectomy or lymph node sampling was performed in only 38% (287 of 748) of patients with endometrial stromal sarcoma. Lymph node metastasis was higher in high-grade than low-grade endometrial stromal sarcoma, 18% and 7% respectively, (P=.013). Lymph node dissection was more frequently performed in high-grade (45.3%) compared with low-grade endometrial stromal sarcoma (26.1%) (P<.001). However, in both high-grade (n=126) and low-grade (n=91) endometrial stromal sarcoma, the extent of lymph nodes resected (mean number 12) was similar (Table 2).
Of the 384 low-grade endometrial stromal sarcoma patients, 100 patients had lymph node dissection performed. The age distribution for these low-grade endometrial stromal sarcoma patients who underwent lymph node dissection were 46% with age younger than 44 years, 41% between ages 45–54 years, and 28.6% older than age 55 years (P=.133). Of those patients with low-grade endometrial stromal sarcoma and disease confined to the uterus, 23.6% (63 of 266) of patients underwent lymph node dissection, and overall, the distribution of stages was not statistically different between those who underwent lymph node dissection and those who did not (P=.264). The performance of lymph node dissection alone, regardless of the finding of any microscopic positive nodes, did not have an effect on overall survival in low-grade endometrial stromal sarcoma. However, in high-grade endometrial stromal sarcoma patients who had lymph node dissection (n=143), those with lymph node metastasis (n=26) had a worse survival than those without lymph node metastasis, with a median survival time of 8 months and 24 months, respectively (P<.001). Unlike high-grade endometrial stromal sarcoma, the presence of lymph node metastasis in low-grade endometrial stromal sarcoma did not portend a poorer survival; lymph node-positive 5-year survival rate was 85.7%, compared with the lymph node-negative 5-year survival rate of 95.2%, (P=.234) (Fig 2B).
The presence or absence of bilateral salpingo-oophorectomy at the time of hysterectomy was ascertained in 304 of the 384 patients with low-grade endometrial stromal sarcoma. Of these patients, only 15.1% (n=46) had their ovaries preserved at the time of hysterectomy. Patients with low-grade endometrial stromal sarcoma with ovarian preservation had extrauterine disease in 18.6% of cases. The majority of patients in both groups, those that underwent BSO and those that had ovarian preservation, were premenopausal, 76% and 91.3%, respectively. There was no statistically significant difference in survival among low-grade endometrial stromal sarcoma patients who underwent BSO compared with those that did not (median survival not yet reached, 5-year survival 92% compared with 94%, P=.267). Particularly in stage I low-grade endometrial stromal sarcoma, where all patients had an excellent survival time of more than 190 months, there was no difference between those that had ovaries removed and those with ovarian preservation (P=.672) (Fig. 2C). Also, among the 199 high-grade endometrial stromal sarcoma patients who had documentation of ovarian preservation or removal, only 11 patients had their ovaries retained. No difference in survival was noted between high-grade endometrial stromal sarcoma patients who underwent BSO and those who underwent ovarian preservation surgery (P=.279).
In our multivariable analysis, high grade was the most significant independent prognostic factor (hazard ratio 8.7, 95% confidence interval 5.9–12.6; P<.001). Among low-grade endometrial stromal sarcoma patients, multivariable analysis revealed younger age (P=.043), white race (P=.001), and earlier stage (P<.001) were independent prognostic factors for improved survival (Table 3). In this cohort of low-grade endometrial stromal sarcoma patients, other prognostic factors, including year of diagnosis, performance of salpingo-oophorectomy, adjuvant radiotherapy, performance of lymphadenectomy, and the presence of lymph node metastasis, were not significant.
In this retrospective population-based study, low-grade endometrial stromal sarcoma have an overall excellent prognosis. The traditional division of low-grade endometrial stromal sarcoma and high-grade endometrial stromal sarcoma based on pathologic findings categorize clinically distinct entities.14 Although some pathologists have found mitotic count to be prognostic,15 others have shown mitotic count is less important than stage.16 This study revealed the markedly shorter survival time among patients with high-grade endometrial stromal sarcoma than with low-grade endometrial stromal sarcoma and lends further support to the reclassification of high-grade endometrial stromal sarcoma into its own category of undifferentiated uterine sarcomas, not endometrial stromal sarcoma.15,17–20 Similarly, the term endometrial stromal sarcoma may more accurately represent what previously has been reported as low-grade endometrial stromal sarcoma or endolymphatic stromal myosis.19
Treatment of endometrial stromal sarcoma is primarily surgical, with several authors commenting on the correlation between extent of surgery and recurrence or metastasis.8,9,14,21–23 A characteristic feature of these tumors is spread into blood and lymphatic vessels.18,24,25 Recurrences develop in more than one third of patients, with hematogenous spread common, with the most frequent distant site being the lungs.14,26,27
Since uterine sarcomas are rare malignancies, the prevalence and prognostic significance of lymphadenectomy and lymph node dissection remains unclear. Riopel et al22 reported the incidence of lymph node metastasis in low-grade endometrial stromal sarcoma to be higher than expected at 33% (5 of 15). Reich et al21 reported lymph node metastasis in 45% (5 of 11) of endometrial stromal sarcoma patients who underwent lymphadenectomy either initially or at the time of recurrence. Recently, Amant et al8 reported on 34 patients with endometrial stromal sarcoma. Among the six patients who underwent lymphadenectomy, only one patient had grossly obvious lymph node involvement. Studies of uterine sarcomas commonly combine not only low-grade endometrial stromal sarcoma and high-grade endometrial stromal sarcoma but also different tumor types, ranging from carcinosarcoma to leiomyosarcoma.24,28–30 The combination of grades and tumor types as well as their small sample size limit the interpretation of these reports.
The current study with 384 low-grade endometrial stromal sarcoma patients represented a comparatively large number of patients. Twenty-six percent of low-grade endometrial stromal sarcoma patients (100 of 383) had lymphadenectomy. The incidence of lymph node metastasis among low-grade endometrial stromal sarcoma was 7% (7 of 100) (Table 2). Unlike Goff et al,10 who reported no lymph node metastasis in 10 endometrial stromal sarcoma (five high grade and five low grade) patients without extrauterine disease, in our patient population three of the seven patients with lymph node metastasis had disease confined to the uterus. This information may be particularly valuable when counseling patients considering a reexploration/staging procedure after serendipitously finding low-grade endometrial stromal sarcoma after hysterectomy for suspected benign reasons such as leiomyomata.
Compared with other more common uterine sarcomas, such as carcinosarcoma or leiomyosarcoma, low-grade endometrial stromal sarcoma seems to have the highest incidence of estrogen and progesterone receptors.31,32 The hormonally sensitive nature of these tumors has led to the hypothesis of an increase in recurrence and poorer overall outcome in patients who underwent surgery without oophorectomy.33,34 Berchuck et al9 reported on 19 unstaged patients with uterine-confined endolymphatic stromal myosis and found that patients who underwent ovarian preservation had a higher rate of recurrence than those that had their ovaries removed, 100% (6 of 6) compared with 43% (6 of 13), respectively. Based on the hormonal sensitivity of endometrial stromal sarcoma and risk for recurrence, some authors have advocated considering reexploration for removal of ovaries.9,10
Conversely, Li et al12 performed a case-control study of 36 patients with low-grade endometrial stromal sarcoma and found that ovarian preservation had no effect on recurrences or overall survival.12 Similarly, other investigators have advocated consideration for ovarian preservation in young women, unable to find an improvement in outcomes in premenopausal women who underwent BSO.8 All these studies are limited by a small sample size that prevents any multivariable analysis for confounding factors. In our study, the effect of ovarian preservation did not affect survival (Fig. 2C), with both groups having an excellent overall 5-year survival rate of more than 90%. Although it is possible that a patient may have had a BSO before hysterectomy, the chance of this is small. Considering the adverse effects of early surgical menopause, it seems appropriate, especially in young premenopausal women, to consider ovarian preservation as an option in women with low-grade endometrial stromal sarcoma.
The major limitations of this study include a lack of information on adjuvant chemotherapy or hormonal therapy, central pathology review, the time and site of recurrence, location of lymph nodes (pelvic or aortic), surgeon subspecialty, and surgical morbidity. Despite this sample size, detection of a difference of 20% in survival among low-grade endometrial stromal sarcoma patients with lymph node metastasis requires more than 2,100 patients. However, among high-grade endometrial stromal sarcomas, a power of 99% was achieved. Previous reports support the reliability of the SEER program in reporting surgical procedures.35 Compared with prior single-institutional studies, the results of the present study may have better external validity, because the SEER program currently covers about 26% of the total population of the United States.36 Because our cohorts are extracted after 1988, we believe the understanding of surgical staging and approach to adjuvant therapy, particularly hormonal responsiveness, may have been appreciated during this time period.
As commonly postulated in discussions of surgical staging in endometrial adenocarcinoma, the attitude that one must approach treatment with knowledge of extent of disease may be critically important in cancer treatment planning. (Orr JW Jr, Naumann WR, Escobar P. “Attitude is a little thing that makes a big difference” Winston Churchill [letter]. Gynecol Oncol 2008;109:147–51). This may still be applicable in low-grade endometrial stromal sarcoma. In light of the lack of effect on survival of lymph node metastasis and salpingo-oophorectomy, the vigor for comprehensive surgical staging must be tempered as always by the potential complications of lymphadenectomy and surgical castration.
In summary, our study confirmed the worse prognosis for high-grade endometrial stromal sarcoma compared with low-grade endometrial stromal sarcoma patients. Specifically in low-grade endometrial stromal sarcoma patients, we found that lymph node metastasis does occur, and although surgical staging may be performed, lymph node metastasis and ovarian preservation do not have a statistically significant effect on survival.
1. Echt G, Jepson J, Steel J, Langholz B, Luxton G, Hernandez W, et al. Treatment of uterine sarcomas. Cancer 1990;66:35–9.
2. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, et al. Cancer statistics, 2008. CA Cancer J Clin 2008; 58:71–96.
3. Kempson RL, Hendrickson MR. Smooth muscle, endometrial stromal, and mixed Müllerian tumors of the uterus. Mod Pathol 2000;13:328–42.
4. Hasiakos D, Papakonstantinou K, Kondi-Paphiti A, Fotiou S. Low-grade endometrial stromal sarcoma of the endocervix. Report of a case and review of the literature. Eur J Gynaecol Oncol 2007;28:483–6.
5. Baker P, Oliva E. Endometrial stromal tumours of the uterus: a practical approach using conventional morphology and ancillary techniques. J Clin Pathol 2007;60:235–43.
6. Scully R, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkinson EJ. Histological typing of female genital tract tumors. In: International histological classification of tumours. 2nd ed. Berlin (Germany): Springer; 1994. p 39–50.
7. Creasman WT, Morrow CP, Bundy BN, Homesley HD, Graham JE, Heller PB. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group Study. Cancer 1987;60 suppl:2035–41.
8. Amant F, De Knijf A, Van Calster B, Leunen K, Neven P, Berteloot P, et al. Clinical study investigating the role of lymphadenectomy, surgical castration and adjuvant hormonal treatment in endometrial stromal sarcoma. Br J Cancer 2007;97:1194–9.
9. Berchuck A, Rubin SC, Hoskins WJ, Saigo PE, Pierce VK, Lewis JL Jr. Treatment of endometrial stromal tumors. Gynecol Oncol 1990;36:60–5.
10. Goff BA, Rice LW, Fleischhacker D, Muntz HG, Falkenberry SS, Nikrui N, et al. Uterine leiomyosarcoma and endometrial stromal sarcoma: lymph node metastases and sites of recurrence. Gynecol Oncol 1993;50:105–9.
11. Norris HJ, Taylor HB. Mesenchymal tumors of the uterus. I. A clinical and pathological study of 53 endometrial stromal tumors. Cancer 1966;19:755–66.
12. Li AJ, Giuntoli RL 2nd, Drake R, Byun SY, Rojas F, Barbuto D, et al. Ovarian preservation in stage I low-grade endometrial stromal sarcomas. Obstet Gynecol 2005;106:1304–8.
13. Mansi JL, Ramachandra S, Wiltshaw E, Fisher C. Endometrial stromal sarcomas. Gynecol Oncol 1990;36:113–8.
14. Gadducci A, Sartori E, Landoni F, Zola P, Maggino T, Urgesi A, et al. Endometrial stromal sarcoma: analysis of treatment failures and survival. Gynecol Oncol 1996;63:247–53.
15. Bodner K, Bodner-Adler B, Obermair A, Windbichler G, Petru E, Mayerhofer S, et al. Prognostic parameters in endometrial stromal sarcoma: a clinicopathologic study in 31 patients. Gynecol Oncol 2001;81:160–5.
16. Chang KL, Crabtree GS, Lim-Tan SK, Kempson RL, Hendrickson MR. Primary uterine endometrial stromal neoplasms. A clinicopathologic study of 117 cases. Am J Surg Pathol 1990;14:415–38.
18. Zaloudek C, Hendrickson MR. Mesenchymal tumors of the uterus. In: Kurman RJ, editor. Blaustein's pathology of the female genital tract. 5th ed. New York (NY): Springer-Verlag; 2002. p. 561–616.
19. Leath CA 3rd, Huh WK, Hyde J Jr, Cohn DE, Resnick KE, Taylor NP, et al. A multi-institutional review of outcomes of endometrial stromal sarcoma. Gynecol Oncol 2007;105:630–4.
20. Geller MA, Argenta P, Bradley W, Dusenbery KE, Brooker D, Downs LS Jr, et al. Treatment and recurrence patterns in endometrial stromal sarcomas and the relation to c-kit expression. Gynecol Oncol 2004;95:632–6.
21. Reich O, Winter R, Regauer S. Should lymphadenectomy be performed in patients with endometrial stromal sarcoma? Gynecol Oncol 2005;97:982.
22. Riopel J, Plante M, Renaud MC, Roy M, Tetu B. Lymph node metastases in low-grade endometrial stromal sarcoma. Gynecol Oncol 2005;96:402–6.
23. Diesing D, Cordes T, Finas D, Loning M, Mayer K, Diedrich K, et al. Endometrial stromal sarcomas—a retrospective analysis of 11 patients. Anticancer Res 2006;26:655–61.
24. Ayhan A, Tuncer ZS, Tanir M, Yuce K, Ayhan A. Uterine sarcoma: the Hacettepe hospital experience of 88 consecutive patients. Eur J Gynaecol Oncol 1997;18:146–8.
25. Lo KW, Yu MY, Cheung TH. Low-grade endometrial stromal sarcoma with florid intravenous component. Gynecol Obstet Invest 2008;66:8–11.
26. Alkasi O, Meinhold-Heerlein I, Zaki R, Fasching P, Maass N, Jonat W, et al. Long-term disease-free survival after hormonal therapy of a patient with recurrent low grade endometrial stromal sarcoma: a case report. Arch Gynecol Obstet 2008 Apr 1. [Epub ahead of print].
27. Krumholz BA, Lobovsky FY, Halitsky V. Endolymphatic stromal myosis with pulmonary metastases. Remission with progestin therapy: report of a case. J Reprod Med 1973;10:85–9.
28. Rose PG, Piver MS, Tsukada Y, Lau T. Patterns of metastasis in uterine sarcoma. An autopsy study. Cancer 1989;63:935–8.
29. Chen SS. Propensity of retroperitoneal lymph node metastasis in patients with stage I sarcoma of the uterus. Gynecol Oncol 1989;32:215–7.
30. Major FJ, Blessing JA, Silverberg SG, Morrow CP, Creasman WT, Currie JL, et al. Prognostic factors in early-stage uterine sarcoma. A Gynecologic Oncology Group study. Cancer 1993;71 suppl:1702–9.
31. Sutton GP, Stehman FB, Michael H, Young PC, Ehrlich CE. Estrogen and progesterone receptors in uterine sarcomas. Obstet Gynecol 1986;68:709–14.
32. Reich O, Regauer S. Hormonal therapy of endometrial stromal sarcoma. Curr Opin Oncol 2007;19:347–52.
33. Li N, Wu LY, Zhang HT, An JS, Li XG, Ma SK. Treatment options in stage I endometrial stromal sarcoma: a retrospective analysis of 53 cases. Gynecol Oncol 2008;108:306–11.
34. Pink D, Lindner T, Mrozek A, Kretzschmar A, Thuss-Patience PC, Dorken B, et al. Harm or benefit of hormonal treatment in metastatic low-grade endometrial stromal sarcoma: single center experience with 10 cases and review of the literature. Gynecol Oncol 2006;101:464–9.
35. Cooper GS, Virnig B, Klabunde CN, Schussler N, Freeman J, Warren JL. Use of SEER-Medicare data for measuring cancer surgery. Med Care 2002;40 suppl IV-43–8.
© 2008 The American College of Obstetricians and Gynecologists
36. Harlan LC, Hankey BF. The surveillance, epidemiology, and end-results program database as a resource for conducting descriptive epidemiologic and clinical studies. J Clin Oncol 2003;21:2232–3.