Objective: The aim of this study was to investigate and validate circulating growth differentiation factor-15 (GDF-15) as a discriminating biomarker between highly malignant uterine sarcomas and benign uterine leiomyomas. In addition, we investigated whether GDF-15 differed between uterine sarcomas and benign adnexal tumors, ovarian or endometrial cancer, and borderline tumors of the ovary.
Materials and Methods: Preoperative blood samples from 19 women with a diagnosis of uterine sarcoma were analyzed for GDF-15 with immunoassay and compared with samples from 50 patients operated on for leiomyoma uteri and with samples from 20 premenopausal and 20 postmenopausal controls. Our previously presented preoperative GDF-15 concentrations in women with borderline (n = 43), benign (n = 144), and malignant ovarian tumors (n = 125), as well as endometrial cancer (n = 510), were used for comparison.
Results: The median circulating GDF-15 concentration was elevated in the uterine sarcoma group (943 ng/L) compared with the myoma uteri group (647 ng/L), the premenopausal and postmenopausal controls (363 and 545 ng/L), and the women with benign ovarian tumors (591 ng/L, all P ≤ 0.007) but was not significantly different from the ovarian borderline tumor (718 ng/L) or ovarian (1242 ng/L) or endometrial cancer (1076 ng/L) groups.
High GDF-15 levels were significantly associated with leiomyosarcomas (P = 0.036), advanced disease (International Federation of Gynecology and Obstetrics stage III/IV, P = 0.013), large tumors (≥10 cm, P = 0.009), and poor survival (P = 0.022).
Conclusions: Circulating GDF-15 may be a promising novel biomarker for the preoperative identification of malignant pelvic disease. Further large prospective studies are needed to evaluate the clinical usefulness of GDF-15 as a discriminator between benign leiomyomas and aggressive sarcomas and as a marker to guide surgical and systemic therapy.
*Department of Obstetrics and Gynecology, Haukeland University Hospital, and †Department of Clinical Medicine, The University of Bergen, Bergen, Norway; ‡Gynecologic Oncology, University Hospitals Leuven, Leuven, Belgium; §Gynecologic Oncology, University Hospitals Leuven & Department of Oncology, KU Leuven, Leuven, Belgium; ∥Department of Gynecology, Oslo University Hospital, Ulleval, and ¶Faculty of Medicine, University of Oslo, Oslo, Norway.
Address correspondence and reprint requests to Jone Trovik, MD, PhD, Department of Obstetrics and Gynecology, Haukeland University Hospital, Jonas Liesv 72, 5021 Hakeland Hospital, Bergen, Norway. E-mail: Jone.Trovik@helse-bergen.no; Jone.Trovik@med.uib.no.
Jone Trovik has been supported by Helse Vest Research Fund no. 911371 as a PhD research fellow. Anne Cathrine Staff has received funding from Oslo University Hospital. Helga Salvesen has received funding from the Norwegian Research Council, The University of Bergen, The Meltzer Foundation, and The Norwegian Cancer Society (The Harald Andersen’s legacy).
Frederic Amant is a senior clinical researcher at the Flandern Research Foundation (F.W.O.).
The authors declare no conflicts of interest.
Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal’s Web site (www.ijgc.net).
Received September 2, 2013
Received in revised form October 11, 2013
Accepted October 17, 2013
Uterine sarcoma is a rare cancer form, comprising 3% of all uterine cancers, with a very poor prognosis.1 Clinically, it is challenging to distinguish sarcoma from leiomyoma preoperatively because no pathognomonic signs on ultrasonography, computed tomography, or magnetic resonance imaging (MRI) correctly discriminate. With leiomyoma as the preoperative diagnosis, sarcoma turned out as the final diagnosis in only 0.2% to 0.5% of premenopausal patients,2,3 rising to 1.7% in postmenopausal women.3,4 In the search of the few malignant sarcomas, many women with benign leiomyoma may undergo a potentially unnecessary hysterectomy. More importantly, some women with sarcomas may receive inferior treatment (conservative or surgical) because of the false assumption of benign leiomyoma and therefore risk less favorable survival.5,6 Hence, there is a need for the identification of new biomarkers improving diagnosis, treatment, and follow-up of uterine sarcomas.
Hitherto, no clinically applicable circulating biomarkers for uterine sarcoma have been identified.7 Preoperative serum CA-125 has not been helpful either in the differential diagnosis of myoma and uterine sarcoma or in assessing the clinical stage of uterine sarcomas.8
Biomarkers can suggest new targets for therapy and may define risk groups for individualized therapy. Growth differentiation factor-15 (GDF-15), also known as macrophage-inhibitory cytokine-1, a distant member of the transforming growth factor-β superfamily, a secreted p53-regulated cytokine shown to have a role in carcinogenesis,9 has been identified as a biomarker for ovarian10 and endometrial11 cancer phenotypes. High GDF-15 predicted presence of lymph node metastasis and poor survival in endometrial cancer.11
Using a prospectively collected cohort of patients with uterine sarcomas, we have investigated whether circulating GDF-15 levels can serve as a biomarker to help distinguish between uterine leiomyomas and sarcomas and, furthermore, whether the determination of GDF-15 levels reflects the extent of disease and prognosis for sarcoma patients.
MATERIALS AND METHODS
EDTA-plasma or serum samples were collected from a total of 19 patients before primary surgery for sarcoma (leiomyosarcoma, n = 13; endometrial stromal sarcoma, n = 4; adenosarcoma, n = 1; and undifferentiated sarcoma, n = 1) at Haukeland University Hospital (Norway, n = 16) from 2001 to 2010 and at the University Hospital of Leuven (Belgium, n = 3) from 2005 to 2009. None of these included patients had a prior diagnosis of sarcoma or had received treatment of the present disease. Extensive clinical information, using International Federation of Gynecology and Obstetrics (FIGO) 2009 staging classification,12 was collected at inclusion directly from the patient, from medical charts, and perioperatively. Follow-up information was retrieved from hospital medical records. Follow-up time ranged from 1 to 106 months, with a mean of 46 months for survivors.
Similarly, preoperative blood samples from 50 women operated on for benign uterine myomas were collected for comparison from Haukeland (n = 16), Ullevål (n = 13), and Leuven (n = 21). The sarcoma and myoma patients from Haukeland and Leuven were all patients treated at these institutions with a preoperative suspicion of uterine sarcoma and thus had blood samples stored in the cancer biobank, whereas the samples from the myoma patients from Ullevål were stored as controls in their general biobank.
Plasma from healthy premenopausal and postmenopausal hospital employees at Oslo University Hospital, Ullevål (Oslo, n = 40), collected from 2003 to 2009 were included as controls, as previously presented, assessed by a different GDF-15 assay.10,11,13 In the present publication, we reanalyzed GDF-15 levels for these healthy controls with the new immunoassay to secure consistency in the method applied.
Growth differentiation factor-15 in plasma or serum was measured by a commercially available enzyme-linked immunosorbent assay (ELISA) and Human GDF-15 Immunoassay (Quantikine; DGD150, R&D Systems, Minneapolis, MN). All analyses were performed in duplicate and blinded for diagnosis and other clinical data. Comparable results have been obtained from plasma and serum samples investigated in parallel for healthy volunteers applied as controls (http://www.rndsystems.com/pdf/DGD150.pdf.). In addition, investigations have supported that the GDF-15 concentration was independent of the added anticoagulant matrix.14 The GDF-15 levels measured in plasma (n = 29; median, 607 ng/L) compared with serum samples (n = 21; median, 706 ng/L) for the leiomyoma patients in our study were not significantly different (P = 0.2), justifying merging the results from GDF-15 measurements in all samples with plasma or serum available for investigations.
We compared the results from the present study with previously published results from women operated on in Oslo during 2003 to 2009 for borderline ovarian tumor (BOT), ovarian cancer, or benign adnexal tumor.10,13 EDTA plasma was collected and stored as previously described.13 In addition, previously published results from women operated on in Oslo or Bergen for endometrial cancer during 2003 to 2009 were used for comparison.11 The GDF-15 assay used in these studies was a radioimmunometric assay,10,11 using the same antibody from the same commercial company that provided the ELISA for the present study.
Plasma from all healthy premenopausal and postmenopausal controls and a subset of patients with ovarian and endometrial cancers (in total, n = 76) was analyzed using both methods for comparison. The level of GDF-15 achieved by the present ELISA method was highly significantly correlated with the previous analyses by the radioimmunometric assay (Pearson coefficient of 0.974, P < 0.001; Fig. 1, Supplemental Digital Content, available at http://links.lww.com/IGC/A192), confirming comparability for the different assays.
The women participating in this study gave written informed consent, and the biobank studies were approved by the Regional Committee of Medical and Health Research Ethics (REK) Eastern Norway (for Oslo), by REK Western Norway (for Bergen, NSD 15501), and the institutional review board at the University Hospital of Leuven.
Statistical analysis was performed applying the IBM Statistical Package for the Social Sciences (SPSS, version 20.0, New York). Probability of less than 0.05 was considered statistically significant. The clinical characteristics are presented as median values and range or percentage of patients; and the GDF-15 results, as medians (and 95% confidence interval [CI] of the median) and interquartile range for the main patient groups. A nonparametric Mann-Whitney U test was used when comparing continuous data between the study groups. For the categorical data, the χ2 test was used. Univariate survival analyses were performed, using the Kaplan-Meier method and the log-rank test, grouping low versus high concentration of GDF-15. Cutoff values for categorization were based on quartiles according to the frequency distribution of the marker, the size of the subgroups, and the number of events in each category. Groups with similar survival were merged, that is, the 3 lower GDF-15 quartiles. Multivariate survival analyses were performed with Cox regression. Receiver operating characteristic (ROC) curve was constructed for circulating GDF-15 as a discriminator between women with uterine sarcoma and benign leiomyoma as well as healthy premenopausal or postmenopausal women.
Circulating GDF-15 Is Elevated in Patients With Uterine Sarcomas Compared With Myomas
Clinical characteristics for the sarcoma and leiomyoma patient groups are presented in Table 1, together with the previously published clinical data from the healthy premenopausal and postmenopausal controls and patients with other diagnoses for comparisons.10,11 Although there is considerable overlap (Fig. 1), the median plasma concentration of GDF-15 was significantly higher in the sarcoma patients (943 ng/L) as compared with the women with leiomyomas (647 ng/L, P < 0.001). The GDF-15 level was significantly higher for the subgroup of leiomyosarcomas (1397 ng/L) compared with leiomyomas (P = 0.001) as well as with the other histological types of uterine sarcoma (680 ng/L, P = 0.036, Table 2). In addition, the median GDF-15 concentration among the women with uterine sarcoma was elevated compared with the control groups of healthy premenopausal and postmenopausal women and compared with the patients operated on for benign adnexal tumor (Table 1). The median circulating GDF-15 concentration for the sarcoma group was comparable with that for the ovarian and endometrial cancer groups (1242 and 1076 ng/L, P = 0.5 and P = 0.6) and with a tendency to a higher level in sarcomas compared with BOTs (718 ng/L, P = 0.08).
An ROC curve for GDF-15 as a discriminator between women with uterine sarcoma and leiomyomas gave an area under the curve (AUC) of 0.71 (95% CI, 0.55–0.88, P = 0.007, Fig. 2). For the patients with sarcomas compared with the healthy premenopausal or postmenopausal controls, the AUC for circulating GDF-15 was 0.83 (95% CI, 0.68–0.98; P < 0.001). For leiomyosarcomas versus myomas, the GDF-15 ROC curve yielded an AUC of 0.80 (95% CI, 0.60–0.99; P = 0.011). Thus, circulating GDF-15 significantly discriminates uterine sarcoma patients from leiomyoma patients as well as from healthy women.
High GDF-15 Levels Are Associated With Aggressive Uterine Sarcomas
Growth differentiation factor-15 concentrations were significantly higher in the women with metastatic disease (FIGO stage III or IV), with large tumor diameter (≥10 cm), and with leiomyosarcomas as compared with other histological types (endometrial stroma sarcoma, adenosarcoma, or undifferentiated sarcoma). Age, parity, or menopausal status was not correlated with GDF-15 level in this small patient series (Table 2). Using the cutoff for tumor diameter of greater than 5 cm, in agreement with FIGO IA compared with IB in the 2009 FIGO classification,12 no significant correlation between tumor size and GDF-15 concentration was seen.
The sarcoma patients with high GDF-15 concentration, defined as belonging to the upper GDF-15 quartile of 2139 ng/L or greater, had significantly shorter survival because no patients survived more than 47 months compared with 58% 5-year disease-specific survival for the patients with GDF-15 levels lower than the upper quartile, as shown in Figure 3 panel A (log-rank test P = 0.02). Analyzing leiomyosarcomas separately demonstrates a similar significant difference with P = 0.01 (Fig. 3, panel B). In univariate survival analyses, age, parity, tumor size, type of surgical treatment (radical surgery vs debulking), or whether any adjuvant treatment was given was not of statistical significance, although the small sample set should call for caution in the conclusion. In Cox analyses, high GDF-15 was a significant factor predicting poor survival for uterine sarcomas even when adjusting for FIGO stage (hazard ratio, 4.0; 95% CI, 1.02–15.49; P = 0.047). The Cox analysis for leiomyosarcomas separately is presented in Table 3.
Our present study is, to the best of our knowledge, the first report on circulating GDF-15 in uterine sarcoma patients and in women with benign leiomyomas.
We find that plasma or serum GDF-15 is higher in patients with uterine sarcomas as compared with healthy premenopausal and postmenopausal controls, women with benign adnexal tumors, and women with benign myomas. Still, further and larger studies are needed to explore the applicability of GDF-15 measurements for risk stratification discriminating benign leiomyomas from malignant pelvic masses and sarcoma/leiomyosarcoma in particular and as a prognostic marker in sarcomas. Because our study sample size of sarcomas and leiomyoma patients is limited, we consider our results as exploratory but promising, thus needing further verification in larger cohorts.
Some of the plasma samples have been stored up to 11 years before analysis, with unknown effects on GDF-15 protein degradation. However, when exploring storage time as a possible confounding factor analyzing the endometrial cancer samples, we did not find any significant time trend changes.15 Still, prospective validation studies of freshly collected patient samples would be needed to rule out time effects from storing.
High plasma concentration of GDF-15 has been linked to all-cause mortality in apparently healthy elderly women associated with cardiovascular disease even after adjusting for inflammation markers such as interleukin 6 and C-reactive protein (CRP).16 Elevated levels in otherwise apparently healthy elderly women have also been linked to higher risk for future cardiovascular disease.17 Growth differentiation factor-15 is normally highly expressed in the placenta,18 but further augmented circulating GDF-15 in pregnancies have been linked to increased occurrence of preeclampsia and diabetes mellitus, both conditions correlated to increased risk for cardiovascular disease.19
We did not find significantly higher GDF-15 concentrations in the higher age groups, suggesting only weak effects of potentially confounding factors related to comorbidities known to increase with age. We also applied disease-specific survival as outcome measure to minimize any confounding effects related to cardiovascular deaths in our study.
Similar to our study of endometrial carcinomas,11 we find an association between tumor characteristics (aggressive subtypes and advanced FIGO stage) and high levels of circulating GDF-15. Inflammatory cytokines such as interleukin 1 and tumor necrosis factor α may have activated macrophages and induced GDF-15 production.20 Inflammation is now recognized as one of the “hallmarks of cancer”.21 Our findings of elevated plasma GDF-15 levels in cancer patients compared with controls may be due to cancer-related activation of inflammatory processes. Interestingly, one study22 has identified neutrophil to lymphocyte ratio as a significant discriminator between myomas and sarcomas, as well as prognosis for the latter group.
Growth differentiation factor-15 increases in cancer as well as in acute injury and inflammation,23 and overexpression has been reported in other nongynecologic cancers such as prostate, pancreatic, colonic, and thyroid cancers.9,24,25 Growth differentiation factor-15 has been found to be induced by the tumor suppressor gene p53 and to have a p53 response element in its promoter, suggesting that GDF-15 may be a downstream target of pathways regulating cell cycle arrest and apoptosis26 and thus important for proliferation, migration, invasion, metastases, and treatment resistance in cancer.9
For the clinicians, the key clinical challenge is that uterine sarcoma is a very rare disease, with no accurate preoperative diagnostic tools available to distinguish from benign leiomyoma. A clear illustration of this is the series published from 1253 women with presumed benign myomas; all were investigated by MRI and 453 had transcervical biopsies performed to diagnose only 7 cases of sarcomas.3 Studies investigating CA-125 are sparse, encompass relatively few sarcoma cases, and do not support this biomarker as a good discriminator. Two studies failed to identify preoperative CA-125 as a significant discriminator of sarcomas, one study including 42 sarcomas compared with 84 myomas27 and another study of 26 sarcomas of 2382 presumed myoma patients.8
A study by Kim et al22 identified neutrophil to lymphocyte ratio as a better predictor than CA-125 in diagnosing sarcomas in a case-matched study of 55 sarcoma patients (including 21 with carcinosarcomas) as compared with 165 leiomyomas and 165 adenomyosis patients.
Lactate dehydrogenase (LDH), an enzyme linked to tissue breakdown, has also been suggested as a biomarker for sarcoma, but studies are small. Goto et al28 demonstrated LDH to be elevated not only in all 10 sarcoma patients but also in 16 of 130 patients with degenerative myoma in their study combining LDH with MRI, whereas Nagamatsu et al29 likewise found 4 sarcoma patients to have elevated LDH and 2 of 15 in the control group with leiomyoma in a study investigating positron emission tomography scan as a diagnostic tool for uterine sarcoma.
We show that circulating GDF-15 correlates with aggressive tumor characteristics including tumor diameter of 10 cm or greater, whereas a cutoff at 5 cm as applied by FIGO in the distinction between IA and IB stages was not significantly correlated with GDF-15 level. Apparently somewhat in line with our finding, Abeler et al30 reported in their Norwegian population-based study of 419 uterine sarcoma patients a tumor size of 10 cm and not 5 cm as an independent predictor of poor survival.
In conclusion, although our findings suggest that GDF-15 levels may improve the preoperative discrimination between benign leiomyoma and leiomyosarcoma, further larger and prospective studies are needed to explore the clinical applicability of GDF-15 measurements as a diagnostic tool. A more reliable preoperative biomarker in this context would be of great clinical value
The authors thank Brit Edvardsen, Haukeland University Hospital, and Lisa Levy, Oslo University Hospital, for their help in patient recruitment and technical assistance.
1. Amant F, Coosemans A, Debiec-Rychter M, et al. Clinical management of uterine sarcomas. Lancet Oncol. 2009; 10: 1188–1198.
2. Parker WH, Fu YS, Berek JS. Uterine sarcoma in patients operated on for presumed leiomyoma and rapidly growing leiomyoma. Obstet Gynecol. 1994; 83: 414–418.
3. Kawamura N, Ichimura T, Ito F, et al. Transcervical needle biopsy for the differential diagnosis between uterine sarcoma and leiomyoma. Cancer. 2002; 94: 1713–1720.
4. Leibsohn S, d’Ablaing G, Mishell DR Jr, et al. Leiomyosarcoma in a series of hysterectomies performed for presumed uterine leiomyomas. Am J Obstet Gynecol. 1990; 162: 968–974; discussion 74–6.
5. Park JY, Park SK, Kim DY, et al. The impact of tumor morcellation during surgery on the prognosis of patients with apparently early uterine leiomyosarcoma. Gynecol Oncol. 2011; 122: 255–259.
6. Perri T, Korach J, Sadetzki S, et al. Uterine leiomyosarcoma: does the primary surgical procedure matter? Int J Gynecol Cancer. 2009; 19: 257–260.
7. Van den Bosch T, Coosemans A, Morina M, et al. Screening for uterine tumours. Best Pract Res Clin Obstet Gynaecol. 2012; 26: 257–266.
8. Yilmaz N, Sahin I, Kilic S, et al. Assessment of the predictivity of preoperative serum CA 125 in the differential diagnosis of uterine leiomyoma and uterine sarcoma in the Turkish female population. Eur J Gynaecol Oncol. 2009; 30: 412–414.
9. Bauskin AR, Brown DA, Kuffner T, et al. Role of macrophage inhibitory cytokine-1 in tumorigenesis and diagnosis of cancer. Cancer Res. 2006; 66: 4983–4986.
10. Staff AC, Bock AJ, Becker C, et al. Growth differentiation factor-15 as a prognostic biomarker in ovarian cancer. Gynecol Oncol. 2010; 118: 237–243.
11. Staff AC, Trovik J, Eriksson AG, et al. Elevated plasma growth differentiation factor-15 correlates with lymph node metastases and poor survival in endometrial cancer. Clin Cancer Res. 2011; 17: 4825–4833.
12. Prat J. FIGO staging for uterine sarcomas. Int J Gynaecol Obstet. 2009; 104: 177–178.
13. Odegaard E, Davidson B, Elgaaen BV, et al. Circulating calprotectin in ovarian carcinomas and borderline tumors of the ovary. Am J Obstet Gynecol. 2008; 198: e1–e7.
14. Kempf T, Horn-Wichmann R, Brabant G, et al. Circulating concentrations of growth-differentiation factor 15 in apparently healthy elderly individuals and patients with chronic heart failure as assessed by a new immunoradiometric sandwich assay. Clin Chem. 2007; 53: 284–291.
15. Trovik J. Endometrial Carcinoma; Can Biomarkers Aid in the Prediction of Aggressive Disease? Bergen, Norway: University of Bergen, Norway; 2012. ISBN: 978-82-308-2019-3.
16. Wiklund FE, Bennet AM, Magnusson PK, et al. Macrophage inhibitory cytokine-1 (MIC-1/GDF15): a new marker of all-cause mortality. Aging Cell. 2010; 9: 1057–1064.
17. Brown DA, Breit SN, Buring J, et al. Concentration in plasma of macrophage inhibitory cytokine-1 and risk of cardiovascular events in women: a nested case-control study. Lancet. 2002; 359: 2159–2163.
18. Lawton LN, Bonaldo MF, Jelenc PC, et al. Identification of a novel member of the TGF-beta superfamily highly expressed in human placenta. Gene. 1997; 203: 17–26.
19. Sugulle M, Dechend R, Herse F, et al. Circulating and placental growth-differentiation factor 15 in preeclampsia and in pregnancy complicated by diabetes mellitus. Hypertension. 2009; 54: 106–112.
20. Bootcov MR, Bauskin AR, Valenzuela SM, et al. MIC-1, a novel macrophage inhibitory cytokine, is a divergent member of the TGF-beta superfamily. Proc Natl Acad Sci U S A. 1997; 94: 11514–11519.
21. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011; 144: 646–674.
22. Kim HS, Han KH, Chung HH, et al. Neutrophil to lymphocyte ratio for preoperative diagnosis of uterine sarcomas: a case-matched comparison. Eur J Surg Oncol. 2010; 36: 691–698.
23. Zimmers TA, Jin X, Hsiao EC, et al. Growth differentiation factor-15/macrophage inhibitory cytokine-1 induction after kidney and lung injury. Shock. 2005; 23: 543–548.
24. Brown DA, Stephan C, Ward RL, et al. Measurement of serum levels of macrophage inhibitory cytokine 1 combined with prostate-specific antigen improves prostate cancer diagnosis. Clin Cancer Res. 2006; 12: 89–96.
25. Koopmann J, Buckhaults P, Brown DA, et al. Serum macrophage inhibitory cytokine 1 as a marker of pancreatic and other periampullary cancers. Clin Cancer Res. 2004; 10: 2386–2392.
26. Agarwal MK, Hastak K, Jackson MW, et al. Macrophage inhibitory cytokine 1 mediates a p53-dependent protective arrest in S phase in response to starvation for DNA precursors. Proc Natl Acad Sci U S A. 2006; 103: 16278–16283.
27. Juang CM, Yen MS, Horng HC, et al. Potential role of preoperative serum CA125 for the differential diagnosis between uterine leiomyoma and uterine leiomyosarcoma. Eur J Gynaecol Oncol. 2006; 27: 370–374.
28. Goto A, Takeuchi S, Sugimura K, et al. Usefulness of Gd-DTPA contrast-enhanced dynamic MRI and serum determination of LDH and its isozymes in the differential diagnosis of leiomyosarcoma from degenerated leiomyoma of the uterus. Int J Gynecol Cancer. 2002; 12: 354–361.
29. Nagamatsu A, Umesaki N, Li L, et al. Use of 18F-fluorodeoxyglucose positron emission tomography for diagnosis of uterine sarcomas. Oncol Rep. 2010; 23: 1069–1076.
30. Abeler VM, Royne O, Thoresen S, et al. Uterine sarcomas in Norway. A histopathological and prognostic survey of a total population from 1970 to 2000 including 419 patients. Histopathology. 2009; 54: 355–364.
Uterine sarcoma; Leiomyosarcoma; Biomarker; Growth differentiation factor-15
Supplemental Digital Content
© 2014 by the International Gynecologic Cancer Society and the European Society of Gynaecological Oncology.