Retrospective analysis of 105 cases with uterine sarcoma

ARTICLE

Auteur(s) : Adnan Yoney, Bekir Eren, Sukran Eskici, Adile Salman, Mustafa Unsal

Okmeydani Training and Research Hospital, Department of Radiation Oncology, 8 Cadde Inci 3 daire 8, Atasehir-Kadikoy, 34578 Istanbul, Turkey

Uterine sarcomas make up 1% of all gynecologic malignancies and 3-7% of uterine malignancies [1]. There is no one treatment modality, which is formed as a result of randomized studies, or is widely accepted in uterine sarcomas, which are very rarely seen and constitute a heterogeneous group in pathological terms. Surgery alone can be curative in malignancies limited to the uterus. On the other hand, the value of pelvic radiotherapy has not been definitive. The efficacy of adjuvant chemotherapy following full resection has not been shown yet with a randomized study in stages I and II [2]. However, adjuvant chemotherapy alone or in combination with radiotherapy has been shown to increase survival in non-randomized studies [3-5].

Because large series cannot be formed due to the low incidence, it is very important that each clinical study group analyze their own experience both for explaining various prognostic factors and for evaluating treatment results by comparing with other series.

In this study carried out with this aim, we retrospectively evaluated the general features, treatments applied and the results obtained in 105 patients present at our clinic with the diagnosis of uterine sarcoma between 1995 and 2003.

Patients and methods

Patient selection

Patients were divided into two groups as under and above 50 years of age in order to evaluate the effect of age on prognosis.

Patients were collected in 3 main groups according to histological diagnosis as: 1) malignant mullerian mixed tumors (MMMT) (adenosarcomas were also included in this group); 2) leiomyosarcoma (LMS); 3) endometrial stromal sarcoma (ESS). Pathology reports in patient files were examined and the grades of the tumor were determined in those patients with sufficient data and were divided into 3 groups as low, moderate and high grade. Moreover 49 patients with the numbers of mitosis reported in their pathology reports were evaluated in 3 groups as: number of mitosis < 5, number of mitosis 5-10, number of mitosis > 10.

Staging

Treatment

A total of 50.5% of cases underwent RT, and a linear accelerator (18 MV) was used in 10 patients (19%) and Co60 unit was used in 40 (81%) patients. Forty-nine (92.5%) patients undergoing RT received external pelvic RT (ERT) and 4 (7.5%) received both intracavitary RT (ICRT) and ERT. External treatment was applied using the box technique in 4 (8%) patients and fixed SSD technique from opposite L5-pelvis regions in 49 (92%) patients. HDR brachytherapy units with Co60 sources were used for ICRT. Total RT doses median 54.8 Gy (range, 40-70 Gy). Four (7.1%) patients were given < 50 Gy, 44 (78.6%) 50-60 Gy and 5 (8.9%) > 60 Gy. Four cases (stage III: 3 cases and stage I: 1 case) undergoing ERT + ICRT received 1470-2000 cGy intracavitary RT following 50 Gy external dose.

Nineteen (29.5%) patients underwent CT and 13 (12.4%) CT + RT. Most commonly used chemotherapy combinations were those containing ifosphamide (IFO), adriamycin (ADR), cisplatin (CDDP) and cyclophosphamide (CTX). The number of chemotherapy treatments applied varied between median 4 (range 2-6).
Table 1 Distribution of patients according to stages and treatment modalities

Follow-up

Statistical methods

Results

Patient characteristics

Fourty-six (43.8%) patients were LSM, 30 (28.6%) were ESS and 29 (27.6%) were MMMT. Other histological subgroups were not seen. Fifty-eight (55.2%) patients were stage I, 14 (13.3%) were stage II, 20 (19.0%) were stage III and 10 (10.5%) were stage IV. We couldn’t get enough stage information about 2 (1.9%) patients.

Twenty-nine (27.6%) patients had grade I disease, 11 (10.5%) had grade II, and 28 (26.7%) had grade III, 37 (35.2%) patients’ grades was unknown.

The percents of the patients, whose grades were known, were 42.6% grade I, 16.2% grade II and 41.2% grade III.

The number of mitosis was 1-5 in 14 (13.3%) patients, 5-10 in 13 (12.4%) and > 10 in 22 (21%). Fifty-six (53.3%) patients’ the number of mitosis was unknown.

Fifty-five (52.4%) patients were pre-menopausal and 50 (47.6%) were postmenopausal. There was no patient with a history of pelvic radiotherapy: 58 (55.2%) were under 50 years of age and 47 (44.8%) were above 50.

The presenting symptom (in some patients there was more than one symptom) was bleeding in 69 (65.7%) patients, pain in 38 (36.2%), mass-swelling in 19 (18.1%) and discharge in 17 (16.2%). In two patients, breast cancer was found as concomitant second malignancy.

There were statistically excessively significant differences between median ages according to pathology types (p < 0.01). Median age of patients with pathology type of MMMT was statistically significantly higher than that of patients with pathology type of LMS (p = 0.001). There were no statistically significant differences between median ages according to other pathology types (p > 0.05). Although there was a difference in median ages according to the stages, this did not reach statistical significance (p > 0.05). There was also no statistically significant difference between distribution rates of the pathologic diagnoses according to the stages (p > 0.05).

Recurrences and metastases

When failure rates between treatment groups are compared according to the stages, patients whose stage is unknown, stage IV patients and adjuvant therapy patients were not taken into evaluation.

Local recurrence was not seen in 6 patients with suboptimal surgery; but metastases developed in 3 (3/6) patients. There was no significant difference when adjuvant treatment groups were compared in terms of local recurrences and distant metastases according to the stages (p > 0.05).

As cases with a local relapse were studied, we marked 6 patients who had relapses after receiving radiotherapy with a dose range of 50-54 Grays, while no relapses were depicted in those who received higher doses of radiotherapy.
Table 2 Site of first recurrence after type of treatment

Survival

Ninety-two patients in the study were followed up median 30 (range, 4-134) months, 44 patients (41.9%) died because of the disease, 88.63% of the deaths occurred within the first 2 years, 93.18% within the first 3 years and all occurred within 5 years. After 5 years, there were no disease-related deaths. Three-year disease-free survival rate was 54.46 ± 5.42% and 5-year disease-free survival rate was 49.88 ± 5.95% (figure 2).

At univariate analysis, we analyzed the impact of surgery, age, menopausal status, histology, grade, number of mitosis, tumor size, stage and treatment groups on survival. Grade, stage, age, menopausal status and presence of RT in treatment modality demonstrated statistical significance, but histology, number of mitosis, tumor size demonstrated not significance.

In our trials, patients with TAH + BSO demonstrated a long median survival time (73.18 and 15.17 months) than patients with suboptimal surgery (p = 0.015) (figure 3).

Histological grade demonstrated a worse, statistically significant prognosis for high grade than for intermediate and low-grade tumors (p = 0.0106) (figure 4).

Univariate analysis of survival of stages I, II, III and IV, demonstrated 29, 46, 65 and 100% mortality, respectively at the time of analysis. These results remained unchanged when the analysis was performed by stage I conferring the best prognosis (p = 0.001) (figure 5).

Patients over 50 years of age at diagnosis appeared to have a higher death rate than patients aged less than 50 (p = 0.021), as shown in figure 6. We also analysed our series by menopausal status and noted that postmenopause patients had a worse prognosis than premenopausal patients (p = 0.0064) (figure 7).

Presence of radiotherapy in treatment modality improve the OS (p = 0.0001) (figure 8).

With respect to histology MMT had the worst prognosis followed by LMS and ESS at 5 years (figure 9) but, it was not statistically significant (p = 0.09).

Also, univariate analysis of tumor size (0-10, 10-20 and > 20 cm) (p = 0.349) and number of mitosis (1-5, 5-10 and >10) (p = 0.243) demonstrated not statistical significance.

In multivariate analysis, when age above or under 50 years, menopause status, grade, histology, tumor size, surgery type and treatment modalities were evaluated with backward stepwise Cox regression analysis in terms overall survival time and deaths, the stage was found to have a statistically significant (p < 0.05) effect on death at the end of step 1. Stage III increases death by 4.28 fold.

Premenopause (p = 0.011), grades I or II (p = 0.0016), tumor size < 30 cm (p = 0.047), stages I or II (p = 0.001), presence of RT in the treatment modality (p = 0.0001) were found to be favourable prognostic factors for disease-free survival in the univariate analysis.

In multivariate analysis, when state, menopause status, grade, histology, tumor size and treatment modalities were analyzed with backward stepwise Cox regression analysis in terms of disease-free survival and recurrences, only menopause status was found to have a significant effect on recurrence at the end of the 4th step and had a 3.21 fold increasing effect on recurrence.

Discussion

Only little is known about the epidemiology of uterine sarcomas [5]. No other known clear risk factor exists besides pelvic radiotherapy. A previous pelvic irradiation history is reported in 2-14% of patients with uterine sarcoma, notably with a higher ratio in those with MMMT [1]. We had no patients with such history in our series.

A second primary malignant tumor is reported in 5-11% of patients with uterine sarcoma. Those second malignancies either previously occurred or simultaneous are mostly breast, ovarian or colon cancers. We detected two breast cancer cases as a second primary malignancy in our study.

Currently there exists no unique, standardized staging system for uterine sarcomas. “FIGO’s staging” for the adenocarcinoma of the corpus uteri is referred. 50-60% of the patients enroll as “stage I” disease as the other patients are distributed among the other stages between 10-20%, while slightly less as “stage II” disease than the others. In convenience with the given data stage I patients’ share is 55.2% in our study. The total of stages I and II diseases is concluded as “early stage”, mentioning the so called “disease limited to the organ” stage is 68.5%.

Being such a heterogeneous group; uterine sarcomas once had been mostly described as LMS as pathological type in elder publications, while more recently 2/3 of them had been cited as mixed mesodermal tumors [7, 8]. Unifying 14 different series in literature and thus forming a group of 1440 patients. Lurain et al. [7] reported the incidence of pathological types as follows: 43% LMS, 39% MMMT, 14% ESS and 4% the others. In one of the major retrospective series ever published including 423 cases of uterine sarcoma between years 1967-1981 Olah et al. [8] reported the incidences as 51% LMS, 36% MMMT and 6% ESS.

In our results the most frequent histological type is (43.8%) LMS which is compatible with the early literature and the following types are 27.6% MMMT and 8.6% ESS.

In the studies comparing the median age in patients with different histological types of tumors, it has been seen that younger patients rather had LMS. The age range of patients with LMS (50-55 years old) at the time of diagnosis is approximately 10 years younger than those patients with MMMT (60-65 years old) [6, 9]. As we also compared the median age at the time of diagnosis in our patients we found out that our patients with LMS had this disease at younger ages (46) which showed a statistically significant difference (p < 0.01) supporting the other issues in the literature. The difference between the median ages of histological types such as LMS and MMMT is found to be close to that reported in the literature (10.89 ages). However, the median age values for each different histological type of disease were smaller in concordance with the age at the time of diagnosis (MMMT: 57, ESS: 1, LMS: 46).

There exists no statistically different overall survival value concerning the different histological types [10]. Salazar et al. [11] found a longer OS time in LMS than that in MMMT. In the retrospective study having 423 patients Olah et al. [8] had reported that the prognosis was worse in patients with LMS after having done the correction regarding the stage, the age and the grade. The difference in OS in patients with ESS is found to be significantly high in the study of Atalar et al. [12]. When considering our study, despite the difference which has been seen in number of the three main histological types of diseases differing from those in literature, no statistical difference has been observed in OS curve probably due to disproportional distribution in number of our patients 5-years OS rates in MMMT, in LMS and in EES are calculated to be 20.92, 44.02 and 65.81% respectably (p = 0.0689).

The prognostic factors in uterine sarcoma which is a rare, yet a very aggressive disease had been studied. Age (menapausal status), histologic type, surgical stage, tumor grade, mitotic index, and p53 expression were prognostic factors of the overall survival of patients with uterine sarcoma [6, 9, 13-15].

In this study the patients are also classified in two groups as those who are 50 years old or younger and those who are over 50 years old in order to observe the effect of the age factor on the prognosis and we have seen a specifically significant difference regarding the effect of age on prognosis same as the data in literature. 5-years OS rates in two groups (50 years old or less and over 50 years old) are 57.70 ± 7.63% and 31.51 ± 11.04% respectively (p = 0.021). Poor prognosis in patients who are over 50 years old and in postmenopausal period can be defined as that the average age of menopause in Turkish women is 50.

When OS rates were investigated regarding the factors such as menopausal status and stage of disease, we have seen a specifically significant difference in favor of those with early stage disease. Three-years and 5-years separate OS rates for each stage I-III diseases are; (stage I) 68.43, (stage-II) 68.38 and 34.19%, (stage III) 40 and 32% respectively. Median survival is 13 months in stage-IV patients and we had no follow-up patients at 3 and 5 years periods (p <0.01).

Hormone receptors have not been studied in these patients since the relation between prognosis and menopausal stage has not been clearly described. The relation of receptors with relapses and overall survival rates could not have been evaluated due to lack of data concerning hormone receptors.

In 68 out of 105 patients’ pathology reports ‘tumor grade’ was noted. As for the effect of tumor grade on OS and on DFS; a significant difference is seen in both aforementioned survival curves correlated with the different grades of tumors like low (grade I), medium (grade II) and high (grade III). OS (p = 0,0106) and DFS (p = 0.0016) rates of grade III tumors are found to be significantly poor compared to grade I and grade II tumors as also cited in the literature.

Mitotic activity is closely related with the prognosis in uterine sarcoma. In endometrial stromal tumors such as endometrial stromal nodules and endolymphatic stromal myosis; the mitotic activity is less than 10 MF/10HPF (10 mitotic figures in each 10 high power field), while it is over 10 MF/10 HPF in ESS. A mitotic activity higher than 10 MF/10 HPF in LMS indicates poor prognosis. Malignity of tumors having a value between 5-10 MF/10 HPF is more difficult to asses and they can either present local failures or distant metastases.

Those with a level less than 5 MF/10 HPF are generally benign tumors [7, 16]. Controversially Evans et al. [17] and Chang et al. [18] have reported that the mitotic count and the grade were not prognostic factors. Studying the activity in 49 of our patients we revealed that no significant relation existed neither between mitotic count and DFS (p = 0.892) and nor between mitotic count and OS (p = 0.243) as in studies of Evens and Chang.

34-64% of uterine sarcoma cases are reported to present recurrences [6, 10]. 60% of the recurrences occur within the first year succeeding the therapy, while the great majority adds up to relapse within the second year [19]. Also as seen in the literature the local failures all occurred within the first two years in our study. Still considering our study; 90% of the distant metastases had occurred in the first two years and added up to 100% within five years, as they confirm the time periods and rates given in literature. The data concerning the recurrences in the published issues are alike. It is reported that the total of the pelvic recurrences range between 14-43% as isolated pelvic relapse portion compromises 7-14% while distant metastases range between 25-52%. 80% of the first incidences occur as distant metastases most commonly in lungs (28-32%) and in abdominal sites (the peritoneum and the serosa) (13-24%) and together form 80% of all metastases. Less common sites of metastases are in order; liver (10-14%), bones (7-10%), skin and brain. Distant metastases and pelvic recurrence ratio is reported to be approximately 3/1 [4, 6, 7]. In our trial distant metastases (28.6%) were about twice the local failures (16.7%) and the coexistence rate of both incidents was 5%. When concerning the sites of distant metastases in our trial; they mostly occurred in the lungs (12.4%) and then in order in the bones (3.8%), in the liver (2.9%), in the other sites (5.7%) and as multiple metastases in diverse sites (3.8%) and they corresponded the rates given in the literature.

Surgery is the primary therapy of choice in uterine sarcomas. TAH + BSO is accepted as the standard surgery [4, 20, 21]. Olah et al. [6] reported a 50% OS rate at 5 years in 181 stage I or stage II cases who could have undergone a total resection. There also exist researchers who suggest “bilateral pelvic lymphadenectomy” due to frequent local dissemination probabilities especially in early stage tumors [19]. Despite the reports of Olah et al. [6] and Salazar et al. [11], which state that cases without any tumor rest following TAH + BSO is a factor favoring good prognosis in any stage, it has been seen that “debulking” pelvic surgery did not improve the OS depending on the results investigated following the correction regarding the stage and the grade [9]. The fact of presence of recurrences (with more than half the cases as distant metastases) nearly up to 50% even in tumors limited in the organ, imposed to us the necessity of the adjuvant therapy [4, 5, 10]. 83.3% of the 105 patients in our trial had adjuvant therapy. Of these 105 patients, 53 (50.5%) patients received radiotherapy, 13 (12.4%) patients received combined modality therapy and 31 (20.9%) patients received chemotherapy.

Perez et al. [22] reported that preoperative, external and intracavitary radiotherapy especially with doses over 60 Grays has lowered the pelvic recurrence rates from 50 to 17% in stage I disease. Monk et al. [23] have emphasized the feasibility of preoperative radiotherapy also in stage II disease. Several trials report that postoperative radiotherapy improves the local control, prolongs the time to recurrence, yet has no effect on OS [10, 11]. In some other few trials it has been shown that radiotherapy improves the OS in early stages of MMMT and ESS [9, 19]. Belgrad et al. [19] have reported that preoperative and postoperative radiotherapy have improved the OS rates in MMMT and in ESS from 20 to 35% and from 37 to 57% respectively. Radiotherapy is generally applied up to 40 or 60 Grays [6, 9]. Also, pre or postoperative radiotherapy have been adviced at French clinical practice guidelines: 2006 update [24].

In our trial, 66 (62.9%) out of 105 patients have received radiotherapy with a median dose of 54.8 Grays. Despite our slightly lower local recurrence rates (16.7%) as compared to data in the literature (14-43%); we had also seen that radiotherapy decreased the local recurrences (p = 0.0001) when compared with the group of patients who did not receive radiotherapy in our trial and did have a similar effect like in Belgrad’s study [19] as to improve the OS (p = 0.0001).

Several trials had been carried out in which adjuvant chemotherapy was used due to matter of frequently occurring distant metastases and poor prognosis of early stage uterine sarcoma. There exist trials with controversial results, yet no outstanding aim is clearly reached. In a randomized trial, composed of 156 patients with stages I and II disease, Omura et al. [2, 25] have seen that no improvement has been attained in the recurrence and the disease-free survival rates after postoperative adriamycin administration, where though Berchuck et al. [26] had reported an improvement in DFS rates in relatively small series with the same agent. Despite two other trials reporting better survival rates with a vincristine, actinomycin-D and cyclophosphamide (VAC) regimen, there exist trials reporting no decreased recurrence rates. In their trial Rose et al. [27] have applied 12 cycles of VAC to 64 patients with stage I disease and have observed no improvement in recurrence rates. Nagell et al. [3] have reported that relapses have occurred in two (28%) out of 7 patients with stage I disease to whom 6 cycles of VAC were applied. They also cited that chemotherapy has decreased the number of recurrences when compared with those patients who were treated without chemotherapy before in the same institute.

Using radiotherapy simultaneously with VAC, cisplatinum and etoposide combination Töre et al. [28] reported a significant improvement in OS at 3-years, which is not sustained in long term follow-up. Cyclophosphamide, vincristine, adriamycin, dacarbazine (Cyvadi) and ifosfamide or cyclophosphamide, adriamycin and ifosfamide have also been used but still without any improvement in OS rates [29]. Gadducci et al. [4] and Papadimitriou et al. [5] have suggested a potential role for anthracycline- and ifosfamide-containing chemotherapy in the adjuvant setting for early-stage uterine sarcomas. Cisplatinum, VAC, adriamycin + cisplatinum + cyclophosphamide and DTIC have been also used in patients with either advanced or metastatic disease and partial response rates have been reached as 20-50% [30]. Chemotherapy is applied to 44 (41.9%) patients in our trial. Thirty-one patients received chemotherapy alone, while 13 patients had radiotherapy + chemotherapy. They received 2 to 6 cycles of chemotherapy (median 4); most commonly with adriamycin, CDDP and CTX including combinations. The DFS and the OS rates of patients who received chemotherapy alone are found significantly low (p = 0.0001) compared to both other arms as radiotherapy alone and chemo-radiotherapy in our trial when evaluated in regard of the different sights considering chemotherapy in the literature.

The OS rates at 5-years range between 30-39% in several different series [6, 8, 9]. Yet, the specific characters of each group of patients should also be taken in account, as these OS rates are evaluated. Some OS rates as high as 73% in stage I disease are recorded at 5 years [29]. Besides, even higher OS rates are reported in premenopausal ,young patients with low grade tumors. Our OS rates at 5 years are found to be 51.09 ± 5,95% and 68.43% in the whole group and in stage I disease respectively.

The stage, grade, menopausal status, modality of therapy and the age; each separately are found to be one of the prognostic factors; “the stage” being the major fact determining the prognosis among the others.

The overall survival times are less satisfying with those patients who had undergone a suboptimal surgical procedure referring to the results in our study. Therefore carrying out the standard surgical procedure TAH + BSO seems to be primordial in such cases. The contribution of a nodal dissection accompanying to this intervention is not defined.

RT with chemotherapy or RT alone as adjuvant therapies improve the results both with OS and DFS; whereas chemotherapy alone doesn’t ameliorate the outcomes.

Consequently in an adjuvant therapy setting; RT alone over 54 grays or with chemotherapy must be applied.

Références

2 Omura GA, Blessing JA, Major F, Lifshitz S, Ehrlich CE, Mangan C, et al. A randomized clinical trial of adjuvant adriamycin in uterine sarcomas: a Gynecologic Oncology Group Study. J Clin Oncol 1985; 3: 1240-5.

3 van Nagell JR, Hanson MB, Donaldson ES, Gallion HH. Adjuvant vincristine, dactinomycin, and cyclophosphamide therapy in stage I uterine sarcomas: a pilot study. Cancer 1986; 57: 1451-4.

4 Gadducci A, Cosio S, Romanini A, Genazzani AR. The management of patients with uterine sarcoma: a debated clinical challenge. Crit Rev Oncol Hematol 2008; [Epub ahead of print].

5 Papadimitriou CA, Zorzou MP, Markaki S, Rodolakis A, Voulgaris Z, Bozas G, et al. Anthracycline-based adjuvant chemotherapy in early-stage uterine sarcomas: long-term results of a single institution experience. Eur J Gynaecol Oncol 2007; 28: 109-16.

6 Olah KS, Gee H, Blunt S, Dunn JA, Kelly K, Chan KK. Retrospective analysis of 318 cases of uterine sarcoma. Eur J Cancer 1991; 27: 1095-9.

7 Lurain JR, Piver MS. Uterin sarcomas: clinical features and management. In: Coppleson M, ed. Gynecolojic oncology. 2nd ed. Edinburgh: Churchill Livingstone, 1992.

8 Olah KS, Dunn JA, Gee H. Leiomyosarcomas have a poorer prognosis than mixed mesodermal tumours when adjusting for known prognostic factors: the result of a retrospective study of 423 cases of uterine sarcoma. Br J Obstet Gynaecol 1992; 99: 590-4.

9 Moskovic E, MacSweeney E, Law M, Prize A. Survival, patterns of spread and prognostic factors in uterine sarcoma: a study of 76 patients. Br J Radiol 1993; 66: 1009-15.

10 Salazar OM, Bonfiglio TA, Patten SF, Keller BE, Feldstein ML, Dunne ME, et al. Uterine sarcomas: analysis of failures with special emphasis on the use of adjuvant radiation therapy. Cancer 1978; 42: 1161-70.

11 Salazar OM, Bonfiglio TA, Patten SF, Keller BE, Feldstein ML, Dunne ME, et al. Uterine sarcomas: natural history, treatment and prognosis. Cancer 1978; 42: 1152-60.

12 Atalar B, Okkan S, Meral G, Sahinler I, Calay Z, Koca S, et al. Postoperative radiotherapy in the treatment of uterine sarcomas: Long-term results and analysis of prognostic factors. J Clin Oncol 2004; 22: 9045.

13 Akahira JI, Tokunaga H, Toyoshima M, Takano T, Nagase S, Yoshinaga K, et al. Prognoses and prognostic factors of carcinosarcoma, endometrial stromal sarcoma and uterine leiomyosarcoma: a comparison with uterine endometrial adenocarcinoma. Oncology 2007; 71: 333-40.

14 Leath CA, Huh WK, Hyde J, Cohn DE, Resnick KE, Taylor NP, et al. A multi-institutional review of outcomes of endometrial stromal sarcoma. Gynecol Oncol 2007; 105: 630-4.

15 Kim SH, Kim JW, Kim YT, Kim JH, Yoon BS, Ryu HS. Prognostic factors and expression of p53 and mdm-2 in uterine sarcomas. Int J Gynaecol Obstet 2006; 95: 272-7.

16 Chauveinc L, Deniaud E, Plancher C, Sastre X, Amsani F, de la Rochefordiere A, et al. Uterine sarcomas: the Curie Institut experience. Prognosis factors and adjuvant treatments. Gynecol Oncol 1999; 72: 232-7.

17 Evans HL, Chawla SP, Simpson C, Finn KP. Smooth muscle neoplasms of the uterus other than ordinary leiomyoma. A study of 46 cases, with emphasis on diagnostic criteria and prognostic factors. Cancer 1988; 62: 2239-47.

18 Chang KL, Crabtree GS, Lim-Tan SK, Kempson RL, Hendrickson MR. Primary uterine endometrial stromal neoplasms. Am J Surg Pathol 1990; 14: 415-38.

19 Belgrad R, Elbadawi N, Rubin P. Uterine sarcoma. Radiology 1975; 114: 181-8.

20 O’Hanlan KA, Pinto RA, O’Holleran M. Total laparoscopic hysterectomy with and without lymph node dissection for uterine neoplasia. J Minim Invasive Gynecol 2007; 14: 449-52.

21 Moinfar F, Azodi M, Tavassoli F. Uterine sarcomas. Pathology 2007; 39: 55-71.

22 Perez CA, Askin F, Baglan RJ, Kao MS, Krauss FT, Perez BM, et al. Effects of irradiation on mixed mullerian tumors of the uterus. Cancer 1979; 43: 1274-84.

23 Monk BJ, Solh S, Johnson MT, Montz FJ. Radical hysterectomy after pelvic irradiation in patients with high risk cervical cancer or uterin sarcoma: morbidity and outcome. Eur J Gynecol Oncol 1993; 14: 506-11.

24 Le Péchoux C, Pautier P, Delannes M, Bui BN, Bonichon F, Bonvalot S, et al. Clinical practice guidelines: 2006 update of recommendations for the radiotherapeutic management of patients with soft tissue sarcoma (sarcoma of the extremity, uterine sarcoma and retroperitoneal sarcoma. Cancer Radiother 2006; 10: 185-207.

25 Hornback NB, Omura G, Major FJ. Observations on the use of adjuvant radiation therapy in patients with stage I and II uterine sarcoma. Int J Radiat Oncol Biol Phys 1986; 12: 2127-30.

26 Berchuck A, Rubin SC, Hoskins WJ, Saigo PE, Pierce VK, Lewis JL. Treatment of uterine leiomyosarcoma. Obstet Gynecol 1988; 71: 845-50.

27 Rose PG, Boutselis JG, Sachs L. Adjuvant therapy for stage I uterine sarcoma. Am J Obstet Gynecol 1987; 156: 660-2.

28 Töre G, Topuz E, Bilge N, Aslay I, Dincer M, Elgin A. The role of adjuvant chemotherapy in the treatment of uterine sarcoma patients. Eur J Gynecol Oncol 1990; 11: 307-12.

29 Hempling RE, Piver MS, Baker TR. Impact on progression-free survival of adjuvant cyclophosphamide, vincristine, doxorubicin (adriamycin), and dacarbazine (Cyvadic) chemotherapy for stage I uterine sarcoma. A prospective trial. Am J Clin Oncol 1995; 18: 282-6.

30 Muss HB, Bundy B, DiSaia PJ, Homesley HD, Fowler WC, Creasman W, et al. Treatment of recurrent or advanced uterine sarcoma: a randomized trial of doxorubicin versus doxorubicin and cyclophosphamide (a phase III trial of the Gynecologic Oncology Group). Cancer 1985; 55: 1648-53.