Uterine sarcomas comprise leiomyosarcoma, endometrial stromal sarcoma, adenosarcoma, undifferentiated endometrial sarcoma, and their variants. Carcinosarcoma is historically classified as sarcoma, but it is now regarded as a metaplastic carcinoma. Uterine sarcomas are rare, and are traditionally staged in the same way as endometrial carcinoma. Because of their different clinical and biological behaviours, the International Federation of Gynecology and Obstetrics introduced a new staging system in 2009 for leiomyosarcoma, endometrial stromal sarcoma and adenosarcoma, and carcinosarcoma, respectively. Following an extensive literature review no good evidence was found to support the modification of the staging system. This is mainly because of the rarity of the sarcomas and the heterogeneity of the reports, the different diagnostic criteria and treatments changing over the decades the retrospective nature and small sample size in most studies, and the lack of uniform pathological review even in large studies. Currently, evidence is still lacking about the use of preoperative imaging for staging purpose, and uterine sarcomas remain to be surgically staged. Total hysterectomy is the cornerstone for both staging and treatment. Newer evidence shows that routine lymphadenectomy and bilateral salpingo-oophorectomy may not be necessary, unless in the presence of extra-uterine spread, suspicious ovaries or lymph nodes, and certain poor histological types, such as undifferentiated endometrial sarcoma and adenosarcoma with sarcomatous overgrowth. More research and data collection are definitely needed in order to verify and further revise the current staging systems.
Introduction
Uterine sarcomas are rare, accounting for 3–8% of all uterine malignancies. They consist of a heterogeneous group of tumours, mainly leiomyosarcoma (LMS), endometrial stromal sarcoma (ESS), adenosarcoma, undifferentiated endometrial sarcoma (UES), and their variants. Uterine carcinosarcoma, also known as malignant mixed Müllerian tumour, is now regarded as a variant of endometrial carcinoma rather than a pure sarcoma, The sarcomatous component is the result of the metaplastic or de-differentiated change of the epithelial component. It is important to have a reliable and objective staging system to stratify women into different groups so that their prognosis can be predicted and treatment planned accordingly. A universally agreed staging system also allows different centres to compare their treatment outcomes. Such a system is developed on the basis of different clinical prognostic factors that are measurable, reproducible and easily available. In the 1988 staging system of the International Federation of Gynecology and Obstetrics (FIGO), uterine sarcomas were staged under the same system of endometrial carcinoma ( Table 1 ). This received much criticism because uterine sarcoma and endometrial carcinoma are separate disease entities with different clinical and biological behaviours. In 2009, the FIGO committee introduced a new staging system for uterine sarcomas with specification to LMS, ESS and adenosarcoma, and carcinosarcoma, respectively ( Table 2 ). Surgical staging is still the mode of assessment. Imaging is not incorporated into the staging procedure because it may not be available in developing countries, and the interpretation of the image results can be subjective. More importantly, at least half of the sarcomas are diagnosed as benign leiomyomas preoperatively, and there is no strong evidence to suggest that imaging modalities can accurately detect uterine sarcomas and metastasis. In a recent retrospective study of 92 women with sarcoma, only 9% had a change in surgical or post-surgical treatment as a result of their imaging studies. Many studies on uterine sarcomas are small because of the rarity of the tumours. It is also not clearly documented how the new FIGO system was derived. For this chapter, we conducted a search on Pubmed, and reviewed references of relevant articles and the evidence behind the modification of the staging system, with an emphasis on LMS and ESS. Information on adenosarcoma, however, is lacking and carcinosarcoma is regarded as a variant of endometrial carcinoma.
| Stage∗ | |
|---|---|
| I | Tumour confined to the corpus |
| Ia | Tumour limited to the endometrium |
| Ib | Invasion to less than half of the myometrium |
| Ic | Invasion equal to or more than half of the myometrium |
| II | Tumour involving cervix |
| IIa | Endocervical glandular involvement only |
| IIb | Cervical stromal invasion |
| III | Tumour limited to the pelvis |
| IIIa | Tumour invades the serosa of the corpus uteri, adnexae, or both, and/or positive cytological findings |
| IIIb | Vaginal metastases |
| IIIc | Metastases to pelvic, para-aortic lymph nodes, or both |
| IV | Tumour invades local structures or metastases in distant sites |
| IVa | Tumour invasion of bladder, bowel mucosa, or both |
| IVb | Distant metastases, including intra-abdominal metastasis, inguinal lymph nodes, or both |
| Leiomyosarcoma and endometrial stromal sarcoma∗ | |
|---|---|
| Stage | |
| I | Tumour limited to uterus |
| Ia | Less than 5 cm |
| Ib | Over 5 cm |
| II | Tumour extends to the pelvis |
| IIa | Adnexal involvement |
| IIb | Tumour extends to extrauterine pelvis tissue |
| III | Tumour invades abdominal tissues (not just protruding into the abdomen) |
| IIIa | One site |
| IIIb | More than one site |
| IIIc | Metastases to pelvic, para-aortic lymph nodes, or both |
| IV | |
| IVa | Tumour invades bladder, rectum, or both |
| IVb | Distant metastases |
| Adenosarcomas | |
| Stage | |
| I | Tumour limited to uterus |
| Ia | Tumour limited to endometrium or endocervix with no myometrial invasion |
| Ib | Less than or equal to half myometrial invasion |
| 1c | More than half myometrial invasion |
| II | Tumour extends to the pelvis |
| IIa | Adnexal involvement |
| IIb | Tumour extends to extrauterine pelvis tissue |
| III | Tumour invades abdominal tissues (not just protruding into the abdomen) |
| IIIa | One site |
| IIIb | More than one site |
| IIIc | Metastases to pelvic, para-aortic lymph nodes, or both |
| IV | |
| IVa | Tumour invades bladder, rectum, or both |
| IVb | Distant metastases |
| Carcinosarcoma | |
| Carcinosarcoma should be staged as carcinomas of the endometrium. | |
Leiomyosarcoma
Leiomyosarcoma accounts for 22.5 to 44% of all uterine sarcomas, and is the most common type of uterine sarcoma after carcinosarcoma. It is an aggressive tumour, with a 5-year survival rate from 18.8 to 68%. Although many studies show that stage is a strong and independent prognostic factor in LMS, the 5-year survival rate varies widely (40–85%) and recurrence rate is 38–50%, even for stage I–II disease. The wide range of reported survival rates and high recurrence rates of early staged disease are most likely a result of small sample sizes as well as the different diagnostic and inclusion criteria used in most studies. This also implies that the staging system is not ideal in categorising individuals into different prognostic groups. Before the development of the new staging system, most centres modified the 1988 FIGO staging system for LMS; stage I included tumour confined to the uterus; stage II included tumours involving the cervix; stage III included invasion of uterine serosa, positive peritoneal cytology, spread to pelvic organs and lymph nodes; and stage IV meant distant metastasis. Other soft-tissue sarcomas, however, are staged by the system developed by the American Joint Committee on Cancer (AJCC), which incorporates size and grade of tumours ( Table 3 ). The AJCC system regards all retroperitoneal and pelvic sarcomas as deep tumours. Hence, uterine LMS does not have T1a and T2a, and stages II and III differ by the size of the tumours only (≤5 cm v >5 cm). Two recent studies compared the predictive values of the modified FIGO and AJCC staging systems in uterine LMS, and both studies showed that neither of these two systems could classify women into four clinically meaningful subsets. For instance, in a study by Raut et al., involving 230 women, a significant difference was reported in progression-free survival only between women diagnosed with FIGO stage I and stage III (adjusted P = 0.0029 where adjusted P value <0.017 was significant) or stage IV uterine LMS (adjusted P < 0.0001) but not between other stages. Similarly, significant difference in overall survival was only observed between FIGO stages I and IV (adjusted P < 0.0001), with no difference between other stages.
| Stage | TNM∗ | Grade | ||
|---|---|---|---|---|
| I | T1a/b T2a/b | N0 | M0 | G1 |
| II | T1a/b T2a | N0 | M0 | G2-3 |
| III | T2b | N0 | M0 | G2–3 |
| IV | Any T | N1 | M0 | Any G |
| N0 | M1 | |||
Substantial changes have been made in the FIGO staging system for LMS. Not all, however, are evidence-based.
Stage I
Leiomyosarcoma arises from the smooth muscle of the uterus and so stage 1a in the 1988 FIGO staging system literally does not exist. Some investigators have attempted to determine the association of the depth of myometrial invasion with survival in LMS.
In a small study by Bodner et al., involving 21 women with LMS, univariate analysis showed that deep myometrial invasion was associated with poor overall survival ( P = 0.006). In another study, Hsieh et al. found no significant difference in 5-year, disease-free survival between women with and without myometrial invasion (37% v 88%; P = 0.129). Mostly, however, the assessment of myometrial invasion is considered to be difficult and may not carry any clinical relevance, especially in big tumours.
Instead of sub-staging stage I tumours according to the myometrial invasion, the current FIGO staging system divides it into stage Ia and stage Ib on the basis of tumour size, using 5 cm as the cut-off. Nevertheless, little evidence supports this. First, only 10–23% of the tumours are 5 cm or less, and the median size is 7–9 cm. Second, only a few studies suggested that tumour size was important. Nordal et al. reviewed 70 women with LMS, of whom 10% had tumours 5 cm or less. They found that 5-year survival reduced from 86% to 42% when the tumour increased from 5 cm to 10 cm (relative hazard [RH] 1.51, 95% CI 1.08 to 2.11; P = 0.016). Giuntoli et al. examined 208 women with LMS whose median tumour size was 9 cm. They showed that tumour size had significant prognostic value in disease-specific survival in the univariate analysis, where the median survival was over 30 years for 33 individual with tumours 5 cm or less, and reduced to 3.5 years for another 128 individuals with tumours greater than 5 cm ( P = 0.007). Tumour size was not evaluated in the multivariate analysis owing to the lack of reporting in some cases. In a Norwegian study involving 259 women with LMS and its variants, tumour size of over 10 cm was an independent prognostic factor in 5-year crude survival (RH 2.7; P < 0.001). Wu et al. also found in 51 women, that tumour size using 11 cm as cut-off correlated with 5-year overall survival (RR 11.63, 95% CI 2.14 to 63.12; P = 0.004) and relapse-free survival (RR 5.69, 95% CI 1.68 to 19.25; P = 0.005). Two other studies also showed that tumour size was a significant prognostic indicator for survival, but they mixed all sarcomas in the analysis, so the result must be interpreted with caution.
Contrary evidence has been presented in a number of studies. Pelmus et al. studied 72 women with LMS stage I–II, and found that tumour size using 7 cm as cut-off was not a prognostic factor in 5-year overall survival. In the AJCC staging system, which takes tumour size into account, Raut et al. showed no difference in progression-free survival and overall survival between stage II (≤5 cm) and stage III (>5 cm) ( P = 0.4162 and P = 0.5867, respectively). Zivanovic et al. also found an overlap in the survival outcomes between the AJCC stages II and III, where the 5-year progression-free survival rate was 24% and 16%, and the 5-year overall survival rate was 45% and 48% in stages II and III, respectively.
In the United States and Canadian Academy of Pathology annual meeting in 2010, Soslow et al. presented 123 women with stage I LMS. They found that only 23 people had tumours 5 cm or less. The overall survival for people with tumours 5 cm or less was 77 months (95% CI 0 to 169.2 months), compared with 73.9 months (95% CI 55.9 to 91.9 months) for larger tumours ( P = 0.3). The investigators challenged that their data did not support the sub-staging of stage I for LMS in the 2009 FIGO system.
Stage II
The new FIGO staging system does not take cervical involvement into consideration as does the AJCC system. Some investigators have considered the invasion of the cervix might in fact be the extension of myometrium instead of genuine invasion as in endometrial carcinoma. Also, the incidence of cervical involvement is only 0–14%, and the reported 5-year survival rate ranges from 25–100%. In a recent report from the Surveillance, Epidemiology and End Results (SEER) database, which included 1396 people with LMS, only 3.1% had cervical involvement. In addition, Raut et al. showed no difference in progression-free survival and overall survival in stage II from those in stage I or III. Zivanovic et al. also reported a prognostic overlap between FIGO stages II and III, where the 5-year progression-free survival rate was 8% in both stages, and the 5-year overall survival rate in stage II was even worse than in stage III (29% and 35%, respectively).
The FIGO revised staging system downstages adnexal spread from stage IIIa to IIa. In general, women in the previous stage III category have a poor prognosis, with 5-year survival rate ranging from 0–44.9%. The incidence of adnexal metastasis, however, is only about 3.5%. The survival outcomes of pure adnexal metastasis alone are not clearly known. In the Gynecologic Oncology Group study published in 1993, multivariate analysis showed that adnexal spread was not a significant prognostic factor related to progression-free interval. Similarly, few data could be found on extra-uterine pelvic spread. Raut et al. reported that the incidence of uterine serosal and vaginal invasion was 19.1% and 0.4%, respectively, and serosal involvement accounted for 94.6% of all women with stage III disease in their study. Because no significant difference was found in progression-free survival and overall survival between stages III and IV (adjusted P = 0.5153 and adjusted P = 0.0346), they suggested that women with serosal involvement had similar prognosis as individuals with stage IV diseases. On the basis of these limited data, it is difficult to draw conclusions about whether it is appropriate to place ovarian and extra-uterine spread in stage II; further studies are needed.
Stage III
As in carcinoma of corpus, the FIGO committee also removed peritoneal cytology from the staging system for sarcoma. Benito et al. studied 89 women with uterine sarcoma. Among the 20 women with LMS, only 5% had positive peritoneal cytology, 30% had negative cytology, and 65% did not have any record at all. Similarly, Leitao et al. found that, out of 41 women of whom peritoneal washing was taken (4.9%), only two women had positive peritoneal cytology. When studying all the sarcomas as a group, peritoneal cytology had no effect in overall survival in the multivariate analysis. Major et al. in their Gynecologic Oncology Group study, showed that positive peritoneal cytology was significantly related to progression-free survival in carcinosarcoma but not LMS.
We could not locate any information about the incidence of metastasis to abdominal tissues (excluding lymph node and distant metastasis), as transcoelomic spread is not a main route of metastasis in LMS. Similarly, the importance of the number of abdominal spread, as in stages IIIa and IIIb in the revised staging system, is also not known.
Retroperitoneal lymph-node metastasis remains to be staged as IIIc. The reported incidence of lymph-node metastasis varies from 0–28%. Many studies are limited by the small number of women receiving lymphadenectomy, and most of them do not define the extent of lymphadenectomy. In the largest study from the SEER database, which involved 348 women undergoing lymphadenectomy, lymph-node metastasis was identified in 23 women (6.6%). Those who had negative lymph nodes had a better 5-year, disease-specific survival rate than those with positive lymph nodes (64.2% v 26.0%; P < 0.001). Also, women with one positive lymph node had better prognosis than those with more than two positive lymph nodes (42.4% v 0%; P = 0.001). More studies are needed to ascertain whether the number of positive lymph nodes is important.
Stage IV
Stage IV includes bladder, rectum and all distant metastasis as in other cancer types. In general, women with stage IV disease have poor prognosis, with a 5-year survival rate of 0–28.7%.
Staging procedure
Leiomyosarcoma is surgically staged, and so the primary procedure should have staging and ideally curative effects. The gold standard is total hysterectomy and removal of as much tumour as possible, and this has been proven to have survival benefit. Although several case reports on myomectomy have been published on subtotal hysterectomy and trachelectomy, these are still not regarded as a standard procedure, and failure has been reported in women receiving myomectomy alone. In the past, bilateral salpingo-oophorectomy was frequently carried out. Recent evidence, however, shows that routine bilateral salpingo-oophorectomy and lymphadenectomy are not necessary.
Bilateral salpingo-oophorectomy
As mentioned previously, adnexal spread is uncommon. Leitao et al. found that the incidence of ovarian metastasis was similar between clinically stage I–II and stage III–IV diseases (2.8% v 5.4%). They also reported that the presence of lymphovascular invasion, tumour grade, serosal and cervical involvement could not predict ovarian metastasis. This implies that no clinical predictor can guide clinicians on whether to perform bilateral salpingo-oophorectomy. Nevertheless, Nordal et al. showed that no micrometastasis could be found in macroscopically normal ovaries.
Bilateral salpingo-oophorectomy in primary procedure may not have any therapeutic effect. Only one case report has been published of regression of pulmonary recurrence in two women with low-grade LMS after bilateral salpingo-oophorectomy. On the other hand, a retrospective study by Giuntoli et al. showed that bilateral salpingo-oophorectomy at primary surgery seemed to have an adverse effect on disease-specific survival (RR 3.25; P = 0.006). The investigators carried out a further case-control analysis, which included 25 pre-menopausal women with ovarian preservation and 25 control women with bilateral salpingo-oophorectomy matched bu stage, grade and age. No difference in disease-specific survival (median survival greater than 30 years in both arms; P = 0.489) and recurrence-free survival (14.3 years without bilateral salpingo-oophorectomy v greater than 25 years with BSO; P = 0.970) was found. In addition, Kapp et al. showed no difference in 5-year disease-specific survival in women who did not undergo oophorectomy compared with those who did (72.3% v 66.2%; P = 0.15). Among women younger than 50 years with stage I–II disease, the 5-year disease-specific survival was also similar between those who had and who had not undergone oophorectomy (83.2% v 83.2%; P = 0.445). Gadducci et al. showed that, among women diagnosed as stage I who were younger than 50 years, and who had undergone total hysterectomy, disease recurred in seven out of 21 women (33.3%) with bilateral salpingo-oophoectomy compared with five out of 21 women (23.8%) who retained one or two ovaries ( P > 0.05).
Although no clear predictor for ovarian metastasis exists, ovarian preservation can be considered in premenopausal women because the incidence of ovarian metastasis is low and the premature menopause carries no strong therapeutic benefit.
Lymph-node evaluation
Lymph-node metastasis is also rare, and some investigators have suggested lymph-node sampling only in selected cases. Most studies have not specified the extent of lymphadenectomy or lymph-node sampling, and the numbers of lymph nodes taken in different centres vary. Giuntoli et al. reported four women with lymph-node metastases out of 36 women (11%) who had lymph-node evaluation. Among these four women, one had no extra-uterine disease or enlarged lymph nodes. In the Gynecologic Oncology Group study, Goff et al. found that none of the nine women with clinically stage I disease had positive lymph nodes, compared with four out of six (66.6%) women with stage IV or recurrent diseases. This result was consistent with other subsequent studies, which found no lymph-node metastasis in tumours clinically confined to the uterus and cervix. In particular, Leitao et al. showed that extra-uterine diseases were a predictor of lymph-node metastasis, where none of the 27 women (0%) with stage I–II disease had lymph-node metastasis, compared with three out of 10 (30%) women diagnosed with stage III–IV disease who had clinically suspicious and enlarged lymph nodes ( P = 0.015); the presence of lymphovascular invasion, tumour grade, serosal and cervical involvement could not predict lymph-node metastasis.
Several studies have shown that lymphadenectomy carries no therapeutic effect. The largest study by Kapp et al. included 348 who had undergone lymphadenectomy and 1047 women who had not undergone lymphadenectomy. The investigators showed that lymphadenectomy failed to have any prognostic significance, as the 5-year disease-specific survival was similar between those who had and had not undergone lymphadenectomy (61.9% v 66.9%; P = 0.249). Another study by Ayhan et al. showed no difference in median disease-free survival (2.51 v 2.36 years; P = 0.4) and median overall survival (2.44 v 3.16 years; P = 0.7) between women who had and had not undergone lymphadectomy. The number of the resected lymph nodes was also not significant for both disease-free and overall survival. Giuntoli et al. found no difference in disease-specific survival between women diagnosed with stage I disease with histologically proven negative lymph nodes (with lymph-node sampling) and those with clinically negative lymph nodes (without lymph-node sampling).
On the basis of the evidence above, lymph-node sampling should only be carried out in women with clinically advanced diseases. Leitao et al. suggested that, if a diagnosis of LMS has been made by an intra-operative frozen section, lymph-node dissection should be carried out because of the possibility of carcinosarcoma on final pathological review.
Leiomyosarcoma
Leiomyosarcoma accounts for 22.5 to 44% of all uterine sarcomas, and is the most common type of uterine sarcoma after carcinosarcoma. It is an aggressive tumour, with a 5-year survival rate from 18.8 to 68%. Although many studies show that stage is a strong and independent prognostic factor in LMS, the 5-year survival rate varies widely (40–85%) and recurrence rate is 38–50%, even for stage I–II disease. The wide range of reported survival rates and high recurrence rates of early staged disease are most likely a result of small sample sizes as well as the different diagnostic and inclusion criteria used in most studies. This also implies that the staging system is not ideal in categorising individuals into different prognostic groups. Before the development of the new staging system, most centres modified the 1988 FIGO staging system for LMS; stage I included tumour confined to the uterus; stage II included tumours involving the cervix; stage III included invasion of uterine serosa, positive peritoneal cytology, spread to pelvic organs and lymph nodes; and stage IV meant distant metastasis. Other soft-tissue sarcomas, however, are staged by the system developed by the American Joint Committee on Cancer (AJCC), which incorporates size and grade of tumours ( Table 3 ). The AJCC system regards all retroperitoneal and pelvic sarcomas as deep tumours. Hence, uterine LMS does not have T1a and T2a, and stages II and III differ by the size of the tumours only (≤5 cm v >5 cm). Two recent studies compared the predictive values of the modified FIGO and AJCC staging systems in uterine LMS, and both studies showed that neither of these two systems could classify women into four clinically meaningful subsets. For instance, in a study by Raut et al., involving 230 women, a significant difference was reported in progression-free survival only between women diagnosed with FIGO stage I and stage III (adjusted P = 0.0029 where adjusted P value <0.017 was significant) or stage IV uterine LMS (adjusted P < 0.0001) but not between other stages. Similarly, significant difference in overall survival was only observed between FIGO stages I and IV (adjusted P < 0.0001), with no difference between other stages.
| Stage | TNM∗ | Grade | ||
|---|---|---|---|---|
| I | T1a/b T2a/b | N0 | M0 | G1 |
| II | T1a/b T2a | N0 | M0 | G2-3 |
| III | T2b | N0 | M0 | G2–3 |
| IV | Any T | N1 | M0 | Any G |
| N0 | M1 | |||
Substantial changes have been made in the FIGO staging system for LMS. Not all, however, are evidence-based.
Stage I
Leiomyosarcoma arises from the smooth muscle of the uterus and so stage 1a in the 1988 FIGO staging system literally does not exist. Some investigators have attempted to determine the association of the depth of myometrial invasion with survival in LMS.
In a small study by Bodner et al., involving 21 women with LMS, univariate analysis showed that deep myometrial invasion was associated with poor overall survival ( P = 0.006). In another study, Hsieh et al. found no significant difference in 5-year, disease-free survival between women with and without myometrial invasion (37% v 88%; P = 0.129). Mostly, however, the assessment of myometrial invasion is considered to be difficult and may not carry any clinical relevance, especially in big tumours.
Instead of sub-staging stage I tumours according to the myometrial invasion, the current FIGO staging system divides it into stage Ia and stage Ib on the basis of tumour size, using 5 cm as the cut-off. Nevertheless, little evidence supports this. First, only 10–23% of the tumours are 5 cm or less, and the median size is 7–9 cm. Second, only a few studies suggested that tumour size was important. Nordal et al. reviewed 70 women with LMS, of whom 10% had tumours 5 cm or less. They found that 5-year survival reduced from 86% to 42% when the tumour increased from 5 cm to 10 cm (relative hazard [RH] 1.51, 95% CI 1.08 to 2.11; P = 0.016). Giuntoli et al. examined 208 women with LMS whose median tumour size was 9 cm. They showed that tumour size had significant prognostic value in disease-specific survival in the univariate analysis, where the median survival was over 30 years for 33 individual with tumours 5 cm or less, and reduced to 3.5 years for another 128 individuals with tumours greater than 5 cm ( P = 0.007). Tumour size was not evaluated in the multivariate analysis owing to the lack of reporting in some cases. In a Norwegian study involving 259 women with LMS and its variants, tumour size of over 10 cm was an independent prognostic factor in 5-year crude survival (RH 2.7; P < 0.001). Wu et al. also found in 51 women, that tumour size using 11 cm as cut-off correlated with 5-year overall survival (RR 11.63, 95% CI 2.14 to 63.12; P = 0.004) and relapse-free survival (RR 5.69, 95% CI 1.68 to 19.25; P = 0.005). Two other studies also showed that tumour size was a significant prognostic indicator for survival, but they mixed all sarcomas in the analysis, so the result must be interpreted with caution.
Contrary evidence has been presented in a number of studies. Pelmus et al. studied 72 women with LMS stage I–II, and found that tumour size using 7 cm as cut-off was not a prognostic factor in 5-year overall survival. In the AJCC staging system, which takes tumour size into account, Raut et al. showed no difference in progression-free survival and overall survival between stage II (≤5 cm) and stage III (>5 cm) ( P = 0.4162 and P = 0.5867, respectively). Zivanovic et al. also found an overlap in the survival outcomes between the AJCC stages II and III, where the 5-year progression-free survival rate was 24% and 16%, and the 5-year overall survival rate was 45% and 48% in stages II and III, respectively.
In the United States and Canadian Academy of Pathology annual meeting in 2010, Soslow et al. presented 123 women with stage I LMS. They found that only 23 people had tumours 5 cm or less. The overall survival for people with tumours 5 cm or less was 77 months (95% CI 0 to 169.2 months), compared with 73.9 months (95% CI 55.9 to 91.9 months) for larger tumours ( P = 0.3). The investigators challenged that their data did not support the sub-staging of stage I for LMS in the 2009 FIGO system.
Stage II
The new FIGO staging system does not take cervical involvement into consideration as does the AJCC system. Some investigators have considered the invasion of the cervix might in fact be the extension of myometrium instead of genuine invasion as in endometrial carcinoma. Also, the incidence of cervical involvement is only 0–14%, and the reported 5-year survival rate ranges from 25–100%. In a recent report from the Surveillance, Epidemiology and End Results (SEER) database, which included 1396 people with LMS, only 3.1% had cervical involvement. In addition, Raut et al. showed no difference in progression-free survival and overall survival in stage II from those in stage I or III. Zivanovic et al. also reported a prognostic overlap between FIGO stages II and III, where the 5-year progression-free survival rate was 8% in both stages, and the 5-year overall survival rate in stage II was even worse than in stage III (29% and 35%, respectively).
The FIGO revised staging system downstages adnexal spread from stage IIIa to IIa. In general, women in the previous stage III category have a poor prognosis, with 5-year survival rate ranging from 0–44.9%. The incidence of adnexal metastasis, however, is only about 3.5%. The survival outcomes of pure adnexal metastasis alone are not clearly known. In the Gynecologic Oncology Group study published in 1993, multivariate analysis showed that adnexal spread was not a significant prognostic factor related to progression-free interval. Similarly, few data could be found on extra-uterine pelvic spread. Raut et al. reported that the incidence of uterine serosal and vaginal invasion was 19.1% and 0.4%, respectively, and serosal involvement accounted for 94.6% of all women with stage III disease in their study. Because no significant difference was found in progression-free survival and overall survival between stages III and IV (adjusted P = 0.5153 and adjusted P = 0.0346), they suggested that women with serosal involvement had similar prognosis as individuals with stage IV diseases. On the basis of these limited data, it is difficult to draw conclusions about whether it is appropriate to place ovarian and extra-uterine spread in stage II; further studies are needed.
Stage III
As in carcinoma of corpus, the FIGO committee also removed peritoneal cytology from the staging system for sarcoma. Benito et al. studied 89 women with uterine sarcoma. Among the 20 women with LMS, only 5% had positive peritoneal cytology, 30% had negative cytology, and 65% did not have any record at all. Similarly, Leitao et al. found that, out of 41 women of whom peritoneal washing was taken (4.9%), only two women had positive peritoneal cytology. When studying all the sarcomas as a group, peritoneal cytology had no effect in overall survival in the multivariate analysis. Major et al. in their Gynecologic Oncology Group study, showed that positive peritoneal cytology was significantly related to progression-free survival in carcinosarcoma but not LMS.
We could not locate any information about the incidence of metastasis to abdominal tissues (excluding lymph node and distant metastasis), as transcoelomic spread is not a main route of metastasis in LMS. Similarly, the importance of the number of abdominal spread, as in stages IIIa and IIIb in the revised staging system, is also not known.
Retroperitoneal lymph-node metastasis remains to be staged as IIIc. The reported incidence of lymph-node metastasis varies from 0–28%. Many studies are limited by the small number of women receiving lymphadenectomy, and most of them do not define the extent of lymphadenectomy. In the largest study from the SEER database, which involved 348 women undergoing lymphadenectomy, lymph-node metastasis was identified in 23 women (6.6%). Those who had negative lymph nodes had a better 5-year, disease-specific survival rate than those with positive lymph nodes (64.2% v 26.0%; P < 0.001). Also, women with one positive lymph node had better prognosis than those with more than two positive lymph nodes (42.4% v 0%; P = 0.001). More studies are needed to ascertain whether the number of positive lymph nodes is important.
Stage IV
Stage IV includes bladder, rectum and all distant metastasis as in other cancer types. In general, women with stage IV disease have poor prognosis, with a 5-year survival rate of 0–28.7%.
Staging procedure
Leiomyosarcoma is surgically staged, and so the primary procedure should have staging and ideally curative effects. The gold standard is total hysterectomy and removal of as much tumour as possible, and this has been proven to have survival benefit. Although several case reports on myomectomy have been published on subtotal hysterectomy and trachelectomy, these are still not regarded as a standard procedure, and failure has been reported in women receiving myomectomy alone. In the past, bilateral salpingo-oophorectomy was frequently carried out. Recent evidence, however, shows that routine bilateral salpingo-oophorectomy and lymphadenectomy are not necessary.
Bilateral salpingo-oophorectomy
As mentioned previously, adnexal spread is uncommon. Leitao et al. found that the incidence of ovarian metastasis was similar between clinically stage I–II and stage III–IV diseases (2.8% v 5.4%). They also reported that the presence of lymphovascular invasion, tumour grade, serosal and cervical involvement could not predict ovarian metastasis. This implies that no clinical predictor can guide clinicians on whether to perform bilateral salpingo-oophorectomy. Nevertheless, Nordal et al. showed that no micrometastasis could be found in macroscopically normal ovaries.
Bilateral salpingo-oophorectomy in primary procedure may not have any therapeutic effect. Only one case report has been published of regression of pulmonary recurrence in two women with low-grade LMS after bilateral salpingo-oophorectomy. On the other hand, a retrospective study by Giuntoli et al. showed that bilateral salpingo-oophorectomy at primary surgery seemed to have an adverse effect on disease-specific survival (RR 3.25; P = 0.006). The investigators carried out a further case-control analysis, which included 25 pre-menopausal women with ovarian preservation and 25 control women with bilateral salpingo-oophorectomy matched bu stage, grade and age. No difference in disease-specific survival (median survival greater than 30 years in both arms; P = 0.489) and recurrence-free survival (14.3 years without bilateral salpingo-oophorectomy v greater than 25 years with BSO; P = 0.970) was found. In addition, Kapp et al. showed no difference in 5-year disease-specific survival in women who did not undergo oophorectomy compared with those who did (72.3% v 66.2%; P = 0.15). Among women younger than 50 years with stage I–II disease, the 5-year disease-specific survival was also similar between those who had and who had not undergone oophorectomy (83.2% v 83.2%; P = 0.445). Gadducci et al. showed that, among women diagnosed as stage I who were younger than 50 years, and who had undergone total hysterectomy, disease recurred in seven out of 21 women (33.3%) with bilateral salpingo-oophoectomy compared with five out of 21 women (23.8%) who retained one or two ovaries ( P > 0.05).
Although no clear predictor for ovarian metastasis exists, ovarian preservation can be considered in premenopausal women because the incidence of ovarian metastasis is low and the premature menopause carries no strong therapeutic benefit.
Lymph-node evaluation
Lymph-node metastasis is also rare, and some investigators have suggested lymph-node sampling only in selected cases. Most studies have not specified the extent of lymphadenectomy or lymph-node sampling, and the numbers of lymph nodes taken in different centres vary. Giuntoli et al. reported four women with lymph-node metastases out of 36 women (11%) who had lymph-node evaluation. Among these four women, one had no extra-uterine disease or enlarged lymph nodes. In the Gynecologic Oncology Group study, Goff et al. found that none of the nine women with clinically stage I disease had positive lymph nodes, compared with four out of six (66.6%) women with stage IV or recurrent diseases. This result was consistent with other subsequent studies, which found no lymph-node metastasis in tumours clinically confined to the uterus and cervix. In particular, Leitao et al. showed that extra-uterine diseases were a predictor of lymph-node metastasis, where none of the 27 women (0%) with stage I–II disease had lymph-node metastasis, compared with three out of 10 (30%) women diagnosed with stage III–IV disease who had clinically suspicious and enlarged lymph nodes ( P = 0.015); the presence of lymphovascular invasion, tumour grade, serosal and cervical involvement could not predict lymph-node metastasis.
Several studies have shown that lymphadenectomy carries no therapeutic effect. The largest study by Kapp et al. included 348 who had undergone lymphadenectomy and 1047 women who had not undergone lymphadenectomy. The investigators showed that lymphadenectomy failed to have any prognostic significance, as the 5-year disease-specific survival was similar between those who had and had not undergone lymphadenectomy (61.9% v 66.9%; P = 0.249). Another study by Ayhan et al. showed no difference in median disease-free survival (2.51 v 2.36 years; P = 0.4) and median overall survival (2.44 v 3.16 years; P = 0.7) between women who had and had not undergone lymphadectomy. The number of the resected lymph nodes was also not significant for both disease-free and overall survival. Giuntoli et al. found no difference in disease-specific survival between women diagnosed with stage I disease with histologically proven negative lymph nodes (with lymph-node sampling) and those with clinically negative lymph nodes (without lymph-node sampling).
On the basis of the evidence above, lymph-node sampling should only be carried out in women with clinically advanced diseases. Leitao et al. suggested that, if a diagnosis of LMS has been made by an intra-operative frozen section, lymph-node dissection should be carried out because of the possibility of carcinosarcoma on final pathological review.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
