Uterine sarcomas are rare and, consequently, data supporting the use of adjuvant radiotherapy in uterine sarcomas consist of few randomised studies and multiple single-institution retrospective reports. It is becoming increasingly clear that each histologic subtype of uterine sarcoma is a distinct entity for which tailored treatment recommendations are needed. In this review, we analysed the effect of adjuvant radiotherapy for the main histologic subtypes of uterine sarcomas. When grouping all histologies, adjuvant radiotherapy has been shown in most studies to reduce local-regional failure without an overall survival advantage, as distant failure is the predominant pattern of relapse. Carcinosarcomas have the strongest indication for adjuvant radiotherapy, especially in early stage disease. Women with leiomyosarcomas and endometrial stromal sarcomas receiving adjuvant radiotherapy have improved local control compared with women undergoing surgery alone. As distant failure rates decrease with improvements in systemic therapies, there may be a broader indication for adjuvant radiotherapy.
Introduction
The role of adjuvant radiotherapy in the curative management of uterine sarcoma remains an area of active debate. Larger studies have reported 5-year local control rates of 38–43%, suggesting that current treatment approaches may not be sufficient in completely eradicating pelvic disease. As these tumours are rare, constituting only 2–5% of all uterine neoplasms, data guiding the use of adjuvant radiotherapy are limited. Only two randomised-controlled phase III trials have been published with adjuvant radiotherapy as the study question, and both were plagued with limited participant numbers and slow accrual. The remaining data consist of retrospective single-institution studies and large database analyses, which are subject to their inherent biases. The three distinct histological subtypes of uterine sarcoma are carcinosarcoma, leiomyosarcoma (LMS), and endometrial stromal sarcoma (ESS). Each has a different clinical behaviour for which tailored therapy is critical. Owing to small participant numbers, most reports have grouped these subtypes into a single category, making it difficult for clinicians to interpret the data.
Adjuvant radiotherapy usually consists of external beam radiotherapy to the pelvis, to address microscopic disease in the pelvic lymph nodes and surgical bed after total abdominal hysterectomy (TAH), bilateral salpingo-oopherectomy (BSO), and pelvic or para-aortic lymph-node sampling. This is commonly accompanied by intra-cavitary brachytherapy to the vaginal cuff, which can be administered by a cylinder applicator using either a low-dose or high-dose rate technique.
In this review, we describe the role of adjuvant radiotherapy in uterine sarcoma, emphasising the results from randomised-controlled trials and highlighting retrospective studies with substantial participant numbers. For clarity and to aid clinical decision-making, we will address the application of adjuvant radiotherapy as it pertains to each of the main histological subtypes of uterine sarcoma.
Carcinosarcomas
Carcinosarcoma, formerly designated as malignant mixed Müllerian tumour, is the most common and well-studied histological subtype of uterine sarcoma pertaining to adjuvant radiotherapy. Two large reviews of the Survey, Epidemiology and End Results cancer registry by Smith et al. and Nemani et al. both reported the lack of overall survival benefit associated with adjuvant radiotherapy in stage I–III carcinosarcoma, whereas Smith et al. showed an overall survival benefit in women with stage IV disease. Several retrospective non-randomised-controlled studies that included all subtypes have indicated that there may be an overall survival advantage with adjuvant radiotherapy, whereas others do not show this to be the case. In a study of 3650 women from the National Oncology Database, Sampath et al. also failed to show adjuvant radiotherapy to be predictive for overall survival on multivariate analysis. Given the high propensity for early haematogenous spread in uterine sarcomas, it is not surprising that adjuvant radiotherapy would not translate into an overall survival benefit. Randomised and non-randomised studies that included all uterine sarcoma subtypes are presented in Table 1 . Studies reporting local outcomes, specifically in carcinosarcomas, are presented in Table 2 .
| Study | Stages | Years | N | Radiotherapy details | Median follow up (years) | Lymphadenectomy | Overall survival | Local failure | Distant metastases | |
|---|---|---|---|---|---|---|---|---|---|---|
| Randomised studies | ||||||||||
| Reed et al. | No radiotherapy | I–II | 1988–2001 | 109 | 50.4 Gy | Not given | 25% | 63%; 5 years | 36%; 5 years | 34%; 5 years |
| Radiotherapy | 110 | 63%; 5 years | 19%; 5 years; P = 0.0013 | 45%; 5 years; P = not significant | ||||||
| Non-randomised studies | ||||||||||
| Sampath et al. | No radiotherapy | 1980–2005 | 1422 | External beam radiotherapy with or without brachytherapy | 5 | Not given | 37%; 5 years | 15%; 5 years | Not given | |
| Radiotherapy | I–IV | 784 | 37%; 5 years | 7%; 5 years; P < 0.05 | ||||||
| Brooks et al. | No radiotherapy | I–IV | 1989–1999 | 1365 | Not given | Not given | Not given | (Stage II) 31% 55%; P < 0.01 | Not given | Not given |
| Radiotherapy | 1312 | (Stage III–IV) 25% 33%; P < 0.05 | ||||||||
| Livi et al. | No radiotherapy | I–IV | 1974–2001 | 36 | 3 | 0% | Not given | a 62%; 3 years, 57%; 5-years | Not given | |
| External beam raditheapy | 37 | 36–60 Gy | 31%; 3 years; 30%; 5 years | |||||||
| External beam radiotherapy and brachytherapy | 27 | 45 Gy + 8 Gy ovoid | 28%; 3 years; 32%; 5 years | |||||||
a 5-year local failure results for women diagnosed with stage 1 uterine sarcomas.
| Study | Stages | Years | N | Radiotherapy details | Lymphadenectomy | Median follow up (years) | Overall survival | Local failure | Distant metastases | |
|---|---|---|---|---|---|---|---|---|---|---|
| Randomised prospective studies | ||||||||||
| Reed et al. | No radiotherapy | I–II | 1988–2001 | 45 | 25% (all women) | Not given | Not given | 47% a | 29% a | |
| Radiotherapy | 46 | 50.4 Gy external beam radiotherapy | 24% a | 35% a | ||||||
| Wolfson et al. | No radiotherapy (chemotherapy) | I–IV | 1993–2005 | 101 | Not given | 5.3 | 45% a | 24% a | 53% a | |
| Radiotherapy | 105 | Whole abdomen 30 Gy; 49.8 Gy pelvis | 35% | 17% a | 56% | |||||
| Non-randomised retrospective studies | ||||||||||
| Sampath et al. | No radiotherapy | I–IV | 1980–2005 | 638 | External beam radiotherapy with or without brachytherapy | Not given | 5 | Not given | 20%, 5 years | Not given |
| Radiotherapy | 490 | 10%, 5 years; P < 0.05 | ||||||||
| Sorbe et al. | No radiotherapy | I–II | 1975–2003 | 4 | 19.6–50 Gy external beam radiotherapy | 7% (all women) | 10 (minimum) | Not given | 25% a | Not given |
| Radiotherapy | 36 | 22% a | ||||||||
| Major et al. | Noradiotherapy | I–II | 1979–1988 | 182 | Not given | 95% (all women) | Not given | Not given | 25% a | 52% a |
| Radiotherapy | 119 | 17% a | 42% a | |||||||
| Smith et al. | No radiotherapy | I–IV | 1973–2003 | 1571 | Not given | Not given | 3.9 | 33%, 5 years | Not given | |
| Radiotherapy | 890 | 42%, 5 years | ||||||||
| Chi et al. | No radiotherapy | I–II | 1975–1993 | 10 | 40–50.4 Gy external beam radiotherapy and brachytherapy | 45% | 6.25 | 60%, 5 years | 50%, 5 years | 40%, 5 years |
| Radiotherapy | 28 | 59%, 5 years | 21%, 5 years; P = 0.09 | 43%, 5 years | ||||||
| Gerszten et al. | No radiotherapy | I–III | 1977–1992 | 31 | 45–50.4 Gy external beam radiotherapy | 73% | 2.7 | 53%, 5–year estimate | 55%, 3–year estimate | Not given |
| Radiotherapy | 29 | 82%, 5–year estimate | 3%, 3–year estimate | |||||||
| Nemani et al. | No radiotherapy | I–IV | Not given | 939 | Not given | 57% | Not given | 33%, 5 years | Not given | Not given |
| Radiotherapy | 653 | 36%, 5 years | ||||||||
| Vongtama et al. | Noradiotherapy | I–II | 1945–1972 | 10 | 30–60 Gy | Not given | Not given | 36%, 5 years | Not given | Not given |
| Radiotherapy | 14 | 57%, 5 years | ||||||||
| Echt et al. | No radiotherapy | I–IV | 1963–1984 | 8 | Mixed | Not given | Not given | Not given | 25% a | Not given |
| Radiotherapy | 22 | 0% a | ||||||||
| Callister et al. | No radiotherapy | I–IV | 1953–1998 | 113 | Mixed | 41% | 9 | 27%, 5 years | 48%, 5 years | Not given |
| Radiotherapy | 160 | 36%, 5 years, P = 0.07 | 28%,5 years; P = 0.001 | |||||||
Several reports have been published by national cooperative groups on the efficacy of radiotherapy. In the Gynaecological Oncology Group prospective surgical–pathological study by Major et al., 301 women with carcinosarcoma received TAH, BSO and pelvic lymph-node sampling. Adjuvant radiotherapy was not randomised and given at clinician discretion. When reporting site of first recurrence, the local failure rate with adjuvant radiotherapy was 17% (43 out of 182) compared with 24% (20 out of 119) with surgery alone. Statistical comparisons were not carried out. The number of pelvic failures declined with adjuvant radiotherapy. Higher rates of distant organs were found as first sites of failure in the adjuvant radiotherapy groups, leading the investigators to conclude that adjuvant radiotherapy should not be routinely given.
The first reported phase III trial in uterine sarcoma involving adjuvant radiotherapy was Gynecologic Oncology Group 150. Two hundred and six women with stages I–IV carcinosarcoma with less than 1 cm 3 residual in the abdomen underwent TAH and BSO, and randomised to whole abdominal-pelvic irradiation (WAI) or three cycles of cisplatin, ifosfamide and mesna (CIM). WAI consisted of 30.6 Gy to the whole abdomen and pelvis, followed by a 19.8 Gy pelvic boost. No brachytherapy was given. This trial opened in 1993, and completed accrual in 2005, encompassing a wide variety of radiotherapy delivery techniques. No statistically significant difference was found in overall survival (45% CIM v 35% WAI) or disease-free survival (58% WAI v 52% CIM) between the two arms. WAI led to a decline in the rate of vaginal failures (4% v 10%), whereas no difference was observed in the rate of pelvic failures (13%). Women receiving WAI had a higher rate of abdominal relapse (28% v 19%). No statistical tests comparing sites of failure were carried out owing to small subgroups. The investigators concluded that, owing to similar pelvic failure and overall survival rates, one may consider eliminating pelvic external beam radiotherapy and limit radiotherapy to vaginal cuff brachytherapy. Several caveats in this trial are worth mentioning. First, the 30.6 Gy dose to address microscopic disease in the abdomen was likely to be inadequate. Typically, doses to the pelvis are in the range of 45–50 Gy to treat microscopic tumour. This may help explain the higher rates of abdominal relapse in the WAI arm, given that 46% of the women in this trial were stage III. Second, the protracted accrual time resulted in modification of the radiotherapy fractionation schedule in the protocol from a twice-daily to a once-daily treatment. Given this heterogeneity of treatment delivery in the radiotherapy arm, the results should be interpreted with caution. The quality of the centralised review of the radiation portals, treatment centres, or both, needs to be seriously questioned, given the two fatal events of radiation hepatitis in the radiotherapy arm. Although the protocol specified that the liver should not be shielded during the whole-abdomen treatment, for 2 grade 5 events to occur where the prescribed dose to liver was specified to be under an established tolerance level is indicative of poor or older radiotherapy techniques, or both. Newer techniques such as intensity-modulated radiation therapy have been shown in gynaecological cancers to confer superior sparing of critical organs compared with conventional techniques, leading to significant decreases in grade 2 gastrointestinal toxicity. Also, small patient numbers precluded adequate statistical comparisons of patient subgroups. Finally, with chemotherapy regimens, agents and number of cycles have changed since this trial first opened and, therefore, these data have limited applicability to a current patients seen in the clinic.
The European Organization for the Research and Treatment of Cancer Gynaecological Cancer Group (EORTC–GCG) recently reported the results of a phase III randomised trial in uterine sarcoma involving radiotherapy, which included a post-surgery observation arm. A total of 224 women received TAH and BSO with optional lymph-node sampling, and were randomised to no further treatment or 50.4 Gy to the pelvis.Similar to GOG 150, this trial took 13 years to accrue, from 1988–2001. With a median follow-up of 6.8 years, no difference was found in overall survival or disease-free survival. Adjuvant radiotherapy conferred a significant reduction in local relapse; 44% v 24%; P = 0.004. Women in the adjuvant radiotherapy arm had fewer isolated local relapses than the observation arm (3 out of 110 [ 3%] v 20 out of 109 [18%]). In addition, the number of women having an initial local recurrence followed by distant failure declined (7 out of 109 [6%] v 1 out of 101 [1%]). Analysis of failures was also carried out by histology. In the carcinosarcoma group, a marked decrease in the rate of local-regional failures was observed (47% v 24%). A slightly higher rate of distant metastases was reported (35%) in the adjuvant radiotherapy arm. Owing to small subgroups, no statistical tests were conducted. The local–regional benefit with adjuvant radiotherapy in carcinosarcoma is analogous to the benefit with adjuvant radiotherapy shown in the early stage endometrial cancer randomised trials. This lends support to the argument that carcinosarcoma is a poorly differentiated endometrial cancer for which pelvic radiotherapy should be given in early stage disease. One important caveat is the low prevalence of lymph-node dissection (LND) (26%) in the EORTC-GCG study. If more thorough lymph-node evaluation was carried out, especially in carcinosarcoma, where the risk for occult pelvic and para-aortic disease is comparable to advanced stage endometrial cancer, one could argue that the additional cytoreduction could have diminished the additional local benefit of adjuvant radiotherapy. Unfortunately, specific locations of failures within the pelvis (nodes v tumour bed) were not reported in the EORTC study and would have been helpful in addressing this issue.
The effect of adjuvant radiotherapy in the context of carrying out a pelvic,para-aortic LND, or both, is poorly studied. Nemani et al. reported a significant overall survival benefit associated with LND, with a 5-year overall survival of 49%, compared with 35% for women who had not undergone LND. No overall survival difference, however, was detected in women receiving radiotherapy with or without LND. As this was a Survey, Epidemiology and End Results study, local failure data were unavailable. Two prospective randomised trials have shown the lack of a survival benefit with LND in endometrial cancer, indicating that the role of LND may be limited to prognostic purposes only. It may be reasonable to extrapolate from those trials and conclude that LND may have a similar role for carcinosarcoma. Further research, however, is needed.
Adjuvant radiotherapy has been shown in many studies of uterine sarcoma to improve local–regional disease control compared with surgery alone. Several reports have indicated that adjuvant radiotherapy does not add a local benefit. Sampath et al. conducted a retrospective database analysis of 3650 women with uterine sarcomas, the largest published to date. In their multivariate analysis for local–regional failure-free survival, adjuvant radiotherapy was associated with a significant 60% risk reduction compared with surgery alone, with a hazard ratio of 0.4, 95% confidence interval (0.3 to 0.6); P < 0.001. When analysing the women with carcinosarcoma ( n = 1877), the 5-year actuarial local–regional failure free survival was 90% v 80% ( P < 0.05), in favour of adjuvant radiotherapy. Several other retrospective studies also confirm an improvement in local relapse rates with adjuvant radiotherapy, specifically in carcinosarcoma ( Table 2 ), whereas others report better local control when including all types of uterine sarcomas ( Table 1 ). In a single-institution study of 273 women with carcinosarcoma, Callister et al. reported a highly significant local control advantage for adding radiation to surgery compared with surgery alone. The 5-year local failure rates were 48% and 28%, in favour of radiotherapy ( P = 0.0001). A trend was also observed for improved overall survival with radiotherapy at 5 years (36% v 27%; P = 0.07). Women receiving radiotherapy had a higher rate of initial failure at distant sites (37–38%) compared with the surgery-alone group (19%), which lends support to the idea that adjuvant radiotherapy can alter the failure pattern for these tumours.
Carcinosarcomas
Carcinosarcoma, formerly designated as malignant mixed Müllerian tumour, is the most common and well-studied histological subtype of uterine sarcoma pertaining to adjuvant radiotherapy. Two large reviews of the Survey, Epidemiology and End Results cancer registry by Smith et al. and Nemani et al. both reported the lack of overall survival benefit associated with adjuvant radiotherapy in stage I–III carcinosarcoma, whereas Smith et al. showed an overall survival benefit in women with stage IV disease. Several retrospective non-randomised-controlled studies that included all subtypes have indicated that there may be an overall survival advantage with adjuvant radiotherapy, whereas others do not show this to be the case. In a study of 3650 women from the National Oncology Database, Sampath et al. also failed to show adjuvant radiotherapy to be predictive for overall survival on multivariate analysis. Given the high propensity for early haematogenous spread in uterine sarcomas, it is not surprising that adjuvant radiotherapy would not translate into an overall survival benefit. Randomised and non-randomised studies that included all uterine sarcoma subtypes are presented in Table 1 . Studies reporting local outcomes, specifically in carcinosarcomas, are presented in Table 2 .
| Study | Stages | Years | N | Radiotherapy details | Median follow up (years) | Lymphadenectomy | Overall survival | Local failure | Distant metastases | |
|---|---|---|---|---|---|---|---|---|---|---|
| Randomised studies | ||||||||||
| Reed et al. | No radiotherapy | I–II | 1988–2001 | 109 | 50.4 Gy | Not given | 25% | 63%; 5 years | 36%; 5 years | 34%; 5 years |
| Radiotherapy | 110 | 63%; 5 years | 19%; 5 years; P = 0.0013 | 45%; 5 years; P = not significant | ||||||
| Non-randomised studies | ||||||||||
| Sampath et al. | No radiotherapy | 1980–2005 | 1422 | External beam radiotherapy with or without brachytherapy | 5 | Not given | 37%; 5 years | 15%; 5 years | Not given | |
| Radiotherapy | I–IV | 784 | 37%; 5 years | 7%; 5 years; P < 0.05 | ||||||
| Brooks et al. | No radiotherapy | I–IV | 1989–1999 | 1365 | Not given | Not given | Not given | (Stage II) 31% 55%; P < 0.01 | Not given | Not given |
| Radiotherapy | 1312 | (Stage III–IV) 25% 33%; P < 0.05 | ||||||||
| Livi et al. | No radiotherapy | I–IV | 1974–2001 | 36 | 3 | 0% | Not given | a 62%; 3 years, 57%; 5-years | Not given | |
| External beam raditheapy | 37 | 36–60 Gy | 31%; 3 years; 30%; 5 years | |||||||
| External beam radiotherapy and brachytherapy | 27 | 45 Gy + 8 Gy ovoid | 28%; 3 years; 32%; 5 years | |||||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
