INTRODUCTION

Ewing's sarcomas are malignant round cell tumours of the bone,1,2 with a neural cell origin.3 The pelvis is the second commonest site involved after the femur, accounting for 17.7% in the Mayo clinic series.4 Advances in multimodal treatment has improved the long-term cure rates from <10% to ≥50%,5,6 which may be attributed to the combination of aggressive, multi-agent chemotherapy and local control by radiotherapy and surgery.

Ewing's sarcomas in the pelvis have a worse prognosis than those in extremities,7,8 because in the pelvis there is no major anatomic barrier to tumour spread. Moreover, proximity to visceral organs and neurovascular bundles makes local control difficult.9,10

The optimal management of pelvic Ewing's sarcoma is controversial. Chemotherapy remains the mainstay of treatment, with increased survival in patients receiving ifosamide.11-13 Radiotherapy is useful when the surgical margins are inadequate or the histological response to chemotherapy is poor.9 Aggressive surgery has also been recommended.14-16

We reviewed various treatment modalities on the outcome of patients with pelvic Ewing's sarcomas from the Scottish Bone Tumour Registry.

MATERIALS AND METHODS

Between 1948 and 2004, 18 male and 15 female patients aged 3 to 48 (mean, 20) years with pelvic Ewing's sarcomas were retrospectively reviewed from the Scottish Bone Tumour Registry. Pelvic and chest radiography, bone scintigraphy and ultrasonography (since 1970), computed tomography and magnetic resonance imaging (since 1975), blood and histological (including PAS-staining) tests, as well as immunohistochemistry and bone marrow aspiration/biopsy (since early 1980s) were performed for diagnosis and evaluation of treatment. The definitive diagnosis was confirmed histopathologically by the presence of a small round cell tumour.

Treatments included surgery, radiotherapy, chemotherapy, or any of them in combination. Event-free (from presentation to recurrence) and overall (from presentation to death/latest followup) survival rates were calculated using the Kaplan- Meier method. Recurrences were defined as local, systemic, or both. Influence of various factors (age at diagnosis [categorised by decade], gender, symptom duration [cut-off at 6 months], tumour site, metastasis at presentation, surgery, radiotherapy, and type of chemotherapy) on survival was assessed using the log rank (Mantel-Cox) test. A p value of <0.05 was considered statistically significant.

RESULTS

The ilium was the commonest site involved (n=25), followed by the pubis (n=7) and the sacrum (n=1) [Fig. and Table 1]. Two patients with lumbar infiltration were initially diagnosed to have prolapsed lumbar intervertebral disc. Pain was the main presenting symptom (n=29; mean duration, 5 months), followed by swelling (n=12), restriction of hip movements (n=11), and fever (n=6). Age (p=0.09), gender (p=0.95), delay in presentation (6 months from symptom onset) [p=0.31], and tumour site (p=0.9) were not predictive of survival in the univariate analysis (Table 2).

According to the status of metastasis at presentation, 6 patients underwent surgery including hemipelvictomy (n=1), prosthetic hemipelvis replacement (n=1), and local resection (n=4). 25 patients underwent conventional radiotherapy (before 1980) or hyperfractionated irradiation (after 1980) of both primary (n=25) and secondary (n=11) sites. 29 patients underwent chemotherapy; 2 of them dropped out or had gaps in treatment and were excluded. 18 of the remaining 27 patients received ifosfamide/etoposide (after 1986) [Table 3]; 6 of them had metastasis at presentation of whom only one survived; the remaining 12 had no metastasis at presentation of whom 6 survived.

Response to treatment of all patients was assessed clinically (decrease in pain and swelling), radiologically, and by bone marrow aspiration/biopsy. Of the 9 patients who received chemotherapy and bone marrow biopsy, 5 survived out of 7 who had positive responses. Surgery (p=0.73) and radiotherapy (p=0.23) were not predictive of survival, but chemotherapy using ifosfamide was (p<0.01) [Table 2].

Of the 23 patients who had no metastasis at presentation (M0), 6 survived with a mean follow-up of 100 (range, 54-177) months, the rest were followed up until death (mean, 28 months; range, 6-105 months) [Table 4]. 16 and 7 patients developed metastasis in the lungs and bones, respectively; 6 in both sites; another 6 had no metastasis and survived (Table 5). The mean times to metastasise to the lungs and bones were 13 and 20 months, respectively. There were 5 local and 17 systemic (metastatic) recurrences.

Of the 10 patients who had metastasis at presentation (M1), one survived with a mean followup of 21 months, the rest were followed up until death (mean, 13 months; range, 6-26 months) [Table 4]. Three and 4 patients developed metastasis in the lungs and bones, respectively; 3 in both sites (Table 5). The mean times to metastasise to the lungs and bones were 6 and 10 months, respectively.

Metastasis at presentation was predictive of survival (p<0.01, Table 2). In 3 of the 10 patients, metastases were extensive and necessitated palliative therapy only. The overall survival (M0+M1) was 20% and the 5-year event-free survival (M0) was 35%, which was comparable to those in other studies.5,6

DISCUSSION

65% pelvic Ewing's sarcomas occur in the second decade of life, with a slight male preponderance.17 Patients with these tumours presenting with metastasis have a poor prognosis.15,18 Early presentation and diagnosis before metastasis appear to improve long-term survival. In our study, survival increased significantly in patients undergoing chemotherapy with ifosfamide, but this may be attributed to other treatment-related factors prevailing after 1986.

Treatment protocols should be customised to each patient based on size, site, stage of the tumour and histopathologic response to chemotherapy. The role of surgery in pelvic Ewing's sarcomas is controversial, because local control is difficult to achieve and mortality is high,15,18-20 and Ewing's sarcomas in the pelvis are more prone to recur than in other sites.18 In 2 studies from the Mayo clinic, patients undergoing surgery with or without adjuvant chemotherapy had higher survival rates than those treated with radiotherapy or chemotherapy alone.15,16 Compared to those treated with radiotherapy and chemotherapy, there was a 2-fold increase in survival rates in patients undergoing surgery in addition to radiation and chemotherapy.20 In 2 intergroup studies, surgery resulted in a better prognosis and lower recurrence rate in patients with primary pelvic tumours.14 In our study, the number of patients undergoing surgery was small and insufficient to make any valid conclusion. Heterogeneity between retrospective studies and variations in selection criteria make comparison of the effect of surgery on survival difficult to evaluate; prospective randomised studies are needed to assess its usefulness.

The usefulness of ifosfamide in the treatment of pelvic Ewing's sarcomas is controversial. In one study of 144 patients, it was considered to be of no benefit21; others claimed it was more effective for Ewing's sarcomas occurring inside, rather than outside, the pelvis.11 In a non-randomised study, the French Society of Pediatric Oncology reported no benefit from ifosfamide treatment of Ewing's sarcomas.13

Before 1986, chemotherapy comprised varying regimens of induction and maintenance therapy including vincristine, adriamycin, cyclophosfamide and actinomycin D. After 1986, ifosfamide replaced cyclophosphamide with the addition of etoposide to the cycles of maintenance chemotherapy. This practice was based on the recommendations set out in the ET-2/CESS-86 (Cooperative Ewing Sarcoma Study, Conducted by the German Society of Pediatric Oncology and Hematology).12 The difference in chemotherapy protocols between the 2 periods was associated with improved treatment outcomes. Nonetheless, other changes may be relevant, such as early pick-up, good supportive care and radiotherapy planning. Bone marrow aspiration and immunohistochemical markers help determine the response to chemotherapy (tumour regression), which is an important prognostic factor.

In our study, patients, diagnostic modalities, and treatment protocols were heterogeneous, with no accurate documentation of the tumour volume and lactic acid dehydrogenase levels. Patients in the earlier part of the study were not benefited from modern imaging techniques. At that time appraisal of histological response to chemotherapy, central radiation planning, and bone marrow aspiration/biopsy were not available. Assessment of treatment effects is difficult because of the low incidence of pelvic Ewing's sarcomas, long study period, and variations of surgical margins, tumour sites, and follow-up durations. Our study has limitation as the data spanned over 57 years. The positive effects of ifosfamide (implemented after 1986) certainly confounded with the time effect, which is possible due to a better understanding of the pathology of the disease leading to better treatment. More investigations on the effect of ifosfamide are therefore warranted.

Local control becomes more important, as long-term survival increases following the advent of modern chemotherapy, and patients may have more time to develop local recurrences. The 5 local recurrences in our series could be attributed to dropouts from chemotherapy or palliative treatment. We recommend intense multi-agent neo-adjuvant chemotherapy (to decrease tumour size) followed by wide excision and postoperative radiation.22 Local therapy should not take precedence over or interfere with systemic chemotherapy. Although surgery is emphasised in some studies, we recommend early aggressive multidrug chemotherapy including the use of ifosfamide.

© 2008 Western Pacific Orthopaedic Association Provided by ProQuest LLC. All Rights Reserved.