DISCUSSION
Soft tissue sarcomas (STS) in children are rare tumours, approximately 1000 cases are diagnosed annually in the United States of America.11,12 These tumours are heterogenous in nature and can occur in any part of the body.13 This study documented 112 paediatric soft tissue sarcomas over a 26 year period, representing roughly 4.3 cases every year. This result is similar to an earlier reported figure of 4.7 per year from our centre. This study however differs with Dauda et al. (Jos Nigeria) and Kachanovet al. (Moscow) who reported 8.9 and 6.8 per year respectively in children less than 15 years of age.14–16 This difference can be attributed to different study intervals. The studies by the afore mentioned authors (Dauda et al. and Kachanovet al. ) each covered a 10-year period compared to this present study. Another reason for this difference may be due to different disease demographics among different study populations.
Soft tissue sarcomas (STS) are classified broadly into Rhabdomyosarcomas( RMS) and non-rhabdomyosarcomas (NRMS) otherwise termed non-rhabdomyosarcoma soft tissue sarcomas (NRSTS).1,11 Most studies report rhabdomyosarcoma (RMS) as the commonest paediatric STS similar to this study2,11,15–17 The predominance of embryonal and alveolar RMS supports the childhood occurrence of these sarcoma subtypes while the very few pleomorphic RMS supports the rare occurrence of these tumours in children.12,13 Other adult-type STS seen in this study differ somewhat in spectrum from those seen elsewhere.18 Leiomyosarcoma and liposarcoma reported by Ferrari et al . were not seen in this study.18 Also, Ferrari et al. reported synovial sarcoma as the commonest NRSTS tumours, accounting for 34% of their cases contrary to vascular sarcomas seen in this study as the commonest NRSTS.18 Other researchers have documented other sarcoma subtypes as being the commonest among the NRSTS supporting the heterogeneity of these tumours.1 Occurrence of these NRSTS in the paediatric age suggests similar yet to-be identified etiologic factors.
No difference in tumour occurrence was seen between male and female gender in this study suggesting a common pathogenetic pathway. Studies have suggested genetic alterations such as Li Fraumeni syndrome and exposures to radiation as possible causes of RMS while the causes and risk factors for NRSTS are not known.11 The modal age in our study differs from those of Dauda et al. in Nigeria and Kachanov et al. in Moscow which documented older age groups.15,16 This difference is due to the differences in distribution of tumour subtypes across the age groups in these studies. In the study by Kachanov et al . RMS constituted 23% of tumours in the age group 10-14 years while the NRSTS tumours were predominant in this age group although RMS accounted for 54.4% of all the STS.16 Likewise, alveolar RMS and Embryonal RMS occurred with higher frequency in the age group 12-15 years in the study by Dauda et al , compared to the 5-9 years seen in this present study.15
Anatomic location for STS varies in different studies. We documented head and neck as the commonest location. This is in contrast to Daudaet al . whose study reported extremities as the commonest location.15 This may suggest different pathogenesis. For example, embryonic cell arrest during migration in caudal-cephalad direction with malignant transformation may explain the higher frequency of embryonal and alveolar rhabdomyosarcoma in such regions.
The known prognostic factors for STS we studied include tumour size, grade and stage, histological subtype and location. Ferrari et al . found among NRSTS, increasing tumour depth and size to be more strongly predictive of survival among children compared to their adult counterparts.18 Deep tumours and tumour size greater than 5cm were associated with poorer survival while tumours of the head and neck showed better survival.18 Tumour size not only determines tumour burden but also surgical resection as a modality of treatment and this factor affects tumour recurrence.12 Rhabdomyosarcomas are by their nature high grade tumours although very chemo- and radiosensitive.13 Overall, most STS in this study were of high grade status. The mean tumour size of 7.4cm, higher frequency of grade 3 tumours and tumour stage suggests worse prognosis for tumours seen in our practice environment. On the other hand, the higher frequency of head and neck location of STS might suggest a favourable prognosis in addition to the predictability of response to therapy if offered to these patients.12 Patients with deep tumours or tumour size >5cm may therefore need to be treated with multimodal approach.19
Absence of follow up data in this study has made it difficult to examine the effects the poorer clinicopathological parameters with patient overall survival. Prospective studies should incorporate data on adjuvant therapies and survival as these may suggest additional biological behaviours of tumours in our environment.