Abstract
Objectives
To assess the affects of a delay in diagnosis on the survival rates of Bone Sarcoma (BS) using size and symptom duration as measures of delay.
Methods
All patients diagnosed with a Primary BS from 1970 to 2005 were included. Demographic data concerning age, sex, diagnosis and tumour site were recorded. The data were collected retrospectively from a prospective database, with 2573 patients included. Symptom duration is defined as the length of time the patient was experiencing symptoms for until the date of diagnosis. Survival analysis was performed using Kaplan-Meier curves and Cox Regression was carried out to identify variables affecting outcome.
Results
2489 patients with Primary BS that matched out inclusion and exclusion criteria were identified. Osteosarcoma (40.4%), Chondrosarcoma (20.6%) and Ewing's Sarcoma (17.6%) were the most common diagnoses, with a median size of 10cm at presentation. Overall 5 year survival was 61.2%, extremity sited tumours carried a statistically significant better prognosis compared to axial tumours (5yr survival 63.2% vs. 53.3%) and patients with axial tumours had significantly longer symptom duration (Mann-Whitney U test p = 0.000 31 and 12 weeks). Overall analysis of all BS diagnoses showed lengthier duration of symptoms was associated with a better outcome (Hazard Ratio = 0.998, p = 0.000), which suggests that 1 extra week of symptoms confers an increased survival of 0.2%. Larger tumour size was associated with a poorer survival rate (p=0.000 to 3 sig. fig, Hazard Ratio 1.059 respectively) which suggests that for every 1cm increase in size there is a decreased survival of 5.9%.
Conclusions
Although many factors have been associated with survival, the limited literature concerning symptom interval and outcome is varied. This large study has shown that regardless of diagnosis, overall, shorter symptom duration and a large tumour size are associated with a poorer prognosis.
