Delay in diagnosis of primary osteosarcoma of bone in children: Have we improved in the last 15 years and what is the impact of delay on diagnosis?

Abstract

Aims: Examine: (1) If length of symptoms (LS) of children with osteosarcoma has improved in the last 15 years (2) Is delay in diagnosis related to the presence of metastases at presentation? (3) The impact of delay in diagnosis on prognosis.

Methods: 250 consecutive patients with a diagnosis of osteosarcoma of bone treated at a national bone tumor treatment center between 2004 and 2018 were studied retrospectively. Three groups comprising those diagnosed over a five-year period (Group 1: 2004-2008, Group 2: 2009-2013; Group 3: 2014-2018) were studied.

Results: There were 126 males and 124 females with a mean age 12.2 years. The median LS for all patients was eight weeks. The median LS for Group 3 was significantly shorter than that for other groups. Development of metastasis during follow-up period was significantly less in Group 3 compared to the other groups. Overall survival gradually improved over the whole study period. There was no difference in the proportion presenting with metastases at diagnosis between the three groups. The survival rates in patients with LS shorter than 4 weeks was better than those with LS longer than 4 weeks, irrespective of the study time period.

Conclusions: There has been an improvement in the LS in patients diagnosed with osteosarcoma over the last 15 years. The development of metastasis during follow-up has reduced and the overall survival in the last 15 years has improved. LS longer than 4 weeks is associated with a poorer prognosis.

Keywords: Delayed diagnosis; Osteosarcoma; Pediatric osteosarcoma; Prognostic factor; Symptom length.

Fig. 1

Flowchart of this study.

Fig. 2

Comparison of length of symptoms (LS) grouped by presence or absence of metastasis at diagnosis. There were no significant differences in the LS of those with or without metastases at diagnosis.

Fig. 3

Kaplan–Meier curves of time to dead of disease grouped by length of symptoms (LS). Cut-off value; a 8 weeks, b 4 weeks. There were no significant differences of survival rate between two groups divided by LS at eight weeks (a). However, the survival rates in the group with LS shorter than 4 weeks were better than those in groups with longer than 4 weeks (b).

Fig. 4

Comparison of length of symptoms (LS) grouped by involved site (Extremity or Pelvis). The LS of pelvic cases trend to longer than extremity cases (it was not significant).