Table 3. Multivariate analysis using Cox proportional hazard model in all RMS sample.
| Parametera | No. of samplesb | HR (CI 95%) OS | OS P-value |
|---|---|---|---|
| Primary tumour location | 0.002 | ||
| Favourable | 50 | 1 | |
| Unfavourable | 69 | 3.1 (1.45–7.33) | |
| Bone or bone marrow metastasis | 0.002 | ||
| No | 101 | 1 | |
| Yes | 18 | 3.3 (1.65–6.78) | |
| Fusion gene | 0.009 | ||
| None | 74 | 1 | |
| PAX7-FOXO1 | 10 | 1.2 (0.37–3.71) | |
| PAX3-FOXO1 | 35 | 3.0 (1.43–6.17) | |
| miR-206 expression | 0.078 | ||
| High | 96 | 1 | |
| Low | 23 | 1.9 (0.96–3.87) | |
Abbreviations: ARMS=alveolar RMS; CI=confidence interval; ERMS=embryonal RMS; OS=overall survival; RMS= rhabdomyosarcoma.
Parameters used in the stepwise Cox proportional hazard model: primary tumour location (favourable, unfavourable), presence of metastasis at diagnosis, presence of bone or bone marrow metastasis at diagnosis, age (favourable <10 years, unfavourable), fusion gene (none, PAX7-FOXO1, PAX3-FOXO1), histology (ERMS, ARMS), miR-206 expression (low, high). In particular, miR-206 expression was defined low if within the first quartile and high otherwise.
The stepwise analysis was performed on 116 patients who had information for all the parameters considered in our model as well as the follow-up, whereas the final coefficient values were calculated using the number of patients reported in each column.