Table 1

Performance of the classifier. Performance of the 13 clusters classifier algorithm re-training and validation on the separate institutional populations using both leave-one-out and multiple 10-fold cross-validations.

Training set (total/events)
Validation set (total/events)
Hazard ratio (95%CI)
Log rank p value
Distant Metastases Free Survival for low risk group only#









DMFS at 3 years
DMFS at 5 years
DMFS at 10 years

OXFT (99/19)
KIT/GUYT(156/48)
2.17 (1.2–3.91)
<0.00001
91%
87%
79%
KIT (69/20)
OXFT/GUYT (186/47)
4.07 (2.23–7.41)
<0.00001
96%
92%
88%
GUYT (87/28)
OXFT/KIT (168/39)
5.93 (3.0–11.75)
<0.00001
93%
89%
82%
KIT/GUYT (156/48)
OXFT (99/19)
14.59 (5.38–39.5)
<0.00001
97%
94%
91%
OXFT/GUYT (186/47)
KIT (69/20)
3.44 (1.36–8.67)
0.005
96%
92%
84%
OXFT/KIT(168/39)
GUYT (87/28)
2.23 (1.05–4.71)
0.03
96%
92%
84%
Leave-one-out cross validation (255/67)*
3.86 (2.32–6.41)
<0.0001
94%
91%
84%
Multiple 10-fold cross-validation (255/67)
3.23 (2.66–3.84)
<0.0001
94%
90%
83%

# as estimated by Kaplan Meier survival curves.

Patients samples obtained from: OXFT: John Radcliffe Hospital, Oxford, UK; KIT Uppsala University hospital, Uppsala, Sweden; GUYT Guys hospital, London, UK.

• Reported results.

Loi et al. BMC Genomics 2008 9:239   doi:10.1186/1471-2164-9-239