Table 1.
Marker | Control | PanIN1 | P Value | PanIN2/3 | P Value (vs Con) | P Value (vs PanIN1) | Invasive CA | P Value (vs Con) | P Value (vs PanIN1) |
CXCL16 | 0.65 ng/ml (N = 9) | 0.69 ng/ml (N = 5) | .753* | 1.45 ng/ml (N = 8) | .0004 | .0152 | 6.96 ng/ml (N = 6) | .0001 | .0001 |
TIMP1 | 2.9 ng/ml (N = 6) | 4.6 ng/ml (N = 6) | .035 | 6.6 ng/ml (N = 10) | .0021 | .0805* | 35.8 ng/ml (N = 6) | .0123 | .0162 |
LCN2 | 165 ng/ml (N = 17) | 337 ng/ml (N = 16) | .0008 | 1937 ng/ml (N = 10) | .0001 | .0001 | 4249 ng/ml (N = 9) | <.0001 | .0001 |
ELISA tests were performed according to the recommendations by the manufacturer for CXCL16, TIMP1, and LCN2/NGAL (R&D Systems). The ELISA plate was measured on an Emax plate reader (Molecular Devices, Munich Germany) and then analyzed with SOFTmax Pro (Version 3-0) software.
To determine whether our assays show differences between healthy (wild-type) and affected (transgenic) animals, we evaluated serum protein levels from each of the ELISAs using an unpaired t test.
Values of P < .05 were considered to be statistically significant.
Not statistically significant.