TABLE 4.
Adjusted Cox proportional hazards analysis for the prediction of DCGF a .
| Variables b | Hazard ratio (95% confidence interval) | p-value | Data (n) |
|---|---|---|---|
| Model 1 (biopsy variables) | 72 | ||
| IRRAT | 1.93 (0.96–3.89) | 0.067 | |
| CI3comp (ci+ct+[cgx2]) | 1.97 (0.97–3.99) | 0.059 | |
| ciprob | 1.24 (0.54–2.83) | 0.61 | |
| Model 2 (biopsy and clinical variables) | 72 | ||
| IRRAT | 1.44 (0.66–3.14) | 0.36 | |
| CI3comp (ci+ct+[cgx2]) | 1.36 (0.64–2.86) | 0.42 | |
| ciprob | 0.96 (0.42–2.19) | 0.92 | |
| Recipient age (years) | 0.54 (0.31–0.95) | 0.033 | |
| eGFR (mL/min/1.73 m2) | 0.32 (0.15–0.70) | 0.005 | |
| UPCR at Bx (mg/g) | 1.87 (0.87–4.03) | 0.11 |
cg, glomerular double contours; ci, interstitial fibrosis; ciprob, molecular classifier reflecting the probability of histologic ci lesion score >1; ct, tubular atrophy; cv, intimal fibrous thickening; DCGF, death-censored graft survival; eGFR, estimated glomerular filtration rate; IRRAT, transcript set associated with injury-repair response; UPCR, urinary protein/creatinine ratio.
Adjusted models (model 1: biopsy variables; model 2: biopsy plus clinical variables) included variables (morphologic indices, molecular scores and/or clinical parameters) associated with DCGF, in univariable analysis (see Table 3).
For continuous variables and ordinal categorical variables, hazard ratios were calculated per increase from the first to the third quartile.