Table 5.
Novel biomarkers accurately predict peak ALT of more than 100 U/L in patients who presented with a staggered paracetamol overdose with normal ALT and INR at hospital presentation
|
Derivation cohort (MAPP) n/N=13/207 |
Validation cohort (BIOPAR) n/N=4/54 |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ROC–AUC | p value | Specificity | Sensitivity | PPV | NPV | ROC–AUC | p value | Specificity | Sensitivity | PPV | NPV | |
| ALT | 0·63 (0·48–0·77) | 0·1342 | 0·95 | 0·38 (0·14–0·68) | 29·4 | 95·8 | 0·57 (0·23–0·91) | 0·6203 | 0·95 | 0·50 (0·22–0·71) | 50·0 | 96·0 |
| Paracetamol concentration | 0·57 (0·38–0·77) | 0·3451 | 0·95 | 0·15 (0·02–0·45) | 20·0 | 94·5 | 0·67 (0·52–0·81) | 0·1019 | 0·95 | 0·25 (0·13–0·55) | 25·0 | 94·0 |
| miR-122 | 1·00 (1·00–1·00) | <0·0001 | 0·95 | 1·00 (0·75–1·00) | 54·2 | 100·0 | 1·00 (1·00–1·00) | <0·0001 | 0·95 | 1·00 (0·75–1·00) | 66·7 | 100·0 |
| HMGB1 | 1·00 (1·00–1·00) | <0·0001 | 0·95 | 1·00 (0·75–1·00) | 59·1 | 100·0 | 0·98 (0·94–1·00) | <0·0001 | 0·95 | 1·00 (0·75–1·00) | 57·2 | 100·0 |
| Full-length K18 | 0·99 (0·98–1·00) | <0·0001 | 0·95 | 0·92 (0·64–0·99) | 54·5 | 99·4 | 0·76 (0·37–1·00) | 0·0800 | 0·95 | 0·75 (0·44–0·92) | 50·0 | 97·9 |
| Caspase-cleaved K18 | 0·77 (0·59–0·95) | 0·0011 | 0·95 | 0·62 (0·32–0·86) | 44·4 | 97·3 | 0 63 (0·22–1·00) | 0·3905 | 0·95 | 0·50 (0·22–0·74) | 40·0 | 95·9 |
| GLDH | 0·78 (0·62–0·93) | 0·0009 | 0·95 | 0·53 (0·25–0·81) | 36·8 | 96·8 | 0·70 (0·47–0·93) | 0·1757 | 0·95 | 0·25 (0·10–0·46) | 25·0 | 94·0 |
ROC–AUC (95% CI), sensitivity at 95% specificity (95% CI), PPV, and NPV predictive values were calculated to identify the potential of novel and established stratification biomarkers to predict the development of acute liver injury. INR=international normalised ratio. ROC=receiver operator characteristic. AUC=area under the curve. PPV=positive predictive value. NPV=negative predictive value. ALT=alanine aminotransferase. HMGB1=high mobility group box-1. GLDH=glutamate dehydrogenase.