It is an important cause of fatal self-poisoning in some countries, particularly in South-East Asia (Gunnell et al., 2007). The outcome of paraquat poisoning is variable but in large cohort studies typically between 40 and 60% of cases die, most within 24–72 h from multi-organ failure (Dawson et al., 2010, Gil et al., 2008 and Senarathna et al., 2009). However, patients with smaller exposures may die over the following weeks from respiratory failure secondary to progressive pulmonary fibrosis. Better prognostic indicators to identify this group would be very useful as ongoing interventions are most likely to be beneficial for this group with delayed toxicity. Paraquat produces free radicals which induce

cellular toxicity (Eddleston et al., 2003). Many treatments have been proposed and trialled, including extracorporeal elimination, immunosuppressants and antioxidants, Venetoclax research buy but the mortality remains high even in centres using all these treatments (Gil et al., 2008) (and JL Lin, unpublished observation 2010). A very strong predictor of death CHIR-99021 mouse in large cohort studies is the volume of paraquat consumed (Wilks et al., 2008 and Wilks et al., 2011), but estimates of this are often unreliable in individual patients. The concentration of paraquat in blood or urine can be used as a surrogate for ingested dose to predict survival or death using a nomogram. These have a

positive predictive value for death of 92–96% (Senarathna et al., 2009). Unfortunately paraquat assays are not PJ34 HCl widely available, particularly in the developing world, and the time of ingestion may be unknown, so alternative biomarkers are required which should ideally be able to be interpreted independent of the time of exposure. A range of alternative clinical and biochemical investigations for prognosis following acute paraquat poisoning have been assessed, but inadequately validated (Eddleston et al., 2003). For example, acute kidney injury is a prominent manifestation of acute paraquat poisoning which has prompted research into renal biomarkers (Gil et al., 2009 and Ragoucy-Sengler and Pileire, 1996). One small study (n = 18) suggested that an increase in creatinine of >3 μmol/L/h

(dCr/dt) predicts death ( Ragoucy-Sengler and Pileire, 1996). The rise in creatinine is probably due to progressive renal impairment and a direct reflection of organ toxicity ( Pond et al., 1993). However, paraquat interferes with some creatinine assays that utilise the Jaffe (picric-acid) method ( Aitken et al., 1994, Fairshter et al., 1986, Price et al., 1995 and Webb and Davies, 1981). Therefore, the increase in creatinine may reflect both exposure and toxicity. The apparent creatinine concentration increases with increasing paraquat concentrations ( Aitken et al., 1994, Fairshter et al., 1986, Price et al., 1995 and Webb and Davies, 1981), although minimally with concentrations less than 10 mg/L, in contrast to concentrations greater than 100 mg/L where interference is marked ( Fairshter et al.