Resistin-Can it be a new early marker for prognosis in patients who survive after a cardiac arrest? A pilot study.

AIM

The aim of our study was to evaluate the potential role of resistin in estimating the 30 days prognosis in patients with hypoxic-ischemic organ injury who survived after a cardiac arrest (CA).

MATERIALS AND METHODS

The study included 40 patients resuscitated after a non-traumatic out-of-hospital CA admitted in Emergency Department (ED). All patients were followed for 30 days after CA or until death. Clinical data on admission were recorded. Blood samples were collected on admission in ED (0-time interval), and at 6, 12, 24, 48- and 72-hours following resuscitation. Serum concentrations of resistin, S100B and neuron specific enolase (NSE) were measured. Several predictive scores for the mortality at 30 days were created with logistic regressions.

RESULTS

At each time interval, median serum levels of resistin and S100 B were significantly higher in non-survivors compared to survivors. For NSE, plasma levels were significantly lower in survivors as compared to non-survivors at 48 and 72 hours, respectively. Accurate predictive scores for 30-days mortality were the ones which included the values of resistin and S100B measured at 12 hours after admittance [AUC 0.938 (0.813-0.989), sensitivity 85.71% (67.3%- 96%), specificity 91.67% (61.5%'99.8%), p<0.001], which included the values of all three markers measured at 12 hours after admittance [AUC 0.955 (0.839-0.995), sensitivity 82.14% (63.1%'93.9%), specificity 100.00% (73.5%'100.0%), p<0.001] and the that included the values of resistin and S-100B at 6 hours together with serum lactate on admission [AUC = 0.994 (0.901-1.0), sensitivity 96.4% (81.7%'99.9%), specificity 100.00% (73.5%'100.0%), p<0.001].

CONCLUSION

In our study, serum levels of resistin or a combination of resistin with S-100B or resistin with S-100B and lactate, were highly predictive for 30 days mortality in resuscitated patients after CA. Further studies on large number of patients are needed to confirm our data.