A new method for predicting splenic „switch-off“ sign: Splenic T2-mapping

Purpose/Objectives In a relevant proportion of patients undergoing myocardial first­pass perfusion cardiac magnetic resonance (FPP­CMR) inadequate adenosine stress may occur. This may cause false­negative results and a suboptimal clinical management. “Splenic switch­off” has recently been proposed as a sign to detect effective adenosine response after first­pass imaging is performed. Novel methods are emerging to detect reduction of splenic blood volume during adenosine stress and prior to contrast administration. We sought to determine the feasibility of using splenic T2­mapping as a marker to assess adenosine­stress effect. Methods and Materials Fifty­five consecutive patients (29 males, mean age 56.1±11.5) with clinical indications underwent FPP­CMR at 1.5T (n=16) and 3T (n=39). Splenic T1­ and T2­values were assessed, using respectively MOLLI and FLASH sequences, performed at rest and during adenosine stress (140μg/kg/min, 4 min) along one mid­ventricular short­axis slice. Changes of T1­ (ΔT1) and T2­values (ΔT2) were calculated and expressed as percentages. FPP­CMR was performed after intravenous injection of gadobutrol (0.1mmol/kg body­weight). Images were analyzed with the use of a semiautomatic software (Circle CMR42, Calgary, Canada) to quantify splenic and myocardial enhancement. Spleen­to­myocardium perfusion ratio (s/mΔSI) was calculated to obtain splenic relative enhancement during FPP­CMR. The presence of “switch­ off” sign was evaluated in consensus by two readers and used as standard of reference. Accuracy of s/mΔSI, ΔT1 and ΔT2 in predicting the splenic switch­off sign was assessed. Results Mean splenic rest native T1­ and T2­values were respectively 1256.7±96.5(ms) and 60.3±11(ms) at 3T, and 1149.1±75.1(ms) and 80±10.1(ms) at 1.5T, versus stress native T1­ and T2­values of 1198.5±85.8(ms) and 54.1±11.2(ms) at 3T, and 1100.4±85.1(ms) and 73.3±10.9(ms) at 1.5T (all p<0.001). Splenic relative enhancement (s/mΔSI) showed a correlation of 0.701 and 0.504 respectively with ΔT2 and ΔT1 (all p<0.001). The best accuracy for predicting “switch­off” sign was obtained with ΔT2 (AUC=0.987), followed by s/mΔSI (AUC=0.962) and ΔT1 (AUC=0.912) (all p<0.001) (Fig.1). The largest difference in terms of accuracy was observed between ΔT2 and ΔT1 (0.0732; p=0.076). Cut­offs were 8.3% for ΔT2 (sensitivity: 95%; specificity 94%) and 3.6% for ΔT1 (sensitivity: 84%; specificity: 83%). Conclusion Splenic T2­mapping may be a helpful tool to predict splenic “switch­off” sign in FPP­CMR. This may be useful to predict an effective adenosine response and, eventually, to allow for preventive dose adaption and less false­negative results.