Submitted on November 20, 2006
Accepted on March 21, 2007
Left ventricular volume measurement in mice by conductance catheter. Evaluation and optimization of calibration
Jan Møller Nielsen1*, Steen Buus Kristiansen1, Steffen Ringgaard2, Torsten Toftegård Nielsen1, Allan Flyvbjerg3, Andrew N. Redington4, and Hans Erik Botker5
1 Cardiology, Aarhus university hospital, Skejby, Aarhus, Aarhus N, Denmark; Department of Cardiology, University Hospital in Aarhus, DK-8200 Aarhus N, United States
2 The magnetic resonance centre, Aarhus university hospital, Aarhus, Aarhus N, Denmark
3 Medical department M, Aarhus university hospital, Aarhus, Aarhus N, Denmark
4 Cardiology, University of Toronto, Toronto, Canada
5 Department of Cardiology, University Hospital in Aarhus, DK-8200 Aarhus N, United States
* To whom correspondence should be addressed. E-mail: janmn{at}dadlnet.dk.
The conductance catheter (CC) allows thorough evaluation of cardiac function as it provides simultaneously measurements of pressure and volume. Calibration of the volume signal remains challenging. Using different calibration techniques, in-vivo left ventricular volumes (Vcc) were measured in mice (n=52) with a Millar CC (SPR-839) and compared to MRI derived volumes (VMRI). Significant correlations between Vcc and VMRI (end-diastolic volume (EDV): R2=0.85, p<0.01, end-systolic volume (ESV):R2=0.88, p<0.01) were found when injection of hypertonic saline in the pulmonary artery was used to calibrate for parallel conductance and volume conversion was done by individual cylinder calibration. However, a significant underestimation was observed (EDV = -17.3 µL (-22.7 - -11.9µL), ESV =-8.8µL (-12.5 - -5.1µL)). Intravenous injection of the hypertonic saline bolus was inferior to injection into the pulmonary artery as calibration method. Calibration with an independent measurement of stroke volume decreased the agreement with VMRI. Correction for an increase in blood-conductivity during the in-vivo experiments improved estimation of EDV. The dual frequency method for estimation of parallel conductance failed to produce CC volumes that correlated with MRI volumes. We conclude that selection of the calibration procedure for the CC has significant implications on the accuracy and precision of volume estimation and pressure-volume loop derived variables like myocardial contractility. Although CC volumes may be underestimated compared to MRI, optimized calibration techniques enable reliable volume estimation with the CC in mice.
