Measurement of Cerebral Blood Flow at the Bedside

Abstract

*University of BonnBonn, Germany.wietasch@mailer.meb.uni-bonn.deIn Reply:—We appreciate the interest of Dr. Mélot et al. in our recent report on a new method of bedside measurement of cerebral blood flow. 1In principle, we agree with Dr. Mélot et al. that, according to their publication, 2the accuracy of their methodology for measurement of cerebral blood flow is better, and we congratulate them on their impressive results. However, we would like to point out some principal differences between the method of cerebral blood flow measurement in the jugular bulb by continuous thermodilution, as described by Mélot et al. , 2and the method of cerebral blood flow measurement by transcerebral thermodilution, as applied in our investigation. The technique of Dr. Mélot et al. 2measures jugular bulb flow in a manner similar to that developed previously for coronary sinus outflow measurements. 3This methodology is based on the principle of mass conservation 4and, therefore, yields blood flow measurements in absolute terms (ml/min). To convert this blood flow measurement into physiologically and pathophysiologically relevant organ-specific blood flow (i.e. , into ml · min−1· 100 g−1), brain weight must be estimated. In the investigation of Dr. Mélot et al. , 2this was done by assuming a proportional relation to body height, which was assumed to be different for men and women, according to the work of Spann et al. 5However, as described in the same investigation, the weight of the brain varies significantly interindividually. Spann et al. 5also present several cases in which the brain weight was 6.1 g/cm in one individual and 11.6 g/cm in another individual. Thus, by measuring absolute flow in the jugular vein and converting this flow to organ blood flow in terms of ml · min · 100 g−1based on an estimated brain weight, this variability should contribute to the accuracy of the methodology. Therefore, we opted for a methodology that is based on a transit time principle and, therefore, measures weight-normalized organ blood flow directly. With use of an intravascular tracer and a diffusible tracer simultaneously, cerebral blood volume also can be determined principally.Dr. Mélot et al. correctly point out that the limits of agreement with the Kety-Schmidt method, which we used as a reference method, were not as good as the method used for continuous jugular thermodilution in their investigation. Some of the possible explanations for the observed scatter have been discussed in the paper in more detail. 1In comparison with the work of Dr. Mélot et al. , we would like to add some further comments. The reference Kety-Schmidt method, which they used in their study, is slightly different from our Kety-Schmidt methodology. We used Argon as a tracer and a sampling system (Unita I; B. Braun, Melsungen, Germany), which draws blood continuously from arterial and jugular bulb catheters, thereby averaging the concentration time courses at these sites “in the syringe.” The advantage of this approach is less analytical effort as opposed to serial blood samples, which are necessitated by the classic Kety-Schmidt method. However, it seems that the price paid for this sparing of blood samples might be less accuracy as compared with the classic Kety-Schmidt methodology, in particular when viewed with the clearly better results of Dr. Mélot et al. 2On the other hand, we clearly pointed out in our article that the reference method used in our investigation is most likely a significant source for the scatter between methods, which has to be taken into account.Another limitation of blood flow measurement with use of transcerebral thermodilution, pointed out by Dr. Mélot et al. , is the limited accuracy at low blood flow rates, which has also been addressed in the publication. We agree that, in some critically ill patients, low blood flow rates might be of particular interest, and, therefore, we tried to improve the methodology of transcerebral thermodilution in this respect. The crucial problem is the duration of data sampling, which was only 5 min in our investigation. For low blood flow rates, the sampling periods should be prolonged, an option that is being investigated in our department.We thank Dr. Mélot et al. for their interest in our work. As with all clinical methods of measurements, each methodology has its advantages and disadvantages. We believe that, particularly in combination with transcranial Doppler measurements of blood flow velocities, transcerebral double-indicator dilution might add to the armentarium of cerebral monitoring, especially when longer sampling periods are used.