| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gastrointestinal Imaging |
1 From the Departments of Radiology (A.J.L., M.J.B., T.S., S.M.N., R.F.) and General, Visceral, and Transplant Surgery (U.S., P.N.) Charité Universitätsmedizin Berlin, Campus Virchow-Klinikum, Augustenburger Platz 1, D-13353 Berlin, Germany. Received December 6, 2004; revision requested February 4, 2005; revision received June 3; accepted July 1; final version accepted January 6, 2006. Supported by a grant from Deutsche Forschungsgemeinschaft. Address correspondence to A.J.L. (e-mail: lemke{at}charite.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Materials and Methods: After medical ethics committee and state medical board approval, informed consent was obtained from eight female and eight male living donors (age range, 18–63 years) for participation in preoperative CT volumetric measurement of the right liver lobes by using the summation-of-area method. Intraoperatively, the graft was weighed, and the volume of the graft was determined by means of water displacement. Distributions of pre- and intraoperative data were depicted as Tukey box-and-whisker diagrams. Then, linear regressions were calculated, and the results were depicted as scatterplots. On the basis of intraoperative data, physical density of the parenchyma was calculated by dividing weight by volume of the graft.
Results: Preoperative measurement of grafts resulted in a mean volume of 929 mL ± 176 (standard deviation); intraoperative mean weight and volume of the grafts were 774 g ± 138 and 697 mL ± 139, respectively. All corresponding pre- and intraoperative data correlated significantly (P < .001) with each other. Intraoperatively expected volume (Vintraop) in millilliters and weight (Wintraop) in grams can be calculated with the equations Vintraop = (0.656 · Vpreop) + 87.629 mL and Wintraop = (0.678 g/mL · Vpreop) + 143.704 g, respectively, where preoperative volume is Vpreop in milliliters. Physical density of transplanted liver lobes was 1.1172 g/mL ± 0.1015.
Conclusion: By using two equations developed from the data obtained in this study, expected intraoperative weight and volume can properly be determined from CT volumetric measurements.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
At our transplantation center, the planning of living related liver transplantation requires accurate knowledge of the volumes of the entire liver and of both of its lobes measured by using computed tomographic (CT) volumetric techniques prior to surgery. This volumetric determination is of crucial importance for both donor and recipient. To prevent posttransplant complications for the recipient because of small-for-size syndrome (4), for example, various similar considerations have been published (5–9). These published studies considered a minimum volume of hepatic tissue, which was adapted to the needs of the individual recipient, that would not put the donor at risk from the loss of too much liver tissue.
At some transplantation centers similar to ours, the equation used for the calculation of the minimum weight of the drained liver lobe needed for the graft includes the ratio of the graft weight to the recipient's body weight (10) and the recipient's body weight. Other transplantation centers consider a minimum percentage of the recipient's standard liver volume, which is a function of the body surface area (11). The determination of volume or weight is of interest, depending on the preference of each transplantation center.
Besides reporting on morphologic characteristics (eg, focal lesions or steatosis hepatis) and analyzing individual vascular anatomy, it is the task of the clinician who uses an all-inclusive preoperative CT examination to determine the approximate weight of the liver lobe transplant by using a special examination protocol. It is widely assumed that the mean density of healthy liver tissue is 1.00 g/mL so that preoperatively calculated volumes of the entire living donor's liver and both of its lobes have been equated with their respective weights (12,13).
Volumetric analyses with phantoms showed strong congruity between volumetric data that were based on CT calculations and those based on liquid displacement (14–17). In contrast, the literature provides differing information about the congruity between volumetric data of hepatic tissue determined preoperatively by using CT and data determined intraoperatively (18,19). These publications generally only included weight determination as an indirect procedure for the determination of hepatic tissue volume with the use of the assumed physical density of 1.00 g/mL for conversion into volume. At present, there is increasing evidence that the preoperatively measured volume of hepatic tissue is substantially larger than the corresponding intraoperatively acquired and measured volume (19). Possible explanations have been perioperative loss of blood (12) or lack of perfusion (20), inaccurate assignment of hepatic tissue to the corresponding liver segments at the borders of these segments when CT was used (21), and deviations between physical density of the liver and the assumed mean physical density of 1.00 g/mL (12,13).
Thus, the purpose of our study was to prospectively develop equations for the calculation of expected intraoperative weight and volume of a living donor's right liver lobe by using preoperative CT for volumetric measurement.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Inclusion criteria for the selection of prospective donors who were participating in the study were all aspects that are part of a standardized preoperative evaluation program and are relevant for a living donor donation. Furthermore the person's psychosocial motivation for a voluntary donation, anamnesis, status of clinical examination, blood chemistry levels, and serologic test results were evaluated. A contrast material–enhanced CT examination of the abdomen was included in the evaluation and was required for the analysis of morphologic characteristics, the vascular status of the liver, and the evaluation of the hepatic parenchyma. Because in individual cases, despite a successful evaluation, donors had to be excluded for other relevant reasons, only donors who actually underwent surgery for resection of the right liver lobe were included in this study.
CT Imaging
Examinations of the abdomen were performed with a 16-section spiral CT scanner (LightSpeed 16; GE Medical Systems, Milwaukee, Wis) by using a four-phase contrast-enhanced technique. The performance of an examination with a nonenhanced phase mainly served to exclude hepatic steatosis (22), because fatty degeneration of the liver with parenchymal damage of more than 10% is an exclusion criterion for the liver to be considered a donor organ (23).
Contrast-enhanced phases consisting of hepatic arterial, portal venous, and hepatic venous phases (24) were used to analyze surgically relevant anatomic variations by means of two- or three-dimensional reconstruction of the corresponding vascular trees (Table). All four phases were used to exclude or characterize focal hepatic lesions. The hepatic venous phase was used for preoperative CT volumetric measurement of the donor liver because, in this phase, the determining hepatic veins are depicted with maximum contrast (25). Iopromide (Ultravist 370; Schering, Berlin, Germany), 100 mL, as an iodinated contrast agent, was injected intravenously and automatically with a power injector (Spectris Solaris; Medrad, Indianola, Pa) at a flow rate of 4 mL/sec. The examination always took place in deep inspiration and included the entire liver and the upper abdomen to the lower pole of the more caudally situated kidney.
|
In our study, living related liver transplantation in all patients was performed by one of two surgeons (U.S., with 12 years of experience, and P.N., with 22 years of experience) of our transplantation center who are well experienced in regard to liver transplantation of any type. In agreement with the surgeons who considered the middle hepatic vein controversy (26) and as described in another context (27), a straight line was drawn section by section between the right and left liver lobes for the determination of the dissection area. For the drawing of the line, we used the center of the inferior vena cava as an orientation point and, depending on the table position, cut caudally and centrally through the middle hepatic vein before ending in the middle of the gallbladder bed. After calibration of the user software concerning the voxel size (determined with section thickness and side lengths of pixels as given in the Digital Imaging and Communications in Medicine header), it is possible to preoperatively calculate the total volume of the donor liver, the volume of the right liver lobe graft, and also the volume of the left liver lobe remaining in situ by using the summation-of-area method (16). By means of this method, the total volume of an object is calculated with the aid of several cylindric partial volumes. The partial volumes result from segmentation of each complete object into adjoining parallel sections of identical thickness. Each partial volume is calculated as a product of its base and the constant section thickness selected or, in the case of incomplete or overlapping techniques, of the section-to-section distance.
Instead of a single CT volumetric measurement acquired in the clinical routine, two methodically identical calculations that were independent from each other were performed by the two radiologists on the pool of donors on the day of the CT examination. Within 24 hours after the CT examination, the surgeons who were conducting the transplantation preoperatively received both the calculation-derived mean value of the total volume of the donor liver, and the mean values of both of the lobes of the liver.
Intraoperative Data Acquisition
During surgery, which on average took place 31.63 days ± 22.54 (median time, 24.5 days; range, 4–82 days) after the CT examination, performed by either of the two surgeons, the right liver lobe was flushed through the severed conducting structures at the hilus immediately after resection for the purpose of preservation. This flushing was performed by using an organ perfusion solution (Custodiol; Köhler Pharma, Alsbach, Germany) cooled to 4°C, which was infused through the severed nonligated main branches of the main hepatic artery and the portal vein at the hilus of the transplant. Since the severed hepatic vein branches also were not ligated, all intrahepatic liquid media capable of clotting, such as blood, bile, lymphatic liquid, and organ perfusion solution from the transplant, could drain to a great extent.
The weight of the resected right liver lobe thus prepared for the period of cold ischemia was determined by one individual (U.S. or P.N.) with electronic laboratory scales (accuracy to 10 g). In the next step, a special container was used to determine the volume of the liver lobe by means of liquid displacement by one individual (M.J.B. or T.S.). This special cylindric 6-L glass container was open at the top and was equipped with a glass tube for overflow at the middle and on the side of the container (Fig 1). The container could be sterilized and was therefore reusable. In the operating room, the container was filled with sterile 4°C physiologic saline to the level of the overflow. After immersion of the liver lobe in saline by the surgeon (U.S. or P.N.), the saline that escaped via the overflow was gathered in a graduated cylinder by the radiologist (M.J.B.) or the properly instructed medical information scientist (T.S.) with sterile conditions. The volume of the displaced liquid that was measured in the graduated cylinder corresponded to the volume of the liver lobe that was immersed in the saline.
|
All statistical analyses of pre- and intraoperative data about volumetric and weight determinations were performed by using software (SPSS for Windows, version 11.0; SPSS, Chicago Ill).
For the descriptive data analysis, the distribution of pre- and intraoperative raw data was graphically depicted as a box plot diagram. The mean 95% confidence intervals were estimated as that range of data in which 95% of all measured or calculated values are included; 2.5% of all values are below that interval and 2.5% of all values are above that interval.
For the comparison between pre- and intraoperative data, the ratios of each corresponding CT volumetric measurement and both of the intraoperative measurements were calculated, and then the distribution of the measurements was graphically depicted as a box plot with identical characteristics as well. In this box plot, both of the intraoperative measurements ought to be considered as a reference quantity of preoperative CT volumetric measurements.
In a second step, two regression analyses of the data were performed, and these analyses included preoperative CT volume, on the one hand, and intraoperative volume and weight, on the other hand, because linear correlation was assumed (27). Both of these regression analyses led to two regression lines, and each was described by a linear equation (Eqq [1, 2], as presented in Results). From the preoperative CT volumetric measurement of the right hepatic lobe graft, the intraoperative volume of the graft could be calculated with the use of Equation (1) and the intraoperative weight of the graft could be calculated with the use of Equation (2). Another regression analysis was performed with all corresponding weight and volumetric data obtained intraoperatively, because linear regression was assumed in this case as well. The physical density of transplanted liver tissue was calculated by using the ratio of weight and volume, and the statistical distribution was described. With all tests, correlations with a difference of P < .05 were considered to be statistically significant.
After the calculation of the physical density of the transplanted liver lobes with the ratio of intraoperatively obtained corresponding weight and volumetric data, the statistical distribution of these liver lobes was analyzed by using the software to analyze potential interindividual variations of physical density.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Intraoperative Data Acquisition
The following volumetric data of the right liver lobe resulted from the liquid displacement measurement during the surgical operation: The volume ranged from 550 to 1030 mL, with a mean value of 697 mL ± 139. The 95% confidence interval ranged between 623 and 771 mL, the 25th percentile value amounted to 598 mL, the median amounted to 650 mL, and the 75th percentile value amounted to 770 mL.
The intraoperatively measured weight of the transplanted right liver lobe yielded the following statistical data: The weight span was 580–1100 g, and the mean weight was 774 g ± 138. The 25th percentile value amounted to 665 g, the median amounted to 750 g, and the 75th percentile value amounted to 875 g, and the 95% confidence interval ranged between 700 and 847 g.
Descriptive Data Analysis
The three groups are depicted in a box plot (Fig 2).
|
|
With regard to the right liver lobe graft, the mean values of preoperative CT volumetric measurements correlated in the regression analysis with the volumes measured intraoperatively by using liquid displacement, with r = 0.834 (P < .001, two-tailed test). The power analysis showed that zpwr = 4.551, which corresponds to a statistical power of greater than .95 (28). From the data received, the regression line could be derived with Equation (1) (Fig 4) as follows:
 | (1) |
|
 | (2) |
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Influence of Perfusion on Liver Volume
In an experimental study in porcine livers Frericks et al (20) were the first, to our knowledge, to point out explicitly the difference in examination conditions with which both volumetric measurements had been performed in some of the studies on the subject. According to their findings, the state of perfusion of the right liver lobe differs to some extent, which must be taken into consideration in discussions concerning substantial deviations in volume. Preoperative CT examination is regarded as a physiologic in situ situation during which the right liver lobe is exposed to physiologic arterial perfusion pressure. The situation during surgery, however, immediately after resection and perfusion of the liver lobe by using a perfusion solution and with nonligated vascular structures at the severed areas is considered a nonphysiologic ex situ situation, where arterial perfusion pressure is immaterial.
Thus, gravity and the retraction energies of the inner tissue cause the right liver lobe to collapse and the inner liquid media such as blood, bile, and lymph to dissipate through the nonligated vascular structures. In their porcine study, Frericks et al (20) showed that this measurement discrepancy between the measurements with the liquid displacement method and those with CT volumetric techniques performed with simulated physiologic perfusion conditions and without perfusion pressure reaches a mean value of 33%.
In regard to the interpretation of the data obtained in our study, this means that, analogous to the experimental animal study by Frericks et al, the substantial differences in volume at both measurements must primarily be seen in connection with the various states of perfusion during the examinations. Consequently, especially when we take into account the excellent volumetric congruity in phantom studies (14–17), the differences in volume should not be interpreted mainly as measurement errors in the course of volumetric determination. Essentially, the discrepancy results from differences in perfusion, and only on a secondary level are actual measurement errors involved. Thus, the sources of errors that have been discussed, such as examination technique, partial volume effects, interindividual and nonconstant hepatic physical density, inaccurate recognition of contour and segments, and hepatic volume deviations, are of lesser importance.
When one looks at the results in this manner, one can consider that the difference in volume of a physiologically perfused liver lobe corresponds approximately to the total volume of liquid media in a physiologically perfused liver lobe. The difference in volume debated with regard to preoperative CT volumetric measurement would then be the approximate volumetric portion of liquid media of the total volume of physiologically perfused liver tissue. Compared with the 33% volumetric portion of liquid media in the porcine study (20), the volumetric portion of liquid media in human hepatic tissue would, based on the existing data, amount to approximately 34%. In this calculation, however, actual measurement errors have not been taken into account. The 1% difference in liquid medium volume between porcine and human livers most likely depends on microanatomic characteristics, which have a different distribution over the entire volume of the liver for hepatocytes, cells of the reticuloendothelial system, connective tissue, and vascular structures in each case. Another characteristic such as different elasticities of each kind of liver tissue should be considered as well.
Physical Density of Healthy Liver Tissue
In general, studies in which preoperative and intraoperative volumetric measurements of livers or liver lobes are compared were based on the assumption that the density value was on the order of 1.00 g/mL, which facilitates the conversion from volumetric value into weight value. In contrast to general opinion, measurements performed on the pool of donors in our study showed that, on the one hand, mean density was approximately 12% higher than suspected, whereas, on the other hand, substantial interindividual variations of density occurred.
Limitations of the Study
The sample size of 16 patients in this study might be considered small and, for that reason, a weakness of the study. However, while we (A.J.L., M.J.B., S.M.N. and T.S.) performed the data and statistical analysis, we found, by a power analysis, that all tests we used led us to statistically significant results and that an increase of the sample size would not substantially have led us to statistically more significant results.
The influence of other factors such as the interval between CT and surgery or potential problems with volumetric determination at CT volumetric measurement was not evaluated in the present study but should be addressed in further studies.
Conclusion
On the basis of this knowledge of certain variations in physical density, in our study, it was possible to demonstrate that preoperative CT volumetric measurement and intraoperative volumetric measurement correlate somewhat less strongly than do preoperative volumetric measurement and intraoperative weight determination.
By using our study data obtained from preoperative volumetric measurement and intraoperative weight and volume determination of the grafts, two linear equations were developed. In a clinical routine, preoperative CT volumetric measurement can be considered sufficiently accurate for determination of weight and volume of a right liver lobe graft for surgery by using two linear equations.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, A.J.L., R.F.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, M.J.B.; clinical studies, A.J.L., M.J.B., T.S., S.M.N., U.S., P.N.; statistical analysis, M.J.B., S.M.N.; and manuscript editing, A.J.L., M.J.B., R.F.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) below or above the inner limit but within the outer limit (three times the box) are indicated.
