| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular Imaging |
1 From the Institute of Diagnostic Radiology (G.F., L.W.P., L.B.F., B.W., U.M.) and Departments of General Surgery (J.S.a.E., S.B.H., A.K., C.F.E., W.T.K.), Cardiothoracic Surgery (M.K., E.G.), and Hemostaseology and Transfusion Medicine (V.S., M.S.), Heinrich-Heine-University of Duesseldorf, Moorenstr 5, 40225 Duesseldorf, Germany. Received April 7, 2006; revision requested June 5; final revision received June 19; accepted August 24. Address correspondence to L.B.F. (e-mail: ben{at}fritz.md).
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Materials and Methods: The study was approved by the institutional ethics committee; informed consent was obtained. Thirteen patients underwent PVE of liver segments I and IV-VIII to stimulate hepatic regeneration prior to extended right hepatectomy. In six patients (three men, three women; mean age, 61 years; range, 46–72 years) with a future liver remnant volume (FLRV) below 25% and/or limited quality of hepatic parenchyma, PVE alone did not promise adequate proliferation. These patients underwent BMSC administration to segments II and III (group I). In seven patients (three men, four women; mean age, 69 years; range, 63–75 years) with an FLRV below 25%, PVE alone was performed (group II). Two radiologists blinded to patients' identity and each other's results measured liver and tumor volumes with helical computed tomography. Absolute, relative, and daily FLRV gains were compared by using the t test or the Wilcoxon test.
Results: The increase of the mean absolute FLRV in group I from 239.3 mL ± 103.5 (standard deviation) to 417.1 mL ± 150.4 was significantly higher than that from 286.3 mL ± 77.1 to 395.9 mL ± 94.1 in group II (P = .049). The relative gain of FLRV after PVE in group I (77.3% ± 38.2) was significantly higher than that in group II (39.1% ± 20.4) (P = .039). The daily hepatic growth rate in group I (9.5 mL/d ± 4.3) was significantly superior to that in group II (4.1 mL/d ± 1.9) (P = .03). Time to surgery was 27 days ± 11 in group I and 45 days ± 21 in group II (P = .057).
Conclusion: In patients with malignant liver lesions, the combination of PVE with CD133+ BMSC administration substantially increased hepatic regeneration compared with PVE alone.
© RSNA, 2007
In up to 45% of patients with primary or secondary hepatobiliary malignancy, extended hepatectomy (more than five segments) is necessary to achieve margin-negative resection (1,2). In these patients, an anticipated future liver remnant volume (FLRV) below 25% of the total liver volume (TLV) leads to an increased risk of postoperative morbidity and mortality (3–6).
Transhepatic percutaneous portal vein embolization (PVE) has gained acceptance for preoperative expansion of the FLRV. Depending on the time given for gain, regeneration rates after PVE range between 4 and 21 mL/d, and relative gains in FLRV range between 29% and 66%, as reported for patients without cirrhosis or diabetes prior to standard right hepatectomy (7,8). In patients requiring resection of liver segments I and IV–VIII, expansion of liver volume after PVE may be substantially lower and time to surgery may be unacceptably long (observed to be up to 150 days), particularly if the left lateral liver segments determining the FLRV are small and the quality of hepatic parenchyma is limited due to steatosis or fibrosis of the liver (7).
Evidence suggests that hematopoietic stem cells participate in hepatic regeneration (9,10) and are helpful to further augment and accelerate hepatic proliferation. It is now accepted that three compartments of cells can respond to the loss of hepatocytes. Mature hepatocytes account for physiologic tissue renewal, hepatic cell replacement after major liver resection, and exposure to toxic chemicals (11,12). At a second level, intrahepatic "oval cells" are considered to be bipotent progenitor cells that can differentiate into hepatocytes and cholangiocytes. This compartment is activated when the proliferation of hepatocytes is inhibited (13).
Bone marrow stem cells (BMSCs) have been shown to be a source of hepatic stem cells, and they are capable of repopulating and repairing the liver after acute and chronic injury (14,15). Mobilization of peripheral, hematopoietic, autologous, CD34+ stem cells (known to bear the capacity for differentiation into a hepatic lineage) has been demonstrated to be 10-fold higher after liver resection than after liver-sparing abdominal surgery (16), which supports the experimental finding that stem cells participate in liver regeneration subsequent to resection (17). Like CD34, CD133 is a highly conserved antigen (5-transmembrane glycoprotein) expressed on hematopoietic stem cells with unknown function. It has been reported that CD133 might be a marker of early progenitors with high engraftment potential and that CD133+ cells may be useful in clinical transplantation protocols (18). Administration of BMSCs enriched for CD133+ cells have already been used therapeutically to promote the regeneration of postinfarction myocardium (19,20). We previously described a concept of portal CD133+ BMSC application after PVE to augment preoperative gain in FLRV (21). Thus, the purpose of the study was to prospectively evaluate the effectiveness of PVE and CD133+ BMSC administration to the liver, compared with PVE alone, to augment hepatic regeneration in patients with large hepatic malignancies.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Between September 2003 and June 2005, 16 consecutive patients were enlisted to undergo extended right hepatectomy. Of these patients, 13 were scheduled for PVE; all fulfilled the following criteria: FLRV below 25% of the TLV, exclusive of tumor volume; technically feasible for trisegmentectomy and without tumor mass in the left lateral liver; and fully patent right portal vein and bile duct. Three patients were excluded because of cholestatic disease and conjugated bilirubin level higher than 3 mg/dL (51.3 µmol/L).
Stem Cell Group: Group I
In six (three women, three men; mean age, 61 years; range, 46–72 years) of the 13 included patients, PVE and portal administration of CD133+ BMSCs prior to the scheduled extended right hepatectomy was performed, because PVE as an isolated technique was questionably sufficient to induce adequate proliferation of the left lateral lobe within a reasonable time (7).
All patients were characterized by large, central liver malignancies and FLRV up to 25% of the TLV (tumor volume, 141.4 mL ± 161.1 [mean ± standard deviation]). A panel of criteria indicating impaired liver regeneration capacity was used for inclusion in this group. Three patients showed limited quality of hepatic parenchyma due to hepatic steatosis (patient 4), hepatic fibrosis (patient 1), or prior hepatotoxic chemotherapy (patient 3), which is known to impair hepatic regeneration potential (22). Patients 2 and 5 showed unusual low basal volume of segments II and III. Patient 3 had a fast progressive liver lesion (doubling of tumor volume in 3 months). Patients 2 and 6 had large tumor volumes of 407 and 177 mL, respectively.
Control Group: Group II
In seven (three women, four men; mean age, 69 years; range, 63–75 years) of 13 consecutive patients, PVE without application of BMSCs prior to the scheduled extended right hepatectomy was performed. PVE as an isolated technique in this group was reasonable to induce adequate proliferation of the left lateral lobe. These patients served as a nonrandomized control group.
Patients in this group were characterized by large central liver malignancies (tumor volume, 150.6 mL ± 121.1) and FLRV below 25% of the TLV. None of the patients had a limited quality of hepatic parenchyma, unusually low basal volume of segments II and III, or fast progressive liver lesions.
Volumetric Assessments
All patients underwent helical computed tomography (CT) to estimate liver volume prior to PVE, 2 weeks after PVE, and then in 1–2-week intervals for up to 5 weeks to determine the degree of induced hypertrophy. If at 2 weeks proliferation generated a FLRV of less than 25% of the TLV, follow-up measurements were performed in 1- to 2-week intervals until FLRV was adequate for extended right hepatectomy. All CT examinations were performed by using a multisection CT scanner with six active detector arrays (Somatom Emotion; Siemens, Erlangen, Germany). The following parameters were used: 2-mm collimation, 3-mm section thickness, 3-mm reconstruction interval, 350 window width and 50 HU window level, and soft-tissue kernel. Transverse scans were obtained in the portal venous phase to measure the TLV, the FLRV (segments II and III), and the tumor volume. Measurements were performed manually with respect to the hepatic segmentation on the basis of the distribution of the portal pedicles and the location of the hepatic veins (Couinaud classification system). CT volumetry was performed by two independent radiologists (G.F. [observer 1] with 17 years and L.W.P. [observer 2] with 8 years of CT experience), both blinded to patients' identity and the results of the other observer. Measurement results of observer 1 were included for volumetric evaluation. Differential absolute, relative, and daily gains of FLRV after PVE and time to surgery were calculated.
PVE Procedures
PVE was performed with a digital subtraction angiographic device (Multidiagnost 4; Philips, Hamburg, Germany) by the same two radiologists; both had extensive experience (15 and 5 years) with complex embolization procedures. A transileocolic portal venous approach with direct cannulation and general anesthesia was used in all patients of group I and in three patients of group II. The abdominal wall was closed subsequently to surgical placement of a 5-F vascular sheath (Terumo, Leuven, Belgium) into the portal vein. In group I, the ileocolic cannulation approach with general anesthesia was preferred to the percutaneous approach, because safe catheter placement and harvesting of bone marrow could be performed during the same intervention. In three patients of group II, ileocolic cannulation was used to prevent an access pathway through a large right hepatic tumor mass and/or an injury to the future liver remnant. In the other four patients of group II, the portal system was accessed by using a percutaneous transhepatic approach, in two patients from the left side and in two patients from the right. Fluoroscopic or ultrasonographic guidance was used to insert a 6-F vascular sheath (Terumo).
Anterioposterior, right and left anterior oblique, and craniocaudal projections were obtained to delineate the portal venous anatomy. Selective right portal vein injections were performed to place 5-F Cobra or Simmons catheters (Terumo) for embolization of segments V–VIII. Since extended right hepatectomy was planned in all patients, segments I and IV were embolized in an effort to cover the entire tumor-bearing liver and to maximize FLRV hypertrophy. For these segments, a microcatheter (Tracker; Target Therapeutics, Freemont, Calif) was placed coaxially through a 5-F selective catheter. Polyvinyl alcohol particles (Contour; Boston Scientific/Target Vascular, Freemont, Calif) ranging from 300 to 500 µm and microcoils (Platinum Microcoils; Target Therapeutics) were used to embolize segments I, IVa, and IVb. For PVE of segments V–VIII, small polyvinyl alcohol particles were used first to occlude the distal smaller portal branches. PVE was performed until stasis or near stasis was achieved. Subsequently, a 1:2 cyanoacrylate-to–iodinized oil mixture (Braun, Tuttlingen, Germany; Guerbet, Roissy, France) was used to obtain complete occlusion of these segments, because in former patients not included in this study, we observed insufficient volume response when polyvinyl alcohol particles alone were used. The number of particles and coils and the volume of cyanoacrylate-to–iodinized oil mixture applied depended on the diameter and number of vessels treated. PVE was stopped when portography demonstrated an absolute flow stop. PVE procedures were identical for both groups. Successful PVE with final flush portography in all patients and peri- and postinterventional complications were documented.
Preparation and Characterization of CD133+ BMSCs
In accordance with the institutional ethical regulations and after obtaining patients' informed consent, the procedure of harvesting bone marrow (M.S.) for readministration of selected cells (G.F., J.S.a.E., W.T.K) was performed in a closed system. Autologous bone marrow aspirated from the posterior iliac crest was drawn in heparin-coated syringes after the induction of anesthesia directly prior to minilaparatomy for portal catheter placement. Bone marrow cells were prepared simultaneously with PVE (M.K. M.S.). The cell suspension was filtered to remove bone spicula and then was processed by using a good manufacturing practice–grade cell-suspension unit (CliniMACS; Miltenyi Biotec, Bergisch-Gladbach, Germany) to enrich CD133+ cells as previously described (20,23). After an average of 160 minutes, the enriched mononuclear cells were ready for intrahepatic application. Cells were resuspended in a total volume of 80 mL of phosphate-buffered saline solution. Aliquots from the bone marrow aspirate and the injected cell fraction were collected for cytofluorometric analyses, as reported in earlier studies (20,23,24). The number of mononuclear cells was determined (M.K., M.S.) by using a cell counter (Sysmex, Duesseldorf, Germany). The aspirated total bone marrow volume in milliliters, the CD133+ purity of cells for application in percentages, and the absolute number of CD133+ cells applied were calculated.
Application of CD133+ BMSCs
In both groups, a 5-F cobra catheter (Terumo, Leuven, Belgium) was introduced into the segment II and segment III branches with fluoroscopic guidance (Exposcop 8000, Ziehm, Germany) 2–4 hours after PVE. CD133+ cells were selectively applied to the nonoccluded segment II and segment III portal branches; this procedure took an average time of 8 minutes. After intrahepatic placement of cells, the portal catheter and the sheath were removed. The time needed for the entire procedure ranged from 3.5 to 5.5 hours. Time to surgery (in days) was calculated from the day of PVE (exclusive) to the day of the extended right hemihepatectomy (inclusive). No special medication was required after BMSC application.
Clinical Chemistry
The total serum bilirubin level, the international normalized ratio (INR), and the aspartate aminotransferase and alanine aminotransferase levels were assessed prior to PVE and on day 1, days 2 or 3, and days 4 or 5 after PVE as chemical markers of liver metabolism, coagulation, and hepatocellular damage, respectively. Means and standard deviations were calculated.
Statistical Analysis
The effects of combined PVE and BMSC therapy were defined according to differential absolute, relative, and daily gains of FLRV after PVE; time to surgery; global in-hospital mortality and morbidity; the kinetics of post-PVE serum bilirubin, aspartate aminotransferase, and alanine aminotransferase levels; and the INR. These and related variables were compared for the two groups with respect to the expected values. All measured values were nonnegative. Therefore, to test the significance, we used the t test if the coefficient of variation (ratio of the standard deviation) over the average was less then of the mean value. If it was not, we used the Wilcoxon test as a nonparametric procedure. For calculating the comparability of the two groups, we used two-sided tests; for posttreatment values (eg, relative, absolute, and daily gain of liver volume), we used one-sided tests. A P value less than .05 was considered to indicate a significant difference. For the analysis of time to surgery, we used the nonparametric log-rank test (survival analysis) and the Wilcoxon test. For interobserver reliability of the liver volume estimations, regression analysis was used and the squared correlation coefficient r2 was calculated. It was proved whether the intercept was significantly different from the offspring and whether the slope included the value 1 (an intercept of 0 and a slope of 1 is expected for no bias between the results of the two observers). Linear regression analysis was also used for correlation of the number of applied BMSCs with the gains of the FLRV after PVE. All statistics were performed by using SAS software (version 9, 2005; SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Characterization of CD133+ Cell Preparations
The aspirated total bone marrow volume, the CD133+ purity of cells for application, and the absolute number of CD133+ cells applied are presented in Table 1.
|
|
|
|
|
|
|
|
In group II, resection was never performed in patient 2 owing to the development of a tumor mass in the left lateral segments. Extended hemihepatectomy was performed in six patients. In patients 1, 3, 4, and 7, Roux-en-Y and bile duct reconstruction were required; patient 7 developed bile leakage and was successfully treated with open reintervention. In patient 4, extended right hepatectomy and pancreatoduodenectomy were performed to warrant resection margins to be tumor free. This patient was dependent on high positive end-expiratory pressure levels in the early course after surgery because of poor respiratory performance. Portal vein thrombosis, most likely attributable to positive end-expiratory pressure–triggered compromised hepatic outflow, was resolved with revision of the portal vein. However, multiorgan failure led to death 10 days after resection.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
It has not yet been clarified which environmental factors initiate mobilization and homing of adult stem cells. Likewise, the factors that induce these stem cells to differentiate along the appropriate organ-specific lineage are only partially known. Several receptor-ligand interactions that may facilitate trafficking and adhesion of stem cells to their destination are a subject of investigation. It is known that a damaged liver expresses chemokines and chemotactants such as stroma-derived factor 1. These factors are thought to participate in the homing of extrahepatic progenitor cells (26). The process of hepatic stem cell engraftment is probably accelerated in a case of liver damage. In our study, we used PVE of liver segments I and IV–VIII as a strong proliferation stimulus to the nonembolized left lateral segments II and III, to which CD133+ BMSCs were applied. The direct portal administration of high concentrations of CD133+ BMSCs may ease the homing to the target segments II and III. The rationale for this application mode is supported by a study in which a high percentage of first-pass entrapment of BMSCs to the liver when applied to the portal vein was reported (27).
In our series, we found more than a twofold higher mean daily hepatic growth rate in patients treated with PVE and BMSCs compared with patients who underwent PVE alone. This resulted in a reduction of the time to surgery by an average of 18 days. Although the mean left lateral liver volume before PVE in the BMSC patients was smaller than that in patients treated with PVE alone, the relative gain of the left lateral liver volume was almost double in the BMSC group. This supports the effect of BMSC application, because owing to the smaller volume of hepatic tissue, the initial number of hepatocytes able to proliferate was probably smaller in the BMSC group. For both groups, minimal transient elevations of the routinely assessed markers (total bilirubin level, INR, and aspartate aminotransferase and alanine aminotransferase levels) normalized to their preinterventional levels 4 or 5 days after the procedure, which indicated no lasting effect on liver metabolism, hepatic synthesis capacity, and hepatocellular integrity.
Our study had limitations. We did not evaluate the mechanisms involved in hepatic volume effects; in particular, we did not investigate the fate of the injected adult stem cells. Although we were able to show that FLRV is significantly (P < .05) higher in patients after administration of CD133+ cells than in patients without stem cell application, we have not experimentally clarified that stem cells are the inducing factor of enforced in vivo liver regeneration. Currently discussed mechanisms of the stem cell–induced hepatic regeneration are cell-cell fusion (28,29), stem cell conversion to liver cells as transdifferentiation without fusion (30,31), and endogenous hepatic regeneration triggered by stem cell–provided trophic factors such as interleukin-6 (32).
Another limitation of our study was the difference in the selection criteria between groups I and II. We used a panel of criteria indicating impaired regeneration capacity and a high risk of inadequate volume response to PVE alone or limited time to gain due to locally advanced hepatic tumor. These included FLRV below 25% of TLV, reduced quality of hepatic parenchyma due to hepatic fibrosis, hepatic steatosis, history of hepatotoxic chemotherapy, respectively marked tumor growth rate, and volume, each serving as an inclusion criterion for group I. Presence of one or more of these criteria bears an increased risk for exclusion from potentially curative surgical treatment (7,22). We cannot exclude (positive or negative) the influence of these inhomogeneous inclusion criteria among the two groups on the differential gain in FLRV.
Clinical liver regeneration is not linear in the 1st weeks after intervention. Therefore, limitation of comparability may result when time for gains is different between the two groups. Because of very fast hepatic proliferation, early surgery was scheduled for two patients of group I; consequently, CT volumetry was performed only on days 7 and 14 after intervention. This resulted in a 6-day shorter mean time to gain in group I compared with that in group II and may have a slight oppositional effect on daily versus total volume gains. However, in our study, both total and daily gain rates were significantly (P < .05) superior in group I than in group II.
Despite the small number of patients and the lack of a randomized reference group, our data suggest that the concept of PVE with CD133+ BMSC administration to the liver bears the potential to accelerate and augment the proliferation of the FLRV more than does PVE alone in preparation for extensive liver resection. The modality seems to be safe and suitable for clinical routine. In particular, CD133+ stem cells may be a powerful adjunct to PVE in patients with very small left lateral segments, a large and fast-progressing tumor mass, and limited quality of hepatic parenchyma. A controlled trial is needed to validate our findings in a larger number of patients.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: BMSC = bone marrow stem cell • FLRV = future liver remnant volume • INR = international normalized ratio • PVE = portal vein embolization • TLV = total liver volume
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, G.F., W.T.K.; 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, G.F., J.S.a.E., L.W.P., S.B.H., M.K., B.W., C.F.E., U.M., W.T.K.; clinical studies, G.F., J.S.a.E., L.W.P., S.B.H., L.B.F., M.K., B.W., M.S., C.F.E., W.T.K.; statistical analysis, G.F., E.G., B.W.; and manuscript editing, G.F., J.S.a.E., L.W.P., S.B.H., L.B.F., V.S., U.M., W.T.K.
|
References |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Related Articles
This article has been cited by other articles:
|
|
 |
|
  F. Ozkan, B. Peynircioglu, B. E. Cil, G. Furst, J. S. a. Esch, L. B. Fritz, E. Godehardt, U. Modder, and W. T. Knoefel Portal Vein Embolization and Stem Cell Administration Radiology, February 1, 2008; 246(2): 646 - 647. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
