Abstract
The only presently viable treatment for end-stage liver disease is whole organ transplantation. However, there are insufficient livers available. The aim of the present study is to provide autologous bone marrow-derived stem cells as a potential therapeutic for patients with end-stage cirrhosis. This is a retrospective chart review of autologous stem cell treatment in 48 patients, 36 with chronic end-stage hepatitis C-induced liver disease and 12 with end-stage autoimmune liver disease. For all patients, granulocyte colony-stimulating factor was administered to mobilize their hematopoietic stem cells. Following leukapheresis, CD34+ stem cells were isolated, amplified, and partially differentiated in culture, then reinjected into each subject via their hepatic artery or portal vein. Treatment was generally well tolerated with the expected moderate but transient bone pain from G-CSF in less than half of the patients. Three patients had serious treatment-related complications, and only 20.8% of these end-stage liver disease patients died during 12 months of follow up. For all patients there was a statistically significant decrease in ascites. There was clinical and biochemical improvement in a large percentage of patients who received the transplantation. In the viral group, there were marked changes in albumin (p = 0.0003), bilirubin (p = 0.04), INR (p = 0.0003), and ALT levels (p = 0.02). In the autoimmune group, values also improved significantly for albumin (p = 0.001), bilirubin (p = 0.002), INR (p = .0005), and ALT levels (p = 0.003). These results suggest that autologous CD34+ stem cell transplantation may be safely administered and appears to offer some therapeutic benefit to patients with both viral and autoimmune-induced end-stage liver disease.
Introduction
Chronic liver disease in Egypt is recognized as a serious health problem affecting greater than 20% of the population, approximately 16 million people (5,22), and it occurs at a rate that is four times higher than in the US (21). Having been iatrogenically spread during a mass treatment campaign for schistosomiasis infections from 1960 to 1980, Egyptian hepatitis C virus (HCV) cases tend to be advanced, and HCV-related mortality in Egypt is expected to at least double in the next 20 years (5). Patients in Egypt with end-stage liver failure are generally without treatment options apart from supportive measures, except for living-related lobe transplantation, albeit not a realistic one for most because of the high cost and paucity of eligible donors. Thus, cell-based therapies offer an option for patients with advanced disease.
One population of stem cells supporting liver cell renewal is thought to reside in the bone marrow. Hematopoietic stem cells (HSC) may not be the primary contributors to hepatocyte populations in normal circumstances, but they are presumed to make a significant contribution to regeneration after severe injury. It is further suggested that they may have the potential to restore normal function, as shown by Lagasse et al. (12) in a murine model of hereditary tyrosinemia. Other research has suggested that HSC may be more plastic than previously believed (1). Removed from their usual niche, they have been shown in vitro and in vivo to support many types of tissue, including skeletal and cardiac muscle, endothelia, and a variety of epithelia including neuronal cells, pneumocytes, and hepatocytes.
In 1999, Petersen et al. (15) first showed that hepatic cells derived from circulating bone marrow cells regenerated the livers of lethally irradiated rats. Theise et al. (26) reported that in normal conditions, 1—2% of hepatocytes in the murine liver might be derived from bone marrow transplantation. This observation suggests that the bone marrow contributes both to normal wear and tear renewal and to regeneration following injury. The engraftment rate and effectiveness of bone marrow-derived cell transplantation into the liver in animals and in humans is controversial (17,28). However, there is reported evidence that liver insufficiency patients may benefit from bone marrow and HSC infusions (2,7,13,25). For example, small pilot studies have been done previously employing bone marrow-derived cells to treat patients with advanced liver disease. Habib and coworkers (7) performed a feasibly study on five patients with liver disease using selected and augmented autologous CD34+ cells that were infused into the portal vein, with no adverse effects. Although these patients received a relatively low number of cells and no evidence of clinical benefit was sought, moderate improvements in biochemical parameters were recorded at 45 days. Am Esch et al. employed autologous CD133+ cells that were amplified and reinfused via the portal vein into three partial liver donors with a resultant 2.5-fold increase in mean cellular proliferation rates compared to a control group of partial liver donors (2). A 2007 study from Lyra et al. tested the feasibility and safety of autologous bone marrow mononuclear cell transplantation in 10 patients on the waiting list for liver transplantation (13). They infused 1 × 108 bone marrow cells into the hepatic artery. Patients showed improved synthetic function over 4 months. In another study conducted by Terai et al. (25), it was shown that intravenous injection of 400 ml of bone marrow aspirate yielded promising results in nine patients with liver cell failure from cirrhosis.
Clearly, there is a need for alternate treatment options in patients with end-stage liver disease, and autologous bone marrow-derived stem cell transplantation appears to be a viable option. Previous pilot studies employed very small numbers of patients; thus, an analysis of a larger number of patients is desirable. In the present analysis, we investigated the effects of the transplantation of partially differentiated and augmented autologous CD34+ stem cells into 48 patients with advanced liver disease. Our results suggest a positive effect of this cell transplantation approach in patients with end-stage liver disease.
Patients and Methods
Patients
This report is an evaluation of innovative care and thus is neither prospective in nature nor does it adhere to a specific study protocol. It is a retrospective analysis of 36 patients with chronic end-stage hepatitis C (HCV) and 12 patients with chronic end-stage autoimmune liver disease. After approval of the ethics committee of the participating institutions, patients were recruited from hepatology outpatient clinics at Kasr El Aini hospitals, Cairo University, and the National Cancer Institute, Cairo, Egypt. To be eligible for treatment intervention, the patients had to have a diagnosis of chronic advanced liver disease as determined by abnormal serum albumin, bilirubin, and INR values, be unable to receive a liver transplantation, have a Childs-Pugh score of at least B, have a World Health Organization (WHO) performance status of 3 or higher, and demonstrate the ability to give consent to the procedure. The patients were not given this treatment option if they had recent GI bleeding or spontaneous bacterial peritonitis, evidence of HIV or other life-threatening infections, hepatocellular carcinoma, a history of alcohol ingestion or of hypersensitivity to granulocyte colony stimulatory factor (G-CSF), were pregnant or lactating, or were unable to comprehend the procedures or give consent. Patients complying with the full entry criteria were asked to give their written consent and were invited to participate.
Once patients had met the criteria for treatment and agreed to participate, they had a full medical history and physical examination, and blood was drawn. Laboratory investigations included a complete blood count (hemoglobin, white blood cell count, platelet count), clotting screen (prothrombin time, APPT), liver function tests (albumin, bilirubin, AST, ALT, and alkaline phosphatase), and infectious disease profile (HBsAg, HBcAb, HIV-1, and syphilis). Radiological assessment of the abdomen was done before and following stem cell transplantation with abdominal ultrasound.
All 48 patients had cell transplantation and were followed for at least 26 weeks or until they died. None of the 36 patients with HCV received interferon or other therapy within 6 months prior to the cell transplantation or during the follow-up period. All the autoimmune patients had been treated with steroids with or without azathroprine without improvement before stem cell transplantation. Levels of immunosuppressant agents were not changed following transplantation. None of the patients in the autoimmune group had positive markers for HCV or HBV. All 48 patients were tested by ultrasound for hepatic fibrosis caused by schistosomiasis, and they were found to be negative for evidence of the portal tract thickening that is pathogneumonic for the disease.
Clinical Protocol
Days 1 to 5. Patients received a daily subcutaneous injection of 300 g of G-CSF (Neupogen, Roche Pharmaceutical) for 5 days to increase the numbers of circulating hematopoietic stem cells.
Day 6. Patients underwent a leukapheresis procedure in the Department of Hematology, National Cancer Institute. Following the collection, the cells were transferred to the stem cell laboratory for immunomagnetic separation of the CD34+ stem cell population. The CD34+ cells were placed in culture for amplification and differentiation from day 6 to day 13.
Day 13. Patients were admitted for collection of blood for CBC and liver function tests. The patients were infused with approximately 1 billion expanded CD34+ stem cells into either the portal vein, if hepatopedal flow, or hepatic artery, if hepatofugal flow, under ultrasound or CT guidance.
Postinfusion of CD34+ Cells. Patients were asked to return at 1, 3, 8, 12, 24, 36, and 48 weeks after cell transplantation for blood tests (CBC, liver function tests, and INR), and a physical examination, including documentation of ascites.
Laboratory Methodology
Cell Source, Isolation, and Cultivation. G-CSF-mobilized peripheral blood (200 ml) was obtained from leukapheresis, and CD34+ cells were isolated from the leukapheresed blood by adding 1 volume of blood to 5—10 volumes of the lysis buffer (155 mM NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA) for 30 min at 4°C to remove RBCs and to separate mononuclear cells (MNCs). The collected MNCs were washed twice with buffer [phosphate-buffered saline (PBS), pH 7.4, supplemented with 0.5% bovine serum albumin and 5 mM EDTA], and centrifuged at 1500 rpm for 10 min. The MNCs were counted, adjusted to 2 × 109/ml, and centrifuged. CD34+ cells were isolated using the CD34+ positive cell selection kit (CliniMacs, Germany). The cells were cultured in α-MEM/Ham F12 (1:1) containing 15% BSA, 1% penicillin/streptomycin, 1 ng/ml GM-CSF, and growth factor (multiplication stimulating factor: buffalo rat liver extract 20 mg/ml; Sigma Lot #35F-0165) and incubated for 5—7 days at 37°C in 5% CO2. Morphology of the cultured CD34+ cells was assayed by phase contrast microscopy.
RNA Isolation. RNA was isolated from fresh cells immediately after isolation, and from cells that were cultured, using the RNeasy Mini Kit according to the manufacturer's instructions (Qiagen, Crawley, West Sussex, UK). Samples were treated with DNase (Promega, Southampton, UK). The RNA was then purified using the PCR Purification Kit (Qiagen) according to the manufacturer's instructions. Total RNA concentration was assessed by measuring the optical density at 260 nm in a spectrophotometer.
Reverse Transcription (RT) and PCR. RT-PCR was done using the One-Step RT-PCR Kit (Qiagen) according to the manufacturer's instructions. The RT reaction was performed at 50°C for 1 min. The primer sequences of the studied genes and the PCR conditions are illustrated in Table 1. A positive control (HepG2 cell line) for the studied genes, together with negative controls and culture medium controls, was included in each PCR run. The GAPDH gene was used as an internal RNA standard. PCR products were analyzed by electrophoresis in an ethidium bromide-stained 2% agarose gel and visualized under UV light on a transilluminator.
Primer Sequences and PCR Conditions of the Studied Genes
Assessment of Safety
Safety was evaluated in terms of adverse events graded according to the Common Toxicity Criteria grading system and laboratory test results. All adverse events were evaluated for relationship to treatment and as expected and unexpected. A short-term fever developed in 14 patients following transplantation, and subsided within 24 h. Mild—moderate bony aches occurred in 22 patients after receiving G-CSF; they subsided spontaneously and did not affect the cell transplantation. There were three significant complications of transplantation. One patient developed hematemesis 1 week after intraportal injection of the cells. It was endoscopically controlled and considered as likely to have been a result of the procedure. Another patient had hemoperitoneum after portal injection and was stabilized with blood transfusion. One patient died from gastrointestinal hemorrhage 1 week after stem cell transplantation, considered as likely to have been a result of the transplantation procedure.
Assessment of Efficacy
Clinical. Physical examinations to determine the amount of ascites, as well as the evaluation of symptoms, were done by one highly experienced physician (HS). Amount of ascites was scored as none, mild, moderate, and marked.
Biochemical. Albumin, bilirubin, ALT, and INR were measured on the day of stem cell transplantation and then at 1, 3, 8, 12, 24, 36, and 48 weeks after transplantation, as well as during clinical follow-up after 1 year.
Statistical Methods
Means and SDs for bilirubin, ALT, INR, and albumin were calculated for baseline and follow-up visits at approximately 1, 3, 8, 12, 24, 36, and 48 weeks posttransplantation, stratified by viral and nonviral groups. Wilcoxon signed rank tests (for nonparametrically distributed variables) were used to test for changes in each of the lab values and ascites from the date of stem cell transfer (baseline) to each follow-up visit date. Mixed effects regression models were used to estimate the weekly change in each of the lab values over the entire follow-up period. These regression models account for the covariance of repeated measures for each patient. The models were adjusted for age and gender, and were stratified by viral or nonviral group. All analyses were performed using SAS software version 9.1 (SAS Institute, Inc., Cary, NC, USA).
Results
Characterization of CD34+-Derived Hepatocyte-Like Cells
Cell morphology of the cultured CD34+ cells was assayed by phase contrast microscopy at 2 days (Fig. 1A) and at later stages of differentiation, days 7, 10, and 21 (Fig. 1B, C, D). The cells proliferated when cultured and, although they did not display significant morphological change characteristic of hepatocyte-like cells by day 7 in culture, their gene expression profile began to include liver-specific gene expression as assessed by RT-PCR (Fig. 2). Thus, this date was chosen for cell transplantation because the cells were actively proliferating and were becoming more hepatocyte like. At day 7 the cells appeared to be at a progenitor stage, less like adult hepatocytes than at day 21 of culture when their morphologic appearance was distinctly more similar to differentiated hepatocytes. By day 21, the cells were larger with polyhydral contour, binucleation, and inconspicuous nucleoli (Fig. 1D).
CD34+ cells grown as a monolayer in tissue culture flasks at different days of culture. G-CSF-mobilized peripheral blood cells were obtained by leukapheresis, and CD34+ cells were isolated as described in Patients and Methods. The cells were then cultured to amplify and differentiate them. Phase contrast microscopy was performed at (A) 2 days, (B) 7, (C) 10, and (D) 21 days of culture. Cells were harvested for cell transplantation at day 7.
RT-PCR of CD34+ gene expression at different days of culture. CD34+ cells were isolated and cultured as described in Patients and Methods. RNA was isolated after different days in culture and RT-PCR were undertaken.
Clinical Outcomes
Table 2 shows mean values and SDs for the laboratory tests at baseline. The two groups, those with or without viral infection, were rather similar in their laboratory testing. Of note was the advanced stage of their liver disease. Table 3 provides the laboratory testing results for all 36 viral patients who received cell transplantation and were followed for at least 6 months. They demonstrate marked improvement in each parameter that was evaluated, beginning as early as 1 week following transplantation, with a longer delay in the improvement in albumin. All liver function tests improved; bilirubin (p = 0.04), ALT (p = 0.02), INR (p = 0.0003), and albumin (p = 0.0003), adjusted for age and sex. Table 4 provides the laboratory testing results for 12 autoimmune patients who received cell transplantation and were followed for at least 6 months. They demonstrate significant improvement in each parameter, beginning as early as 1 week following transplantation, again with a longer delay in the improvement in albumin. All liver function tests improved; bilirubin (p = 0.02), ALT (p = 0.003), INR (p = 0.0005), and albumin (p = 0.001), adjusted for age and sex. Figure 3 is a graphic representation of these four biochemical parameters, and ascites, in all 48 patients.
Laboratory Tests at Baseline
Biochemical and clinical evaluation of patients with end-stage liver disease transplanted with CD34+ cells. (A—D) Blood tests and (E) determination of ascites in patients at different weeks posttransplantation.
One possible concern with the data is that the improvement in the biochemical parameters over time, as demonstrated in Tables 3 and 4, was caused primarily by the removal of the very sick patients when they died. That criticism would postulate that the patients who remained in the later stages of the study had better laboratory tests because they were less sick initially and thus did not die. But that was not the case. The patients who died had, if anything, slightly better liver function than did the patients who lived when therapy was initiated, and their tests also improved, with the exception of albumin, which remained stable. This is shown in Table 5. Thus, it is clear that there was a real improvement in all these biochemical parameters in the vast majority of patients, an improvement than can best be explained by the cell transplantation.
Lab Values for HCV Patients
Week 0 represents pretreatment baseline values.
p-Value for difference from baseline (Wilcoxon signed rank test).
p-Value for trend over time, adjusted for age and sex (mixed effects regression models).
Lab Values for Autoimmune Patients
Week 0 represents pretreatment baseline values.
p-Value for difference from baseline (Wilcoxon signed rank test).
p-Value for trend over time, adjusted for age and sex (mixed effects regression models).
Lab Values for Patients Who Died During the Study Period
Week 0 represents pretreatment baseline values.
Two of the 48 patients in the analysis had none or minimal ascites at the time that they were transplanted, 43 of 48 (90%) had mild to marked ascites, and three participants were missing this data. We had 6-month follow-up ascites data on 37 of the patients with mild to marked ascites. Twenty-seven of them had improvement of ascites during the first 6 months following transplantation. This occurred without an increase in their diuretic regimen. Four of the mild to marked ascites patients were unchanged, and one worsened. The improvement in ascites was statistically significant starting at week 3 after transplantation (p < 0.05) and remained significant for 24 weeks after transplantation. Ascites (Fig. 3E) continued to improve among the patients who were followed for more than 24 weeks, but due to the small sample size at the longer durations, the improvements were not statistically significant.
Discussion
The only current therapy modality for liver failure is liver transplantation. Unfortunately, relatively few patients who require transplantation are transplanted, due to organ shortages, high costs, etc. The result of this situation is that the majority of untransplanted patients die, as there is no effective liver support device that is comparable to renal dialysis. For this reason it is necessary to develop new therapeutic modalities for patients with chronic liver disease. The use of stem cell therapy is promising; cellular therapies based on stem cells and their derivatives are likely to revolutionize the practice of medicine in the foreseeable future. Human hematopoietic tissue contains two major stem cell populations, the hematopoietic (CD34+) and the mesenchymal (CD34-) stem cells. The CD34+ cells that we have used in the present analysis can be isolated from bone marrow, or from leukapheresed blood when the patient has been treated with G-CSF, in abundant numbers and as a relatively homogeneous cell population. When purified and amplified in a hepatocyte differentiation medium, they can then be infused back into the donor with advanced liver disease with significant improvement in liver function.
We found that treatment with G-CSF was generally well tolerated in these very sick patients. Less than half experienced moderate transient bone pain. The G-CSF treatment over 5 days allowed us to obtain high levels of mononuclear cells in all of these patients with relatively suppressed bone marrows. Thus, we were able to obtain reasonable numbers of CD34+ cells that were then amplified and partially differentiated into hepatocyte precursor cells. This enabled us to transplant as many as 1 billion of these cells per patient. Of note, there are suggestions in the literature that treatment with G-CSF alone may improve liver function in animal models or patients with liver disease (6,27). Thus, it is possible that at least some of the effect of our cell transplantation protocol could be attributed to G-CSF treatment alone. However, the magnitude of the improvement that we saw in these patients has not been shown to occur in any previous reports. Moreover, numerous interferon-treated patients with advanced HCV disease are routinely treated with G-CSF without manifesting the type of positive effects shown in the patients treated with our protocol.
Our results indicate that the treatment protocol led to an acceptable level of complications in a population of very sick patients. Transient fevers and bone pain occurred in a minority of patients and were tolerable. Three of 48 patients experienced serious complications likely due to cell transplantation, requiring hospitalization, including only one death.
One concern is the induction of increased portal hypertension with the cell transplantation. Our experience showed that serious complications of portal hypertension developed in only two of our patients. The potential complication can be reduced in future studies by employing hepatic artery infusion as the sole approach to transplantation.
Ten deaths (20.8%) in a 1-year follow-up in this population of 48 patients is better than expected, given their advanced disease. Unfortunately, serum creatinine levels were not routinely recorded in this patient group; thus, MELD scores cannot be determined. However, given the severe degree of liver disease as manifested by their mean albumin of 2.76, bilirubin of 2.46, and INR of 1.59, and the presence of ascites in essentially all patients, it is quite likely that their MELD scores would be on average above 15. Assuming that almost all the patients had MELD scores between 10 and 19, which translates to a 3-month mortality rate of 26% (11), then our mortality rate of 6.3% over 3 months and 20.8% over 1 year would appear to be extremely good. What is clearly determined by our results is that patients with either HCV or autoimmune liver disease had marked improvement in their synthetic function and in their ascites, a result not expected in patients with this level of disease.
The finding of improvement in ascites in a significant number of the patients is impressive and somewhat surprising. It suggests that cell transplantation might be of some considerable clinical significance beyond mere improvement in laboratory parameters. The improvement was statistically significant, without a change in the diuretic regimen.
The improvement in ascites is probably multifactorial. Improved albumin levels would certainly cause an increase in oncotic pressure. But such a significant improvement in ascites might also suggest a decrease in portal hypertension although there is no direct evidence for this. However, previous studies have suggested that bone marrow-derived cells can enhance metalloproteinase activity (8). Several putative mechanisms for the emergence of improved liver function after bone marrow-derived stem cell transplantation should be considered.
First, bone marrow could theoretically harbor a specialized endodermal stem cell capable of producing hepatocytes and other epithelial cells. This cell would be analogous to the mesenchymal stem cell that produces mesodermal derivatives such as muscle, cartilage, and fat (4,9). Second, hepatocytes and blood cells could be derived from the same stem cell by hierarchical differentiation (19). Third, bone marrow-derived hepatocytes could potentially be derived not from differentiation at all but rather by cell fusion between host hepatocytes and transplanted donor cells. This possibility was raised by the observation that hematopoietic cells could spontaneously fuse with embryonic stem cells in vitro and then produce multiple tissues in chimeric mouse embryos (29,30), as well as many other reports of cell fusion (4,16,29,30). Fourth, bone marrow-derived stem cells could be plastic and home to the injured liver and differentiate along a hepatocyte lineage there (10,19,21,23). Fifth, bone marrow-derived stem cells may act in a paracrine manner to enhance the function and proliferation of endogenous hepatocytes and hepatic progenitor cells (20).
Our clinical analysis in patients with advanced disease, not readily amenable to liver biopsies, does not differentiate among these different mechanisms, and other studies must be undertaken to determine which mechanism(s) might explain our positive results. Of note, all four laboratory values and ascites show linear improvement over time, even at 1 year after transplantation. This result suggests that hepatocyte proliferation is likely to be occurring rather than there being only a short-term paracrine effect of growth factor synthesis, which would likely have a more limited and short-term effect. Certainly the predominance of data using hematopoietic stem cell transplantation, particularly in other organs, suggests that hematopoietic stem cells are not transdifferentiating into hepatocytes. However, in our analysis, the cells that we were transplanting had already begun to develop a hepatocyte phenotype in culture prior to transplantation; thus, it is tempting to speculate that these cells may have acted as hepatocypte-like cells following engraftment, by cell fusion or by transdifferentiation, as well as providing paracrine effects on endogenous hepatocytes.
Our use of G-CSF was successful in mobilizing a sizable number of CD34+ stem cells into the peripheral circulation in order that they could be accessed by leukapheresis. The culturing step that followed allowed for the proliferation of the stem cell population and for their expression of hepatocyte-specific genes. Thus, a highly select population of stem cells with early characteristics of a hepatocyte lineage was reinfused into the patient. These cells are not fully differentiated hepatocytes, but appear to represent a precursor cell with hepatocyte-like activity. Moreover, G-CSF administration will likely enhance a variety of circulating bone marrow-derived cells producing growth factors, which could positively affect liver regeneration and perhaps even have a positive effect on portal hypertension. A series of preclinical studies have been done recently which suggest that bone marrow-derived cells can enhance liver function by increasing hepatocyte proliferation and by inhibiting fibrosis through increasing metalloproteinase production and/ or decreasing the synthesis of tissue inhibitor of metalloproteinase (3,8,14,24,31). The marked improvement in laboratory parameters in subjects with advanced liver disease that was associated with our analysis suggests that such mechanisms might be at work in our transplanted subjects. However, we have no proof of such mechanisms.
In the present analysis we have infused autologous CD34+ stem cells into 48 patients with advanced liver disease and have demonstrated biochemical and clinical evidence of improvement in a very high percentage of the patients with a limited number of major side effects. The mechanism by which the infusion of the cells improves the liver function is yet to be determined. Our next effort will be to establish a prospective, placebo-controlled study to better assess the efficacy of this treatment strategy. In forthcoming trials we will seek to have longer follow-up periods, enter sicker patients to determine whether mortality can be improved, and compare different routes of infusion in an attempt to determine whether peripheral infusion of the transplanted cells will be adequate. The use of peripheral infusion would markedly decrease the costs and risks of this cell transplantation, thus making the treatment approach a more viable option in Egypt, and throughout the world.
Footnotes
Acknowledgment
Partial funding support was provided by the California Institute of Regenerative Medicine, RC1-000359 (MAZ).