A Phase Ii Dose Escalation Study of Intraarterial (Hepatic) Adult Human Bone Marrow Derived, Cultured, Pooled, Allogeneic Mesenchymal Stromal Cells (Stempeucel®) in Patients with Alcoholic Liver Cirrhosis

Background: Alcoholic liver cirrhosis is an end stage alcoholic liver disease with poor prognosis. The denitive treatment of alcoholic liver cirrhosis is orthotopic liver transplantation, which is expensive, requires long term immunosuppression and is limited by supply of organs. Being an unmet medical need, cell therapy is under investigation for alcoholic liver cirrhosis. Aims: This study was designed primarily for assessing the safety and feasibility of administering stempeucel® through the hepatic artery in alcoholic liver cirrhosis and secondarily to assess possible ecacy and dose-response. Methods: Forty patients with alcoholic cirrhosis (18-65 years/Child Pugh class B or C/Model for End-Stage Liver Disease score of minimum 10) were included in 4 groups: 2.5 million cells/kg Body Weight (2.5M Cell) and respective control (2.5M Control); 5 million cells/kg Body Weight (5M Cell) and respective control (5M Control) with 10 patients in each group. Cell groups received stempeucel® administered via hepatic artery by catheterization through femoral artery (Seldinger technique) and Standard Protocol of Care. Control group received Standard Protocol of Care. Patients were followed up at 1 week, 1 month, 3 months and 6 months. Safety evaluations included clinical examination, Electrocardiogram and laboratory investigations. Ecacy evaluations included liver function test, Model for End-Stage Liver Disease score, Child Pugh score, Short Form-36 questionnaire, liver stiffness using Fibroscan (Transient Elastography), and liver volume using Computerized Tomography scan. Results: stempeucel® injection was well tolerated. Common treatment emergent adverse events were in Gastrointestinal disorders, General disorders and administration site conditions and Infections and infestations. Most of the treatment emergent adverse events were unrelated / remotely related to stempeucel®. Thirty serious adverse events occurred in 10 patients (3 in 2.5M Cell, 5 in 5M Cell and one each in control groups). Three patients died due to SAEs: Two in 2.5M and one in 5M Cell group, none were related to stempeucel®. There was no signicant difference in ecacy evaluations at 6 months versus baseline compared to control at both the dose levels of stempeucel®. Statistically signicant improvement was seen in 2.5M group compared to control group in Short Form-36 score: bodily pain, mental component summary, vitality and social functioning. Conclusions: stempeucel® was safe, well tolerated and subjective improvement in few component scores of Short Form-36 was seen 2.5M cell group. hepatitis B (26) and Umbilical Cord Mesenchymal Stem Cells (UCMSC) in hepatitis B (27, 28) with successful outcomes. Most studies used a single dose level, had single arm and involved small sample size. We designed a dose nding study with the primary objective of evaluating the safety and feasibility of intraarterial (hepatic) administration of 3 dose levels (2.5, 5 and 7.5 million cells/kg body weight) of stempeucel® in ALC. was removed. The catheter was guided into the coeliac axis. The catheter was then selectively negotiated into the hepatic artery. After accessing the hepatic artery (which is conrmed by injecting a non-ionic contrast agent), stempeucel® was injected in the artery at a rate of 1 ml/min using auto-injector. Standard protocol of care was as per the investigator discretion and was given to patients both in stempeucel® and control group. They included diuretics, antacids, multi vitamin preparations, laxatives, beta blockers, hepato-protectors, antidiarrheals, and systemic antibacterial medications. Control group did not receive any placebo injection due to invasive nature of administration. demonstrate histopathology biopsy cirrhotic in albumin, total protein, AFP and proliferating cell nuclear antigen in liver biopsy after 4 weeks of autologous BMMNC infusion therapy et in ascites, liver function, MELD score in addition to decrease in liver brosis et of hepatic progenitor cell compartment, hepatic progenitor cell improvement in Child Pugh scores 80% increase in liver volume per Magnetic et al seen histological improvement in 6 out of 11 patients who received BMSC through the hepatic artery for ALC Enhanced angiogenesis was seen in follow-up liver biopsy specimens after boost infusions of mobilized peripheral blood stem cells in decompensated alcoholic cirrhosis in a study by et the present study, we did not conduct liver biopsy at 6 months follow-up. small tissue adequate representation of liver intra-observer variation 67). Fibroscan Elastography) non-invasive stiffness validated liver cirrhosis reproducible and reasonably safe to administer stempeucel® through intraarterial route (hepatic artery) in ALC at a dose of up to 2.5 million cells/kg body weight. Higher doses of stempeucel® administered through the intraarterial (hepatic artery) route may not be justied in ALC owing to the limited ecacy seen at this dose and higher risk of complications at 5 million cells/kg dose group. Ecacy was limited to improvement in quality of life in 2.5 million cells/kg dose group. Future studies have to be done to identify the appropriate dose and route of administering cell therapy for ALC.

MSC (14,15). Fusion of donor BMMSC to recipient hepatocytes has also been cited as a reason for hepatic regeneration (16,17). BMMSC reduce oxidative stress and inhibit apoptosis of hepatocytes (18). Vascular Endothelial Growth Factor secreted by MSCs induce angiogenesis and thus contribute to healing of target organ (19). Zhang et al. (20) reported that the expression of human albumin, alfa fetoprotein (AFP), cytokeratin 18 and cytokeratin 19 were detected in the liver tissue of brotic and cirrhotic rats after MSC transplantation, suggesting that transplanted MSCs could migrate into the injured liver, where they could differentiate into hepatocyte-like cells. Furthermore, they also demonstrated that MSCs did not directly differentiate into functional hepatocytes; they rst differentiated into epithelial cell-like cells and subsequently differentiated into hepatocyte-like cells. All the above ndings indicated that MSCs could differentiate into hepatocyte-like cells through exposure to the liver brosis microenvironment both in vitro and in vivo.
Preclinical studies evaluating MSC in liver brosis are mostly in conducted in carbon tetrachloride induced liver brosis experimental model. Luo et al administered human BMMSC to rats through portal vein and found improvements in liver function and reduction in brosis associated with differentiation into hepatocyte-like cells (21). Tanimoto et al injected BMMSC through the tail vein into mice and found reduction in brosis, enhanced expression of MMP-9 and decreased expression of alpha smooth muscle actin (α-SMA), TNF alpha, transforming growth factor (TGF) beta (22).
In clinical setting, MSCs have been evaluated in various types of liver cirrhosis using different routes of administration. Jang et al administered 50 million autologous BMMSC through the hepatic artery in alcoholic cirrhosis and found improvements in Child Pugh score, decrease of TGF-beta1, collagen type 1 and α-SMA in addition to histological improvements (23). There are reports of studies involving BMMSC in chronic hepatitis C (24,25), hepatitis B (26) and Umbilical Cord Mesenchymal Stem Cells (UCMSC) in hepatitis B (27,28) with successful outcomes. Most studies used a single dose level, had single arm and involved small sample size. We designed a dose nding study with the primary objective of evaluating the safety and feasibility of intraarterial (hepatic) administration of 3 dose levels (2.5, 5 and 7.5 million cells/kg body weight) of stempeucel® in ALC. "We present the following article in accordance with the CONSORT reporting checklist."

Study design
This trial was designed as a phase II, dose nding, parallel group, randomized, open label study. The protocol was approved by the Drugs Controller General of India (Indian FDA) and ethics committees of all 9 participating clinical trial sites. The study was conducted as per International Council for Harmonization Good Clinical Practice (ICH-GCP) guidelines, Principles of Declaration of Helsinki, Schedule Y of Drugs and Cosmetic Act, 1945, and Ethical guidelines for biomedical research on human participants, Indian Council of Medical Research, 2006. An independent data safety monitoring board (DSMB) was constituted to review the data of patients at prede ned intervals and adhoc whenever required. Study was conducted in India. Informed consent was obtained from each patient before screening. The eligibility criteria are mentioned in Table 1. Patients were randomized to either cell or control group at each dose level (2.5 or 5 million cells/kg) as per prede ned randomization code concealed from the investigators. At each dose level, 20 patients were randomized into 5 blocks with the block size of 4 patients. Study was registered in clinicaltrials.gov website (NCT01591200). The bone marrow aspiration protocol was approved by the Institutional Ethics Committee. Healthy consenting voluntary donors in the age group 18-40 years, who were not Human Leucocyte Antigen matched to recipients were screened according to Indian Council of Medical Research guideline for healthy bone marrow donor screening. Sixty ml of bone marrow was aspirated from posterior superior iliac spine of both sides of each volunteer. It was diluted (1:1) with knockout Dulbecco's modi ed Eagle's medium (KO-DMEM; Gibco-Invitrogen, Grand island, New York, USA), and centrifuged at 1,800 (g) for 10 min to remove the anti-coagulant. Bone marrow Mononuclear Cells (BMMNCs) were separated by density gradient centrifugation (1.077g/ml) as described earlier (29). Plastic adherence was used to isolate BMMSCs from the donor BMMNC and cultured till passage 1. Donor master cell bank containing individual donor's BMMSCs was created and cryopreserved. Subsequently, a working cell bank (WCB) was prepared by combining MSCs from three individual donors and cryopreserved. The WCB was used for manufacturing stempeucel® and further expanding the pooled WCB for additional passages (US patent number 8956862 dated 02/17/2015). For the current clinical trial, pooled BMMSCs were cultured, harvested and characterized at passage 5 and cryopreserved in liquid nitrogen as the nal product stempeucel®, which was the investigational medicinal product (IMP). Speci cations and release criteria of stempeucel® are the same as published by us earlier (30). Stempeucel® (200 million ± 10%) was stored in 15 ml of PLASMA-LYTE A (multiple electrolytes injection, type 1, United States Pharmacopeia) containing 5% human serum albumin (Baxter Healthcare, California, USA) and 10% dimethyl sulfoxide (Sigma -Aldrich, Irvine, United Kingdom) in a cryocyte bag (MacoPharma, Mouvaux, France). Stempeucel® was shipped to clinical trial sites in liquid nitrogen (-185 to -196 O C) for each patient in the cell group.

Reconstitution of stempeucel® at clinical trial sites
Reconstitution of stempeucel® was done by a trained person independent of investigator's team. The procedure was done under a validated biosafety cabinet. Cryocyte bag containing stempeucel® was thawed in a water bath at 37 O C. Cell suspension was diluted to 50 ml using PLASMA-LYTE A using 50 ml centrifuge tube. Representative cell suspension sample was taken for performance of cell count. Based on cell count and body weight of the patient, nal cell suspension was made in a total volume of not more than 50 ml of PLASMA-LYTE A and transferred to a new cryocyte bag. The cryocyte bag was placed in a temperature controlled, validated transport box at 2-8 O C and shipped to the cath lab for intraarterial injection.

Intraarterial injection protocol
Before injecting stempeucel®, patients were pre-medicated using 100 mg Inj. Hydrocortisone and 45.5 mg Inj. Pheniramine maleate (both administered intravenously). Seldinger technique was used for hepatic artery cannulation (31). Femoral artery was accessed using an introducer needle under local anesthesia. A soft tipped guide wire with a diameter of 0.038 cm was passed through the needle and the needle was removed. A dilator of a diameter of 6 French was passed over the guide wire. Dilator was removed and catheter (100-120 cm long with diameter of 5 French) was passed over wire and wire was removed. The catheter was guided into the coeliac axis. The catheter was then selectively negotiated into the hepatic artery. After accessing the hepatic artery (which is con rmed by injecting a non-ionic contrast agent), stempeucel® was injected in the artery at a rate of 1 ml/min using auto-injector.
Standard protocol of care was as per the investigator discretion and was given to patients both in stempeucel® and control group. They included diuretics, antacids, multi vitamin preparations, laxatives, beta blockers, hepato-protectors, antidiarrheals, and systemic antibacterial medications. Control group did not receive any placebo injection due to invasive nature of administration.

Patient follow-up
All patients were observed for at least 24 hours after administration of stempeucel® before discharging from the hospital. Follow-up evaluations were done at 1 week, 1 month, 3 months and 6 months. Telephonic follow-up was done on 15th day for knowing the well-being of patients.

Study endpoints
Primary endpoint was safety and tolerability, assessed by occurrence and type of treatment emergent adverse events (TEAEs) (onset on or after the IMP administration visit [for cell group] and randomization visit [for control group]), electrocardiogram parameters, hematological and biochemical values, physical examination and vital signs. Secondary endpoints included assessment of e cacy by liver function tests (LFT), Model for End-Stage Liver Disease (MELD) and Child Pugh score, Quality of life as per Short Form-36 (SF-36) questionnaire, liver stiffness measured by Fibroscan and volume of liver measured by Computerized Tomography (CT) scan.

Data collection
Electronic case record form was used for data collection. Data was veri ed with source notes by third party monitors independent of investigators.

Data Safety Monitoring Board
An independent DSMB was constituted comprising of drug safety physicians, pharmacovigilance expert and gastroenterologist. The DSMB rst met during protocol nalization. Second meeting was held to review the 1 week follow-up data of 20 patients in the 2.5 million cells/kg dose level (both cell and control group). Third meeting was held to review the data of 20 patients in the 5 million cells/kg dose level (both cell and control group) along with the cumulative data of all patients in the 2.5 million cells/kg body weight dose group. Final meeting was planned to review of safety data of rst ve patients in the cell group of 7.5 million cells/kg, along with other cumulative data before dosing remaining patients at this dose level.
Statistical analysis SAS package (SAS Institute Inc, USA, version 9.2) was used for statistical analysis. Data are presented as mean ± SD. TEAEs are summarized descriptively by total number of AEs in each group by system organ class (SOC). Normality of continuous data was tested using Kolmogorov-Smirnov test. E cacy data was analyzed using Analysis of Covariance or Nonparametric Wilcoxon Rank Sum test as appropriate. P < 0.05 was considered statistically signi cant. E cacy data are presented for modi ed intention to treat population, which represents the patients who had at least one post baseline e cacy data point.

Results
Of the 69 patients screened, 40 were enrolled in the study. At each dose level, 20 patients were randomized to either cell group or control group. The CONSORT diagram ( Fig. 1) shows the number of patients screened, enrolled to each group and completing the 6 months followup.

Patient characteristics
The demographics and baseline characteristics of the patients are given in Table 2. It is seen that the groups are matched in terms of baseline characteristics. Data are expressed as Mean ± SD Safety of the procedure Most patients tolerated the intraarterial injection of stempeucel® without any safety issues except one patient, who developed dissection of hepatic artery during catheterization. Procedure was aborted immediately and stempeucel® was not administered. This patient was observed for 24 hours before discharging the next day without further complications. However, the patient died due to complications associated with liver cirrhosis 1 month later.
TEAEs during the follow-up period are summarized in Table 3. Common TEAEs were in the SOC Gastrointestinal disorders, General disorders and administration site conditions, and Infections and infestations. Most of the TEAEs were unrelated / remotely related to the IMP. Electrocardiogram parameters, hematological and biochemical values, physical examination and vital signs did not reveal any signi cant abnormalities compared to baseline (data not presented).  (5) 24 (6) 32 (6) 13 (6) Blood and lymphatic system disorders General disorders and administration site conditions 9 (4) 5 Three patients died due to SAEs: Two in 2.5M Cell group (both due to hepatic encephalopathy and associated complications) and one in 5M Cell group (due to bilateral bronchopneumonia with sepsis with renal failure), none of which were related to stempeucel® as per the investigators.

DSMB recommendation
Upon review of data during second meeting, the DSMB opined that there was no safety concerns noted and recommended that the next higher dose of 5 million cells/kg body weight of the IMP may be administered as per the protocol. In the third meeting, the DSMB observed that the patients in the cell groups of the study, and more so who received 5 million cells/kg body weight have experienced more SAEs in SOC of infections than those in the control group. The DSMB recommended that the study be stopped with no further IMP administration; however the monitoring of the patients had to be continued as per protocol. Cause of infection was not attributed to the IMP by the DSMB. As per the investigators, infections were not related to the IMP except fever in one patient (which occurred after 2 weeks following injection of 2.5 million cells/kg body weight, labeled as possibly related) and bacteremia in another patient (which occurred after 4 months following 5 million cells/kg body weight, labeled probably related  (Table 4) and liver volume measured by CT scan (Table 5) did not show any difference between cell and control groups at both the dose levels.

Discussion
To our knowledge, this is the rst dose nding study using allogeneic BMMSC in ALC. Study has shown that it is feasible to administer stempeucel® at a dose of 2.5 million & 5 million cells/kg body weight through the hepatic artery in ALC. The procedure was reasonably tolerated well in majority of patients though one patient in the 5 million cells/kg dose group developed hepatic artery dissection during catheterization. There were higher incidents of infections in patients who received 5 million cells/kg body weight compared to control group even though not all were attributed to stempeucel® by the investigators.
Hepatic artery catheterization has been in practice since 1960s for administering anticancer chemotherapy (32,33) and hepatic arterial dissection is a known complication of this procedure. In a report by Habbe et al, six incidents of hepatic artery dissections were observed in 100 attempted hepatic arterial catheter placements for administering chemotherapy (34). In another study of chemotherapy for hepatic malignancy, there was one incidence of hepatic artery dissection out of 28 patients (35). Similarly, intraarterial chemotherapy resulted in one case of hepatic artery dissection out of 30 patients (36). In a phase 1 study of bone marrow mononuclear therapy in cirrhosis in 8 patients, there was one incidence of hepatic artery dissection (37). Thus, the incidence of this complication in our study (one in 20 patients) is comparable to that reported in studies involving hepatic artery catheterization including those intended for stem cell delivery.
The DSMB opined that there is increased incidence of infection in the cell group compared to that of control. Theoretically, MSC can lead to enhanced susceptibility to infection through immunomodulatory function, particularly in patients with liver cirrhosis due to immunosuppression. It is possible that the increased rate of infection in this study was also because of preexisting immunosuppression due to cirrhosis in these patients. Infection and its complications were seen in similar studies involving administration of stem cells in liver cirrhosis. In a study by Sharma et al, one patient died on the 88th day post CD34 + cell transplantation due to development of sepsis and hepatorenal syndrome (38). In a case report by Gasbarrini et al, infusion of CD34 + resulted in fatal outcome due to multiorgan failure secondary to bacterial infection (39). Autologous bone marrow cell infusion in patients with liver cirrhosis resulted in fever in all recipients in a study by Terai S et al (40). Two patients developed self-limiting fever within 2-6 hours after UCMSC administration in acute-on-chronic liver failure patients (28). Hence it appears that infections and its complications are a common place in interventions involving cell therapy in morbid liver conditions.
Infections following stem cell administration have been seen in non-cirrhotic conditions also. In a phase I study using autologous BMMSC for therapy of allograft rejection following renal transplantation, 3 out of 6 patients developed opportunistic viral infection (41). This has been speculated to be due to the immunosuppressive effects of MSC (42). In Graft versus host disease patients, MSC therapy has raised concerns over infections as a complication (43). However, a meta-analysis of MSC studies showed that there was no difference between MSC and control groups in terms of occurrence of infection (44). The same report revealed a signi cant increase in transient fever in MSC group compared to control probably due to acute in ammatory reactions to particular MSC preparations.
Paradoxically, BMMSCs are thought to be protective against infectious diseases through direct effects on pathogens or indirect effect on the host. While they reduce proin ammatory cytokine and chemokine induction and reduce the migration of proin ammatory cells into sites of injury in the host, they also exert antimicrobial effect on the infectious agents (45). Mechanism of antimicrobial effects include indoleamine 2,3-dioxygenase expression induced by in ammatory cytokines (46) and secretion of cathelicidin LL-37 antimicrobial peptide (47). Antifungal effect has also been demonstrated by IL-17 producing subset of MSC (48). Bene cial role of MSC has also been discussed in tuberculosis, through immunomodulatory functions favorable to the host and down regulation of host susceptibility to infection (49). Sepsis, which is a deranged response of host immune mechanism to microbial invasion, results in organs damage. MSC has been considered a suitable agent to be tested for sepsis because of its antibacterial, immunomodulatory effect, antiapoptosis and regenerative response (50). Extensive preclinical studies have demonstrated e cacy of MSC in animal models of sepsis (51,52). One clinical trial has also been initiated using MSC in septic shock (53). Arango-Rodriguez has reviewed the mechanisms through which MSCs can facilitate infection in the recipient as well as literature suggesting that MSC may reduce infection (54). Con icting opinions about role of MSC in infection are probably because of the heterogenicity in MSC with respect to its source, dose, route of administration and the disease condition in which it is administered.
Cell therapy can be administered to liver cirrhosis patients through different routes: peripheral vein, portal vein, spleen and hepatic artery. Intravenous delivery has been commonly used for administration of cells in liver cirrhosis patients. Portal vein catheterization has technical challenges due to ascites in these patients and additional risk of portal vein thrombosis and subsequent variceal bleed. Intrasplenic route has been used by few studies for cell administration in liver cirrhosis (25,55). Hepatic artery catheterization was chosen in our study, owing to higher proportion of cells possibly lodging in liver. Other than dissection, the potential complication of this route of delivery is worsening of liver function due to embolization. Deterioration of liver function was not seen following cell administration in this study, ruling out liver damage due to cell embolization. There were no other immediate complications directly attributed to stempeucel®. Study by Mohamadnejad et al was prematurely stopped because patients developed complications renal failure and radio-contrast nephropathy, and concluded that injection of CD34 + cells through hepatic artery was probably unsafe (56). However, later studies involving CD34 + cells infusion through hepatic artery did not show such safety issues with administration of CD3 + cells through hepatic (38,(57)(58)(59).
This study was designed primarily for assessing the safety and feasibility of administering stempeucel® through the hepatic artery in ALC. As a secondary objective, we also explored possible e cacy and dose-response. E cacy was seen only in quality of life of patients who received 2.5 million cells/kg dose of stempeucel® compared to control as seen in few mental component scores of SF-36. This may be partially due to open label nature of the study. SF-36 improvement was seen in a study by Salama et al following haematopoietic stem cells therapy in end-stage liver disease patients (60). Quality of life improvement associated with clinical improvement was seen in a study by Kim et al evaluating autologous bone marrow infusion in advanced liver cirrhosis (61). The lack of obvious clinical e cacy seen in this study may be attributed to three reasons: Firstly, the eligibility criteria included patients in the higher severity of liver cirrhosis (Child Pugh class B and C). It is possible that these patients were already in advanced stage of the disease and not amenable for cell therapy. Probably cell therapy has to be attempted at an early stage of disease like alcoholic hepatitis, rather than at a late stage when cirrhosis has already set in. The alcoholic hepatitis stage may help better homing of cells because of the local in ammation. In the advanced stage, it may be di cult for the cells to home to the site of action since there is no active in ammation. Secondly, the starting dose selected for this study (2.5 million cells/kg body weight) might be in the upper end of therapeutic range; higher doses potentially leading to deleterious effects. Lastly, in-spite of strongly conveying the need of alcohol abstinence, some patients might have consumed alcohol during the study, which might have negatively affected the clinical outcome. While most pilot studies involving stem cells in liver cirrhosis had successful outcomes, few studies had negative results. Mohamadnejad, who pioneered the MSC administration in liver cirrhosis with several successful pilot studies, found no bene t of intravenous BMMSC administration compared to placebo in a randomized trial (62). Recently, a double blind study by the same group using BMMNC administered through portal vein in decompensated cirrhosis showed overall no bene t albeit a transient bene t at 3 months (63).
Evidence of e cacy of stem cells requires demonstration of tissue regeneration in addition to proving clinical bene t. Tucker et al recommended a triad of outcome measures for cell therapy trials: demonstration of mechanism of action in terms of cellular response, clinical evidence of improvement and structural bene t (64). This translates to clinical and biochemical improvements in liver cirrhosis, which are easier to demonstrate and structural changes through histopathology of liver tissue, which is a complex procedure. Liver biopsy is challenging especially in cirrhotic patients though several studies have included this procedure. Terai et al has shown that there was improvement in serum albumin, total protein, AFP and proliferating cell nuclear antigen in liver biopsy after 4 weeks of autologous BMMNC infusion therapy (40). Zhang et al showed improvement in ascites, liver function, MELD score in addition to decrease in liver brosis markers (27). Kim et al have demonstrated increasing activation of hepatic progenitor cell compartment, hepatic progenitor cell differentiation, and improvement in Child Pugh scores (61). Interestingly, 80% patients showed increase in liver volume as per Magnetic Resonance Imaging. Jang et al have seen histological improvement in 6 out of 11 patients who received BMSC through the hepatic artery for ALC (23). Enhanced angiogenesis was seen in follow-up liver biopsy specimens after boost infusions of mobilized peripheral blood stem cells in decompensated alcoholic cirrhosis in a study by Yannaki et al (65). In the present study, we did not conduct liver biopsy at 6 months follow-up. It is debated that small tissue sample may not be adequate representation of liver pathology and may be subject to sampling error and intra-observer variation (66,67). Hence, Fibroscan (Transient Elastography) which is a non-invasive technique of assessment of liver stiffness was employed in this study. This method has been validated for diagnosis of liver cirrhosis (68) and was found to be reproducible in patients with chronic liver disease (69). Hence Fibroscan is considered to be an option instead of liver biopsy (70). In this study, there was no change from baseline in Continuous Attenuation Parameter and liver stiffness indicating there is no worsening or improvement in liver brosis. To our knowledge, Fibroscan has not been used in published studies involving cell therapy in liver cirrhosis, though the REALISTIC study protocol evaluating CD133 + cells incorporates this technique (71).
Several approaches have been evaluated for improving e cacy of cell therapy in liver cirrhosis. Animal studies have shown that pretreatment of MSC with injured liver cells has improved the ability of MSC for homing and hepatic differentiation (72). Amer et al differentiated the MSC towards hepatocytes by pretreating them with HGF before infusion via intrahepatic or intrasplenic routes in patients with end-stage liver cell failure due to chronic hepatitis C and found improvement in cell treated group (55). However, El-Ansary et al found no difference in e cacy between MSCs differentiated to hepatocytes and undifferentiated MSCs in hepatitis C virus induced liver cirrhosis (24). Salama et al administered granulocyte-colony stimulating factor (G-CSF) daily for 5 days before administering MSC through the peripheral vein (73). Recently, co-administration of MSC with PPAR gamma agonists has been tried with encouraging results (74). Repeat injection has been tried in cell therapy studies. Jang et al administered autologous BMMSC at baseline, and again after 4 weeks and found that histological improvement was seen in 6 out of 11 patients and Child Pugh score improved in ten patients (23). In a study by Zekri et al, liver cirrhosis patients were randomized to receive one session of autologous haematopoietic stem cells followed by MSC, two sessions of similar treatment separated by 4 months or control (75). It was observed that while one session group showed improvement in serum albumin, bilirubin and INR values till 6 months, two session group sustained improvement till 12 months. Zhang et al have tried UCMSC administration thrice, using peripheral vein and showed clinical improvements and MELD scores (27). REALISTIC study aims to evaluate improvement in disease severity using GCSF alone or G-CSF followed by repeated infusion of CD-133 + cells compared to standard protocol of care alone (71).
The transplantation of MSC showed therapeutic potential for liver function improvement according to recent experimental studies and human studies. Although they remain unclear, the major potential mechanisms have been proposed as a twofold; one is the improvement of the microenvironments through paracrine effects, and the other is the replacement of functional hepatocytes (76).
Dose response relationship has to be established in any drug product development. However, it is still unclear whether classical doseresponse exists with cell therapy. Review of cell therapy studies in heart disease has shown that the dose response was inconsistent and contradictory in terms of dose of administered cells and clinical response, both in preclinical and clinical setting (77). Most clinical trials apply MSCs according to the body weight of patients (n = 9, 0.5-3 × 10(6)/kg as a single dose), while others apply MSCs according to the total quantity of cells (n = 7, 1-20 × 10 (7)) (78). In liver cirrhosis trial using bone marrow cells, Lyra et al noted that there was no correlation of number of cells and clinical improvement (79). However, in a dose ranging study using 5X10(5), 1X10(6) and 2X10(6) cells/kg of CD34 + cells, Nakamura et al have observed improvement in patients who received middle or higher dose (59). It is important to accept that conclusion on dose response cannot be drawn from different trials. Among the recent clinical trials involving applying MSCs to treat liver diseases, the total number of MSCs used was from ~ 10(7) − ~ 10(9), regardless of which method was chosen to deliver MSCs (78).
There are few limitations for this study. First, we have not traced the cells within the body using radioactive technology owing to the inherent complexity of the procedure. Second, we do not have follow-up liver biopsy data. Third, the control group did not receive sham intervention or placebo injection due to ethical reasons. Fourth, the sample size was limited to 20 patients in each dose level, which is insu cient for meaningful detection of e cacy. Lastly, incorporation of a biomarker of liver regeneration like AFP would have provided insights into possible mechanism of action.

Conclusion
The present study has shown that it may be feasible and reasonably safe to administer stempeucel® through intraarterial route (hepatic artery) in ALC at a dose of up to 2.5 million cells/kg body weight. Higher doses of stempeucel® administered through the intraarterial (hepatic artery) route may not be justi ed in ALC owing to the limited e cacy seen at this dose and higher risk of complications at 5 million cells/kg dose group. E cacy was limited to improvement in quality of life in 2.5 million cells/kg dose group. Future studies have to be done to identify the appropriate dose and route of administering cell therapy for ALC. The trial was conducted in accordance with the Declaration of Helsinki. The study was approved by institutional ethics committee and informed consent was taken from all individual participants.

Consent for publication
Written informed consent for publication of patient clinical details and/or clinical images was obtained from the patient/guardian of the patient. A copy of the consent form is available for review by the Editor of this journal Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests