Preliminary Report on the Effect of Mesenchymal Stem Cell Therapy in Patients with Chronic Lung Allograft Dysfunction

Lung transplantation is the best option of care for a variety of end-stage lung conditions no longer responding to standard options of medical or surgical care [1]. The most common causes of early morbidity and mortality after transplant include primary graft dysfunction [2], acute infectious complications [3], acute cellular rejection [4], or acute antibody-mediated rejection [5]. Long-term survival of lung transplant recipients is compromised by chronic lung allograft dysfunction (CLAD) [6], either obstructive (or bronchiolitis obliterans syndrome [BOS]) [7] or restrictive (restrictive allograft syndrome) [8].


Introduction
Lung transplantation is the best option of care for a variety of end-stage lung conditions no longer responding to standard options of medical or surgical care [1]. The most common causes of early morbidity and mortality after transplant include primary graft dysfunction [2], acute infectious complications [3], acute cellular rejection [4], or acute antibody-mediated rejection [5]. Long-term survival of lung transplant recipients is compromised by chronic lung allograft dysfunction (CLAD) [6], either obstructive (or bronchiolitis obliterans syndrome [BOS]) [7] or restrictive (restrictive allograft syndrome) [8]. FEV 1 achieved, not concomitantly related to acute rejection, airway stenosis, or acute infection [7]. This effect is likely predisposed by repeated previous episodes of acute rejection; ischemic and vascular injury; bacterial, fungal, or viral infections; active gastroesophageal re lux; environmental exposures; and others [9]. Although some patients stabilize or improve after enhanced immunosuppression [10,11], addition of azithromycin [12], medical or surgical treatment of gastro esophageal re lux disease [13], or extracorporeal photopheresis, a large portion of these patients will continue to have severe advanced disability or death, unless they receive a repeat transplant [14]. Unfortunately, retransplant has a lower rate of survival than initial transplant [15], and not every patient will qualify to receive a second transplant. Alternatives of care for patients with refractory BOS are limited; therefore, continued exploration of further options of care are justi ied. Mesenchymal stem cells (MSCs) are known to modulate the cellular immune system suppressing effector T cells [16]. They are known to shift immune responses toward anti-in lammatory and tolerogenic phenotypes, shifting from T helper (Th) 1 to Th2 immune response [17]. MSCs also reduce antibody production by B cells [18]. In addition, MSCs has been shown to be naturally and preferentially trapped in the lung [19], providing the cells easy access into lung tissue, particularly if there is a process of active in lammation present [20]. Clinical trials have evaluated autologous and allogeneic MSCs in the treatment of in lammatory conditions, such as graft-versus-host disease [21], and preliminary results have shown potential, bene icial functional effect using MSCs to treat lung transplant recipients with obstructive CLAD in a small group of patients [22,23].
Our group and others have also demonstrated that intravenous administration of bone marrow-derived allogeneic MSC is safe and well tolerated by lung transplant recipients diagnosed with moderate or severe BOS [22,24].
In our phase 1 study, we evaluated the functional outcome in 9 patients diagnosed with moderate BOS, refractory to standard medical therapy during the irst year after receiving MSC therapy. Patients received a single infusion of bone marrow-derived allogeneic MSCs using a dose escalation model. In addition, we measured biomarkers in blood samples of these patients to evaluate changes to immune cell, cytokine, and growth factor levels before and after intravenous infusion of MSCs.

Materials and Methods
This study was approved by the Mayo Clinic Institutional Review Board (Protocol#14-000025; Investigational New Drug#15897). The trial was registered at ClinicalTrials.gov [25].

Patient population
Study patients were recruited from the lung transplant program at Mayo Clinic in Jacksonville, Florida. Patients diagnosed with moderate BOS, refractory to standard medical therapy, who did not qualify for retransplant were offered study participation and enrolled after providing informed consent. Inclusion and exclusion criteria, process of enrollment, manufacturing, transportation, and preparations for infusion of allogeneic bone marrow-derived MSCs have all been reported in a previous publication [24]. All patients received MSC manufactured by Waisman Biomanufacturing at the University of Wisconsin-Madison. Bone marrow was harvested from a young healthy male (18-45 years), blood type O who completed a standardized Donor Health History Questionnaire and had communicable disease testing in accordance with FDA standard 21 CFR part 1271-subpart C Donor Eligibility Final Rule. Allogeneic bone marrow derived MSC were manufactured using standard manual tissue culture lasks, cryopreserved bags were stored and shipped to our facility where they remained frozen until infusion day where the frozen MSCs underwent standard protocol for thawing, 5x dilution and infusion [24]. Patients were enrolled from September 1, 2014, through October 31, 2015, and each patient was followed for up to a year after infusion of MSCs.
Patients were assigned to a dose-escalation model to receive a single infusion of bone marrow-derived MSCs obtained from a single healthy donor as previously described [24]. The irst 3 patients were assigned to receive 1×10 6 MSC/kg (Group 1), the next 3 received 2×10 6 MSC/kg (Group 2), and the inal 3 received 4×10 6 MSC/kg (Group 3).

Biomarkers monitoring
The day of infusion was considered Day 0. Blood samples were collected at baseline (Days: 1 to 7 or 0 prior to infusion) and on Days 1 and 7 after infusion to enumerate immune cells (B cells, natural killer [NK] cells, regulatory T cells [Tregs]) using low cytometry and cytokines and growth factors using multiplex assay. These patients were not having any signs of active infections at the time of the Cytokine assay blood draw.

Flow cytometry
Antibody cocktails were used to detect positive and negative MSC surface markers by following International Society for Cellular Therapy guidelines (positive markers: CD90, CD73, and CD105; negative markers: CD45, CD34, CD14 or CD11b, CD79α or CD19, and HLA-DR) (

Cytokine multiplex assay
Plasma was analyzed for the presence and concentration of 42 different cytokines and growth factors using the Human Cytokine Array/Chemokine Array 42-plex Discovery Assay and performed by Eve Technologies. We used the multiplex cytokine assay to assess proin lammatory (Th1) and tolerogenic (Th2) cytokine levels and proangiogenic factors such as VEGF (vascular endothelial growth factor).

Clinical monitoring
Patient hemoglobin, white blood cell and platelet counts, blood urea nitrogen, serum creatinine, and glucose levels were measured for all patients prior to infusion (Days 1-7) and on Days 30 and 365 after infusion. Arterial blood gases were obtained prior to infusion and repeated on Days 30 and 365 after infusion. Spirometric values (forced vital capacity [FVC] and FEV 1 ) were obtained on Days -365 and -180 prior to infusion, on Day 0 (just prior MSC to infusion), and on Days 180 and 365 after infusion. Spirometric values obtained on Days -365, -180, 180 and 365 were compared to the values obtained on Day 0 just prior infusion of MSCs. All patients had routine post-lung transplant monitoring with scheduled clinic visits every 3 months and when clinically indicated.

Statistical analysis
All variables are summarized as Mean ± SD or median (range). Individual variables were compared over time using paired t test. All results were completed using SAS, version 9.4 (SAS Institute Inc.)

Patient population
Demographic characteristics of the patient population are described on Table 1. All enrolled patients had failed to adequately respond to standard medical therapy and were not candidates for repeat transplant. Nine patients (7 men and 2 women) who were recipients of either double-lung (n=5) or single-lung transplant (n=4) were included. They were of advanced age (69±5 years), likely related to the study design excluding younger patients who could qualify for retransplant. Indications for double-lung transplant were idiopathic pulmonary ibrosis (n=2), chronic obstructive pulmonary disease (n=1), primary ciliary dyskinesia (n=1), and a retransplant due to CLAD (n=1). Indications for single-lung transplant were chronic obstructive pulmonary disease (n=3) and idiopathic pulmonary ibrosis (n=1). All had moderate BOS diagnosed a mean of 4.8±2.2 years after transplant. Their mean FEV 1 prior to MSC infusion was 1.62±0.51 L which was 56.8%±3.2% of the peak FEV 1 achieved posttransplant. This represented a 43.2±3.4% drop in post-transplant FEV 1 consistent with moderate BOS. These patients received their MSC infusion at 6.6±3.1 years after lung transplant.

Fate of Infused MSC
To determine the fate of the infused MSCs, we performed low cytometric analysis of mononuclear cells from collected blood samples. Figure 1 shows the number of MSCs (Mean ± SD) at baseline (before MSC infusion) and Days 1 and 7 after infusion for all 9 patients. There was no statistical difference in the number of circulating MSCs, suggesting that all intravenously infused MSCs were trapped in the lungs because we observed no increase in circulating MSCs by 24 hours after infusion.

Effect of MSC therapy on immune cells
We performed low cytometric analysis and evaluated the immune effector cells (B, NK, and Treg) in the blood samples, comparing them to baseline levels Figure 2.There was considerable variability at baseline. In Group 1 (low dose), the mean percentage Treg levels increased 2.3 fold relative to baseline by Day 7, this pattern appeared different than in the other groups but without reaching statistical difference. Analysis on B cells and NK cells did not disclose a discernible pattern

Effect of MSC therapy on cytokine profi le
Plasma was analyzed for the presence and concentration of 42 secreted cytokines and chemokines. The results of this analysis are summarized in Figure 3. Changes in cytokine levels greater or less than 1 SD from mean baseline levels were considered signi icant. Overall, compared with baseline, more cytokines and growth factors appeared to be affected on patients receiving low doses of MSC (Group 1). Promoters of cell proliferation, like epidermal growth factor, which have antiapoptotic and proangiogenic properties, were signi icantly increased in Group 1. Toleranceinducing Th2 cytokines, such as interleukin (IL)-4, signi icantly increased in Groups 1 and 2. Proin lammatory Th1 cytokines (IL-1α, IL-6, and IL-8) and proin lammatory chemokines (monocyte chemoattractant protein (MCP-1), macrophage in lammatory protein-1α (MIP-1α), and MIP-1β) were signi icantly decreased in Group 1. Overall, the most bene icial biologic effect of MSC therapy appeared to occur in patients receiving the lowest MSC dose compared to the other 2 groups.  Table 2 displays results from blood work obtained on Day 0 prior to MSC infusion, and then on Days 7, 30 and 365 after MSC infusion. Baseline values revealed a population with normal mean hemoglobin (13.2±1.2 g/dL), white blood cell (6.5±2.0×10 9 cells/L), and platelet counts (173±39×10 9 cells/L). Our cohort of 9 patients had a baseline mildto-moderate degree of chronic kidney disease (mean estimated glomerular iltration rate 55.0±14.2 mL/min/1.73m 2 ) and mild elevation of glucose levels (132±29 mg/ dL). There were no signi icant changes observed in hemoglobin, white blood cell, and platelet counts; blood urea nitrogen; creatinine; estimated glomerular iltration rate; or glucose levels when comparing baseline values to those obtained on Day 365 after MSC infusion. Table 3 displays variables derived from arterial blood gases obtained on Day-7 prior to MSC infusion, and then on Days 7, 30 and 365 after infusion. All patients except 1 had arterial blood gases obtained on room air. One patient from Group 1 used 2 L/minute oxygen low via a nasal cannula (estimated FIO 2 =0.28) from baseline and

Important clinical events following MSC infusion
There were no major adverse events during, immediately after, or up to a month after MSC infusion in any patient; these observations have been reported by our group in a previous publication [24]. There were a variety of clinical events observed between Day 30 and Day 365 after MSC infusion, reported in Table 5. Since we do not have a control population, it is impossible to de ine whether or not observed clinical events were related to MSC infusion. Three patients had clinical events following trauma, 2 patients developed skin cancer, and 2 patients developed complications related to atherosclerosis. One patient developed an episode of symptomatic minimal acute rejection with positive transbronchial biopsies 2 months after MSC infusion.   At the time of this episode, she developed increased expression of class II human leukocyte antigen antibodies, with a total panel reactive antibody of 83%; 1 of these antibodies (DQ5) was speci ic to the lung allograft donor. We could not ind donorspeci ic antibodies to the MSC donor in HLA antibody testing. She was treated with pulse corticosteroids and thymoglobulin, with clinical resolution of symptoms. Human leukocyte antigen pro ile repeated 3 months after treatment was negative.

Discussion
The best method to deliver MSCs to sites of injury has been debated, and whether MSCs need to be in direct contact with target cells or if their effects are mediated by altering the tissue microenvironment via secretion of soluble factors, including growth factors and cytokines [26], or by releasing lipid microvesicles is still being studied [27]. Intravenous administration of MSCs is known to produce the irst-pass effect, resulting in cells being trapped by lung tissue [19]. This effect was not a concern in our study because the site of injury was in the lungs; therefore, the intravenous injection of MSCs was expected to deliver the cells to the target tissue. Our intravenously infused MSCs indeed appeared to be trapped in the lungs, as we did not see an increase in circulating MSCs 24 hours after infusion, nor did we observe an increase in circulating MSCs with increasing doses. We did not have the opportunity to measure biochemical markers in bronchoalveolar lavage samples, which could have demonstrated the presence of cells in the alveolar space, or at least, the biochemical impact of their presence. Studies have demonstrated that MSCs respond to chemokine's, such as SDF-1 [28] and MCF-3 [29]. In lammatory cytokines, such as tumor necrosis factor α, interferon, and IL-1α, are needed to activate MSCs [16]. Therefore, the infused cells conceivably were chemotactically attracted to the in lamed lung in the setting of moderate obstructive CLAD.
One MSC characteristic is the ability to suppress mitogen-induced lymphocyte (B, NK, and Treg) proliferation in vitro [30] and in vivo [31], although data is limited on systemic antiproliferative effect of MSCs in humans. We observed inconsistent effects of MSC therapy on levels of circulating B and NK cells, but the number of Tregs appeared to increase, nearly doubling by Day 7 in patients who received low or intermediate MSC doses (Groups 1 and 2).
In our study, we observed a signi icant decrease in proin lammatory cytokines, IL-6 and IL-8, in Group 1, and an increase in tolerogenic cytokine, IL-4, in Groups 1 and 2. In addition, serum analysis of patients in Group 1 showed a signi icant increase in epidermal growth factor levels, which have an important role in MSC-induced wound healing [32] and tissue regeneration [33]. Analysis of biomarkers suggests that lower doses of MSCs appeared to have more favorable biologic effects. Other clinical studies have suggested similar observations. In a clinical study of ischemic heart failure, the lowest dose was associated with the greatest improvement in left ventricular ejection fraction, and these effects were not present at higher doses [34]. A 32-patient clinical trial treating cases of graft-versus-host disease showed no difference in ef icacy or safety among patients receiving 2×10 6 MSC/kg or 8×10 6 MSC/kg [35]. These results suggest that lower MSC doses may be as effective as or better than larger doses, and future studies should de ine whether even lower doses may be as effective, or if there may be speci ic doses better suited for different clinical conditions. Our phase 1 trial was designed mostly to de ine safety and tolerance of administration of MSCs, as reported elsewhere [24]. The extended observations reported here, up to a year after infusion, con irm the absence of anemia, pancytopenia, hyperglycemia, or deterioration in renal function following MSC therapy. The clinical events we observed were not unexpected in an elderly population, many years after lung transplant, subjected to the chronic adverse events related to long-term immunosuppressive therapy. Development of acute rejection and positive donor-speci ic antibodies to lung allograft donor (not to MSC donor) in 1 patient raises the possibility that this event may have been related to cell therapy, and events like these will need to be de ined in larger trials. This particular case was also the only 1 with continued and relentless decline in FEV 1 after MSC infusion.
Patients with BOS having mostly progressive air low obstruction without alveolar disease will have progressive dyspnea on exertion, but not necessarily hypoxemia in early stages. Hypoxemia will eventually appear when BOS becomes severe, with near complete obliteration of airways in larger areas, producing major ventilationto-perfusion mismatch [36]. Our patients were not hypoxic or hypercapnic prior to MSC therapy, nor did they develop any deterioration in oxygenation or ventilation parameters during the year following MSC infusion. This inding corresponds with the stabilization in lung function observed in the pulmonary function tests. This stabilization of lung function for this cohort was achieved in 5 patients, having a mild improvement in lung volumes and lows following therapy. Another 3 continued to decline in function, but at a lesser rate than prior to MSC infusion. Only 1 continued to have relentless loss in lung function, similar to the steep decline in function observed in the year prior to MSC therapy, despite receiving standard therapy for BOS. Our observations match those reported by Chambers et al. [22], where 8 of 10 patients with moderate or severe obstructive, rapidly progressive CLAD showed a slowing in the rate of decline in FEV 1 . Their patients received a much larger dose of MSCs compared to our population (total of 8×10 6 MSC/kg in 4 divided doses of 2×10 6 MSC/kg). Our study included a less diverse population compared to Chambers et al. [22]; all our patients had BOS-2, had a slower rate of decline in FEV 1 and received bone marrow-derived MSCs from a single healthy donor, compared to 5 different donors in the Chambers et al., trial. Despite these differences, both studies showed that, regardless of higher or lower MSC doses, this therapy was safe, well tolerated, and appeared to modify the rate of decline in lung function in some patients who failed standard therapy. Several clinical studies are now reporting the safe administration of cell therapy in patients with an array of lung conditions, such as graft-versus-host disease [37], chronic obstructive pulmonary disease [38] and adult respiratory distress syndrome [39]. To our knowledge, the study from Chambers et al. [22], and our study are the irst to recognize a bene icial therapeutic effect in transplant-related BOS.
Our study has limitations. As a phase 1 trial, the patient population is too small to make de inite conclusions, and they were further subdivided into 3 groups according to dosage. A signi icantly more robust protocol of biochemical markers, not only from blood samples, but also from bronchoalveolar lavage samples, is needed to better de ine the mechanisms by which MSCs exert their effects and the extent of the irst pass effect. Subsequent trials with larger numbers of study patients, including control subjects, should give us a better understanding of possible sort and long-term functional as well as adverse events. Detailed human leukocyte antigen screening panels from MSC donors and monitoring patients regularly after infusion of MSCs are also necessary to better understand any clinical effects derived from the development (if any) of donor-speci ic antibodies.

Conclusion
It is safe and feasible to administer allogeneic bone marrow-derived MSCs to lung transplant recipients with treatment-refractory moderate BOS. Intravenous administration of MSCs may produce a bene icial effect in these patients, by either slowing the rate of functional decline or by actually improving air low in some patients. These preliminary observations justify a larger, randomized, double-blind study, including control subjects, to better de ine if indeed there may be a de ined therapeutic role of MSC therapy in lung transplant recipients with CLAD.