Stem cells in animal asthma models: a systematic review

Cytotherapy, 2014; 16: 1629e1642

Stem cells in animal asthma models: a systematic review 6,7,8  NADIM SROUR1,2,3,4,5 & BERNARD THEBAUD 1

Universite de Sherbrooke, Faculte de Medecine et des Sciences de la Sante, Department of Medicine, Division of Pulmonology, Sherbrooke, Canada, 2Hôpital Charles-LeMoyne, Department of Medicine, Division of Pulmonology, Montreal, Canada, 3McGill University, Department of Medicine, Montreal, Canada, 4Mount Sinai Hospital Centre, Montreal, Canada, 5The Ottawa Hospital Research Institute, Clinical Epidemiology Program, Ottawa, Canada, 6The Ottawa Hospital Research Institute, Regenerative Medicine Program, Ottawa, Canada, 7Children’s Hospital of Eastern Ontario, Ottawa, Canada, and 8The University of Ottawa, Faculty of Medicine, Ottawa, Canada Abstract Background aims. Asthma control frequently falls short of the goals set in international guidelines. Treatment options for patients with poorly controlled asthma despite inhaled corticosteroids and long-acting b-agonists are limited, and new therapeutic options are needed. Stem cell therapy is promising for a variety of disorders but there has been no human clinical trial of stem cell therapy for asthma. We aimed to systematically review the literature regarding the potential benefits of stem cell therapy in animal models of asthma to determine whether a human trial is warranted. Methods. The MEDLINE and Embase databases were searched for original studies of stem cell therapy in animal asthma models. Results. Nineteen studies were selected. They were found to be heterogeneous in their design. Mesenchymal stromal cells were used before sensitization with an allergen, before challenge with the allergen and after challenge, most frequently with ovalbumin, and mainly in BALB/c mice. Stem cell therapy resulted in a reduction of bronchoalveolar lavage fluid inflammation and eosinophilia as well as Th2 cytokines such as interleukin-4 and interleukin-5. Improvement in histopathology such as peribronchial and perivascular inflammation, epithelial thickness, goblet cell hyperplasia and smooth muscle layer thickening was universal. Several studies showed a reduction in airway hyper-responsiveness. Conclusions. Stem cell therapy decreases eosinophilic and Th2 inflammation and is effective in several phases of the allergic response in animal asthma models. Further study is warranted, up to human clinical trials. Key Words: animal studies, asthma, stem cells, systematic review, therapy

Introduction Asthma is a worldwide problem [1]. In the developed world, 20e30% of people are affected by allergic disorders such as anaphylaxis, hay fever, eczema, and asthma [2]. Furthermore, asthma control frequently falls short of the goals set in international guidelines [1]. Although definitions and estimates vary, approximately 15% of asthmatics may be classified as having severe asthma [3]. Unfortunately, there are few treatment options for patients with poorly controlled asthma already receiving inhaled corticosteroids and long-acting b-agonists. These options include leukotriene receptor antagonists, tiotropium, omalizumab or theophylline, which are either expensive, cumbersome, of modest benefit or marred by potentially serious side effects. New therapeutic options are needed.

There has been much enthusiasm about the therapeutic potential of mesenchymal stromal cells (MSCs) in several clinical disorders such as multiple sclerosis, stroke, myocardial infarction, diabetes, sepsis, hepatic and renal failure, as well as asthma [4,5]. However, to our knowledge, there has been no human clinical trial of MSC therapy for asthma. We therefore aimed to review the literature about the potential benefits of MSC therapy in animal models of asthma. Methods Study selection We sought to include studies of in vivo animal models of asthma, in which the effects of stem cell administration on clinical or biological outcomes relevant to

Correspondence: Nadim Srour, MD, Hôpital Charles-LeMoyne, 3120 boul. Taschereau, Greenfield Park, QC J4V 2H1 Canada. E-mail: nadim.srour@ usherbrooke.ca (Received 9 May 2014; accepted 12 August 2014) http://dx.doi.org/10.1016/j.jcyt.2014.08.008 ISSN 1465-3249 Copyright Ó 2014, International Society for Cellular Therapy. Published by Elsevier Inc. All rights reserved.

1630

N. Srour & B. Thebaud

asthma were compared with the effects of control therapy. We identified studies from two databases: Embase (1996 to 2014 week 24) and MEDLINE (Ovid MEDLINE In-Process & Other Non-Indexed Citations and Ovid MEDLINE, 1996 to June 13, 2014). The search query “(exp Stem cells/ or exp Stromal Cells/ or exp Bone Marrow Cells/ or exp Stem Cell Transplantation/ or exp Bone Marrow Transplantation/) and (exp asthma/ or exp Airway Remodeling/ or exp Bronchial Hyperreactivity/ or exp Bronchoconstriction/ or airway inflammation.ti,ab.)” was run on both databases. We then used Ovid’s deduplication feature to identify unique studies, with higher preference given to the Embase database. Limits were used to identify reviews, editorials and conference abstracts. The remaining abstracts and the full text of selected abstracts were then reviewed for inclusion criteria: (i) an animal model of asthma was used; (ii) there was administration of stem cells or progenitor cells that were not used as a vector for other agents; (iii) the study reported on original data.

“negative” studies (in which the null hypothesis is not refuted) to be less likely to be published than “positive” studies (or to be published faster, in English, etc). For the current systematic review, this is relevant because it might be more interesting for a journal to publish a study in which MSC treatment improved asthma outcomes than a study in which outcomes were not improved. Thus published literature can then overestimate the effect of an intervention or show an effect when in fact there is none. A funnel plot is one technique than can be helpful to detect publication bias. It plots a measure of effect size on the x-axis with a measure of its dispersion on the y-axis. In the absence of publication bias, less precise studies should be scattered symmetrically around more precise studies. This classically leads to a funnel appearance. An asymmetry can indicate publication bias, but there are other possible explanations.

Results Data extraction Data were extracted from the selected studies. In a first step, the following information was recorded: animal model; sensitizing agent, route, dose and time used to sensitize the animal and induce asthma; type of stem cells, dose, route and time of administration; outcomes reported; time of outcome measurement. Outcomes were identified from the Methods section, the Results section, tables and figures and were classified into quantitative, semi-quantitative and qualitative.

The search query returned 1873 entries (Figure 1), 874 of which were not identified as duplicates, reviews, editorials or conference abstracts by use of the Ovid system and were reviewed. Of these, 30 studies were selected for full text review, 19 of which met the inclusion criteria. Two studies were excluded because bone marrowederived mononuclear cells were used rather than stem cells [7,8] and another

Query results: 1873 Duplicates Review Editorial Review/Editorial Conference

Data analysis We had planned for a meta-analysis of the 2 most commonly reported quantitative outcomes, which were bronchoalveolar lavage (BAL) total cell and eosinophil counts. This was not possible, mostly because of heterogeneity in study design. Furthermore, the data were presented graphically, and only a few authors responded to our request for numerical data. The remaining BAL total cell and eosinophil count data were therefore extracted from the published vector or raster graphics. Although a meta-analysis was not performed, the data for these 2 outcomes are presented by use of the ratio of means method [6]. We assessed the risk of bias for these 2 outcomes through the use of funnel plots. For the latter purpose, only 1 experimental group was included for each control group; preference was given to groups that received bone marrowederived cells and syngeneic cells. Graphs were prepared with the use of Review Manager (Version 5.2; Cochrane Collaboration, Oxford, United Kingdom). Publication bias refers to the tendency for

189 408 32 1 369

Abstracts reviewed: 874 4 Duplicates 10 No abstract 1 Chinese 142 Non-original 686 Not of interest Stem cells as vector 1 Full text reviewed: 30 Non-original Not of interest Not stem cells Not in vivo 19 articles selected Figure 1. Study selection.

2 5 3 1

Systematic review: stem cells in animal asthma models because soluble factors from bone marrow cells were used [5]. Study design All studies used mice asthma models (BALB/c mice: 15 studies, C57BL/6 mice: 5 studies, Table 1). The most common allergen used was ovalbumin (15 studies), with other studies that used ragweed [9], toluene diisocyanate [10], house-dust mite [11,12], aspergillus fungal extract [13] and cockroach extract [11]. The studies varied in duration, up to 118 days [14]. Bone marrowederived mesenchymal stromal cells (BM-MSCs) were obtained from murine donors in 14 studies, rats in 1 study [10] and humans in 4 studies [4,15e17]. One of the latter studies also used induced pluripotent stem cells (iPSCs) derived from human fibroblasts [17]. One study used murine iPSCs without c-Myc [18]. One study compared adipose tissue, umbilical cord and BM-MSCs [16]. One other study used adipose tissue MSCs [19] and another used compact bone MSCs [14]. MSC administration occurred before allergen challenge most commonly but also occurred before sensitization [11,17,20,21] or after allergen challenge [12,14,15,18,22] in other studies. MSCs were administered intravenously in all but 3 studies [11,23,24] in which they were administered intratracheally. The two most commonly reported quantitative outcomes were BAL eosinophil count (16 studies, Table II) and BAL total cell count (13 studies, 1 study reporting the differential count but not the total cell count). BAL cytokines were also commonly reported. Histopathology was reported in all but 1 study, sometimes semi-quantitatively (8 studies) or quantitatively (8 studies). Airway hyper-responsiveness was reported in 11 studies that used various techniques. The number of mice included in each experimental group was sometimes reported as a range rather than an exact number.

1631

experiment, MSCs were administered at the beginning of a second challenge at days 76e78, a first challenge having occurred at days 14e16. However, a significant reduction in the BAL total cell count was seen in other studies with repeated challenge such as the Mariñas-Pardo [12] study, in which challenge occurred 3 times weekly for 4 or 6 weeks, and the Ogulur [14] study, in which challenge occurred twice weekly for 12 weeks. The mean cell count ratio in the MSC group compared with the control group is illustrated in Figure 2. Studies with the most precise estimates included some with a treatment effect closer to unity but also some with the largest effect sizes. A funnel plot was used to help detect publication bias. In this case, an example of publication bias would be studies in which MSC treatment led to a smaller or no effect being less likely to be published, skewing the results. The funnel plot for BAL total cell count, which only includes one of several experiments sharing a control group, is not a typical plot but does not suggest obvious publication bias. BAL eosinophil count BAL eosinophil count was reduced by MSC treatment in all but 2 of 16 studies reporting this outcome (P < 0.001 to P  0.05). Although the 95% confidence interval of the ratio of means for the allogeneic treatment group in the Goodwin study includes unity (Figure 3), the treatment effect was reported as significant (P < 0.05). One exception was the Lathrop et al.[13] Th17 inflammation study. The other exception was the experiment in the Mariñas-Pardo [12] study, in which the outcome was assessed 72 h after MSC administration, after 4 weeks of housedust mite instillation. When BAL fluid was obtained 2 weeks after MSC administration, a significant reduction in BAL eosinophils was seen [12]. Again, studies with the most precise estimates include some with a treatment effect closer to unity but also some with the largest effect sizes. The funnel plot appears as expected and does not suggest publication bias.

BAL total cell count

Other biochemical outcomes

All 13 studies reporting on this outcome found that MSC treatment decreased BAL total cell count (P < 0.0001 to P ¼ 0.05), except for 2 experiments. One exception was the group treated before sensitization in the Sun et al. [17] study. However, treatment with MSCs before sensitization did decrease BAL total cell count in the Kavanagh [25] and Goodwin [20] studies. The other exception was the recurrent inflammation group in the Lathrop [13] study [13]. This study used Aspergillus fungal extract to induce Th17 airway inflammation rather than Th2. In one

BAL interleukin (IL)-4 levels were assessed in 11 studies [9,12,13,17e21,23,24,26] and were found to be significantly decreased with MSC treatment in 8 studies (P < 0.05), including with treatment before sensitization [17,20,21], with treatment before allergen challenge [9,17,19,23,26] and with treatment after allergen challenge [18]. No significant effect was seen in the Lathrop [13] study, the Mariñas-Pardo [12] study, one of the Ge studies [24] and in one group of the Sun study. BAL IL-5 levels were assessed in 9 studies [9,13,15e20,26] and were significantly

1632

Table I. Characteristics of selected studies.

Animals

Allergen

Sensitization (days)

Stem cell source

Abreu [26]

6

C57BL/6 mice

Ovalbumin

C57BL/6 BM-MSC

Ahmad [11]

Bonfield [15]

>5 >5 >5 16e24

BALB/c BALB/c BALB/c BALB/c

Ovalbumin Cockroach extract House-dust mite Ovalbumin

Human/murine BM-MSC Human/murine BM-MSC Human/murine BM-MSC Human BM-MSC

0, 2, 4, 6, 8, 10, 12 0, 7, 14 0, 7, 14 0, 7, 14 0

Bonfield [4] Firinci [22]

20e30 10

BALB/c mice BALB/c mice

Ovalbumin Ovalbumin

Human BM-MSC BALB/c BM-MSC

0 0, 14

Ge [24] Ge [23]

10

BALB/c mice BALB/c mice

Ovalbumin Ovalbumin

BALB/c BM-MSC BALB/c BM-MSC

1, 14 0, 7, 14

Goodwin [20]

4e5

Ovalbumin

C57BL/6 BM-MSC

Kavanagh [21]

>15

C57BL/6 and BALB/c mice BALB/c mice

Ovalbumin

Lathrop [13]

16e24

C57BL/6 mice

Lee [10] Mariñas-Pardo [12] Mathias [16]

6 5e8

BALB/c mice BALB/c mice

Aspergillus hyphal extract TDI House-dust mite

FVB/NHanHsd mice BM-MSC C57BL/6 BM-MSC

9e15

BALB/c female mice

Ovalbumin

Nemeth [9]

4e9

C57BL/6 mice

Ragweed

Ogulur [14]

4

BALB/c mice

Ovalbumin

Ou-Yang [25] Park [19] Sun [17]

10 25 4e6

C57BL/6 mice BALB/c female mice BALB/c mice

Ovalbumin Ovalbumin Ovalbumin

Wang [18]

6

BALB/c mice

Ovalbumin

mice mice mice mice

Challenge (days)

MSC injection (days)

Outcome

Timing

40, 43, 46, 47

47

54

C

27 30 29 During 6th week

A A A Aa

14 or 16b 75

32 31 30 During 7th week 18 82 and 89

23 20

27 (PFT), 28 78

C C

0, 7

21e27 27e30 25e29 Every 2 days for 4 weeks starting on day 14 14e18 3x weekly for 8 weeks starting on day 21 24e26 3 weekly for up to 8 weeks starting on day 21 14e16

0 and 7

18

S

0, 7, 14

14, 25e27

7 and 14

28

S

0, 7

14e16  76e78 (recurrent) 14

C

SD rat BM-MSC BALB/c ASC

1e5 None

9e11 5 3 weekly for 4 or 6 weeks 28

19, 81 (recurrent) 13 31 or 42

C A

Human USC/ASC/ BM-MSC C57BL/6J and BALB/c BM-MSC BALB/C compact bone MSC C57BL/6 BM-MSC BALB/c ASC Human iPSC and BM-MSC C57BL/6 iPSCc

0

8e10

5e7

11,12

C

0, 5

14 (IT), 15 (IN)

14

18

C

0, 14, 21

2 weekly for 12 weeks starting on day 26 15e17 21e23 21e27

104

118

A

14 18e20 20/0

18 24 (PFT), 25 29

C C S/C

64

A

1, 8 1, 14 1, 3, 5, 7, 9, 11, 13 0, 14

3 during 4th week  day 35 64

C/Ab A

In the case of discrepancy, the number of animals/group refers to the experiment where BAL cell counts were measured. A, MSC administration after allergen challenge; C, before allergen challenge; S, before allergen sensitization; ASC, adipose tissue MSC; IN, intranasal; IT, intratracheal; IMR90-IPSC and N1-IPSC, induced pluripotent stem cell lines; PFT, pulmonary function tests; TDI, toluene diisocyanate; USC, umbilical cord MSC. a During last week of challenge. b separate experiment, limited data. c Without c-Myc (with Oct-3/4, Sox-2, Klf-4).

N. Srour & B. Thebaud

Animals/ group

Study identifier

Systematic review: stem cells in animal asthma models decreased by MSC treatment in 7 studies [15e20,26] (P < 0.05), including with treatment before sensitization [17,20], before allergen challenge [16,17,19,26] and after allergen challenge [15,18]. No significant effect was seen in 2 studies, including the Lathrop et al. study [9,13]. MSC treatment led to a significant decrease in BAL IL-13 levels in 3 studies [9,15,17] (P < 0.001 to P < 0.05), including with treatment after allergen challenge [15], whereas BAL IL-13 levels were already low in controls in 1 study [25] and there was no significant change in 5 studies [12,13,20,23,24]. One short-term study that reported on IL-9 levels in BAL found a reduction with MSC treatment (P < 0.05) [26]. BAL IL-6 levels were significantly decreased in a syngeneic pre-sensitization model [20], whereas levels were not detectable in the allogeneic model. In the Lathrop study [13], a significant decrease was noted in the recurrent challenge experiment and nonsignificant decrease in the unique challenge experiment. No significant effect was seen in 3 other studies [12,15,19]. MSC treatment before sensitization [21], before allergen challenge [24] and after allergen challenge [18] led to a significant increase in BAL IL10 levels (P < 0.05) in 3 studies, whereas no significant change was noted in 4 studies [9,12,13,23] that used MSC treatment before [9,13,23] and after [12] allergen challenge. The effect of MSC treatment on BAL interferon (IFN)-g was variable with a significant decrease noted in 4 studies [4,15,18,26], including 2 studies with treatment after allergen challenge [15,18] no significant change in 6 studies [9,12,13,17,23,24] and a significant increase in 3 studies [12,17,19]. In the Sun study [17], a significant increase was noted with treatment with BM-MSCs but not with induced progenitor cells. In the Mariñas-Pardo study [12], a significant increased was noted 72 h after MSC administration but not after 2 weeks. Levels were not detectable in the Goodwin study [20]. Serum IFN-g was increased in 2 studies with MSCs administered before or after allergen challenge, but no effect was seen in 2 studies [23,24]. The effect on BAL IL-12 levels was variable [12,13,23,24]. In one study that used ragweed as the allergen [9], MSC treatment before allergen challenge resulted in a significant increase in transforming growth factor (TGF)-b levels (P < 0.05). Production of TGF-b by MSCs in vitro was increased by exposure to BAL fluid from ragweed-sensitized and challenged mice [9]. However, BAL TGF-b was decreased by MSC administration before sensitization [20] and before allergen challenge [19] in 2 other studies. No effect was seen in 1 study with MSC administration before allergen challenge [23].

1633

In the ragweed study, treatment with MSCs resulted in decreased serum immunoglobulin (Ig)E levels (P < 0.05) [9] with either syngeneic or allogeneic MSCs. Treatment with MSCs also resulted in decreased serum IgE levels in another study with administration after allergen challenge (P < 0.05) [15]. However, no significant changes were seen in another study [12]. Five studies reported a decrease in serum allergen-specific IgE levels (P < 0.01 to P < 0.05) [17,18,20,21,23] with MSC treatment, with one study reporting a significant decrease with allogeneic but not syngeneic MSC treatment before sensitization [20]. Serum allergen-specific IgG1 was decreased in some experiments [17,18,20,23] but not others [17,20]. The effect on serum allergenspecific IgG2a was variable [18,20,23]. Treatment with MSCs before allergen challenge was also found to lead to a decrease in lung collagen content in 3 studies in which MSC were administered before allergen challenge (P < 0.05) [10,23,27]. Lung Periodic acideSchiff (PAS) cells were decreased in 6 studies [10,12,14,17,22,23], including MSC administration before sensitization [17] before [10,17,23] and after [12,14,22] allergen challenge, except in the experiment in the Mariñas-Pardo [12] study, in which BAL was sampled 2 weeks after MSC administration. Regulatory T cells (Treg) were assessed in 7 studies [9,13,14,16,21,23,24], in the lung tissue [9,13,14,16,21] pulmonary lymph nodes [23,24] and the spleen [13,21]. They were increased by MSC treatment in all but the Lathrop study [13]. Histopathology Histopathology was not assessed in the Wang and Ahmad studies. It was not described in the Sun study. All other studies reported improvement in lung histopathology in mice treated with MSCs, with the improvement described as great, marked or dramatic. Visually, improvements in airway inflammation [4,9,10,15,25,26], peribronchial and perivascular infiltration [9,12,13,16,19,26], epithelial lining thickening [4,15,22], subepithelial smooth muscle layer thickening [22], basement membrane thickening [22] and goblet cell hyperplasia and mucus production [4,9,15,16,24] were seen. These improvements were seen in studies in which MSCs were administered before sensitization [20,21], before [4,9,10,13,16,19,23,24,26,27] or after [12,14,15,22] allergen challenge. In the Mariñas-Pardo [12] study, the improvement in inflammation seen 72 h after MSC administration rebounded 2 weeks after MSC administration. Ten of the studies [9e11,13,17,19,20,23,24,26] used a semi-quantitative index for lung histopathology. MSC treatment before challenge was found to result in a significant reduction of an index of

Study/experiment

BAL Eo

BAL Lymph BAL Macro BAL Neutro BAL IFN-g BAL IL-4 BAL IL-5 BAL IL-6 BAL IL-10 BAL IL-12 BAL IL-13 BAL TGF-b

Y(0.05) Y(0.03)

Y(0.05) Y(0.05)

Y(<0.05) Y(<0.05)

Y(<0.01) Y(<0.01)

Y(<0.01) Y(<0.05) Y(<0.05) e Y(<0.05) e Y(<0.01) e e e e e Y(<0.05) e Y(0.003) Y(0.005) e e Y(0.01) Y(0.001) Y(0.001) Y(<0.05) Y(<0.001)

Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.01) Y(<0.05) e Y(<0.05)

Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.01) Y(<0.01) Y(<0.01) Y(<0.05)

Y(<0.05) Y(0.01) e Y(<0.05) Y(0.002) Y(0.037)

e e

e e Y(<0.05) Y(<0.01) Y(<0.01) e

e [(0.05)

[(0.05) Y(0.05)

Y(<0.05) [(<0.05) [(<0.05) Y(<0.05) e e Y(<0.05) e e

Y(<0.05) e e e Y(0.05) [(0.001) Y(<0.05)

Y(<0.05) Y(<0.05) e e ND ND e e e [(0.031)

Y(<0.05)

e Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) e e

e

Y(<0.05)

[(<0.05) e Y(<0.05) Y(<0.05) e e

e e

[(<0.05) Y(<0.05)

Y(0.05) e

e e e e ? e e

[(0.025) [(0.047)

e e

ND Y(<0.05) eY Y(0.05) e e

[(<0.05) e e e e

e

Y(<0.05)

Y(0.01)

e e Y(<0.05) Y(<0.05) Y(<0.05) e e

e e Y(<0.05) Y(<0.01) Y(<0.05) e Y(<0.05)

Y(<0.05) [(<0.05) [(<0.05) e e e Y(<0.05)

Y(<0.05) Y(<0.01)

e

Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) e Y(<0.05) Y(<0.05)

Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05)

e

Y(<0.01) Y(<0.001)

e

[(<0.01)

Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) Y(<0.05) [(<0.05)

(continued)

N. Srour & B. Thebaud

Abreu [26] Ahmad/Cockroach [11] Ahmad/house-dust mite [11] Ahmad/Ovalbumin [11] Bonfield [15] Bonfield [4] Firinci [22] Ge [24] Ge [23] Goodwin/allogeneic [20] Goodwin/syngeneic [20] Kavanagh [21] Lathrop [13] Lathrop/recurrent [13] Lee [10] Mariñas-Pardo/2 weeks [12] Mariñas-Pardo/72 h [12] Mathias/ASC [16] Mathias/BM-MSC [16] Mathias/USC [16] Nemeth/allogeneic [9] Nemeth/syngeneic [9] Ogulur [14] Ou-Yang [25] Park [19] Sun/BM-MSC [17] Sun/iPSC-iMR90 [17] Sun/iPSC-N1 [17] Sun/pre [17] Wang [18]

BAL Total

1634

Table II. Quantitative outcomes reported in at least 3 studies.

Table II. Continued

Study/experiment

Serum IFN-g

Serum ovalbuminspecific IgG1

Serum ovalbumin-specific IgG2a

Serum ovalbuminspecific IgE

e e

Lung collagen content

Lung Periodic acideSchiff cells

Treg

Y(0.05)

Abreu [26] Ahmad/cockroach [11] Ahmad/house-dust mite [11] Ahmad/ovalbumin [11] Bonfield [15] [(0.043) Bonfield [4] [(0.043) Firinci [22]

AHR Y(<0.05)a Y(<0.05) Y(<0.05) Y(<0.05)

Y(<0.05) Y(early: 0.010, late: 0.045) Y(<0.01) Y(<0.05)

Y(<0.01) [(<0.05)

Y(<0.01) Y(<0.05)

e

[(<0.05)

e

Y(<0.05)

Y(<0.05)b

[(LN: <0.01) [(LN: <0.05)

Y(<0.05) Y(<0.05) Y(<0.05)

Y(<0.01)

Y(<0.05)c

e

Y(<0.05)

Y(<0.05) e

e

Y(<0.05)

[(LungþSpleen: <0.001) Y(<0.05) e(LungþSpleen) Y(0.01) e(LungþSpleen) Y(0.05)d Y(<0.05) Y(<0.05) e [(Lung: 0.001)

Y(0.01) Y(0.05) Y(0.05)

Y(<0.05) Y(<0.05) Y(<0.001)

[(Lung: <0.001) [(Lung: 0.037) Y(<0.01) Y(<0.05)

Y(<0.05) e e e Y(<0.005)

Y(<0.05)

Y(<0.05) Y(<0.01) Y(<0.05) Y(<0.05) Y(<0.05)

Y(<0.01) Y(<0.01) Y(<0.01) Y(<0.05) Y(<0.005)

1635

-, no significant change, Y, no significant change with a decreasing trend, AHR, airway hyper-responsiveness, ASC, amniotic fluid MSC, LN, lymph node, ND, not detectable, PAS, Periodic acid Schiff, Treg, regulatory T cells, USC, umbilical cord MSC. a Resistive pressure rather than airway hyper-responsiveness. b Periodic acideSchiff positive goblet cell score. c Significance not indicated. d Significant for total tissue resistance, not airway resistance.

Systematic review: stem cells in animal asthma models

Ge [24] Ge [23] Goodwin/allogeneic [20] Goodwin/syngeneic [20] Kavanagh [21] Lathrop [13] Lathrop/recurrent [13] Lee [10] Mariñas-Pardo/2 weeks [12] Mariñas-Pardo/72 h [12] Mathias/ASC [16] Mathias/BM-MSC [16] Mathias/USC [16] Nemeth/allogeneic [9] Nemeth/syngeneic [9] Ogulur [14] Ou-Yang [25] Park [19] Sun/BM-MSC [17] Sun/iPSC-iMR90 [17] Sun/iPSC-N1 [17] Sun/pre [17] Wang [18]

Serum IgE

1636

N. Srour & B. Thebaud

A

B

Ratio of Means, 95% CI

Ge [23] Ge [24] Nemeth Syngeneic Nemeth Allogeneic [9] Sun BM-MSC Sun iMR90-iPSC Sun N1-iPSC Sun presensitization [17] Lee [10] Bonfield [15] Ou-Yang [25] Kavanagh [21] Park [19] Lathrop [13] Mariñas-Pardo 2w [12] Mariñas-Pardo 72h [12] Wang [18] Bonfield [4]

0

0.19 [0.15, 0.24] 0.20 [0.15, 0.27] 0.28 [0.13, 0.60] 0.30 [0.17, 0.54] 0.32 [0.17, 0.61] 0.32 [0.19, 0.54] 0.46 [0.28, 0.75] 0.63 [0.39, 1.02] 0.37 [0.20, 0.68] 0.50 [0.29, 0.87] 0.51 [0.27, 0.95] 0.55 [0.31, 0.97] 0.54 [0.34, 0.87] 0.58 [0.40, 0.85] 0.61 [0.38, 0.99] 0.61 [0.46, 0.82] 0.64 [0.41, 1.02] 0.77 [0.59, 0.99] 0.1 0.2 0.5 Favors stem cell

0.1 SE(log[Ratio of Means])

Study or Subgroup

0.2

0.3

0.4

1

0.5 0.1

2 5 10 Favors control

0.2 0.5 Ratio of Means

1

Figure 2. Effect of stem cell treatment on BAL total cell count. (A) Forest plot. Brackets indicate study arms sharing a control group. (B) Funnel plot. IMR90-IPSC and N1-IPSC, induced pluripotent stem cell lines. Except for the pre-sensitization arm that used BM-MSCs, cells were administered before allergen challenge in the Sun et al. study. Periodic acideSchiff.

peribronchial cuffing, perivascular cuffing, goblet cell hyperplasia and interstitial inflammation [P < 0.001 [9]], an index of peribronchial and perivascular inflammation [P < 0.01 [26] and P < 0.05 [19,24]], an index of peribronchial inflammation [P <0.05 [20] P  0.001 [13] and P < 0.0001 [10]] and an index of overall lung inflammation [P < 0.001 [17] and P < 0.05 [11]].

Six studies assessed histopathology quantitatively. One study showed that treatment with MSCs after challenge resulted in a reduction of basement membrane thickness (P ¼ 0.00 and P ¼ 0.00), subepithelial smooth muscle thickness (P ¼ 0.047 and P ¼ 0.00) and number of mast cells (P ¼ 0.039 and P ¼ 0.002) after 1 and 2 weeks, respectively [22]. Epithelial thickness was significantly reduced 2

A

B Ratio of Means, 95% CI

Lee [10] Sun BM-MSC Sun iMR90-iPSC Sun N1-iPSC Sun presensitization [17] Ge [24] Kavanagh [21] Bonfield [4] Nemeth Allogeneic Nemeth Allogeneic [9] Mathias USC Mathias BM Mathias ASC [16] Bonfield [15] Wang [18] Ge [23] Park [19] Goodwin Allogeneic Goodwin Syngeneic [20] Ou-Yang [25] Mariñas-Pardo 2w [12] Lathrop [13] Mariñas-Pardo 72h [12]

0

0.02 [0.01, 0.05] 0.06 [0.03, 0.12] 0.14 [0.06, 0.30] 0.12 [0.05, 0.27] 0.24 [0.14, 0.43] 0.10 [0.06, 0.15] 0.11 [0.08, 0.16] 0.11 [0.04, 0.33] 0.19 [0.04, 0.79] 0.21 [0.06, 0.69] 0.20 [0.12, 0.34] 0.24 [0.14, 0.41] 0.40 [0.28, 0.57] 0.21 [0.08, 0.56] 0.25 [0.02, 4.00] 0.25 [0.20, 0.33] 0.27 [0.13, 0.56] 0.39 [0.10, 1.54] 0.43 [0.26, 0.72] 0.55 [0.34, 0.89] 0.55 [0.36, 0.84] 0.58 [0.33, 1.00] 0.78 [0.37, 1.66] 0.01 0.1 Favors stem cells

0.5 SE(log[Ratio of Means])

Study or Subgroup

1

1.5

1

10 100 Favors control

2 0.01

0.1 Ratio of Means

1

Figure 3. Effect of stem cell treatment on BAL eosinophil count. (A) Forest plot. Brackets indicate study arms sharing a control group. (B) Funnel plot.

Systematic review: stem cells in animal asthma models weeks after treatment (P ¼ 0.002) but not after 1 week (P ¼ 0.059) [22]. MSC administration after allergen challenge decreased epithelial thickness, basement membrane thickness and smooth muscle thickness (P < 0.05) in another study after 2 weeks [14]. The number of inflammatory nuclei was decreased by MSC administration after allergen challenge in another study (P < 0.001) [15]. MSC administration after allergen challenge resulted in a reduction in airway contractile tissue after 2 weeks (P < 0.05) but not after 72 h in one study, with no effect on extracellular matrix mass [12]. In another study, MSC treatment before allergen challenge resulted in reductions in airway inflammation, hyperplasia of goblet cells and subepithelial fibrosis (P < 0.05 for all) [23]. In another study, MSC treatment before challenge resulted in a decrease in alveolar collapse, neutrophilic infiltration and bronchoconstriction index (P < 0.05 for all) [26]. MSC administration before sensitization resulted in a reduction in airway mucin content (P < 0.05) and subepithelial collagen (P < 0.05) [11]. Airway hyper-responsiveness and body weight Twelve studies assessed airway hyper-responsiveness to methacholine. All found a significant reduction with MSC treatment before sensitization [11,20,21] and before [10,13,16,19,24,25,26] and after [12,18] allergen challenge, with the use of a variety of techniques such as invasive ventilation, the enhanced pause technique, forced oscillation and the airway pressure-time index. In the MariñasPardo study [12], the significant improvement was seen 2 weeks after MSC administration but not after 72 h. In the Lathrop study [13], significant improvement in airway resistance was seen in the experiment in which MSCs were administered before a first challenge on day 14, but, in the experiment in which MSCs were administered before a second challenge on day 76, the improvement was significant for total tissue resistance but not for airway resistance. In a study in which mice were challenged repeatedly for 4 weeks, MSC administration at the end of the challenge period attenuated weight loss (P ¼ 0.05) [15]. Treatment group comparison Although most studies used BM-MSCs, MSCs from other sources such as compact bone [14], adipose tissue [12,16,19] and umbilical cord [16] and iPSCs [17,18] were effective as well. Allogeneic MSCs were used in 2 studies. In the Nemeth [9]study, allogeneic MSCs had generally similar effects compared with

1637

syngeneic MSCs, including BAL total cell and eosinophil count, BAL IL-4 and IL-13, as well as serum IgE levels [9]. In the Goodwin [20] study, allogeneic MSCs also had similar effects compared with syngeneic MSCs, including the effects on lung inflammation, airway hyper-responsiveness, BAL eosinophils and BAL IL-4 and IL-5 [20]. Induced PSCs were used in 2 studies before sensitization [17], before [17] and after allergen challenge [18]. They were effective in reducing BAL eosinophilia and generally in reducing Th2 cytokines. Airway hyper-responsiveness was only assessed in the Wang [18] study, in which MSCs were administered after allergen challenge and were found to be significantly reduced. Adipose tissueederived stem MSCs were used before allergen challenge in the Park [19] study and after allergen challenge in the MariñasPardo [12] study. They were compared with MSCs derived from bone marrow and umbilical cord in the Mathias study [16], after allergen challenge. All these treatment groups were effective in reducing BAL eosinophilia and airway hyper-responsiveness; however, the effects were seen after 2 weeks in the Mariñas-Pardo study but not after 72 h. Paraformaldehyde-fixed MSCs were used in the Kavanagh [21] study, with a loss of effect on histology, BAL total cell and eosinophil count, ovalbuminspecific IgE levels and airway hyper-responsiveness. In the Goodwin study, the use of 1-ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDCI) to prevent the release of soluble mediators from MSCs resulted in loss of some effects such as the improvements in histopathology and BAL eosinophil counts, but other effects such as the reduction in BAL IL-4, IL-5 and IL-6 were preserved. Airway hyper-responsiveness was only significantly different from controls at the lowest methacholine dose with EDCI-treated MSCs. Similarly, in the Lathrop [13] study, some treatment effects such as the reduction in IL3, IL12, keratinocyte chemoattractant, CXCL1 and RANTES BAL levels were preserved despite the use of EDCI to prevent the release of soluble mediators, whereas there remained a nonsignificant trend toward improvement in airway hyper-responsiveness. Murine dermal fibroblasts were used in 2 studies. In the Nemeth [9] study, there was still an effect on BAL total cell count and BAL IL-4 and IL-13, but there was no effect on BAL eosinophils or serum IgE with administration before allergen challenge. There was still a decrease in BAL-IL3. In the Goodwin [20] study, fibroblast administration before sensitization did improve airway hyper-responsiveness and decrease the BAL total cell count but did not decrease the BAL eosinophil count or histologic airway inflammation. In one of the Bonfield studies [15], bone marrowederived macrophage administration

1638

N. Srour & B. Thebaud

after allergen challenge did not improve lung inflammation, BAL total cell count or differential, serum IgE or weight loss [15]. In the Abreu [26] study, bone marrowederived mononuclear cell administration before allergen challenge did improve several histological parameters, lung IL-4 and IL-13. Lung TGF-b and vascular endothelial growth factor as well as lung function were improved more than with MSCs. Discussion This review has identified several studies exploring the effects of MSC administration in animal asthma models. Inflammation and particularly eosinophilic inflammation were reduced by MSC administration in all studies. This included short-term and longer studies, in which MSCs were administered before sensitization with an allergen, before allergen challenge and after allergen challenge. Th2 cytokines in BAL fluid such as IL-4 and IL-5 were decreased by MSC treatment, whereas the effects on IL-13 were not as consistent. There was also evidence of a decrease on serum IgE or allergen-specific IgE. All studies reported improvement in histopathology such as peribronchial and perivascular inflammation, epithelial thickness, goblet cell hyperplasia and smooth-muscle layer thickening. A few studies reported on clinically relevant outcomes and showed a reduction in airway hyper-responsiveness. The study design, that is, the point of MSC administration within the experiments, is a crucial issue. One pitfall of many studies is that MSCs were administered before allergen challenge. We would contend that MSC administration before sensitization, before allergen challenge and after allergen challenge are all relevant to human disease. Regarding MSC administration before allergen sensitization, patients will be sensitized to new agents in their daily life from time to time. Perhaps more importantly, the prevention of a deleterious reaction to future allergen challenges is important in asthma control because most asthmatics are repeatedly exposed to allergens. Given the paroxysmal natural history of asthma, it is just as important to prevent further clinical events (such as asthma exacerbations) and further remodeling (MSC administration before allergen challenge) than to try to reverse established damage (MSC administration after allergen challenge). A hypothetical advantage of MSC therapy is the potential efficacy in all phases of the allergic response, possibly by different mechanisms in each case. In the Mariñas-Pardo study, the improvement in histopathology seen 72 h after MSC administration (which occurred after allergen challenge) bounced back after 2 weeks. However, the decrease in BAL

total cell count was still present, whereas a decrease in BAL eosinophil count, an improvement in airway hyper-responsiveness and a decrease in airway contractile mass was only seen after 2 weeks [12]. Furthermore, in the Ogulur study [14], in which challenge occurred repeatedly until day 104, when MSC were administered, and when outcomes were assessed 2 weeks later, an improvement in histopathology was still seen. The Lathrop study [13] used Aspergillus fungal extract to induce Th17 inflammation. In this study, there was no decrease in BAL eosinophils and Th1 cytokines in BAL such as IL-4, IL-5 or IL-13 but decreased BAL IL-17 levels with decrease in lung inflammation and airway hyper-responsiveness. Although most studies used BM-MSCs, it may feasible to use adipose tissue [12,16,19,27], umbilical cord [16] and compact bone [14] as sources, or even iPSCs [17,18], but corroboration is needed. Some studies have used nonestem cell types [9,15,20,26] or even soluble factors [5], but the evidence at this time is not sufficient to support efficacy of these treatments in all phases of the allergic response. Proposed pathways Multiple pathways have been proposed for the effects of MSCs in asthma. Nemeth et al. [9] have shown that TGF-b production in vitro increased when MSCs were co-cultured with serum from ragweedchallenged mice. This phenomenon was not present when IL-4eneutralizing antibodies were added to the medium or when IL-4R knock-out MSCs were used [9]. Furthermore, the phenomenon was not present when STAT6-deficient MSCs were used, illustrating the dependence of this phenomenon on the IL4-Ra/STAT6 pathway [9]. In vivo, the decrease in BAL total cell counts and eosinophil counts seen with MSC treatment before allergen challenge was eliminated when treating the animals with TGF-beneutralizing antibodies or when TGFb1 knock-out MSC or STAT6-deficient MSCs [9] were used. However, TGF-b levels were decreased by MSC administration before sensitization; this discrepancy may result from the timing of MSC administration and suggests that the response of MSC depend on their micro-environment [20]. TGF-b is thought to be involved in Treg differentiation, and an increase in Treg in MSC-treated mice was observed by Nemeth et al. [9], Kavanagh et al. [21] and Ge et al. [23]. Furthermore, Treg depletion through the use of cyclophosphamide nullified the beneficial effects of MSCs on peribronchial inflammation, mucus production, IgE levels or the shift away from a Th2 profile [21]. However, the effect on BAL eosinophilia persisted, which suggests the

Systematic review: stem cells in animal asthma models existence of a Treg-independent pathway. One study co-cultured iPSCs with peripheral blood mononuclear cells of human subjects with allergic rhinitis [28]. Decreased lymphocyte proliferation and a shift away from Th2 and toward Th1 was observed with a decrease in supernatant IL-4, IL-5 and IL-13 and an increase in IFN-g. An increase in supernatant IL-10 with increased Treg expansion with allergen stimulation was also seen. Prostaglandin E2 might be mediating these effects, because inhibition of prostaglandin E2 production with a Cox-2 inhibitor significantly reversed the immunomodulatory effect on lymphocyte proliferation. The effects of MSCs, in a study in which they were injected before sensitization, might also be partly dependant on IFN-g because the effects on BAL eosinophilia and IL-4, IL-5 and IL-13 were not seen in IFN-g knock-out mice [20]. In that study, MSCs did not inhibit in vivo, antigen-specific T-cell proliferation but rather altered antigen-specific T-cell differentiation toward a Th1 response. In an ovalbumininduced Th1-mediated lung inflammation, MSCs had no effect on the Th1-mediated increase in BAL neutrophils and lymphocytes [20]. Similarly, a decrease in BAL and spleen IL-4 with an increase in IFN-g in the Park study suggests a shift away from a Th2 response. Alveolar macrophages appear to be important in mediating the effects of MSCs. In the Mathias study [16], although there was no increase in macrophage TGF-b that would increase the Treg population and no shift in the M1/M2 macrophage profile, depletion of alveolar macrophages led to a loss of effects on airway hyper-responsiveness and histopathology. MSCs exert their effects through a paracrine effect rather than by lung engraftment [12,23,29]. In fact, Abreu et al. [7,8] have shown that bone marrow mononuclear cell administration before allergen challenge also reduced eosinophilic infiltration, Th2 cytokines, airway remodeling and airway hyperresponsiveness whether injected intravenously or intratracheally. With the use of MSCs treated with a cross-linker to prevent release of soluble factors, Goodwin et al. [20] demonstrated that the effects of MSCs were not due to surface antigens either, but to soluble factors. In fact, Ionescu et al. [5] used medium conditioned by the use of plastic-adherent bone marrow cells differentiated along mesenchymal lineages rather than actual cells in murine ovalbumininduced acute and chronic asthma models. This was administered intranasally after each challenge [5]. In both acute and chronic models, conditioned medium prevented airway hyper-responsiveness and inflammation [5]. In the chronic model, conditioned medium prevented airway smooth-muscle thickening and peribronchial inflammation [5]. Significant reductions

1639

in the Th2 cytokines IL-4 and IL-13 and a significant increase in IL-10 were found along with an increase in IL-10esecreting Treg lymphocytes [5]. Adiponectin was found in greater concentrations in bone marrow celleconditioned medium than in fibroblastconditioned medium, and administration of adiponectin rather than conditioned medium did prevent airway hyper-responsiveness, airway smooth muscle thickening and peribronchial inflammation. Furthermore, the effects of conditioned medium from adiponectin knock-out mice or in which adiponectin was neutralized were blunted [5]. However, treatment with a cross-linker did not entirely eradicate the effects of MSC treatment in the Goodwin [20] and Lathrop studies [13]. Thus, it appears that the observed MSC effects are mostly (but not entirely) mediated through soluble factors rather than by MSCs acting as precursors to mesenchymal structural cells. A possible explanation for the partial effects with cross-linkers is effect mediation through cell contact, as proposed by several authors [13,20,30]. Most compelling are data from the study co-culturing iPSCs with peripheral blood mononuclear cells from human subjects with allergic rhinitis [28]. With transwell separation of iPSCs from the mononuclear cells, the effects on lymphocyte proliferation and Treg expansion were lost [28]. Some data suggest that microsomes are involved either in signaling by lung epithelial cells injured by radiation toward marrow-derived cells [31] or as mediators of MSC effects in a pulmonary hypertension model [32]. Ahmad et al. [11] have used rotenone to induce mitochondrial stress in epithelial cells. They demonstrated and produced a movie of mitochondrial transfer from MSCs to epithelial cells through tunneling nanotubes. They produced MSC underor over-expressing Miro1, a mitochondrial transport protein that was decreased by rotenone in this study. In an ovalbumin mouse model, MSC over-expressing Miro1 were more effective in reducing airway hyper-responsiveness, decreasing lung IL-5 and IL13 and decreasing inflammatory cell infiltration, collagen deposition and mucus hypersecretion. In house-dust mite and cockroach extract models, MSCs overexpressing Miro1 were more effective in reducing airway hyper-responsiveness and airway remodeling, whereas MSCs under-expressing Miro1 were less effective than control MSC. Study limitations It was unfortunate that data were often presented in graphs but not in the numerical form that would facilitate a systematic review or a meta-analysis. Furthermore, the sample size was often unclear and reported as a range rather than an exact number. We

1640

N. Srour & B. Thebaud

attempted to obtain the required data from the authors, but few replied. The significance of this is unclear. Sample size and patient selection and group assignment is of utmost importance in clinical studies. Issues such as these in a clinical study would raise concern about selection bias (for instance, was one subject excluded from an experimental group for a reason that could affect the outcome?), but these animal models involve genetically identical individuals with a fairly unchanging environment. Regardless, inadequate reporting has been described as wasteful [33], and more rigorous reporting and compliance with the ARRIVE guidelines [34] should certainly be encouraged. The studies included in this review were heterogeneous in their design, including the type of MSC used (murine, rat, human, induced pluripotent cells, syngeneic or allogeneic cells), the timing of MSC administration (before sensitization with an allergen, before allergen challenge and after allergen challenge), the duration of the study and the sensitizing agent. For all these reasons, a meta-analysis would not have been valid. However, given that eosinophilic inflammation was clearly decreased by means of MSC administration in almost all studies, the heterogeneity in study design, rather than being detrimental, only serves to strengthen this conclusion. The funnel plot for BAL total cell count is atypical and does not suggest obvious publication bias. More importantly, the funnel plot for BAL eosinophil count is as expected and does not suggest publication bias. In any case, the validity of funnel plot analysis is questionable, given the heterogeneity of the study design. For the same reason, the use of the I2 statistic to assess heterogeneity in effect size is irrelevant. Another limitation is that many studies did not use nonestem cell types as controls. There are studies showing partial responses to fibroblasts [9,20], and one study shows a better response to mononuclear cells than to MSCs [26]. However, the only study that used nonestem cells (macrophages) as controls administered after allergen challenge found that they had no effect [15]. Further data are therefore needed to determine whether other cell types could be used instead of MSCs, during all phases of the allergic response. We note that histology was assessed visually, qualitatively or semi-quantitatively in most studies but were only assessed quantitatively in 6 studies. Blinding of the histology evaluator, as would be expected in clinical studies, would have been ideal. Nevertheless, the conclusion that MSC treatment decreases eosinophilic inflammation is supported by the demonstration in several studies of downregulation of Th2

inflammation and by the plausible mechanisms that have been uncovered. In contrast to embryonic stem cells and iPSCs, MSC treatment does not tend to result in teratoma formation [35]. The Wang [18] study used iPSCs without c-myc; no tumor formation was found 6 months after transplantation. Other concerns may include clumping of the stem cells in the pulmonary vasculature when administered intravenously and the introduction of foreign antigens present in culture media. We note that intratracheal administration is feasible because it was used successfully in a few studies [8,11,23,24]. MSC treatment has been evaluated in human clinical trials for several other conditions including acute myocardial infarction [36], stroke, diabetes and multiple sclerosis [37] and for cartilage repair [38], with no particular safety concerns emerging, including tumors. This includes 85 patients with a follow-up of 5 years [39] and 41 patients with follow-up of 5e137 months (mean, 75 months) [38]. Conclusions In summary, MSCs decrease eosinophilic and Th2 inflammation and are effective in several phases of the allergic response. MSCs derived from sources other than the bone marrow may also be effective. The mechanism may be different, depending on the phase. There is some evidence that soluble factors may be used to reproduce the effects of MSCs during the sensitization and challenge phases, but there are no data on their effectiveness during the postallergen challenge phase. Some data suggest that nonestem cell types may also be useful, but further data are needed. We note that a randomized, controlled trial of MSCs for patients with chronic obstructive pulmonary disease has been completed [40]. There was no evidence of clinical efficacy, although the study was not powered for that, but MSC treatment did result in a reduction of C-reactive protein levels and, most importantly, appeared to be safe. This was the first human trial of stem cells for any lung disease. The observed reduction in inflammation in this trial is relevant to asthma and paves the way for future stem cell trials in lung disease, including asthma. Acknowledgments The authors wish to thank Dr Qing-Ling Fu, Dr Eva Mezey, Dr Daniel Weiss, Dr Meagan Goodwin, Dr Melissa Lathrop and D. Anurag Agrawal for providing the numerical data for their respective studies. This study was not funded.

Systematic review: stem cells in animal asthma models Disclosure of interests: The authors have no commercial, proprietary, or financial interest in the products or companies described in this article.

References [1] Rabe KF, Adachi M, Lai CK, Soriano JB, Vermeire PA, Weiss KB, et al. Worldwide severity and control of asthma in children and adults: the global asthma insights and reality surveys. J Allergy Clin Immunol 2004;114:40e7. [2] Grammatikos AP. The genetic and environmental basis of atopic diseases. Ann Med 2008;40:482e95. [3] Bell MC, Busse WW. Severe asthma: an expanding and mounting clinical challenge. J Allergy Clin Immunol 2013;1: 110e21. quiz 22. [4] Bonfield TL, Nolan MT, Lennon DP, Caplan AI. Defining human mesenchymal stem cell efficacy in vivo. J Inflamm 2010:7. http://dx.doi.org/10.1186/1476-9255-7-51. [5] Ionescu LI, Alphonse RS, Arizmendi N, Morgan B, Abel M, Eaton F, et al. Airway delivery of soluble factors from plasticadherent bone marrow cells prevents murine asthma. Am J Respir Cell Mol Biol 2012;46:207e16. [6] Friedrich JO, Adhikari NK, Beyene J. Ratio of means for analyzing continuous outcomes in meta-analysis performed as well as mean difference methods. J Clin Epidemiol 2011; 64:556e64. [7] Abreu SC, Antunes MA, Maron-Gutierrez T, Cruz FF, Carmo LG, Ornellas DS, et al. Effects of bone marrowderived mononuclear cells on airway and lung parenchyma remodeling in a murine model of chronic allergic inflammation. Respir Physiol Neurobiol 2011;175:153e63. [8] Abreu SC, Antunes MA, Maron-Gutierrez T, Cruz FF, Ornellas DS, Silva AL, et al. Bone marrow mononuclear cell therapy in experimental allergic asthma: Intratracheal versus intravenous administration. Respir Physiol Neurobiol 2013; 185:615e24. [9] Nemeth K, Keane-Myers A, Brown JM, Metcalfe DD, Gorham JD, Bundoc VG, et al. Bone marrow stromal cells use TGF-beta to suppress allergic responses in a mouse model of ragweed-induced asthma. Proc Natl Acad Sci U S A 2010;107:5652e7. [10] Lee SH, Jang AS, Kwon JH, Park SK, Won JH, Park CS. Mesenchymal stem cell transfer suppresses airway remodeling in a toluene diisocyanate-induced murine asthma model. Allergy Asthma Immunol Res 2011;3:205e11. [11] Ahmad T, Mukherjee S, Pattnaik B, Kumar M, Singh S, Rehman R, et al. Miro1 regulates intercellular mitochondrial transport and enhances mesenchymal stem cell rescue efficacy. EMBO J 2014;33:994e1010. [12] Marinas-Pardo L, Mirones I, Amor-Carro O, Fraga-Iriso R, Lema-Costa B, Cubillo I, et al. Mesenchymal stem cells regulate airway contractile tissue remodeling in murine experimental asthma. Allergy 2014;69:730e40. [13] Lathrop MJ, Brooks EM, Bonenfant NR, Sokocevic D, Borg ZD, Goodwin M, et al. Mesenchymal stromal cells mediate aspergillus hyphal extract-induced allergic airway inflammation by inhibition of the th17 signaling pathway. Stem Cells Translat Med 2014;3:194e205. [14] Ogulur I, Gurhan G, Aksoy A, Duruksu G, Inci C, Filinte D, et al. Suppressive effect of compact bone-derived mesenchymal stem cells on chronic airway remodeling in murine model of asthma. Int Immunopharmacol 2014;20:101e9. [15] Bonfield TL, Koloze M, Lennon DP, Zuchowski B, Yang SE, Caplan AI. Human mesenchymal stem cells suppress chronic airway inflammation in the murine ovalbumin

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

1641

asthma model. Am J Physiol Lung Cell Mol Physiol 2010; 299:L760e70. Mathias LJ, Khong SM, Spyroglou L, Payne NL, Siatskas C, Thorburn AN, et al. Alveolar macrophages are critical for the inhibition of allergic asthma by mesenchymal stromal cells. J Immunol 2013;191:5914e24. Sun YQ, Deng MX, He J, Zeng QX, Wen W, Wong DS, et al. Human pluripotent stem cell-derived mesenchymal stem cells prevent allergic airway inflammation in mice. Stem Cells 2012;30:2692e9. Wang CY, Chiou GY, Chien Y, Wu CC, Wu TC, Lo WT, et al. Induced pluripotent stem cells without c-myc reduce airway responsiveness and allergic reaction in sensitized mice. Transplantation 2013;96:958e65. Park HK, Cho KS, Park HY, Shin DH, Kim YK, Jung JS, et al. Adipose-derived stromal cells inhibit allergic airway inflammation in mice. Stem Cells Dev 2010;19:1811e8. Goodwin M, Sueblinvong V, Eisenhauer P, Ziats NP, LeClair L, Poynter ME, et al. Bone marrow-derived mesenchymal stromal cells inhibit Th2-mediated allergic airways inflammation in mice. Stem Cells 2011;29:1137e48. Kavanagh H, Mahon BP. Allogeneic mesenchymal stem cells prevent allergic airway inflammation by inducing murine regulatory T cells. Allergy Eur J Allergy Clin Immunol 2011; 66:523e31. Firinci F, Karaman M, Baran Y, Bagriyanik A, Ayyildiz ZA, Kiray M, et al. Mesenchymal stem cells ameliorate the histopathological changes in a murine model of chronic asthma. Int Immunopharmacol 2011;11:1120e6. Ge X, Bai C, Yang J, Lou G, Li Q, Chen R. Effect of mesenchymal stem cells on inhibiting airway remodeling and airway inflammation in chronic asthma. J Cell Biochem 2013;114:1595e605. Ge X, Bai C, Yang J, Lou G, Li Q, Chen R. Intratracheal transplantation of bone marrow-derived mesenchymal stem cells reduced airway inflammation and up-regulated CD4þCD25þ regulatory T cells in asthmatic mouse. Cell Biol Int 2013;37:675e86. Ou-Yang HF, Huang Y, Hu XB, Wu CG. Suppression of allergic airway inflammation in a mouse model of asthma by exogenous mesenchymal stem cells. Exp Biol Med 2011;236: 1461e7. Abreu SC, Antunes MA, de Castro JC, de Oliveira MV, Bandeira E, Ornellas DS, et al. Bone marrow-derived mononuclear cells vs. mesenchymal stromal cells in experimental allergic asthma. Respir Physiol Neurobiol 2013;187: 190e8. Cho KS, Park HK, Park HY, Jung JS, Jeon SG, Kim YK, et al. IFATS collection: Immunomodulatory effects of adipose tissue-derived stem cells in an allergic rhinitis mouse model. Stem Cells 2009;27:259e65. Fu QL, Chow YY, Sun SJ, Zeng QX, Li HB, Shi JB, et al. Mesenchymal stem cells derived from human induced pluripotent stem cells modulate T-cell phenotypes in allergic rhinitis. Allergy 2012;67:1215e22. Lee JW, Fang X, Krasnodembskaya A, Howard JP, Matthay MA. Concise review: Mesenchymal stem cells for acute lung injury: role of paracrine soluble factors. Stem Cells 2011;29:913e9. Nasef A, Chapel A, Mazurier C, Bouchet S, Lopez M, Mathieu N, et al. Identification of IL-10 and TGF-beta transcripts involved in the inhibition of T-lymphocyte proliferation during cell contact with human mesenchymal stem cells. Gene Express 2007;13:217e26. Aliotta JM, Sanchez-Guijo FM, Dooner GJ, Johnson KW, Dooner MS, Greer KA, et al. Alteration of marrow cell gene expression, protein production, and engraftment into lung by

1642

[32]

[33]

[34]

[35]

[36]

N. Srour & B. Thebaud

lung-derived microvesicles: a novel mechanism for phenotype modulation. Stem Cells 2007;25:2245e56. Lee C, Mitsialis SA, Aslam M, Vitali SH, Vergadi E, Konstantinou G, et al. Exosomes mediate the cytoprotective action of mesenchymal stromal cells on hypoxia-induced pulmonary hypertension. Circulation 2012;126:2601e11. Glasziou P, Altman DG, Bossuyt P, Boutron I, Clarke M, Julious S, et al. Reducing waste from incomplete or unusable reports of biomedical research. Lancet 2014;383:267e76. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol 2010;8:e1000412. Knoepfler PS. Deconstructing stem cell tumorigenicity: a roadmap to safe regenerative medicine. Stem Cells 2009;27: 1050e6. Hare JM, Traverse JH, Henry TD, Dib N, Strumpf RK, Schulman SP, et al. A randomized, double-blind, placebocontrolled, dose-escalation study of intravenous adult

[37] [38]

[39]

[40]

human mesenchymal stem cells (prochymal) after acute myocardial infarction. J Am Coll Cardiol 2009;54: 2277e86. Sng J, Lufkin T. Emerging stem cell therapies: treatment, safety, and biology. Stem Cells Int 2012;2012:521343. Wakitani S, Okabe T, Horibe S, Mitsuoka T, Saito M, Koyama T, et al. Safety of autologous bone marrowderived mesenchymal stem cell transplantation for cartilage repair in 41 patients with 45 joints followed for up to 11 years and 5 months. J Tissue Eng Regen Med 2011;5: 146e50. Lee JS, Hong JM, Moon GJ, Lee PH, Ahn YH, Bang OY, et al. A long-term follow-up study of intravenous autologous mesenchymal stem cell transplantation in patients with ischemic stroke. Stem Cells 2010;28:1099e106. Weiss DJ, Casaburi R, Flannery R, LeRoux-Williams M, Tashkin DP. A placebo-controlled, randomized trial of mesenchymal stem cells in COPD. Chest 2013;143: 1590e8.