1. Introduction
Recessive dystrophic epidermolysis bullosa (RDEB) is a rare inherited skin blistering disorder in which total loss or deficiency in functional collagen VII at the dermo-epithelial junction causes excessive skin fragility and progressive multi-organ fibrosis [
1,
2,
3]. Effective systemic curative therapies targeting this underlying genetic defect are not available for routine clinical care so far [
4,
5], pointing to an urgent need for disease-modifying treatments that effectively improve defective wound healing and alleviate severe symptoms such as itch and pain [
6]. In addition to investigation of several small molecule-based drug therapy approaches [
7], the association of RDEB with systemic inflammation beyond skin-limited involvement [
2,
3,
8] has stimulated the development of cell-based therapeutic approaches including allo-transplantation of mesenchymal stromal cells (MSCs) [
9,
10,
11]. Very recently we found that three intravenous infusions of allogeneic skin-derived ABCB5
+ MSCs [
12] to patients with RDEB decreased disease activity, alleviated itch and pain, and facilitated healing of the wounds that were present at baseline [
13,
14].
However, unless a treatment was capable of not only healing existing wounds, but also of restoring deficient collagen VII expression at the dermo-epithelial junction, even minor mechanical forces would continue to provoke new wound development. Consequently, in the absence of causal RDEB cures, there exists an urgent need for maintenance therapies capable of slowing new wound development and/or accelerating new wound healing before the onset of possible wound infection or of wound progression to chronic, more difficult-to-heal stages. Given the pronounced anti-inflammatory and ECM-remodeling effects that have been attributed to ABCB5
+ MSCs in various preclinical and clinical settings [
12], we wondered whether these cells might be a candidate for such a preventative wound reduction maintenance treatment approach. To this end, we conducted an exploratory post-hoc analysis of the above-referenced clinical trial in RDEB patients, specifically focusing on the development and healing kinetics of wounds that were not present at baseline during 12 weeks of systemic treatment with ABCB5
+ MSCs.
4. Discussion
Even though RDEB has been recognized to display features of systemic inflammation leading to progressive multi-organ fibrosis [
2,
3,
8], skin wound closure is consistently identified among the most desired outcomes of disease-modifying treatment [
18,
19]. As recently reported, systemically administered allogeneic ABCB5
+ MSCs have emerged capable of facilitating complete and durable wound closure in patients with RDEB [
14]. However, in view of the complex and highly dynamic RDEB skin wound evolution composed of chronic persistent and recurrent healing/re-opening wounds [
17,
20], an ideal wound closure strategy would not only target already existing wounds, but also induce slowing of the occurrence of newly developing or recurring wounds and/or facilitate and accelerate healing once such wounds have developed, before they enlarge, become infected or become chronic.
An indication that treatment with ABCB5
+ MSCs might indeed have been capable of delaying the occurrence of new (i.e., not present at baseline) wounds is provided by the distribution of the newly occurred wounds over the different time points of observation: After nearly half of these wounds (44 %) had occurred already by day 17, only 28 % of wounds occurred after another 18 days (on day 35), whereas another 28 % of wounds developed only over a comparatively long period of a further 7-week time interval (until week 12) (
Figure 1).
Moreover, when compared to the previously reported healing responses of the baseline wounds to treatment with ABCB5
+ MSCs [
14], the new wounds displayed improved outcomes. Most strikingly, day-17 wounds showed exhibited an approximately two-fold proportion of rapidly healing wounds (56 % within 18 days) compared to baseline wounds (27 % within 17 days), which corresponded to approximately half the median time to first wound closure (18 days versus 35 days for baseline wounds) (
Table 1). In addition, a greater proportion (88 %) of these early-healing new wounds, compared to 74 % of early-healing baseline wounds [
14], remained stably closed over at least 7 weeks (
Figure 1), i.e., they remained closed at least two-fold longer than the typical average time observed for closed wounds to reopen in RDEB (3 weeks) [
17]. For the day-35 wounds, a direct comparison of the wound healing parameters with those of the baseline wounds was impeded by the different lengths of the follow-up periods (see
Table 1), owing to the varying intervals between the trial visits. Nevertheless, while 65 % of the baseline wounds were closed at the end of the 12-week treatment and efficacy phase, 75 % of the day-35 wounds were closed after just over half the time (7 weeks) (
Table 1). This might indicate an earlier treatment response also in the day-35 wounds compared to baseline wounds.
It is important to note that by specifically monitoring the wounds that developed during the treatment period, the present analysis was able to distinguish between the two types of wounds, i.e., chronic persistent and recurrent healing/re-opening wounds, which typically co-exist in RDEB patients [
17,
20]. While the baseline wounds included both types of wounds, including a significant proportion of chronic wounds, the newly developed wounds displayed exclusively the non-chronic, recurrent wound type. The herein observed faster and greater healing successes of new wounds compared to baseline wounds might be explained by the fact that with repeated cell dosing, successive MSCs delivered after the first dosing had homed to new wounds at earlier time points following their first occurrence and before sufficient time had elapsed for such new wounds to significantly increase in size or become chronic.
Additionally, the herein observed delay in the development of new wounds under a regime of repeated ABCB5
+ MSC treatments was noteworthy. Crucial prerequisites for successful local anti-inflammatory and repair-promoting effects by systemically administered cells are efficient recruitment, migration and homing of therapeutically grafted cells in response to chemokine gradients released from sites of injury [
21,
22]. In this regard, for the herein studied skin-derived human ABCB5
+ MSC populations,
in-vivo skin homing and engraftment capabilities have already previously been demonstrated in pre-clinical studies in recipient NSG mice: Systemically grafted human ABCB5
+ MSCs homed to skin wounds and were detectable for at least 14 days, demonstrating a superior engraftment potential compared to side-by-side evaluated bone-marrow derived MSCs [
23]. Moreover, ABCB5+ MSC engraftment into uninjured skin has also been demonstrated in mice: Intravenously infused mouse ABCB5
+ MSCs homed to the skin and survived for at least 17 days against a fully allogeneic barrier (BALB/c ABCB5
+ MSCs grafted into C57/BL6 mice) [
24]. These findings, together with the consideration that the permanent, intrinsic inflammatory environment in RDEB skin is associated with increased expression of genes related to immune system activation [
25], increased neutrophil and CD38
+ (M1) pro-inflammatory macrophage infiltration with high MHC II expression, and defective macrophage phenotype switching [
26], potentially explain why RDEB skin might be prone to preferentially recruit systemically administered ABCB5
+ MSCs to skin sites even before open wound manifestation, and hence our current clinical observation of inhibitory effects on new wound formation by systemically grafted allogeneic ABCB5
+ MSCs.
This possibility is further supported by previous findings that ABCB5
+ MSCs possess potential to ameliorate neutrophil overactivation [
27] and to abrogate M1 macrophage persistence while inducing transition to anti-inflammatory, healing-promoting M2 macrophages [
28]. Of note, this ABCB5
+ MSCs-mediated induction of M2 macrophage phenotype switching has been shown to be associated with a significant reduction of tumor necrosis factor alpha (TNF-α) signaling in skin [
28], while conversely, TNF-α is upregulated and pathophysiologically involved in RDEB [
29,
30,
31]. Therefore, the current observations raise the possibility that systemically grafted ABCB5
+ MSCs, beyond their known wound healing-promoting effects on established wounds [
14,
32,
33], are also recruited to RDEB inflamed skin prior to wound formation and at such sites alleviate inherent M1 macrophage-driven inflammation, with concomitant improvements in extracellular matrix organization and tissue stabilization, and enhancement of damage resistance of RDEB skin [
26] (
Figure 2).
Finally, ABCB5
+ MSCs are also capable of secreting type VII collagen [
23], which raises the possibility that repeatedly grafted allogeneic ABCB5
+ MSCs could additionally contribute to skin integrity improvements also through provision of functional type VII collagen, lack of which is the primary cause of skin fragility in RDEB. In healthy skin, collagen VII trimerizes to form anchoring fibrils that connect the epidermal basement membrane with the dermal extracellular matrix [
34] (
Figure 2). In a collagen VII-hypomorphic mouse model of RDEB [
35], intradermally injected human bone marrow-derived MSCs were capable of depositing collagen VII at the dermal-epidermal junction and enhancing skin resistance to shear forces through de-novo formation of immature anchoring fibrils [
36]. Intravenous MSC administration, as performed in the present trial, would hereby spare patients the discomfort of multiple intradermal injections while at the same time allowing for systemic treatment effects. While it has been postulated that, at least for bone marrow-derived MSCs, intravenous administration might not deliver sufficient numbers of cells to the skin [
36], it is noteworthy that skin-derived ABCB5
+ MSCs home in greater numbers to skin upo0n intravenous infusion and possess superior collagen VII secretion capacity compared to bone marrow-derived MSCs [
23]. Thus, further studies are warranted to determine whether intravenously infused ABCB5
+ MSCs, at currently explored doses, are indeed capable of replacing defective collagen VII in RDEB skin at rates that could potentially confer therapeutic benefit.
Author Contributions
Conceptualization, E.N.-R., K.D., M.H.F. and M.A.K.; methodology, K.D., C.D., S.F. and M.A.K.; formal analysis, E.N.-R.; investigation, K.D., C.D. and S.F.; data curation, K.D., C.D., S.F., M.K. and I.S; writing—original draft preparation, E.N.-R.; writing—review and editing, K.D., S.F., M.K., I.S., C.G., M.H.F. and M.A.K.; visualization, E.N.-R.; supervision, C.G., M.H.F. and M.A.K; project administration: C.D., S.F., M.K. and I.S. All authors have read and agreed to the published version of the manuscript.
Institutional Review Board Statement
The study was conducted in accordance with the Declaration of Helsinki, and approved by the local independent Ethics Committees or Institutional Review Boards: Ethics Committee of the Albert Ludwig University of Freiburg, Freiburg, Germany (167/18 FF-MC, 2 August 2018); Ethics Committee for the State of Salzburg, Salzburg, Austria (415-E/2349/17-2018, 11 October 2018); Comité de Protection des Personnes (CPP), Ile de France 8, Boulogne-Billancourt, France (18 09 49, 7 January 2019); Ethics Committee at the Ospedale Pediatrico Bambino Gesù, Roma, Italy (787, 29 May 2019); North East – York Research Ethics Committee, Newcastle upon Tyne, United Kingdom (18/NE/0240, 17 September 2018); and University of Minnesota Institutional Review Board, Minneapolis, MN, USA (MT2018-14, 17 May 2019).