3. Therapeutic Use
3.1. Feasibility of Allogeneic Use
Mesenchymal stem cell therapy requires safe and efficacious administration of the cell product. First attempts of therapeutic MSC administration used autologous MSCs, derived from the patients themselves, and autologous MSCs are still used in most clinical trials
[17]. For the first clinical trial of ABCB5
+ MSCs, the cells were derived from small skin biopsies of the patients and expanded ex vivo, before being topically applied onto chronic, standard therapy-refractory venous ulcers (CVUs)
[5]. While the clinical outcome of a median 63% wound size reduction at 12 weeks from baseline, also associated with early pain reduction, substantiated the view that autologous ABCB5
+ MSCs could deliver a clinically relevant wound closure strategy, the autologous approach emerged associated with several drawbacks owing to the labor- and time-consuming manufacturing process, which was carried out for each patient individually. This was associated with comparably high costs and weeks- or monthslong delays until the treatment could be started, and generated only limited numbers of cells, which neither enabled treatment of larger wounds
[5] nor would allow for higher cell number-requiring potential systemic therapy approaches. In addition, candidate patients for MSC therapies are often of advanced age and frequently affected by comorbidities. Unfortunately, age and disease state can negatively affect the number and functionality of MSCs
[18]. The ABCB5
+ MSCs, as outlined above, frequently underlie age-dependent declines in the number of cells present in situ
[6][16] and can accumulate DNA damage and enhanced apoptosis
[14] as well as changes in the differentiation potential
[14][15]. Moreover, diabetes, at least in mice, is associated with reduced numbers of ABCB5
+ MSCs along with profound changes in the dermal stromal niche
[19].
These hurdles associated with the use of autologous MSCs prompted additional research on allogeneic treatment strategies. In the allogeneic setting, ABCB5
+ MSCs can be obtained from young, healthy donors and manufactured as a ready-to-use product of consistent quality
[20]. The generally low immunogenicity of resting MSCs, owing to only low or modest expression of MHC class I molecules and lack of MHC class II and co-stimulatory molecules required for full T-cell activation
[21][22][23], and their immunosuppressing capacities contribute to a lower immune rejection response elicited by transplanted allogeneic MSCs compared to other cell types
[24][25]. Therefore, although previous studies that controlled for MHC haplotype of donors and recipients suggested that MHC-mismatched MSCs may not be immune-privileged
[25][26][27] and that about 11.5% of patients may develop donor-specific antibodies upon administration of allogenous MSCs
[28], the transplantation of allogeneic MSCs typically requires no immunosuppressive treatment of the recipient.
A body of studies that directly compared immune-modulatory and regenerative efficacies of allogeneic vs. autologous MSCs in immunocompetent animals or in humans have confirmed that allogeneic MSCs are not inferior to autologous MSCs
[29][30][31][32][33][34][35]. This applies also to ABCB5
+ MSCs: Comparison of clinical trial results using either autologous, patient-derived or allogeneic, donor-derived ABCB5
+ MSCs to treat chronic, treatment-refractory CVUs reveals that allogeneic treatment achieved a median wound size reduction from baseline at 12 weeks of 76% (full analysis set) and 78% (per-protocol set) with 20% and 22%, respectively, of wounds fully closed
[36], as compared with 63% wound size reduction and 17% full wound closure rate in the autologous setting
[5]. Moreover, in fully MHC-mismatched cardiac allotransplantation models, allogeneic (donor-strain or third party-strain) ABCB5
+ MSCs profoundly prolonged graft survival, whereas treatment with syngeneic (recipient-strain) ABCB5
+ MSCs showed only modest prolongation of graft survival, indicating that prolonged enhancement of cardiac allograft survival by treatment with ABCB5
+ MSCs even depends on MSC-dependent allogeneic tolerogenic stimulation
[1].
3.2. Homing and Engraftment
When tissues are damaged, endogenous MSCs migrate to the site of injury to locally trigger mechanisms that promote regeneration
[37]. The resultant rationale behind therapeutic application of MSCs is that transplanted MSCs would also migrate and home to the damaged tissues
[38] (
Figure 1).
Figure 1. Clinical use and effects of ABCB5+ MSCs. Upon topical or systemic administration, ABCB5+ MSCs migrate and home to injured, inflamed and/or hypoxic tissues to exert anti-inflammatory, pro-angiogenic and trophic responses that facilitate wound healing and enhance skin integrity. The mode-of-action pathways that are utilized for routine potency testing of the MSC-based medicinal product are highlighted in yellow. HIF-1α—hypoxia-inducible factor 1α; HOXA3—homeobox A3; M1—M1 macrophage; M2—M2 macrophage; VCAM-1—vascular adhesion molecule 1. Created with BioRender.com.
Upon local administration in spatial proximity to a lesion, the MSCs can be directed to the target tissue via a chemokine gradient released from the site of injury
[39]. In a chronic wound model characterized by excessive pro-inflammatory macrophage activation
[40], human ABCB5
+ MSCs that had been intradermally injected around the wound edges localized within 2 days in the wound bed to close proximity with endogenous macrophages
[2]. This was shown to be a prerequisite for paracrine suppressive activity of ABCB5
+ MSCs directed at overactivated wound macrophages
[2].
Systemic administration of MSCs offers the advantage that grafted cells can migrate to multiple and/or difficult-to-access sites of tissue injury. Upon injection into the systemic circulation, MSCs must undergo a multi-step process that involves travelling in the vascular system and egressing from the circulation in a lesion’s vicinity followed by migration through the interstitium towards a target site
[38][39]. In these regards, intravenously administered mouse dermal ABCB5
+ MSCs have emerged capable of efficiently homing to mouse skin and thymus
[1]. This tropism pattern is consistent with mouse ABCB5
+ MSC expression
[1] of molecules important for homing to the skin (CCR4, CCR10, E-selectin ligands) and thymus (CCR7, E-selectin ligands)
[41][42][43].
Generally, only small percentages of systemically administered MSCs reach the target tissues, which is considered a major bottleneck in reaching the full therapeutic potential of systemic MSC therapies
[38]. It is therefore noteworthy that the capacity of systemically grafted human ABCB5
+ MSCs to home to NSG mouse skin wounds significantly exceeded that of directly compared BM-MSCs
[44]. The superior skin homing properties of ABCB5
+ MSCs over BM-MSCs might relate to their physiological derivation from the skin, as well as to the higher expression of the homeobox gene
HOXA3, a master coordinator of wound repair through regulating migration of endothelial and epithelial cells
[45][46], and of vascular adhesion molecule 1 (
VCAM-1) by ABCB5
+ MSCs compared to BM-MSCs
[44].
It is still controversial whether therapeutically applied MSCs, beyond migrating to their target sites to exert transient effects, also possess longer-term engraftment potential. Nevertheless, some observations point to a distinct ability of ABCB5
+ MSCs to engraft in host target tissues. Upon topical application of human ABCB5
+ MSCs to full-thickness excisional skin wounds in NSG mice, dose-dependent detection of human CD31 in the wound bed on day 13 indicated that the transplanted cells had persisted and differentiated into CD31
+ endothelial-lineage cells in the recipients’ wounds for at least 13 days
[5]. Furthermore, immunohistochemical detection of ABCB5-expressing cells amongst CVU wound-covering cell debris on day 23 following topical application of autologous ABCB5
+ MSCs suggested persistent engraftment also in human wounds, though thus far only in the autologous setting
[5]. Upon systemic, intravenous injection, engrafted mouse ABCB5
+ MSCs persisted in mouse skin and thymus across fully allogeneic barriers for at least 17 days
[1]. Moreover, intravenously grafted human ABCB5
+ MSCs persisted in NSG mouse full-thickness excisional wounds for at least 14 days, exhibiting superior engraftment capacity compared to intravenously grafted human BM-MSCs
[44]. The latter observation opens up potential options for MSC-based central tolerization strategies of thymic immune cells to alloantigens, e.g., in solid organ transplantation
[1], the induction of splenic regulatory T-cells
[1], e.g., in GvHD, and for the systemic treatment of generalized skin diseases, including severe genodermatoses such as epidermolysis bullosa
[44].
3.3. Product Quality
3.3.1. Homogeneity
The ATMPs based on living cells are inherently prone to heterogeneity because the cells’ gene and expression profiles may vary considerably depending on variations in donor characteristics as well as methods and conditions of cell expansion, cell isolation and product formulation
[47][48][49][50]. In the ABCB5
+ MSC manufacturing process, careful donor selection utilizing rigorous in-/exclusion criteria regarding age and health state, strict definitions of all production steps and in-process controls at every stage of production
[20], and, not at least, the use of a unique cell surface marker (ABCB5) specifically selecting for a reproducibly potent MSC subpopulation
[1][2] are critical tools to minimize potential fluctuations.
Indeed, during process validation, comparative gene expression analyses revealed only rare variations between donors as well as between different passages within donors, which indicates high intra- and inter-donor homogeneity of the expanded cells
[5]. As evaluated in periodic GMP product quality reviews
[20], homogeneity of GMP-compliantly manufactured ABCB5
+ MSCs is reliably achieved by close monitoring and continuous refinement of each process step and mandatory release tests ensuring identity, purity, and biological activity of the produced cells. Specified, validated acceptance criteria guarantee that each released cell batch contains ≥90% ABCB5
+ cells (actually 97.7 ± 1.9%) and ≥90% CD90
+ cells (actually 99.5 ± 0.5%) at ≥90% cell vitality (actually 98.6 ± 0.6%) and ≥90% cell viability (actually 99.5 ± 0.6%) (means ± SD of 66 cell batches derived from seven donors
[20].
3.3.2. Potency
A vital component of the quality assessment of cell therapy products is the evaluation of the cells’ actual biological functionality using appropriate tests that predict their clinical effectiveness
[51][52]. A validated potency assay with predefined acceptance criteria guarantees that the product exerts a certain intended effect at a specific dose and ensures that the treated patient receives a potent therapy product
[53]. Considering that ABCB5
+ MSCs exert their therapeutic effects through multiple pathways induced upon interaction with the host microenvironment, three potency assays have been integrated into the manufacturing process to reflect the most clinically relevant biological modes of action of ABCB5
+ MSCs:
-
Secretion of IL-1RA after coculture with M1-polarized macrophages
[5][20] as a predictive measure of the anti-inflammatory potency in M1 macrophage-dominated inflammatory milieus.
-
Secretion of VEGF under hypoxic culture conditions
[20][54] as a predictive measure of the pro-angiogenic bioactivity in ischemic tissue environments.
-
Tube formation on gel matrix
[20][54] as a predictive measure of the endothelial differentiation capacity.
These three potency assays are an integral component of the release test panel for human skin-derived ABCB5
+ MSCs, and only batches that have met the specified, validated acceptance criteria for each assay are released
[20].
3.4. Safety
3.4.1. Product Safety
Ex vivo expansion entails enhanced cell division rates in an artificial environment lacking the physiological mechanisms of negative selection and clearance of altered cells that are active in the whole organism
[55]. Therefore, in order to facilitate early detection of signals of non-physiological growth behavior during cell expansion that might indicate potential deleterious effects on cell biology including tumorigenic transformation, several in-process controls monitoring cell morphology, contact inhibition, time between passages and cell cycle phase distribution are vital components of the routine ABCB5
+ MSC manufacturing process
[20].
To ensure sterility and purity of the manufactured cell products, specified validated control procedures guard, at each production step, against potential contamination of the cultures, cellular intermediate and final products with infectious agents (including aerobic and anaerobic bacteria and fungi, mycoplasma, and endotoxins) and residual impurities
[20]. Only cultures and batches that meet specified release criteria ensuring product sterility and purity are released for further processing, cryostorage, or delivery
[20].
3.4.2. Preclinical Safety Profile
The preclinical safety profile of GMP-compliantly produced, reconstituted (ready-to-use) ABCB5
+ MSCs has been determined in a Good Laboratory Practice-conforming preclinical in vivo study program following the recommendations of the European Medicines Agency
[51] addressing all relevant aspects of biosafety
[56]. All studies were performed in severely immunocompromised (NOD scid or NSG) mice to prevent rejection of the administered cells in the xenogeneic host. In these mice, subcutaneously injected ABCB5
+ MSCs did not significantly migrate to other tissues and organs, indicating confinement of the cells to the application site following local administration. Systemic (intravenous) infusion at a concentration intended for use in clinical trials (1 × 10
7 cells/mL
[20][57]) was well tolerated without any clinical signs or mortality indicative of clinically relevant pulmonary embolus formation by cells mechanically entrapped in the lungs. Repeated subcutaneous or intravenous injections neither provoked any signs of human cell-related tumor development, ectopic tissue formation or micrometastases, nor did it elicit any signal indicative of ABCB5
+ MSC-related toxicity regarding mortality, clinical signs, body weight development, food consumption, ophthalmological examination, urine analysis, hematology, blood chemistry, blood coagulation, and macro-pathological and histopathological examination. Together, these data demonstrated a favorable biosafety profile of GMP-manufactured ABCB5
+ MSCs in terms of distribution to non-target tissues, toxicity, ectopic tissue formation, or tumor development
[56].
In intramuscular local-tolerance studies, injection of ABCB5
+ MSCs into the thigh muscles of NOG mice resulted in a slight increase in microscopically detectable inflammatory and detectable processes in the muscular tissue when compared to vehicle-treated animals at 1 week, whereas at 4 weeks no differences between cell- and vehicle-injected thigh muscles were detectable
[56].
Several lines of evidence of liver safety of ABCB5
+ MSCs have come from studies investigating the effects of ABCB5
+ MSCs, delivered to the liver via the intrasplenic route to bypass pulmonary entrapment, in an
Pfp/Rag2−/− immunodeficient mouse model of liver regeneration after partial (one third) liver resection
[58][59]. In this model, GMP-manufactured ABCB5
+ MSCs did not augment the increase in serum transaminases that occurred following partial hepatectomy over controls at week 7 after MSC application
[59]. In addition, ABCB5
+ MSC did not impact on physiological lipid accumulation, hepatocyte proliferation rate, physiological zonal distribution of metabolism markers (i.e., periportal expression of E-cadherin, and perivenous expression of glutamine synthetase and cytochrome P450 2E1) during seven weeks of liver regeneration
[59]. Moreover, ABCB5
+ MSCs did not induce toxicity regarding liver fibrosis (as assessed by collagen deposition and expression of
Timp1, the gene encoding tissue inhibitor of metalloproteinases 1), inflammation (as assessed by expression of the pro-inflammatory cytokines Il1B and Il6 in liver tissue) or hepatocellular destruction (as assessed by expression of caspases 3 and 9 in liver tissue) as compared with vehicle-treated animals at week 7 after MSC application
[58].
3.4.3. Safety Data from Clinical Trials
At present, the clinical safety of allogenic ABCB5
+ MSCs has been investigated in three completed phase I/IIa clinical trials
[36][54][57]. Overall, 54 adult patients presenting with chronic wounds received, in total, 92 topical doses of 1–2 × 10
6 ABCB5
+ MSCs per cm
2 wound surface area
[36][54], and 16 patients aged 4–36 years suffering from RDEB received 46 intravenous infusions of 2 × 10
6 ABCB5
+ MSCs per kg body weight in total
[57].
Treatment-related adverse events were reported with 3 of 92 (3.3%) topical and 3 of 46 (6.5%) of intravenous applications. Treatment-related AEs reported from topical treatment were related to the treated wounds and were of mild or moderate severity. None of these events were serious, and all resolved without sequelae
[36]. Treatment-related AEs reported from intravenous treatment were one mild lymphadenopathy and two severe hypersensitivity events, which may have resulted from immunological sensitization following the previous ABCB5
+ MSC infusion. These latter two events were considered serious; however, the affected patients recovered without sequelae shortly after withdrawal of treatment
[57]. Given that the risk of hypersensitivity reactions can likely be reduced in future treatments by monitoring potential induction of anti HLA antibodies and premedication with antihistamines, the Trial Data Monitoring Committee evaluated the potential risk of hypersensitivity reactions as being justified by the anticipated treatment benefits for RDEB patients
[57]. Together, the clinical experience to date indicates a favorable safety profile of the cell product with few mild and/or manageable adverse events.