Treating the ischemic heart failure is one of the core targets of new regenerative therapies. However, complexity in ischemic insult has induced various conceptual tools that will inevitably question many of the assumptions of the healing process and the devoted cell or tissue therapeutic strategies . In the recent years, it is urging to better analyze/define the initial status of damaged areas especially in the case of chronic MI. Based on our original small animal SPECT imaging technique, the chronic and transmural MI could be correctly authenticated allowing the central discussion on how the epicardial application of a collagen bio-prosthesis made with MSCs could promote a reverse remodeling process. Outcomes extracted from our multimodal study pointed to a positive impact of MSC-Patch application on the LV reverse remodeling of chronic fibrotic scar including an enhancement of myocardial wall thickness and a reduction of LV dilatation. We also found that these architecture changes were associated with a functional improvement in segmental myocardial perfusion, LV contractility and with tissue reorganization through dense myofibrotic network and angiogenesis.
In the present study, we have adhered to the concept of using collagen in solution as the recipient 3-D structure for seeding stem cells. It is well known now that collagen I is a major component of the cardiac extra-cellular matrix and it provides resistance to deformation . Thus, it porous constitution might facilitate the cell reorganization and the cell exchange of various growth factors and chemokines . Although recent works have suggested that more sophisticated scaffolds forms could be imagined and/or produced with this biomaterial , for now we chose a simple, malleable and easily handled forms dictated by the fact that it allowed to avoid ambiguity due to complex structure and to directly address the question that pertains to the very first efficient repair of chronic MI by MSC-collagen patch. In our implementation, when mixing directly MSCs with Collagen I solution, the cell distribution within collagen scaffold was facilitated as evidenced by their homogenous repartition. Physical consistence of the 3D-articicial tissue was particularly dependent of the initial MSC seeded concentration and we found that a 1x106 cells/150 μL of collagen ratio has allowed obtaining a compact and sewable patch. MSCs have been showed to have a 2-wk survival capacity in 3D-collagen , which is in agreement with our results. However, such a cell culture in liquid Matrigel affected the MSC morphology (Figure 3b) and cell growth. The spheroid MSC appearance observed can be due to the altered cell-cell attachment caused by unstable collagen 3D architecture (see Figure 3) rather than being a consequence of cell death induction . Although growing slowly (doubling time being 7.9 days compared to 2.6 days in conventional 2D cell culture), these MSCs after plating in collagen still preserved their mesenchymal phenotypes and potentiality.
We tested thereafter the relative contribution of the obtained MSC-patch by using several clinical derived imaging and/or hemodynamic assessments so that relevant decision, regarding the relationship between cardiac function and histopathological changes, could be extracted. After engraftment, our original non-invasive co-localization technique has clearly showed the short-term viable patch within the underperfused and infarcted targeted MI (see Figure 2b for example). The double isotopic detection has already used in our group to successfully follow myocardial engraftment of MSCs alone [16, 17]. In this study, this procedure worked well with 3-D tissue and allowed, to our knowledge, the very first in vivo follow-up of engrafted engineered tissue with MSCs thus suggesting the short term survival of these patches within chronic infarcted areas. This is a particularly important issue because the lack of nutrients and oxygen in fibrotic scar might limit direct application of these thick patches. Our in vivo SPECT and immunohistology data have outlined a significant enhancement of myocardial perfusion within the treated areas suggesting that angiogenesis was a preeminent effect in the integration of MSC-patches within chronic MI. Compared with control MI hearts, recipients transplanted with MSC- collagen patch had a higher density of functional microvessels one month after transplantation. We used α-smooth muscle actinin labelling to rather target precapillaries and microvessels networks since it has widely demonstrated that scintigraphic tissue perfusion is strongly correlated with density of functional vessels. It is beyond the scope of this work to identify the specific mechanisms of the short and/or long term inception of angiogenesis, albeit synergic phenomenon resulting from initial surgical inflammatory reaction and angiogenesis due to secretion of cytokines and growth factor (VEGF, FGF.) by MSCs [24, 25] might account for the observed local myocardial perfusion. In a recent clinical study using mononuclear cells to treat chronic myocardial infarction, we have further demonstrated that the cell related angiogenesis could be initiated and amplified by an existence of residual metabolic activity .
Another observation was that epicardial application of MSC collagen patch has yielded a significant augmentation in the LV thickness suggesting a possible “reverse ventricular remodeling” process after MI. According to pathological analysis, a tissue mesenchymalization as witnessed by dense myofibroblastic tissue was observed at the site of patch application. In addition, although no clear presence of neo-cardiomyocytes was evidenced within engrafted areas, the reshape of the MI defect seemed to have positive impact on reducing the LV expansion and dilatation and on cardiac function as witnessed by our hemodynamic assessments in rats treated with patch tissue. There are no clear consensus concerning the contribution of such reverse remodeling after cell or patch application on LV function . Some authors have suggested that regenerative process of cardiac tissue is underlying the improved function. In a porcine model of hibernating and ischemic heart failure -a tissue condition that is different from chronic transmural infarction in our study- Suzuki et al.  have suggested that transplanted MSCs might stimulate proliferation of cardiac progenitors next to the grafted area as well as MSCs differentiation into cardiomyocyte regardless microvascular proliferation. Similarly, Rossini et al.  have demonstrated that bone marrow derived MSCs have a better capacity to migrate and might develop in vivo a sarcomeric component and have cardiac stromal cell morphology. Other works seemed to suggest however that these mechanisms do not occur in sufficiently high frequency to account for the observed functional improvement after MSC administration . There is increasing evidence suggesting that the cardiovascular beneficial effects of stem cell therapy are largely due to the actions of trophic factors and/or paracrine mediators as suggested earlier [24, 25]. Moreover, although beneficial impact of patch graft was either on contraction and relaxation function, the fact that the main effect was observed with LVdP/dt max suggested an improvement in the systolic contractile mechanism. However, as discussed above, we have not clearly noticed an organized myocardial tissue formation in the area of patch engraftment. Two hypotheses might be advanced to explain this apparent discrepancy. First, the observed neoangiogenesis might have a beneficial impact on the border areas of MI where hibernating cardiomyocytes could be better rescued by the enhanced myocardial perfusion. Second, changes in physical and morphological properties of passive myocardium such as shear properties and viscoelastic properties might also have a positive impact on the LV overall improvement including systolic function . Indeed, these latter modifications could facilitate the shortening of sarcomeric fibers, reduce the cardiomyocyte’energy consumption and this especially in the case when the preload is augmented after volume expansion. In this study, the patch treated hearts developed a better response to volume overload test suggesting therefore a better myocardial wall compliance that would explain the enhanced V-P relationship.
Our study had however some limits. First, the harvesting time of MSCs before MI inception had no equivalence in clinical process. Second, we did not address the issues of potential arrhythmogenic effect due to the patches made with MSCs. Indeed, heterogeneous electrical conduction might exist within tissue grafting site especially in the area of border zone and could represent a serious risk. Although we did not observed any suspect death after the patch grafting in this study and noticeable arrhythmogenic episodes during 6-mo follow-up in our recent clinical work where MSCs were injected massively in MI areas, we could not rule out this possibility.