Best Paper by a PhD Candidate or Postdoc Award: Macrophage-to-Myofibroblast Trans-Differentiation to Mesh in a Primate Model

Objectives: Mesh deformations (pore collapse and wrinkling) induced by tensioning alter mesh geometry causing variable stress distributions on the vagina – a risk factor for mesh complications. While the overall host response to implanted polypropylene mesh is that of stress shielding, areas of myofibroblast formation and fibrosis can occur in regions of high material burden and increased tissue stress. Macrophages are known to drive downstream outcomes, however, emerging evidence supports myofibroblasts as key players. The relationship between these two cell types, and potential for trans differentiation of macrophages to myofibroblasts (MTM) is not yet understood. We used an established nonhuman primate sacrocolpopexy model to determine the impact of increased tissue stress induced by high material burden on the MTM transition.

Methods: Forty-eight rhesus macaques with transected level I and II support underwent a total hysterectomy and sacrocolpopexy (IACUC 16088646). A square-pored polypropylene mesh was implanted in 3 geometries resulting in progressively increased material burden: stable (R0, stable open pore, n=10), unstable (R45, pores collapsed n=10), and predeformed (RD, pores collapsed, mesh wrinkled, n=10) and placed on tension (10N) or no tension (RDNT). Sham operated animals were controls (n=8). At 12 weeks, mesh-vagina complexes were excised, and the cellular response measured after subtracting mesh area. Macrophages were phenotyped by co-labelling for CD68, CD86 (M1), and CD206 (M2). Myofibroblasts labeled with vimentin and SMA, and CD68 for MTM transitioning cells. Collagen was quantified using trichrome. ANOVA was used for statistical analysis with appropriate post hoc testing.

Results: All animals had similar weight, gravidity, and parity. Mesh burden progressively increased with each deformation and was highest in nontensioned deformed mesh (R0< R45< RD< RDNT, P < 0.0001). Macrophage (CD68+) labelling also increased with increasing deformation/mesh burden and was highest in RDNT (p=0.002, Figure 1). The ratio of classically activated M1 to M2 macrophages was increased relative to Sham (P = 0.0001) with no differences observed between groups. More macrophages expressed a hybrid M1-M2 phenotype (CD68+CD86+CD206+) as compared to the classic phenotypes (p=0.008, Figure 1). There was no significant difference in myofibroblast number between groups (p=0.1), however, the percentage of transitioning CD68+ myofibroblasts increased with increasing mesh burden (P=0.001, R0 to RDNT, p=0.03, Figure 2). Collagen density decreased with increased material burden (R45, p=0.042), indicative of stress shielding and further decreased in absence of tension (RDNT, p=0.0006).

Conclusions: A novel hybrid phenotype of macrophages with both M1 and M2 surface marker expression was the predominant phenotype in the host response to mesh in all configurations followed by M1 classically activated macrophages. Myofibroblasts co-labelling with CD68 increased with increased material burden consistent with MTM transdifferentiation. The concomitant decrease in collagen suggests degradative properties. Longitudinal data and additional mechanistic studies are needed to define impact on mesh outcomes.