Translational research into novel therapies for pelvic floor disorders
Pelvic floor disorders, including pelvic organ prolapse (POP) and genitourinary syndrome of the menopause (GSM) decrease the quality of life of many women. The mainstay of therapy for POP is a surgical correction, either using native tissues or substitutes, usually referred to as mesh. Surgery may however cause complications, in particular when using mesh. The recommended treatment of GSM is hormonal replacement, yet many patients do not accept this therapy because of the perceived or real risk for cancer. Laser therapy is currently being proposed as an alternative, initially for GSM yet now also for POP and urinary incontinence. The overall aim of this thesis was to investigate novel treatment modalities for POP and GSM. As a translational research group we use different animal models. Rats and rabbits were used for investigation of the host-response to novel degradable electrospun extracellular-matrix-like implants. The latter were chosen because they should improve ingrowth of novel tissue while being resorbed. Sheep were first used for the further characterization of changes in the vagina throughout selected stages of life, in particular menopause. Further ewes were used in two other experiments. One to assess the outcomes of Er:YAG laser treatment of simulated GSM; a second on for evaluation of the host-response to novel electrospun implants for vaginal reconstruction.
To characterize vaginal changes during the life of sheep, we collected specimens during prepubescence, adolescence, after first vaginal delivery, after multiple deliveries and following surgical menopause. Vaginal dimensions increased during adolescence, to reach reproductive levels, with a significant decrease after ovariectomy. Principal biomechanical changes were an increased laxity in the distal vagina one year after first delivery, which later in life was reversed. The thickness of the glycogen containing layer was markedly lower in puberty and after castration, yet not the thickness of epithelium. The lamina propria and muscularis significantly increased in thickness during pre-reproductive age. Semi-quantitative determination of collagen demonstrated lower collagen and higher elastin content after first and multiple deliveries. The changes observed at representative moments during ovine life parallel those observed in women.
Non-ablative Er:YAG laser is frequently used in dermatology and now also in gynecology. Clinical studies document high satisfaction rates, however hard data on the effects at the structural and molecular level are limited. The purpose of non-ablative thermal laser energy is induce a controlled sub-epithelial injury, in turn inducing a tissue response that should alleviate symptoms. We performed a systematic review to summarize objective findings after non-ablative Er:YAG laser (2940nm) on the skin and vaginal wall. Due to the lack of methodological uniformity, no meta-analysis could be performed and therefore results had to be presented as a narrative review. All included studies were prospective yet no randomized controlled trials were identified. We identified three studies reporting ex vivo or in vitro experiments, two studies in rats and 11 clinical studies. Eleven studies were on laser application on the skin (n=11) and only four on the vagina (n=4). The quality of studies was limited, laser settings and read outs used were diverse, i.e. temperature, histology, immunohistochemistry, gene expression, imaging, vaginal pH and maturation index. There were demonstrable effects in all studies. Laser immediately increases superficial temperature, usually partially preserves the epithelium, and induces coagulation of the subepithelial extracellular matrix. Later on, an increase of epithelial thickness, inflammatory response, fibroblast proliferation, an increase of collagen amount and vascularization are documented. In conclusion Er:YAG laser induces changes in the deeper skin or vaginal wall, without unwanted epidermal ablation. Laser energy initiates a process of cell activation, production of extracellular matrix and tissue remodeling. Based on that, it seemed justified to experimentally characterize the effects of Er:YAG laser on the vagina in a representative pre-clinical model. 60 days after ovariectomy, 16 ewes were randomized to undergo either sham or three vaginal Er:YAG laser applications (first application d10, fluence 3 J/cm²; spot size 7 mm²; 4 pulses; 5 passages) at monthly intervals. Primary outcome was vaginal epithelial thickness (at baseline d0 and after laser application d11, d13, d17, d100 of the experiment). Secondary outcomes included markers of atrophy (vaginal health index=VHI), pH, cytology, morphology at the above time points, microcirculation focal depth (FD; d10 and d100), and at sacrifice (d100) vaginal dimensions and active and passive biomechanical testing. Menopausal changes between 60 and 160 days after ovariectomy included progressive decrease in epithelial thickness, in VHI, FD, glycogen, elastin content and vasculature, and an increase in pH and collagen content. In ewes undergoing laser, the first day a few white macroscopic foci were visible and the pH increased, both disappearing within 3 days. Seven days after laser the epithelial thickness increased. At sacrifice d100 after the procedure, there were no differences in vaginal dimensions, morphometry, mitotic and apoptotic activity, active contractility, vaginal compliance, except for a lower blood vessel density in the lamina propria of the midvagina in laser group. Briefly, vaginal Er:YAG laser induced in menopausal ewes a brief increase in epithelial thickness yet no long term changes were demonstrated.
In other preclinical studies we tested a novel electrospun degradable mesh for pelvic floor surgery. In a first study we tested the performance of electrospun polycaprolactone-ureidopyrimidinone (UPy-PCL) mesh ex-vivo and for reinforcement of a primarily sutured defect in the abdominal wall of rats. Outcomes of reconstructions with UPy-PCL were compared with those following repair with light-weight polypropylene and non-injured abdominal wall tissue. Dry UPy-PCL implants were less stiff than polypropylene. In wet conditions both were more compliant. Polypropylene lost some stiffness on cyclic loading. In vivo, both implant types were well incorporated without clinically obvious degradation or herniation. Exposure rates were similar (n=2/12) as well as mesh contraction. The explant conserved natural biomechanical properties at low loads, while, at higher tension, polypropylene explants were much stiffer than UPy-PCL. The latter was initially weaker yet by 42 days it had a compliance similar to native abdominal wall tissue. There were eventually more foreign body giant cells around UPy-PCL fibers, yet the amount of M1 subtype macrophages was higher than in polypropylene explants. There was less neovascularization and collagen deposition in UPy-PCL animals compared to polypropylene. We concluded that abdominal wall reconstruction with electrospun UPy-PCL mesh does not compromise physiologic tissue biomechanical properties, yet provokes a vivid inflammatory reaction.
Thereafter we validated these results in the more biomechanically challenging gap-bridging (overlaid full thickness abdominal wall defect) model in the rat and further in the rabbit over a longer time period again using UPy-PCL and polypropylene as a reference. Animals were sacrificed at 7, 42 and 54 days (rats) and 30 and 90 days (rabbits). Outcomes were again explant compliance and histology, reherniation, mesh degradation and mesh dimensions. High failure rates in the UPy-PCL-group prompted us to provide additional material strength by increasing fiber diameter and mesh thickness, which were not previously tested in the earlier rat studies. Compliance was tested in animals without reherniation. In both rats and rabbits, UPy-PCL-explants were as compliant as native tissue. In rats, yet not in rabbits, polypropylene-explants were stiffer. The contraction rate was similar in UPy-PCL and polypropylene-explants. All UPy-PCL-meshes macroscopically degraded from 30 days onwards leading to reherniation in up to half of the animals. Increased fiber and mesh thickness did not improve outcome. Degradation of UPy-PCL coincided with an abundance of foreign body giant cells until UPy-PCL disappeared. In conclusion, abdominal wall reconstruction with electrospun UPy-PCL meshes failed in 50%. Degradation coincided with a transient vigorous foreign body reaction. Non-failing UPy-PCL-explants were as compliant as native tissue. Despite that, the high failure rate forced us to explore electrospun meshes based on other polymers.
Therefore, we changed the mesh backbone to polycarbonate (PC). Twenty four New-Zealand rabbits were implanted with electrospun PC (UPy-PC) or textile light-weight polypropylene to either reinforce a primary sutured fascial defect (“reinforcement”) or to cover a full thickness abdominal wall defect (“gap bridging”). Rabbits were harvested at 30, 90 and 180 days. Explants were tested for compliance and morphometry. No local complications were observed. In one third of the gap-bridged defects there was mild subclinical herniation. UPy-PC meshes induced a more vigorous foreign body reaction than polypropylene at all time points. The amount of musculofascial tissue tended to be lower in the polypropylene sites, yet this was only significant for connective tissue (fascia) at 30 and 90 days in the reinforcement model and at 180 days in the gap-bridging model. We observed progressively more signs of muscle atrophy and intramuscular fatty infiltration in both types of explants. The infiltration of macrophages across the implant area was significantly higher in UPy-PC explants at 30 days. When used as a reinforcement, at 180 days UPy-PC-explants were stiffer then polypropylene, the latter having the compliance of native tissue. Yet, when the materials were used as a gap-bridging implant, outcomes were comparable and stiffer than native tissue, without obvious herniation.
We then used UPy-PC for reinforcing a posterior vaginal repair in sheep (n=48) compared to polypropylene or to simulated native tissue repair. In this experiment we also used a non-absorbable polyurethane electrospun mesh thanks to an external collaboration. Implants did not compromise vaginal wall contractility or compliance in comparison to native tissue repair. There were no graft related complications. Contraction of durable meshes (polypropylene and polyurethane) was around 30% and comparable. The UPy-PC meshes were progressively degraded. The inflammatory reaction to electrospun meshes was predominated by foreign body giant cells and a M2 response, whereas Polymorphonuclear cells and M1 cells were predominant in polypropylene meshes explants (p < 0.05). Vascularization, innervation and connective tissue were comparable. In conclusion, neither the polypropylene nor the electrospun meshes did cause local complications. Also biomechanical properties were similar. The only measurable difference was in the inflammatory profile. Conceptually three observations emerge: (1) both electrospun implants induced mechanical properties comparable to that of native tissue repair; (2) yet also ultra-light weight polypropylene yielded such outcome; (3) the inflammatory reaction of electrospun implants differs from polypropylene since polymorphonuclear cells and an early M1 response predominates to polypropylene material.
Overall this work led to three main findings. First, throughout life, vaginal changes observed in women can also be observed in sheep. In castrated sheep, we documented progressive changes in several parameters which are also observed in women with GSM. Therefore, we used this model for the first time to document the effects of Er:YAG laser vaginal application. Compared to shams, short term effect were observed, yet not lasting. Third, we tested different resorbable electrospun materials in rats and rabbits. We eventually used resorbable UPy-PC and non-resorbable polyurethane electrospun meshes in the ovine model for vaginal reconstructive surgery, with polypropylene as a reference. With none of these recent implants there were local complications and biomechanical properties were comparable.