Objective
This study was undertaken to test injectable surgical sealants that are biocompatible with fetal membranes and that are to be used eventually for the closure of iatrogenic membrane defects.
Study Design
Dermabond (Ethicon Inc, Norderstedt, Germany), Histoacryl (B. Braun GmbH, Tuttlingen, Germany), and Tissucol (Baxter AG, Volketwil, Switzerland) fibrin glue, and 3 types of in situ forming poly(ethylene glycol)-based polymer hydrogels were tested for acute toxicity on direct contact with fetal membranes for 24 hours. For the determination of elution toxicity, extracts of sealants were incubated on amnion cell cultures for 72 hours. Bonding and toxicity was assessed through morphologic and/or biochemical analysis.
Results
Extracts of all adhesives were nontoxic for cultured cells. However, only Tissucol and 1 type of poly(ethylene glycol)-based hydrogel, which is a mussel-mimetic tissue adhesive, showed efficient, nondisruptive, nontoxic bonding to fetal membranes. Mussel-mimetic tissue adhesive that was applied over membrane defects that were created with a 3.5-mm trocar accomplished leak-proof closure that withstood membrane stretch in an in vitro model.
Conclusion
A synthetic hydrogel-type tissue adhesive that merits further evaluation in vivo emerged as a potential sealing modality for iatrogenic membrane defects.
Invasive diagnostic and therapeutic interventions with large diameter fetoscopes frequently are complicated by amniotic fluid leakage, separation of amnion and chorion, or even frank iatrogenic preterm premature rupture of the fetal membranes (iPPROM). For fetoscopic procedures, rates of iPPROM range from 6-45% ; however, in a trial of fetal endoscopic tracheal occlusion for severe congenital diaphragmatic hernia, a 100% rate was reported. Because these procedures usually are performed in the second trimester of pregnancy, iPPROM usually occurs at an early gestational age. Hence, the associated morbidity and death may compromise the expected benefits of the intervention. IPPROM is therefore a potentially serious complication for prenatal fetal surgery. Clinically, measures of plugging membranes after established rupture and of preventive plugging of fetoscopic access sites have been undertaken, as reviewed previously. For closure after obvious iatrogenic rupture, intraamniotic injection at the puncture site of maternal platelets mixed with fibrin cryoprecipitate (amniopatch) has evolved as a promising route to seal. However, the sudden activation of a large number of platelets in the amniopatch has accounted for otherwise unexplained fetal death in some cases. But increasing efforts have concentrated on taking prophylactic measures before rupture rather than therapy after established or symptomatic rupture of the membranes.
Several preventive plugging methods that use dry collagen and gelatin plugs or liquid blood-derived sealants already have been investigated clinically. Preliminary experience supports this prophylactic intervention for the prevention of iPPROM. A 2006 report on a 27-patient cohort found a 4.2% rate of postoperative PPROM on gelatin plug (Gelfoam; Pfizer, New York, NY) insertion on port retrieval in endoscopic fetal surgery. In another small clinical study, sequential injection of platelets, fibrin glue, and powdered collagen slurried directly to the puncture site and successfully prevented amniotic fluid loss after endoscopic procedure. Still, the positive outcome with these methods await to be reproduced in other centers.
Of note, collagen fleece plugs (Lyostypt; B. Braun GmbH, Tuttlingen, Germany) are now used routinely for prophylactic plugging of iatrogenic membrane defects after fetoscopic endoluminal tracheal occlusion for in utero therapy of congenital diaphragmatic hernia in 1 center in Leuven. Other prophylactic plugging techniques (such as scaffold-type plugs that are manufactured directly from decellularized amnion tissue) have been evaluated so far only in animal models. Further, laser welding, preemptive placement of synthetic surgical sealants before fetoscopic access, direct injection into amniotic fluid of fibrinogen/thrombin-based tissue sealant, and sealing with platelet-rich plasma were evaluated in laboratory settings.
An emerging notion is that spontaneous healing appears to be slow, if not absent, in human fetal membranes. Histologic follow-up evaluation of fetoscopic puncture defects in membranes of human patients several months after the procedure showed that the defects did not close by growth of new tissue. The amnion layer contains few cells and does not contain blood vessels, which makes the healing response in this layer unlikely.
In rabbits, trials of prophylactic plugging of membrane defects with decellularized amnion scaffolds showed effective sealing without detectable signs of biologic repair after 1 week, which is the maximum achievable in this experimental model. Recent studies in the mid gestational rabbit observed signs of early healing of membrane defects on the addition of platelets or amniotic fluid cells to collagen plugs; it is unclear whether this effect could assume relevant degrees of healing long term. The criteria for a prophylactic plug material may be to present an immediate, nontoxic, and ideally durable physical barrier to amniotic fluid and not necessarily the induction of biologic healing. With this strategy in mind, we examined 5 liquid synthetic sealants (namely 2 types of cyanoacrylate glues and 3 polyethylene glycol(PEG)–based hydrogel-type polymers) for their principal aptitude for fetal membrane repair. Repair of defective tissue in moist/wet conditions or even underwater presents a particular challenge. In the present study, we addressed adhesion to moist, intact fetal membranes. Alkyl-cyanoacrylate glues were chosen on the basis of their well-known strong bonding to tissue and their use as tissue adhesives in surgical and traumatic wound repair. Our choice of 3 synthetic PEG-based hydrogel sealants (photopolymerizable gel, mussel-mimetic adhesive, and commercial SprayGel [Confluent Surgical, Inc, Waltham, MA]) was based on data that showed their interfacial bonding to various tissues and the possibility to deliver them in minimally invasive liquid form for gelation in situ. Two types of PEG-based hydrogels under present study, SprayGel and photopolymerized PEG, were already used clinically. SprayGel has been used clinically as bioabsorbable antiadhesion barrier in patients who undergo myomectomy. A clinically approved formulation of photopolymerized PEG hydrogel sealant, FocalSeal-L (Focal Inc, Lexington, MA) sealant, proved successful for closure of pulmonary air leaks in the lung that occur at cardiac operations. For the experimental mussel-mimetic adhesive hydrogel formulation of the present study, no clinical data exist yet. Here we estimated applicability of these synthetic polymers as sealants on fetal membranes based on their bonding to fetal membranes and toxicity in vitro, with the use of the biosurgical Tissucol fibrin glue sealant (Baxter AG, Volketwil, Switzerland) as internal reference.
Materials and Methods
Membrane collection and amnion cell isolation
The Ethical Committee of the District of Zurich approved the protocol (study Stv22/2006). A total of 15 fetal membranes were collected with written patient consent from elective cesarean section deliveries. Mean gestational age was 38 ± 1 weeks in the absence of labor, PROM, chorioamnionitis, or chromosomal abnormalities. Fetal membrane pieces of 150-200 cm 2 were collected. The fetal membranes were cut approximately 2 cm from the placental disc to avoid the “zone of altered morphology” overlying the cervix that is considered to be a naturally predefined breaking site of the membranes. Human amnion epithelial cells (hAECs) and amnion mesenchymal cells (hAMCs) were isolated and cultured, as described previously.
Sealants
Alkyl-cyanoacrylate glue sealants
Dermabond (Ethicon Inc, Norderstedt, Germany) and Histoacryl (B. Braun GmbH, Tuttlingen, Germany) are 2-octyl cyanoacrylate monomer and n-butyl-2-cyanoacrylate monomer, respectively. The formulations possess syrup-like viscosity. These glues act through anionic polymerization of hydroxyl groups from the minute amounts of moisture that normally are present on actual surfaces that are glued, including biologic surfaces. Indeed, cyanoacrylate glues are known to be extremely adhesive to tissue. Water acts as a catalyst to accelerate this polymerization. The polymerization occurs within minutes after application to tissue. The resulting resin is water resistant. Both Dermabond and Histoacryl are marketed as topical skin adhesives to hold skin edges of wounds from surgical incisions. As specified by the manufacturer of Dermabond, it is not for application on wet wounds.
Hydrogel sealants
SprayGel is a sprayable antiadhesion barrier polymer that consists of 2 synthetic liquid precursors that, when mixed together, rapidly cross-link to form a solid absorbable hydrogel in situ. The first precursor is a modified PEG with terminal electrophilic esters groups; the other precursor solution contains PEG that has nucleophilic amine groups. SprayGel is marketed outside the United States for use in abdominal and pelvic surgical procedures and has been clinically tried to reduce adhesion formation after ovarian surgery. SprayGel was deposited on the fetal membranes through the air pump–assisted SprayGel Laparoscopic Sprayer. The gel is formulated to remain adherent on the site of application for approximately 5 days, after which time it is absorbed by way of gradual hydrolysis.
Photopolymerized PEG hydrogel sealant (pPEG) was formed by in situ interfacial photopolymerization of PEG diacrylate precursor of average molecular weight 700 d (Sigma Chemical Company, St. Louis, MO), according to a previously described gelation protocol. Fetal membranes were flushed with a tissue adsorbing photoinitiator eosin Y (1 mmol/L in 10 mmol/L 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, pH 7.4, 0.15 mol/L sodium chloride; HEPES-buffered saline solution. Then solution containing 10% PEG diacrylate and the co-catalysts triethanolamine (13.2 μL/mL) and 1-vinyl-2- pyrrolidine (3.5μL/mL) in HEPES-buffered saline solution was applied to the membranes and photopolymerized by irradiation at 480-520 nm and 75 mW/cm 2 for 1 minute from a portable Cermax xenon fiber optic light source (CXE300; ILC Technology Inc, Sunnyvale, CA).
The mussel-mimetic tissue sealant is a catechol-functionalized PEG (cPEG) whose molecules crosslink into a hydrogel by way of oxidation after addition of sodium periodate. The composition and synthesis of cPEG will be described elsewhere. For gelation, equal volumes of the polymer precursor solution (300 mg/mL in phosphate-buffered saline solution) and the cross-linking solution (12 mg/mL sodium periodate in water) were mixed with the use of a dual syringe applicator device that was equipped with a blending connector with mixer (FibriJet; Micromedics Inc, St. Paul, MN). Hydrogels prepared from cPEG polymer and its derivatives are expected to possess the ability to secure very strong adhesion to almost any surface, even under wet conditions. The presence of catechol in cPEG sealant was inspired by the wet adhesive properties conferred by the catechol side chain of 3,4-dihydroxyphenylalanine (DOPA) amino acid, which is found in high concentrations in the foot proteins of marine or freshwater mussels.
Tissucol Duo S fibrin glue is a biologic 2-component adhesive that forms by the mixing of human plasma cryoprecipitate solution with thrombin solution. The chemical and physical polymerization of the main component of fibrin glue sealant, fibrinogen, mimics the last step of the natural blood clot formation; fibrin glue clinically is applied widely as hemostatic surgical sealant or an adjunct to suture.
Toxicity tests
The toxicity of sealants for fetal membrane cells was evaluated with direct contact and elution tests, as per International Organization for Standardization (ISO) 10993-5 guidelines.
Direct contact cytotoxicity
Direct contact studies were performed with term fetal membranes that were obtained from 3 cases. The amniotic layer was chosen for sealant application ( Figure 1 , A) because this layer was proposed to be the strength-bearing layer of fetal membranes and a major determinant for PPROM. Patches (2 × 1 cm) of freshly harvested fetal membranes were placed into wells of 6-well plates with amnion layer up. The sealants were applied at 50 μL and 200 μL volumes (except cPEG adhesive, which was tested only at 200 μL volume because of limited material). Membranes that were covered with sealant were covered with 3 mL culture medium (Ham’s F-12/DMEM supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin) and cultured for 24 hours at 37°C. Controls were untreated membranes that were processed immediately for histologic findings (control “0”) or cultured for 24 hours (control “24”). After 24 hours, the treated membranes were fixed in 4% formalin, embedded in paraffin, and sectioned for histologic examination. Deparaffinized sections were either stained with hematoxylin-eosin or stained for apoptotic cells with the use of TUNEL technology (Terminal deoxynucleotidyl transferase dUTP nick end labeling; In Situ Cell Death Detection Kit, Fluorescein; Roche Diagnostics GmbH, Mannheim, Germany). For total cell counts, all cell nuclei were counterstained with 4′,6-diamidin-2′-phenylindol-dihydrochlorid (DAPI; Sigma, Buchs, Switzerland). The histologic images were taken with a fluorescent microscope (Zeiss Axiovert 200M; Carl Zeiss, Goettingen, Germany) that was equipped with an c digital camera (Zeiss AxioCam MR; Carl Zeiss), and analyzed with AxioVision Rel software (version 4.5: Carl Zeiss). Apoptotic and total cell counts were acquired from fluorescence micrographs with automated image analysis software (ImageJ, version 1.34s; National Institutes of Health, Bethesda, MD). One tissue section per case was analyzed, with 4 optical fields per section for analysis.
