Characterization of Poly-4-Hydroxybutyrate Mesh for Hernia Repair Applications

journal of surgical research xxx (2013) 1e8

Available online at www.sciencedirect.com

journal homepage: www.JournalofSurgicalResearch.com

Characterization of poly-4-hydroxybutyrate mesh for hernia repair applications

David P. Martin, PhD,a Amit Badhwar, PhD,b Devang V. Shah, PhD,b Said Rizk, MS,a Stephen N. Eldridge, MS,b Darcy H. Gagne, BS,b Amit Ganatra, MS,a Roger E. Darois, BEng,b Simon F. Williams, PhD,a Hsin-Chien Tai, PhD,b and Jeffrey R. Scott, PhDb,c,* a Tepha, Inc, Lexington, Massachusetts b C. R. Bard, Inc (Davol), Warwick, Rhode Island c Department of Molecular Pharmacology, Physiology & Biotechnology, Brown University, Providence, Rhode Island article info abstract

Article history: Background: Phasix mesh is a fully resorbable implant for soft tissue reconstruction made from Received 1 December 2012 knitted poly-4-hydroxybutyrate monofilament fibers. The objectives of this study were to Received in revised form characterize the in vitro and in vivo mechanical and resorption properties of Phasix mesh over 12 March 2013 time, and to assess the functional performance in a porcine model of abdominal hernia repair. Accepted 13 March 2013 Materials and methods: We evaluated accelerated in vitro degradation of Phasix mesh in 3 mol/L Available online xxx HCl through 120 h incubation. We also evaluated functional performance after repair of a surgically created abdominal hernia defect in a porcine model through 72 wk. Mechanical Keywords: and molecular weight (MW) properties were fully characterized in both studies over time. Hernia repair Results: Phasix mesh demonstrated a significant reduction in mechanical strength and MW Phasix mesh over 120 h in the accelerated degradation in vitro test. In vivo, the Phasix mesh repair Poly-4-hydroxybutyrate demonstrated 80%, 65%, 58%, 37%, and 18% greater strength, compared with native Resorbable abdominal wall at 8, 16, 32, and 48 wk post-implantation, respectively, and comparable repair strength at 72 wk post-implantation despite a significant reduction in mesh MW over time. Conclusions: Both in vitro and in vivo data suggest that Phasix mesh provides a durable scaffold for mechanical reinforcement of soft tissue. Furthermore, a Phasix mesh surgical defect repair in a large animal model demonstrated successful transfer of load bearing from the mesh to the repaired abdominal wall, thereby successfully returning the mechanical properties of repaired host tissue to its native state over an extended time period. ª 2013 Elsevier Inc. All rights reserved.

1. Introduction until now, the use of fully resorbable synthetic surgical meshes in hernia repair has resulted in a high frequency of The development of a resorbable mesh that can provide incisional hernias because of the short-term strength reten- an abdominal closure with adequate long-term mechanical tion of those materials [1]. For this reason, fully resorbable stability remains an attractive goal, provided that recurrence meshes have traditionally been used for temporary, short- rates and chronic complications can be minimized. However, term wound support rather than for long-term use in hernia

* Corresponding author. C. R. Bard, Inc (Davol), 100 Crossings Blvd, Warwick, RI 02886. Tel.: (401) 825-8542; fax: (401) 825-8762. E-mail address: [email protected] (J.R. Scott). 0022-4804/$ e see front matter ª 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2013.03.044 2 journal of surgical research xxx (2013) 1e8

Fig. 1 e Scanning electron microscopy photomicrographs of Phasix mesh before (A) and after incubation in 3 mol/L HCl (simulating polymer degradation) for 32 (B), 72 (C), and 96 (D) h. Note the higher magniﬁcation of (D) (scale bar [ 200 versus 500 mm for [AeC]), which better illustrates the near-complete loss of the ﬁber integrity.

repair, and permanent meshes remain the standard for hernia [4HB]) that is normally present in human tissues [3].The repair. metabolite has an in vivo half-life of approximately 30 min [4,5] As an alternative to permanent mesh, our goal was to and is eliminated via the Krebs cycle as carbon dioxide and develop a long-term, resorbable hernia repair mesh that could water [2,6]. The Phasix mesh has a knitted mesh pattern similar provide the necessary support at the repair site during the to traditional Bard polypropylene mesh (C.R. Bard, Inc); before initial wound healing period while allowing tissue ingrowth implantation, the mechanical properties also resemble those of and progressive transfer of mechanical load from the mesh to Bard mesh (Table 1). the host tissue over time. We hypothesized that this gradual The purposes of this study were to evaluate the perfor- load transfer would promote tissue remodeling such that the mance of the Phasix mesh as a buttress to reinforce the primary repaired tissue could eventually provide long-term function repair of an approximate 2.5-cm (or 1-in) circular abdominal wall similar to that of native tissue. Once the repaired tissue is defect in a porcine model, and to correlate the in vivo behavior of capable of supporting the load and the mesh is no longer the Phasix mesh with its in vitro degradation profile. The primary needed, a resorbable mesh can degrade to leave behind healthy end points of the in vivo study were to determine the strength host tissue. Preferably, at the time of implant, the resorbable mesh should also offer comparable properties to traditional hernia mesh materials, such as polypropylene, and similar Table 1 e Comparison of properties of Phasix mesh and wound healing with compliance at the repair site that would Bard mesh. ultimately mature to that of the native abdominal wall (NAW) Mechanical/physical Phasix mesh Bard mesh as the mesh becomes integrated and resorbed over time. properties Phasix mesh (C.R. Bard, Inc [Davol], Warwick, RI) is Material Poly-4-hydroxybutyrate Polypropylene a biosynthetic resorbable monofilament mesh, derived from Pore size (in2) 0.0004 0.0009 poly-4-hydroxybutyrate (P4HB), specifically designed for hernia Thickness (in) 0.02 0.03 repair. Poly-4-hydroxybutyrate is a fully resorbable polymer Area density (g/m2) 182 105 produced by the microorganism Escherichia coli K12 via trans- Ball burst strength (lb 54.11 68.0 genic fermentation techniques [2].Phasixmeshismadeof force with 3/8-in ball) knitted monofilaments of this naturally derived polymer Suture pull out 5.6 8.0 strength (kgf) (Fig. 1A). The mesh provides immediate short-term support similar to traditional nonresorbable meshes, but provides an The Phasix mesh used in all studies was a warp knit mesh design absorbable scaffold that enables the abdominal wall to remodel made from high-strength P4HB monofilament fiber. The material to host tissue over time. Poly-4-hydroxybutyrate is a high- was sterilized using ethylene oxide. The initial mechanical and physical properties of Phasix mesh are compared with non- strength polyester [2] that degrades in a predictable and resorbable monofilament polypropylene Bard mesh. steady manner to a natural metabolite (4-hydroxybutyrate journal of surgical research xxx (2013) 1e8 3 and compliance of the repaired abdominal wall versus the NAW the repair site compared with NAW over time. The study was over the course of degradation of the Phasix mesh. approved by the Institutional Animal Care and Use Committee of DaVinci Biomedical Research Products, Inc (South Lancas- ter, MA) and was conducted in compliance with all regulations 2. Methods regarding the humane treatment of laboratory animals set forth by Institutional Animal Care and Use Committee. 2.1. Mesh specifications We anesthetized 30 male Yucatan swine, weighing 39.7 1.8kgattimeofimplant,with2.5%e4.0% inhalational iso- The Phasix mesh used in all studies was a warp knit mesh design flurane, and maintained them at 0.5%e2.5% throughout the (Fig. 1A) made from high-strength P4HB monofilament fiber procedure. The ventral abdomen was prepared for aseptic (Tepha, Inc, Lexington, MA). The material was sterilized using survival surgery by shaving the entire abdominal region ethylene oxide. The initial mechanical and physical properties and cleaning the operative area with three alternating scrubs of Phasix mesh are presented in Table 1, as compared with the of povidone-iodine solution and 70% alcohol. We applied nonresorbable, monofilament polypropylene Bard mesh. sterile surgical drapes over the entire operative field. After preparing the ventral abdomen for aseptic surgery, we per- 2.2. Phasix mesh accelerated in vitro degradation formed a midline laparotomy (w30 cm). We created a 1-in (2.5-cm)-diameter full fascial defect in the anterior abdom- We cut ethylene oxideesterilized Phasix mesh into 2 2-in inal wall using a preperitoneal approach (the peritoneum square pieces. We randomly selected eight samples for each remained intact). The surgical defect was initially reap- time point, weighed, placed them in 50-mL Falcon tubes, proximated with four interrupted 2-0 Vicryl sutures (Ethicon, incubated them in 45 mL of 3 mol/L HCl solution at 37C, and Somerville, NJ). The abdominal defect was further reinforced shook them at 50 rpm. For each time point (t ¼ 0, 8, 16, 32, 48, with one 3.5-in (8.9-cm)-diameter circular piece of Phasix 72, 96, and 120 h), we prepared mesh specimens as follows: We mesh circumferentially fixated with 16 PermaSorb resorbable decanted the HCl, rinsed the mesh samples twice with fixation device fasteners (C.R. Bard, Inc). The abdominal distilled water, and then quenched them for 5 min with midline was repaired with standard closure techniques. After phosphate-buffered saline solution (pH 7.42). We then rinsed recovering from anesthesia, the pigs were allowed access to the samples two times with distilled water and vacuum-dried food and water ad libitum. The animals were individually them for at least 2 h before testing. Time 0 samples were not housed in pens during the survival portion of the study and subjected to HCl treatment. the abdominal region was inspected daily to assess the For each time point, we measured the mass of the eight condition of both the wound line and subcutaneous tissue samples. We performed scanning electron microscopy (SEM) (seromas and/or hematomas). analysis on one representative sample at each time point. We randomly selected five animals at each time point (0, 8, Burst strength was measured on the remaining samples at 16, 32, 48, and 72 wk post-implantation). We euthanized the each time point using an MTS universal testing machine Q test designated animals by injecting 150 mg/kg sodium pentobar- Elite (Eden Prairie, MN) with a 3/8-in ball burst fixture, 1000 N bital intravenously, in accordance with the American Veteri- load cell, and grip pressure of 70 psi. The mesh specimen was nary Medical Association Panel of Euthanasia. After held on a fixture over a 7/16-in circular hole and a 3/8-in euthanasia, we dissected the abdominal skin from the entire rounded probe was forced through the mesh, perpendicular abdomen and inspected the abdominal wall for evidence of to the mesh surface at 12 in/min until rupture. We recorded hernia or diastases, excised it, and placed it in saline solution the load (measured in Newtons) as a function of extension by (0.9% NaCl) for subsequent mechanical testing. Before mech- the universal testing machine. After mechanical testing, we anical testing, all loosely adhered tissue was removed to allow determined the weight average molecular weight (Mw) of the the mechanical ball burst fixture to be tightly applied around mesh remnants for each time point by gel permeation chro- the porcine abdominal wall to ensure adequate sample grip- matography (GPC) relative to monodisperse polystyrene stan- ping. The repair site tissue was mounted in a ball burst fixture dards using an Agilent 1100 Series high-performance liquid with the peritoneum intact and a 3/8-in rounded probe was chromograph (Agilent Technologies, Inc, Santa Clara, CA) with forced through the repair site, perpendicular to the abdominal a refractive index detector. Samples for GPC were prepared at wall until rupture. We recorded mechanical data using an 1 mg/mL in chloroform; 100 mL of the solutions were injected Instron universal testing machine (Norwood, MA) with a 1500 onto a PLgel column (5 mm, mixed C, 300 7.5 mm) and eluted N load cell and displacement rate of 25.4 mm/min. For each at 1 mL/min. We calculated the percent retention of mass, animal and each time point, we recorded mechanical burst strength, and Mw as the ratio of the testing time point value to force and relative stiffness for the Phasix mesh repair and the time 0 value multiplied by 100%. a representative area of NAW within the same plane and cranial-caudal position of the mesh repair. The mechanical 2.3. Phasix mesh repair strength retention over time ball burst force was defined as the peak load required to (in vivo): porcine model achieve failure of the surgical defect repair or NAW. We calculated relative stiffness from the slope of the load versus We implanted Phasix mesh in a Yucatan swine model of elongation curve. simulated hernia repair to evaluate the functional perfor- To further characterize the degradation of Phasix mesh mance of the repair at 8, 16, 32, 48, and 72 wk. The objective of in vivo, we determined the Mw of explanted mesh specimens by the study was to determine the mechanical burst strength of GPC. After mechanical testing, we dissected a portion of the 4 journal of surgical research xxx (2013) 1e8 repair site free of loose tissue. The repair site tissue was enzymatically digested with collagenase at selected time points (0, 32, 48, and 72 wk post-implantation) using the following enzymatic digestion protocol. Each explanted mesh specimen was placed in a 50-mL Falcon tube containing 25 mL collagenase (type I) solution (1.0 mg/mL) in TESCA buffer (50 mmol/L 2-[Tris(hydroxymethyl)methylamino]-1-ethane- sulfonic acid, 2 mmol/L CaCl2, 10 mmol/L NaN3, pH 7.4, sterile filtered). The tube was placed in a shaker (50 rpm) and incubated at 37C overnight (w17 h) to digest and remove tissue attached to the mesh specimen. After the incubation was complete, we removed the specimen from the tube; manually Fig. 2 e Comparison of the residual mass (black line), ball removed residual tissue from the explant, taking care not to burst strength (blue line), and M (red line), expressed as damage the mesh; and rinsed the mesh with distilled water, w a percentage of the initial (time [ 0) levels from the followed by 70% ethanol. We then blotted dry the mesh spec- accelerated in vitro study. The overall mass of the Phasix imens using a lint-free wipe and determined the Mw by GPC of mesh was relatively stable over the first 48 h, after which it the collagenase-digested mesh remnants. Molecular weight steadily declined. The burst strength declined immediately analysis was not included in the original animal protocol and and rapidly over the first 72 h, after which the burst so was not performed at the earlier time points (8 and 16 wk). strength was negligible. The strength profile followed the same degradation profile as the M , indicating 2.4. Statistical analysis w a correlation between strength and fiber integrity, as indexed by the M . We collected, analyzed, interpreted, and graphically displayed w study data with GraphPad Prism 5.03 statistical software (GraphPad Software, Inc, La Jolla, CA). We used either Student t-test or grouped analysis of variance with Tukey’s post- hoc analysis for multiple comparisons. Data are reported as with Mw retention, whereas substantial mass loss occurs later mean mass (grams) standard error (SE), mean ball burst in the degradation process. Furthermore, the data suggest that force (Newtons) SE, mean relative stiffness (Newtons per by the time bulk structural changes were apparent by SEM, over 90% of strength and M had been lost. millimeter) SE, and/or mean Mw (kilodaltons) SE. Statis- w tical significance was set at P < 0.05. The Mw and ball burst strength data were correlated and graphically represented as a standard curve (Fig. 3). As such, this curve can be used to predict the ball burst strength rela- 3. Results tive to a specific Mw of the Phasix mesh. Importantly, the graph demonstrates that the bulk hydrolysis of the polymer 3.1. Accelerated in vitro testing results in loss of strength of the mesh and ultimately in a complete loss of structural integrity of the mesh at a Mw of We employed accelerated aging to simulate real-time in vivo approximately 50 kDa. resorption under in vitro conditions. This resulted in a resorption profile similar to that seen in vivo, but on a greatly reduced time scale. 3.2. Large animal model Scanning electron microscopy images of Phasix mesh specimens after in vitro incubation in 3 mol/L HCl at 37C We evaluated the functional performance of Phasix mesh in demonstrated visibly intact mesh through 32 h incubation, a buttress repair in a large animal (Yucatan swine) model. bulk structural changes by 72 h, and significant bulk degra- Functional performance was assessed by mechanical ball dation by 96 h, as illustrated in Figure 1BeD. The physical burst and relative stiffness (compliance) of a Phasix mesh changes observed are reflected in the Mw, burst strength, and buttress repair, compared with NAW at 0, 8, 16, 32, 48, and mass retention profiles shown in Figure 2 and summarized in 72 wk post-implantation (n ¼ 5/time point). All animals Table 2. survived the implantation procedure and expected study time Molecular weight and ball burst strength retention frame, with no evidence of hernia or surgical site diastasis decreased significantly (>70%) during the initial 32 h and over time. declined to <10% by 72 h. The mass of mesh, on the other Figure 4 shows the ball burst strength over implant time for hand, remained relatively constant up to about 32 h and the Phasix mesh repair site compared with the NAW. As noted declined to 5.0% remaining mass at 120 h. At 48 h, the mesh previously, all samples were tested with an intact peritoneum. had essentially lost most of its strength and Mw (e.g., 8% The mesh repair site was almost twice as strong as the NAW at strength and 14% Mw retention) just as the mesh initiated the time of implant. The strength of the mesh repair increased mass loss (95% residual). After 48 h, the mass loss continued up to week 16 and then began to decline. At week 16, the mesh through the remainder of the incubation and the mesh was repair was significantly stronger than the NAW. However, at essentially completely degraded at 120 h (5.0% residual mass). later times, the ball burst strength for the mesh repair was These data demonstrate that the strength retention correlates similar to that of the NAW, which had been slowly gaining in journal of surgical research xxx (2013) 1e8 5

Table 2 e Summary of in vitro data. Time (h) Mass (g SE) Mass retention (%) Burst strength (N SE) Strength retention (%) Mw (kDa SE) Mw retention (%)

0 0.381 0.007 100.0 225.5 7.28 100.0 368.3 0.71 100.0 8 0.376 0.006 99.0 196.4 6.74 87.4 204.5 1.40 55.5 16 0.370 0.004 97.2 145.5 6.19 64.8 133.4 1.81 36.2 32 0.390 0.004 102.5 57.7 4.20 25.6 76.1 0.09 20.7 48 0.364 0.007 95.5 18.9 0.98 8.4 50.7 0.31 13.8 72 0.269 0.004 70.6 1.1 0.09 0.5 29.0 0.12 7.9 96 0.146 0.000 38.4 0.0 0.00 0.0 15.6 0.20 4.2 120 0.002 0.000 5.0 0.0 0.00 0.0 10.4 0.04 2.8

Molecular weight and ball burst strength retention decreased signiﬁcantly (>70%) during the initial 32 h and declined to <10% by 72 h. The mass of mesh remained relatively constant up to w32 h and declined to 5.0% remaining mass at 120 h. At 48 h, the mesh had essentially lost most of

its strength and Mw (e.g., 8% strength and 14% Mw retention) as the mesh initiated mass loss (95% residual). After 48 h, the mass loss continued through the remainder of the incubation and the mesh was essentially completely degraded at 120 h (5.0% residual mass). strength over time because of the natural growth of the approximately 20% from week 8 to week 72 (Table 4). As animal. shown in Figure 6B, after 48 wk in vivo the mesh material did Figure 5 shows the relative stiffness over implant time for not demonstrate substantial bulk degradation. However, after the mesh repair and NAW. At time 0, the relative stiffness was 72 wk, the mesh material was essentially completely similar for both the Phasix mesh repair site and NAW. The degraded, and only a small portion of the initial mesh material relative stiffness of the mesh repair increased and was was recoverable (Fig. 6D). The Mw of the remaining mesh significantly higher than NAW up to 32 wk, and then declined remnants was measured (23 kDa) and found to be <10% of to approximately NAW values at 48 and 72 wk post- initial Mw of the starting mesh. We estimated the strength of implantation. the mesh device by the standard curve generated by the The pre-implantation Mw of the Phasix mesh was 256.7 in vitro study (Fig. 3), demonstrated in Table 3. This estimated 3.3 kDa. The Mw declined by over 50% (to 97.0 1.4 kDa) over strength contribution was confirmed by arithmetically sub- the first 32 wk post-implantation and declined an additional tracting the strength of the NAW from the strength of the 40% (to 23.4 1.2 kDa) for recovered mesh remnants by the repair. As such, the increased strength of the repair over the end of the study (Table 3). Figure 6 shows the mesh remnants NAW can be attributed to the Phasix mesh at each time point. resulting from the tissue digestion in preparation for Mw Histologic analysis of the Phasix mesh-reinforced repair testing at representative time points. The images demonstrate demonstrated the presence of a moderate chronic host inflam- the bulk degradation of the mesh that occurred after 48 and matory response surrounding the mesh knots and mono- 72 wk. Concomitant with the bulk degradation, we observed filaments at 8, 16, 32, 48, and 72 wk post-implantation (data a decrease in fiber diameter using morphometric analysis of not shown). This response primarily consisted of macrophages, the tissue sections. The fiber diameter decreased lymphocytes, and occasional giant cells. Further morphometric analysis demonstrated a mild reduction in the Phasix mesh monofilament fiber diameter at 16, 32, 48, and 72 wk post-

Fig. 4 e Ball burst force (strength) of the explanted Phasix e e Fig. 3 Standard curve relating the weight average Mw and mesh repaired abdominal hernia defect (blue line) relative ball burst strength from the in vitro analysis. The central to the strength of the unadulterated NAW. Over the early

(solid) line represents the correlation between Mw and time points post-implantation, the Phasix mesh provided strength; the dashed lines represent the 95% conﬁdence an increased strength of repair relative to NAW. Over the intervals. Using this graph, the strength of a mesh piece later time points, the strength of repair decreased to the with a known Mw can be predicted. level of the NAW. 6 journal of surgical research xxx (2013) 1e8

cross-section at 72 wk post-implantation, the overall amount of recovered mesh remnants was small (Fig. 6). Additional analysis also demonstrated a minor increase in the thickness of host inﬂammatory response surrounding the Phasix mesh at 16, 32, and 48 wk post-implantation, compared with 8 wk, followed by a reduction at 72 wk post-implantation (Table 4).

4. Discussion

Several studies have previously evaluated the use of resorbable meshes for hernia repair. These include a functional assessment of the long-term resorbable (LTS) mesh derived from a glycolide/lactide copolymer described by Fig. 5 e Relative stiffness of the explanted Phasix Klinge et al. [1] and, more recently, an implantation study mesherepaired abdominal hernia defect (blue line) relative of another multifilament mesh (TIGR mesh) derived from e e to the relative stiffness of the unadulterated NAW. Over the a glycolide lactide trimethylene carbonate copolymer [7]. early time points post-implantation, the relative stiffness of Phasix mesh described in this report differs from the LTS and the Phasix mesh repair significantly increased (relative to TIGR meshes in several important aspects. First, Phasix mesh time 0) and demonstrated an increased relative stiffness is made from a monofilament fiber, whereas the LTS and compared with NAW. Over the later time points, the relative TIGR meshes are based on multifilament fibers. This is an stiffness of the repair decreased to the level of the NAW. advantage because monofilament mesh devices have been deemed to be more biocompatible [8,9] and less susceptible to infection [10,11]. Second, Phasix mesh is made from implantation, compared with 8 wk, which is consistent with a biosynthetic polymer (P4HB) that is metabolized in vivo degradation/resorption of the mesh monofilament over time to a natural metabolite (4HB) that is far less acidic (i.e., has (Table 4). Although, we observed the bulk of the fiber diameter in a higher pKa), compared with glycolic and lactic acid

Fig. 6 e Photographic images of a portion of the 48-wk (A and B) and 72-wk (C and D) Phasix mesh repair sites removed from subjects in the porcine study before (A and C) and after (B and D) collagenase digestion to remove attached tissue. (A) Repair site tissue before collagenase digestion, 48 wk. (B) Mesh remnants recovered from tissue after collagenase digestion and tissue removal at 48 wk. (C) Repair site tissue before collagenase digestion, 72 wk. (D) Mesh remnants recovered from tissue after collagenase digestion and tissue removal at 72 wk. journal of surgical research xxx (2013) 1e8 7

Table 3 e Using standard curve (Fig. 3) to estimate Phasix mesh strength over time in vivo. Time post- Phasix mesh NAW strength Estimation of Phasix Mw (kDa SE) Predicted strength of Phasix implantation repair strength (N SE) mesh strength [(Phasix mesh alone (from in (wk) (N SE) mesh repair strength) vitro study) (N)* (NAW strength)] (N)

0 332.4 33.47 169.4 17.93 163.0 256.7 3.33 207 y 8 440.0 37.05 266.6 35.78 173.4 N/A y 16 502.3 71.77 317.5 31.24 184.8 N/A 32 402.5 46.76 294.7 46.97 107.8 97.0 1.40 99 48 362.7 44.82 328.0 50.26 34.7 55.7 1.10 36 72 289.5 29.03 299.7 29.51 (10.2) 23.4 1.17 0

The in vitro Mw and ball burst strength data were correlated and graphically represented as a standard curve (Fig. 3). This curve was used to

predict the in vivo ball burst strength of Phasix mesh relative to a specific measured in vivo Mw of Phasix mesh. The predicted strength of Phasix mesh is highly correlated with the actual experimental in vivo strength (R2 ¼ 0.9532). * Read from Figure 3. y Data not measured. monomers, which could contribute to the increased change up to 32 h. Thereafter, the degradation of the mesh biocompatibility [2]. Furthermore, the 4HB metabolite of fibers became more noticeable, with increasing degradation Phasix mesh is eliminated from the body via the Krebs cycle up to the 72-h point. At 96 and 120 h, the fibers and mesh were as carbon dioxide and water [2,4]. Third, Phasix mesh has substantially degraded and no longer intact. Only small re- initial mechanical properties and handles like polypropylene sidual fragments consisting of 5.0% of the initial mass could be meshes, but is designed to allow the repaired abdominal wall recovered after 120 h. The apparent physical changes were to remodel to native host tissue in a predictable manner. The reflected in the ball burst strength retention data. The ball report here describes one of several studies evaluating Phasix burst strength declined sharply after 32 h. The Mw retention mesh for use in hernia repair procedures, to determine the profile mirrored that of ball burst strength, and these two strength and compliance (e.g., relative stiffness), essentially physical parameters could be correlated to generate a stan- to complete degradation of the mesh. The study was also dard curve relating Mw to burst strength (Fig. 3). This corre- designed to provide an initial assessment of the host infla- lation allows for the prediction of the upper limit of burst mmatory response to the Phasix mesh material in an ab- strength of a piece of mesh with a known Mw. Other factors dominal buttress repair model. in vivo, such as surface erosion, can result in slightly lower Although the initial burst strength of Phasix mesh is an strength values than would be strictly predicted from the important element of its design, it is also vital that the mesh in vitro model. However, this model has shown good correla- degrades in a predictable and steady manner so that the load tion between in vivo and in vitro degradation studies in other at the repair site can be gradually transferred from the mesh preclinical studies (data not shown). As such, this is a valuable to the host tissue during wound healing. Before testing the tool for use in in vivo studies, because the Mw of explanted strength retention of Phasix mesh in animals, we determined mesh can be measured, allowing for an estimation of the the strength retention profile of Phasix mesh in vitro in an strength of the remaining Phasix mesh material in the repair. accelerated degradation model: incubation in 3 mol/L HCl Table 3 compares the arithmetic measurement of the strength solution at 37C. Scanning electron micrographs showed little contribution of the mesh material in the repair and the predicted strength contribution based on the standard curve. There is a high level of similarity between these two estima- Table 4 e Histologic morphometry summary. tors, indicating the utility of the standard curve and the predictable degradation profile of Phasix mesh in vivo. Another Time post- Mean Phasix mesh Mean inflammatory pertinent observation that can be made from the information implantation fiber diameter (mm) thickness (mm)* (wk) in Figure 3 and Table 3 is a correlation factor between in vitro degradation and in vivo degradation. Taken together, 1 wk 8 0.244 26 in vivo is approximately equal to 0.8 h in vivo under the 16 0.224 39 conditions tested. We estimated this correlation factor by 32 0.208 32 48 0.206 34 comparing the percent retention of mass, Mw, and burst 72 0.198 22 strength in both in vivo and in vitro conditions. An investigation into the burst strength of the explanted Morphometric analysis demonstrated a mild reduction in the porcine wall with and without the surgical defect repair Phasix mesh monofilament fiber diameter at 16, 32, 48, and 72 wk post-implantation, compared with 8 wk, which is consistent with indicates that Phasix mesh has two distinct phases in its degradation/resorption of the mesh monofilament over time. life cycle. Initially (0e16 wk), we saw that the Phasix mesh Additional analysis also demonstrated a minor increase in the material offered additional strength to the repair of a surgi- thickness of host inflammatory response surrounding the Phasix cally created hernia defect compared with the strength of mesh at 16, 32, and 48 wk post-implantation, compared to 8 wk, the NAW. This is important because during this early phase followed by a reduction at 72 wk post-implantation. of hernia repair and wound healing, the highest risk of * Surrounding the Phasix mesh fibers. reherniation and dehiscence exists owing to the immaturity 8 journal of surgical research xxx (2013) 1e8 of the repair in the absence of a mesh scaffold [12].Therelative references stiffness (Fig. 5) of the repair was relatively high, which indicates a natural fibrotic response that occurs during remodeling and wound repair. After this early phase, a second phase exists [1] Klinge U, Schumpelick V, Klosterhalfen B. Functional (32e72 wk) in which the Phasix mesh material continues to assessment and tissue response of short- and long-term remodel with native host tissue and the fibers begin to degrade absorbable surgical meshes. Biomaterials 2001;22:1415. [2] Martin DP, Williams SF. Medical applications of poly-4- and resorb (as indicated by the Mw of the remaining mesh). hydroxybutyrate: a strong flexible absorbable biomaterial. The overall strength of repair begins to transfer to the NAW and Biochem Eng J 2003;16:97. the relative stiffness returns to the same level as the NAW. [3] Nelson T, Kaufman E, Kline J, Sokoloff L. The extraneural The mesh fiber remnants that remain at 72 wk are completely distribution of gamma hydroxybutyrate. J Neurochem 1981; nonfunctional with respect to their ability to support the load 37:1345. requirements of the repair. The morphometric data indicate [4] Sendelbeck SL, Girdis CL. Disposition of a 14C-labeled that the fiber remnants that can be found in the 72-wk samples bioerodible polyorthoester and its hydrolysis products, have a smaller diameter than the pre-implant fibers, and gross 4-hydroxybutyrate and cis, trans-1,4-bis(hydroxymethyl) cyclohexane, in rats. Drug Metab Dispos 1985;13:291. imagery (Fig. 6), M , and strength (Fig. 4) indicate that the w [5] Ferrara SD, Zotti S, Tedeschi L, et al. Pharmacokinetics of material is essentially resorbed. This is further supported by gamma-hydroxybutyric acid in alcohol dependent patients the near-complete decrease in the bulk mass of the Phasix after single and repeated oral doses. Br J Clin Pharmacol 1992; mesh material present at this late time. Based on the Mw of the 34:231. explanted mesh at 72 wk, it would closely resemble the 96-h [6] Gibson KM, Nyhan WL. Metabolism of [U-14C]-4- in vitro SEM image (Fig. 1D), which demonstrates a degraded hydroxybutyric acid to intermediates of the tricarboxylic acid cycle in extracts of rat liver and kidney mitochondria. mesh with no discernable organization. Taken together, these Eur J Drug Metab Pharmacokinet 1989;14:61. data demonstrate that during the early phase of hernia repair, [7] Hjort H, Mathisen T, Alves A, Clermont G, Boutrand JP. Three- Phasix mesh provides load support to the repair site, and as the year results from a preclinical implantation study of a long- mesh material begins to degrade and resorb and the host tissue term resorbable surgical mesh with time-dependent remodels, the load transfers to the abdominal wall. By 72 wk, mechanical characteristics. Hernia 2012;16:191. the Phasix mesh material is nearly completely degraded and [8] Nguyen PT, Asarias JR, Pierce LM. Influence of a new resorbed and no longer contributes strength to the repair. monofilament polyester mesh on inflammation and matrix remodeling. J Invest Surg 2012;25:330. In conclusion, the studies described here suggest that [9] Bryan N, Ahswin H, Smart NJ, Bayon Y, Hunt JA. In vitro Phasix mesh offers promise for use in hernia repair proce- activation of human leukocytes in response to contact with dures, and may represent an attractive option to have the synthetic hernia meshes. Clin Biochem 2012;45:672. handling advantages of a polypropylene mesh, yet provide [10] Aydinuraz K, Agalar C, Agalar F, Ceken S, Buruyurek N, a host tissueebased repair free from permanent material Voral T. In vitro S. epidermidis and S. aureus adherence to retention. The strength profile (as measured from the in vitro composite and lightweight polypropylene grafts. J Surg Res and in vivo studies) and biocompatibility of the Phasix mesh 2009;157:e79. [11] Klinge U, Junge K, Spellerberg B, Piroth C, Klosterhalfen B, appear to make it well suited for this purpose. There were no Schumpelick V. Do multifilament alloplastic meshes instances of reherniation, diastasis or dehiscence in this increase the infection rate? Analysis of the polymeric study, which indicates that Phasix mesh provides adequate surface, the bacteria adherence, and the in vivo support in our porcine model of abdominal hernia repair consequences in a rat model. J Biomed Mater Res 2002;63:765. during early times after surgical repair, before transferring the [12] Ebersole GC, Buettmann EG, Macewan MR, et al. support to the NAW. Based on these preclinical findings, Development of novel electrospun absorbable polycaprolactone (PCL) scaffolds for hernia repair appropriate studies in humans will have to confirm the applications. Surg Endosc 2012;26:2717. preclinical outcomes.