Systematic review of the stage of innovation of biological mesh for complex or contaminated abdominal wall closure

Background Achieving stable closure of complex or contaminated abdominal wall incisions remains challenging. This study aimed to characterize the stage of innovation for biological mesh devices used during complex abdominal wall reconstruction and to evaluate the quality of current evidence. Methods A systematic review was performed of published and ongoing studies between January 2000 and September 2017. Eligible studies were those where a biological mesh was used to support fascial closure, either prophylactically after midline laparotomy, or for reinforcement after repair of incisional hernia with midline incision. The primary outcome measure was the IDEAL framework stage of innovation. The key secondary outcome measure was the GRADE criteria for study quality. Results Thirty‐five studies including 2681 patients were included. Four studies considered mesh prophylaxis, 23 considered hernia repair, and eight reported on both. There was one published randomized trial (IDEAL stage 3), none of which was of high quality; the others were non‐randomized studies (IDEAL stage 2a). A detailed description of surgical technique was provided in most studies (27 of 35); however, no study reported outcomes according to the European Hernia Society consensus statement and only two described quality control of surgical technique during the study. From 21 ongoing randomized trials and observational studies, 11 considered repair of incisional hernia and 10 considered prophylaxis (seven in elective settings). Conclusion The evidence base for biological mesh is limited, and better reporting and quality control of surgical techniques are needed. Although results of ongoing trials over the next decade will improve the evidence base, further study is required in the emergency and contaminated settings.


Introduction
Incisional hernias carry a significant burden for both patients and the health service 1 -4 . They prevent return to normal activities and can be painful. Elective repair can be challenging, and emergency repair carries significant clinical risks. Incisional hernia is common, occurring in up to 50 per cent of patients after laparotomy 5,6 , and with the growing number of emergency laparotomies performed in the UK, the number of affected patients is likely to increase 7 .
To limit the number of incisional hernias there has been a focus on the use of prophylactic mesh reinforcement. The cost of mesh is far less than that of major reoperations and emergency admissions 3,8 -10 . Although synthetic meshes are accepted in many cases, they are not used in complex and contaminated settings owing to the risk of infection (as high as 50-90 per cent), pain, fistulation and need for explantation 11 -14 . Biological mesh has evolved to fill this gap, with expected reduced rates of infection leading to safer prophylaxis. Current guidelines, including the Ventral Hernia Working Group expert consensus, and several systematic reviews recommend against the use of synthetic mesh when the risk of wound complications is high, such as in the presence of gross contamination; instead they advocate the use of a biological absorbable mesh 15 -17 . Biological mesh has entered widespread clinical practice, but the quality and scope of the evidence base for use in complex and contaminated abdominal wounds are unclear. This review aimed to determine the quality and stage of innovation of the evidence supporting biological mesh placement during abdominal wall reconstruction with primary fascial closure. The hypothesis was that the evidence base supporting biological mesh use is currently too limited to support routine clinical use outside clinical trials.

Search strategy
A systematic search of PubMed, EMBASE and the Cochrane Library between 1 January 2000 and 27 September 2017 was performed by two independent investigators. The ClinicalTrials.gov database was also queried for ongoing studies. The search terms used were 'laparotomy', 'mesh', 'biologic material', 'abdominal wall', 'hernia', and 'complications', 'contamination', 'infection' or 'surgical site infection', individually or in combination. The 'related articles' function was used to broaden the search, and all citations were considered for relevance. A manual search of reference lists in recent reviews and eligible studies was also undertaken. This paper is reported according to the PRISMA guidelines 18 .

Inclusion and exclusion criteria
Studies were included according to the following criteria: evaluation of the use of a xenograft biological mesh to support primary fascial closure of midline abdominal wounds or repair of incisional hernia with midline incision; study design was an RCT, prospective observational study, retrospective cohort study or case series; study included only patients aged 16 years or more.
The following exclusion criteria were employed: study design was a systematic review, meta-analysis, letter, review, comment or conference abstract; fewer than five patients were included in the study; only synthetic mesh or composite meshes were evaluated; allograft or autograft meshes, including human-derived acellular dermal matrix, were used (availability in Europe across the selected inclusion dates was low until recently, so reporting is likely to be incomplete); study reported bridging repairs (fascial closure not achieved), including studies where outcomes for fascial closure were not reported separately from bridging repair.

Study outcome measures
The primary outcome measure was the stage of innovation, according to the IDEAL framework 19 . The level of evidence in the IDEAL staging system were 1 (case series with high risk of bias), 2a (cohort study), 2b (feasibility RCT), 3 (RCT) and 4 (high-quality prospective registry with long-term monitoring and low risk of bias). All assessments in the present study were carried out independently by two authors; disagreement was resolved by re-examining the relevant article until consensus was achieved.

Secondary outcome measures
The main secondary outcome measure was the quality of evidence assessed using the GRADE system 20 . In the GRADE approach, studies are categorized as of high (randomized trials or double-upgraded observational studies), moderate (downgraded randomized trials or upgraded observational studies), low (double-downgraded randomized trials or observational studies) and very low (triple-downgraded randomized trials, downgraded observational studies or case series/case reports) quality. The other secondary outcome measures of interest were the numbers of studies reporting: outcomes according to the European Hernia Society consensus statement 21 , incidence of incisional hernia, surgical-site infection (SSI) rate, and seroma.

Data extraction
Data extracted included patient demographics, indications and type of biological mesh used. Studies were grouped into those examining prophylactic placement in primary closure of laparotomy only (prophylaxis), repair of incisional hernia only (reinforcement), or both (mixed). Descriptions of procedures performed were collected, including surgical technique, number of procedures previously performed by the surgeon, and monitoring of technique. Degree of contamination (clean-contaminated, contaminated or dirty surgery) was defined according to the US Centers for Disease Control and Prevention (CDC) surgical wounds classification 22 , and the location of biological mesh placement was also evaluated, as either intraperitoneal (intraperitoneal, intraperitoneal onlay mesh, underlay, intra-abdominal) or extraperitoneal (sublay, onlay, inlay, retromuscular, retrorectus, prefascial) 23 .

Results
Of 1304 studies shortlisted, 35 full-text articles 24 -58 met the inclusion criteria ( Fig. 1). Of these, four examined biological mesh for prophylaxis, 23 reported on reinforcement after incisional hernia repair, and eight reported both prophylaxis and incisional hernia repair. Studies of biological mesh for prophylaxis included a total of 85 patients with a median follow-up of 12 (i.q.r. 2-31) months; those used for reinforcement included 1744 patients with a median follow-up of 16 (12-24) months, and those for mixed indications included 852 patients with a median follow-up of 24 (17-48) months.

IDEAL stage of innovation and GRADE quality of evidence
Distribution of IDEAL stage and GRADE quality of included studies are presented in Tables 3 and 4 respectively. Of the four prophylaxis studies, two 24,25 evaluated biological mesh at the time of stoma closure, one 26 following midline laparotomy after abdominal aortic aneurysm (AAA) repair, and one 27 after cytoreduction and hyperthermic intraperitoneal chemotherapy. All four studies included only elective patients and the degrees of contamination were clean-contaminated (2) and contaminated (2). Strattice™ was used in two studies 24,25 with an intraperitoneal placement; the others used bovine pericardium in an extraperitoneal position (1) 26 or Surgisis ® in an intraperitoneal position (1) 27 . One study 24      elective and emergency operations. The eight studies involved a mixture of procedures, with degree of contamination ranging from clean-contaminated to dirty. Mesh placement was intraperitoneal in six studies 51 -56 , extraperitoneal in one study 57 , and a combination in one study 58 . All were IDEAL stage 2a (cohort studies). Evidence was of very low quality in three studies 54,56,57 , low quality in three 51,52,55 , and moderate quality in two 53,58 . The evidence in one study 58  reconstruction with porcine acellular dermal matrix (Permacol™) was of moderate quality owing to reporting of long-term outcomes of at least 5 years.

Outcome reporting
None of the studies in this review reported outcomes according to the European Hernia Society consensus statement 21 , and none reported 'free from hernia' survival times. All four studies 24 -27 in the prophylaxis group reported a definition for detection of incisional hernia, which included a combination of clinical examination and radiological assessment. In the reinforcement group, 13 29,30,32,33,35,36,38,39,41 -44,47 of the 23 studies gave a definition for recurrence of hernia (6 clinical, 7 radiological, none patient-reported). SSI rates were reported in one 25 of the four studies in the prophylaxis group, and in 21 of the 23 studies in the reinforcement group. The incidence of seroma was reported in three prophylaxis and 19 reinforcement studies.

Reporting of surgical technique
Of the 35 studies, 27 provided details of surgical procedures: all four studies in the prophylaxis group, 16 in the reinforcement group, and seven in the mixed group ( Table 5). Only one paper 46 reported the minimum number of procedures performed by the operating surgeons as a requirement.

Ongoing studies
Twenty-one ongoing studies were identified from ClinicalTrials.gov, of which ten were for prophylaxis *Includes prophylaxis only, reinforcement only, and mixed. n.r., Not reported. and 11 for reinforcement. In the prophylaxis group, all were RCTs; four had completed data collection, five were still recruiting, and one had terminated early. Patient groups being studied included emergency midline laparotomy (1 study), elective patients for AAA repair (1), midline laparotomy (1), contaminated abdominal wall defect (1, terminated), abdominoperineal resection (1) and stoma closure (5). Of these ten, the majority studied Strattice™ (4), followed by Permacol™ (1) and Surgisis ® (1). The type of biological mesh was not mentioned in the remaining four studies. In the 11 ongoing trials of reinforcement, nine were RCTs and two were cohort studies. Two studies (1 cohort study of Permacol™ and 1 RCT of XenMatrix™) were in follow-up phase; the remainder were still recruiting patients.

Discussion
This review identified that the evidence base for biological mesh in complex and contaminated settings is still evolving, and highlighted areas for improvement. At present, the quality of the evidence base is generally low, with a few exceptions. The majority of studies included in this review were IDEAL stage 1 or 2 (case series or cohort studies) with a low or very low GRADE quality of evidence, indicating that biological meshes remain in the early stages of evaluation and adoption. This is compounded by a wide variation in mesh types and mesh placement, with little control for surgical technique, making synthesis of evidence ineffective.
There are two key recommendations from the present study. First, the evidence base needs to be improved by testing the efficacy of biological mesh in randomized trials. This should include standardization of techniques and reporting, and inclusion of more emergency cases to establish the limits of indication. Second, future studies should allow consistent reporting of mesh type and exact placement to enable high-quality recommendations to help standardize practice. Until such data are available, use in selected higher-risk patients (such as prophylaxis during abdominal wall closure in contaminated cases at high risk of incisional hernia) should be supported by data capture within controlled trials or registries. Routine clinical use in low-risk patients is not yet justified.
Surgeons and patients will benefit from knowing about mesh performance based on the specific type of mesh, the position it is placed in, and the expected long-term outcome. The present study identified variation in outcome reporting for recurrence rates, SSI and seroma. This variation precludes reliable assessment of outcomes and formation of recommendations. Recently, Blencowe and colleagues 59 proposed a standard approach for the description, standardization and monitoring of the intervention to enable reliable assessment of outcome from this type of study and, importantly, reproducibility of an intervention by surgeons in their clinical practice. In this review, only one study 46 had monitoring of technique by a senior surgeon to allow consistency of mesh placement.
It is plausible that different biological meshes may have varying failure rates, degrees of immunogenicity, biocompatibility and risk profiles 60 . In a rat study 61 of 85 laparoscopic ventral hernia repairs, Strattice™ and Parietex™ (Covidien Surgical, Dublin, Ireland) were seen to grow a new mesothelial layer on their visceral side, whereas microscopic degradation and new collagen formation were seen in the Surgisis ® group. In a mouse model of 135 mice with peritonitis, XCM BIOLOGIC ® (LifeCell, KCI, Branchburg, New Jersey, USA) and Permacol™ showed better incorporation than Strattice™, whereas Strattice™ had fewer strong adhesions 62 . More accurate information from human studies may allow improved selection of mesh for patients in future clinical practice.
The direct advantages of biological mesh remain unproven in widespread practice. First, the long-term durability of biological grafts used for complex abdominal wall reconstruction has been disappointing 36,43 . Rosen and co-workers 43 reported the overall hernia recurrence rate as 31 per cent over a mean follow-up of 21⋅7 (range 1-74) months, and estimated the 3-year recurrence-free survival rate to be 51 per cent. Second, implementation and use of biological mesh in clinical practice depend on the cost, as biological meshes can be up to ten times more expensive than synthetic ones 17,63 . Totten et al. 64 demonstrated that use of biological mesh for hernia repair can cost $21 000 (€17 100; exchange rate 20 April 2018) in comparison with synthetic mesh, which costs $7100 (€5780) for minimal improvement in surgical outcomes such as SSI.
With high costs of abdominal wall reconstruction using biological meshes and limited long-term data, there has been emerging interest in the use of long-term absorbable synthetic materials. These biosynthetic meshes are a clinical alternative to biological meshes and are significantly cheaper. A prospective longitudinal study by Rosen and colleagues 65 , evaluating the use of GORE ® BIO-A ® (W. L. Gore, Newark, Delaware, USA) biosynthetic mesh in CDC class II-IV wounds, demonstrated an SSI rate of 18 per cent and a hernia recurrence rate of 17 per cent at 24 months. In contrast, the RICH trial 36 , which evaluated CDC II-IV wounds with biological mesh, had an SSI rate of 66 per cent and recurrence rate of 28 per cent at 24 months. Although this evidence with biosynthetic meshes is promising, any superiority over biological mesh in clean, clean-contaminated, contaminated or infected wounds remains to be tested in RCTs.
Several ongoing cohort studies and RCTs will improve the evidence base, although they predominantly involve elective patients. Only three studies include both elective and emergency patients for prophylaxis. Future studies in high-risk patients (such as those undergoing emergency surgery, with active sepsis or high BMI) will establish new indications for biological mesh, with potentially greater benefit in these patients. Preventing the need for reoperation in high-risk groups is likely to provide even greater cost savings to health services.
There are weaknesses to this study. Assessment of quality using the GRADE tool is subjective, although this was overcome by discussion between the two authors involved in assessing grade of evidence, and resolving disagreement by re-examining the relevant article until consensus had been achieved. Nevertheless, this scoring system is used widely for assessing strength of evidence in the literature 20 . Biosynthetic resorbable meshes and patients undergoing bridged repairs were not included in the study, as they represent a clinically separate group and are likely to have a different stage of innovation due to timing of introduction.
The evidence base for biological mesh in this clinical context is limited and evolving. Better reporting and quality control of surgical techniques is needed and, although new trial results over the next decade will improve the evidence base, more trials in emergency and contaminated settings are required.

Disclosure
The authors declare no conflict of interest.