Abstract
Background: Many surgeons utilize biologic mesh for elective complex ventral hernia repair (VHR; large hernias, contaminated fields, or patients with comorbid conditions). However, no randomized controlled trials (RCTs) have compared biologic and synthetic mesh. We hypothesize biologic mesh would result in fewer major complications at one-year post-operative compared with synthetic mesh.
Patients and Methods: We performed a single-center, pilot RCT. All eligible patients undergoing complex, open VHR were randomly assigned to receive biologic or synthetic mesh placed in the retromuscular position. Primary outcome was major complications, namely, a composite of mesh infection, recurrence, or re-operation at one-year post-operative. Secondary outcomes included surgical site infections (SSI), seromas, hematomas, wound dehiscence, re-admissions, and Clavien-Dindo complication grade. Outcomes were assessed using Fisher exact test and Bayesian generalized linear models.
Results: Of 87 patients, 44 were randomly assigned to biologic mesh and 43 to synthetic mesh. Most cases were wound class 2–4 (68%) and 75% had a hernia width >4 cm. Most patients were obese (70%) and had an American Society of Anesthesiogists (ASA) score of 3–4 (53%). Compared with patients in the synthetic mesh group, patients in the biologic mesh group had a higher percentage of: major complications at one-year post-operative (42.4% vs. 21.6%; relative risk [RR] = 1.96 [95% confidence interval {CI} = 0.94–4.08]; number needed to harm = 4.8; p = 0.071); SSI (15.9% vs. 9.3%; RR = 1.71 [95% CI = 0.54–5.42]; p = 0.362); wound dehiscence (25.0% vs. 14.0%; RR = 1.79 [95% CI = 0.73–4.41]; p = 0.205); and re-admissions (22.7% vs 9.3%; RR = 2.44 [95% CI = 0.83–7.20]; p = 0.105). Bayesian analysis demonstrated that compared with synthetic mesh, biologic mesh had a 95% probability of increased risk of major complications at one-year post-operative. No clear evidence of a difference was found on seromas, hematomas, or Clavien-Dindo complication grade.
Conclusions: In elective complex open VHR, biologic mesh demonstrated no benefit compared with synthetic mesh in one-year outcomes. Moreover, Bayesian analysis suggests that biologic mesh may have an increased probability of major complications.
Keywords: biologic mesh, polypropylene, randomized controlled trial, synthetic mesh, ventral hernia repair
Evidence supports the use of mesh during elective ventral hernia repair (VHR) given a decrease in the risk of hernia recurrence [1]. Synthetic meshes have been available for decades and multiple randomized controlled trials (RCT) have favored their use for uncomplicated VHR compared with suture repair [2–4]. However, synthetic mesh is associated with increased risk of surgical site infection (SSI), mesh infection, adhesions, bowel obstruction and enterocutaneous fistula, especially in cases of contamination [5,6]. Triggered by synthetic mesh related complications, biologic meshes have been developed and highly adopted for use in complex (large hernias, contaminated fields, or patients with comorbid conditions), open cases.
There has been interest in the role of biologic meshes for VHR and a variety of options are currently available including porcine, bovine, or human sources of tissues such as dermis, intestinal submucosa, and pericardium [7–11]. These tissues are de-cellularized, yielding a matrix of collagen that serves as a supporting scaffold for host cellular integration and neovascularization, followed by new collagen deposition and tissue regeneration. Many surgeons believe biologic meshes are associated with lower risk of SSIs, hernia recurrence, and improved patient abdominal wall function [12]. Based on these perceived benefits, biologic mesh has been widely used for complex open VHR. However, no RCTs assessing the safety or efficacy of biologic mesh for VHR have been published [9,13].
We sought to compare clinical and patient centered outcomes of complex open VHR with biologic versus synthetic mesh and hypothesized that biologic mesh would cause fewer major complications at one-year post-operative.
Patients and Methods
After obtaining Iinstitutional Review Board (IRB) approval, we performed a prospective, single-center, blinded pilot RCT of patients undergoing complex open VHR. Consolidated Standards of Reporting Trials (CONSORT) guidelines were followed, and the trial was registered in clinicaltrials.gov (NCT03091790).
All eligible patients who were determined to be candidates for complex open VHR in hernia clinics between September 2017 and January 2019 were approached. Inclusion criteria included any adult patient (≥18 years old) with a complex primary or incisional ventral hernia deemed most suited for open repair given a large size defined as defects >10 cm wide, contamination, or extensive adhesions. Exclusion criteria included those patients with an acute infection, such as an infected mesh from previous hernia repair with untreated drainage or abscess; patients with severe comorbidities limiting survival beyond two years after the operation based upon surgeon judgment (e.g., metastatic cancer or complicated cirrhosis); patients in whom the surgeon would not normally place a prosthetic mesh (e.g., planned second surgery such as an ostomy takedown); and patients with high likelihood of being lost to follow-up (e.g., resides out of state or no access to telephone). Patients with chronic mesh infections that had been drained and treated with antibiotic agents but with no active/acute mesh infections were eligible for enrollment. A surgical informed consent and a separate research consent were signed by those patients who agreed to be enrolled in the study. No modifications were made to the research protocol after enrollment of patients began.
Baseline demographic information was obtained through a complete medical history and physical examination during pre-operative clinic visits. Patients completed the modified Activity Assessment Scale (mAAS), a validated hernia-specific survey that investigates patients' abdominal wall quality of life (AW-QOL) through 12 questions [14,15]. Responses are measured on a 10-point Likert scale and then the cumulative survey results are normalized to a scale of 1–100. On this scale, a score of one is poor AW-QOL, 100 is perfect, 80 or greater is normal, and the minimally clinically important difference (MCID) is seven points [16,17].
Patients were randomly assigned intra-operatively through computer-generated variable block randomization to either biologic mesh (porcine acellular dermal matrix) or synthetic mesh (medium-density, macroporous polypropylene). Porcine acellular dermal matrix was selected because it is one of the most commonly used biologic meshes in the market and is readily available in our institution, whereas polypropylene was selected because it is the gold standard synthetic mesh used for uncomplicated VHR. At the time of abdominal closure and mesh placement the operating surgeon called a research coordinator not involved in data collection who performed the allocation using sequentially numbered, sealed, and opaque envelopes kept in the research coordinator's office. A one-to-one allocation ratio was used to ensure an equal number of patients in each group and patients were considered randomized when the sealed envelope was opened.
Only experienced hernia surgeons performing more than 50 open VHR per year and with equipoise on using biologic or synthetic mesh for complex ventral hernias participated in the study. Pre-operative skin decontamination with chlorhexidine solution, prophylactic antibiotic therapy, multimodal pain control regimen, and discharge criteria were standardized. To control for differences in operative techniques between surgeons, hernia repairs were performed following recommendations of several surgical societies.
In general, the abdomen was entered via a midline laparotomy. The incision was extended cranially and caudally to provide adequate access to all the margins of the fascial defect. The peritoneum was entered, and the sac contents were reduced into the abdominal cavity. All abdominal wall adhesions were divided. All previously placed mesh, if present, was explanted. When the midline could not be closed without undue tension or when there were lateral hernias, posterior component separation was performed. Prior to closure, we determined if primary fascial closure was achievable. Among patients in whom the fascia (anterior or posterior sheath) could not be closed with suture alone, the defect was bridged with in situ hernia sac or polyglactin 910 mesh.
Posterior fascial closure was performed with running 0-polyglactin 910 suture. The mesh was placed in the retromuscular position and secured with 0-polydioxanone sutures. Jackson Pratt 10-French drains were selectively placed anterior to the mesh based on surgeon judgement. Anterior fascial closure was performed with running 0-polydioxanone suture. All excess skin and subcutaneous tissue were excised, Scarpas and Camper's fascia were approximated with 0-polyglactin 910 suture, skin was closed with staples, and negative pressure wound therapy devices were applied.
The operating surgeon and research coordinator who determined the randomization allocation were not blinded. However, the patient, the statisticians (O.A.O. and C.S.B.), and the remainder of the research team including surgeons assessing post-operative outcomes were all blinded to the allocation group.
All patients were scheduled to return to the clinic after their procedure for follow up at one-month and one-year post-operative. A physical examination with special focus on diagnosing complications such as SSI, seromas, hematomas, wound dehiscence, hernia recurrence, and re-operations was performed by surgeons blinded to the treatment allocation group. Additionally, patients completed the mAAS survey during each visit to determine changes in AW-QOL.
Primary outcome was defined as major complications at one-year post-operative, a composite of mesh infection, hernia recurrence, and re-operation. Hernia recurrence was determined by clinical assessment and on-demand computed tomography (CT) scan. Re-operation was defined as any procedure involving the fascia, mesh, or intra-peritoneal cavity. Secondary outcomes included SSI, surgical site occurrence (SSO), re-admissions, Clavien-Dindo complication grade, days in hospital at 90-days post-operative (including post-operative length of stay and re-admission length of stay), AW-QOL measured through the mAAS, and pain scores measured through the visual analog scale (VAS). Surgical site infection was defined, per the U.S. Centers for Disease Control and Prevention (CDC) guidelines [18]. Surgical site occurrence included seromas, hematomas, and wound dehiscence. Seromas were defined as localized fluid collection around the wound without signs of infection, hematomas were defined as an accumulation of clotted blood around the incision, and wound dehiscence was defined as an opening of the surgical incision of any size. Re-operation was defined as an unplanned return to the operating room for complications associated with the initial surgery. Surgical site infection and SSO were assessed at one-month post-operative clinic visit, whereas electronic medical records were queried to determine re-admissions, re-operations, and other complications up to 90-days after the initial operation.
Primary outcome was analyzed in an intention-to-treat model using both frequentist and Bayesian statistics. Frequentist analysis was performed using a Fisher exact test. Because there are no studies assessing differences in major complications between biologic and synthetic mesh, a neutral, non-informative prior (mean = 0, standard deviation = 1) was used for the Bayesian logistic regression analysis. Secondary outcomes were analyzed with frequentist statistical methods including Fisher exact tests for categorical data, two-tailed Student t-test for parametric continuous data, and Mann-Whitney U test for non-parametric continuous data. Analysis of covariance (ANCOVA) was performed to analyze AW-QOL and pain scores [19]. Subgroup analyses of the primary outcome, planned a priori, were performed for the following subgroup variables one at a time: hernia type, hernia area, and wound class.
Multivariable logistic regression analysis for the primary outcome using the same subgroup variables (hernia type, hernia area, and wound class) was performed to allow risk-adjusted interpretation of the results. All statistical analyses were performed using the computing environment R version 3.6.3 [20].
We aimed to perform the largest feasible pilot trial to obtain the least biased estimate of treatment effect. We initially planned to assess results on 50 patients at one-year post-operative. However, because of an increase in the volume of cases in our center, we obtained IRB approval to increase our sample size to the largest sample feasible, an estimated 50–100 patients.
Results
A total of 190 eligible patients were approached. Of these, 87 patients were randomly assigned: 44 to biologic mesh and 43 to synthetic mesh (Fig. 1). The majority of patients were Hispanic or African American females. Most patients were obese, had at least one comorbid condition, had an incisional hernia, and had a medium or large size hernia (>4 cm width). Patient baseline demographics and hernia characteristics were similar between the two groups (Table 1). Most cases were wound class 2–4, all cases were classified as Ventral Hernia Working Group (VHWG) class 2–4, primary fascial closure was achieved in the majority of patients, and all repairs were performed with retromuscular mesh placement (Table 2).
FIG. 1.
Consolidated Standards of Reporting Trials (CONSORT) flow diagram.
Table 1.
Patient Baseline Demographic Characteristics
| Total |
Biologic mesh |
Synthetic mesh |
|
|---|---|---|---|
| n = 87 | n = 44 | n = 43 | |
| Age (y), mean (SD) | 51 (±11.0) | 51 (±9.9) | 51 (±12.1) |
| Gender | |||
| Male, no. (%) | 38 (43.7%) | 21 (47.7%) | 17 (39.5%) |
| Race, no. (%) | |||
| Hispanic | 58 (66.7%) | 28 (63.6%) | 30 (69.8%) |
| African American | 16 (18.4%) | 9 (20.5%) | 7 (16.3%) |
| White | 8 (9.2%) | 4 (9.1%) | 4 (9.3%) |
| Other | 5 (5.7%) | 3 (6.8%) | 2 (4.6%) |
| BMI, (kg/m2), mean (SD) | 32.3 (±6.3) | 31.8 (±6.4) | 32.9 (±6.1) |
| Recent smoker,a no. (%) | 8 (9.2%) | 5 (11.4%) | 3 (7.0%) |
| Diabetes mellitus, no. (%) | 25 (28.7%) | 15 (34.1%) | 10 (23.3%) |
| ASA class, no. (%) | |||
| 1–2 | 41 (47.1%) | 23 (52.3%) | 18 (41.9%) |
| 3–4 | 46 (52.9%) | 21 (47.7%) | 25 (58.1%) |
| Hernia type, no. (%) | |||
| Primary | 10 (11.5%) | 6 (13.6%) | 4 (9.3%) |
| Incisional | 77 (88.5%) | 38 (86.4%) | 39 (90.7%) |
| Prior ventral hernia repair, no. (%) | 35 (40.2%) | 18 (40.9%) | 17 (39.5%) |
| Hernia width, (cm), median (IQR) | 6.0 (3.2–8.5) | 6.0 (3.1–8.9) | 6.0 (4.0–8.3) |
| Hernia length, (cm), median (IQR) | 6.0 (3.6–15) | 6.0 (3.7–15.0) | 8.0 (3.5–15.0) |
| Hernia area, (cm2), median (IQR) | 40.0 (15.2–120.0) | 35.4 (10.7–157.0) | 45.0 (18.8–92.5) |
| Hernia width, no. (%) | |||
| Small (<4 cm) | 22 (25.3%) | 12 (27.3%) | 10 (23.2%) |
| Medium (4–10 cm) | 48 (55.2%) | 22 (50%) | 26 (60.5%) |
| Large (>10 cm) | 17 (19.5%) | 10 (22.7%) | 7 (16.3%) |
SD = standard deviation; BMI = body mass index; ASA = American Society of Anesthesiologists; IQR = interquartile range.
Recent smoker was defined as current smokers and those who quit within 30 days preceding the surgery date.
Table 2.
Intra-Operative Details
| Total |
Biologic mesh |
Synthetic mesh |
|
|---|---|---|---|
| n = 87 | n = 44 | n = 43 | |
| Wound class 2–4, no. (%) | 59 (67.8%) | 29 (65.9%) | 30 (69.8%) |
| VHWG 2–4, no. (%) | 87 (100%) | 44 (100%) | 43 (100%) |
| Component separation, no. (%) | 32 (36.8%) | 14 (31.8%) | 18 (41.9%) |
| Unilateral PCS | 16 (18.4%) | 5 (11.4%) | 11 (25.6%) |
| Bilateral PCS | 15 (17.2%) | 9 (20.4%) | 6 (14.0%) |
| Bilateral ACS | 1 (1.2%) | 0 (0%) | 1 (2.3%) |
| Concomitant procedures, no. (%) | 44 (50.6%) | 21 (47.7%) | 23 (53.5%) |
| Gastrointestinal | 18 (20.7%) | 11 (25.0%) | 7 (16.3%) |
| Gynecologic | 13 (14.9%) | 4 (9.1%) | 9 (20.9%) |
| Hepatobiliary | 10 (11.5%) | 5 (11.4%) | 5 (11.6%) |
| Urologic | 2 (2.3%) | 1 (2.3%) | 1 (2.3%) |
| Inguinal hernia repair | 1 (1.2%) | 0 (0%) | 1 (2.3%) |
| Primary fascial closure, no. (%) | 84 (96.6%) | 43 (97.7%) | 41 (95.4%) |
| Retromuscular mesh placement, no. (%) | 87 (100%) | 44 (100%) | 43 (100%) |
| OR time, (minutes), mean (SD) | 180 (90.6) | 183 (86) | 177 (97) |
Risk assessment tool for complications after hernia repair which include: grade 2 (history of wound infection, comorbidities); grade 3 (presence of a stoma, violation of the gastrointestinal tract); and grade 4 (presence of mesh infection and septic dehiscence. Of note, no patients with septic dehiscence were enrolled and only patients with actively treated, chronic mesh infections were included per the inclusion/exclusion criteria.
VHWG = Ventral Hernia Working Group classification; PCS = posterior component separation; ACS = anterior component separation; OR = operating room; SD = standard deviation.
Patients were followed up for a median of 12.4 months (interquartile range = 11.4–12.8 months; biologic mesh = 12.4 months versus synthetic mesh = 12.5 months). At one-year post-operative, 70 (81.4%) patients were evaluated, 33 in the biologic mesh group and 37 in the synthetic mesh group. Seventeen patients were unable to be contacted by telephone or e-mail and were therefore excluded from our analysis. Baseline characteristics of patients lost to follow-up were similar compared with those included in the analyses except for the percentage of smokers (Supplementary Table S1). A total of three patients, two in the synthetic mesh group and one in the biologic mesh group, had a hernia so wide that primary fascial closure was not feasible. In these cases, either the posterior or anterior fascia, or both, were bridged with hernia sac. In all three cases, the mesh was still placed in the retromuscular position.
The percentage of patients in the biologic mesh group who had a major complication at one-year post-operative was nearly double that of patients in the synthetic mesh group, (42.4% vs. 21.6%; relative risk [RR] = 1.96 [95% confidence interval {CI} = 0.94–4.08]; number needed to harm = 4.8; p = 0.071) (Table 3). On Bayesian logistic regression analysis, there was a 95% posterior predicted probability of open VHR with biologic mesh resulting in more major complications compared with synthetic mesh. Details regarding patients who had major complications are shown in Supplementary Table S2. There were five patients in the biologic mesh group who developed mesh infections in the early post-operative period. Three of these required re-operation for abdominal wound washout and excisional debridement, and the other two were able to be managed non-operatively with wound care, percutaneous drainage of associated fluid collection, and/or antibiotic therapy. Two patients in the synthetic mesh group experienced mesh infections, one required re-operation and subsequent wound care, and one was amenable to be managed non-operatively with percutaneous drainage and antibiotic therapy.
Table 3.
One-Year and One-Month Post-Operative Outcomes by Mesh Type
| One-year post-operative outcomes | Biologic mesh n = 33 | Synthetic mesh n = 37 | Relative risk (95% CI)a | p | Bayesian posterior probability |
|---|---|---|---|---|---|
| Major complication, no. | |||||
| (%) | |||||
| Overall | 14 (42.4%) | 8 (21.6%) | 1.96 (0.94–4.08) | 0.071 | 95% |
| Mesh infection | 5 (15.2%) | 2 (5.4%) | 2.80 (0.58–13.49) | 0.242 | 86% |
| Recurrence | 10 (30.3%) | 5 (13.5%) | 2.24 (0.85–5.89) | 0.143 | 93% |
| Reoperation | 4 (12.1%) | 3 (8.1%) | 1.49 (0.36–6.19) | 0.699 | 67% |
| One-month post-operative outcomes | Biologic mesh n = 44 | Synthetic mesh n = 43 | Relative risk (95% CI)a | p | Bayesian posterior probability |
| Surgical site infection, no. (%) | 7 (15.9%) | 4 (9.3%) | 1.71 (0.54–5.42) | 0.362 | 79% |
| Seroma, no. (%) | 5 (11.4%) | 5 (11.6%) | 0.98 (0.30–3.14) | 0.977 | 49% |
| Hematoma, no. (%) | 3 (6.8%) | 2 (4.7%) | 1.47 (0.26–8.34) | 0.666 | 63% |
| Wound dehiscence, no. (%) | 11 (25.0%) | 6 (14.0%) | 1.79 (0.73–4.41) | 0.205 | 88% |
| Re-admission, no. (%) | 10 (22.7%) | 4 (9.3%) | 2.44 (0.83–7.20) | 0.105 | 93% |
| Clavien-Dindo | |||||
| complications, no. (%) | |||||
| Overall | 21 (47.7%) | 17 (39.5%) | 1.21 (0.75–1.95) | 0.444 | 76% |
| 1–2 | 16 (36.4%) | 13 (30.2%) | |||
| 3–5 | 5 (11.4%) | 4 (9.3%) | |||
| Days in hospital at 90-days post-operative, (days), median (IQR) | 3.5 (1.8–7.3) | 3.0 (1.0–6.5) | 1.14 (0.95–1.37) | 0.160 | 92% |
CI = confidence interval; IQR = interquartile range.
Synthetic mesh represents the control for relative risk calculation.
Subgroup analysis by hernia type, hernia width, and wound class revealed an increase in the risk of major complications with the use of biologic mesh as opposed to synthetic mesh (Supplementary Table S3). Biologic mesh was associated with nearly double the percentage of hernia recurrence at one-year post-operative (30.3% vs. 13.5%), SSI (15.9% vs. 9.3%), wound dehiscence (25.0% vs. 14.0%), and re-admissions (22.7% vs. 9.3%). There was no clear evidence of differences in the percentages of seromas, hematomas, or other Clavien-Dindo complications. However, results of these subgroup analyses should be interpreted as hypothesis generating because of the lack of statistical power.
At one-year follow up, mean AW-QOL scores of patients who underwent complex open VHR with biologic mesh improved by 20.3 points, whereas on average those in the synthetic mesh group experienced an increase of 27.9 points (Table 4). Adjusting for baseline AW-QOL, the estimated mean difference in follow-up AW-QOL between groups was 7.5 points. A greater percentage of patients in the biologic arm experienced a worsening in AW-QOL (18.2% vs. 13.9%; RR = 1.36 [95% CI = 0.40–4.65]; p = 0.726) whereas a larger percentage of patients in the synthetic group had an improvement in AW-QOL (67% vs. 75%; RR = 0.89 [95% CI = 0.65–1.21]; p = 0.609) (Supplementary Table S4). There was no clear evidence of a difference seen in pain scores between groups at one-year post-operative.
Table 4.
Patient-Centered Outcomes by Mesh Type
| Biologic mesh n = 33 | Synthetic mesh n = 36a | Mean difference (95% CI)b | p | |
|---|---|---|---|---|
| AW-QOL Scores (mAAS), mean (SD) | ||||
| Baseline | 42.3 (29.9) | 42.1 (30.5) | — | |
| Follow-up | 62.6 (27.4) | 70.0 (28.1) | −7.4 (−20.8 to 6.0) | 0.273 |
| Change | 20.3 (31.2) | 27.9 (31.3) | −7.6 (−22.6 to 7.4) | 0.316 |
| ANCOVA | — | −7.5 (−19.7 to 4.7) | 0.226 | |
| AW-QOL change category, No (%) | ||||
| Worsened (<7) | 5 (15.1%) | 4 (11.1%) | 1.36 (0.40–4.65)c | 0.747 |
| Improved (>7) | 22 (66.7%) | 27 (75.0%) | 0.89 (0.65–1.21)c | |
| No change | 6 (18.2%) | 5 (13.9%) | 1.31 (0.44–3.89)c | |
| Pain scores (VAS), mean (SD) | ||||
| Baseline | 6.2 (3.4) | 7.2 (2.9) | — | |
| Follow-up | 4.3 (3.3) | 4.0 (3.5) | 0.3 (−1.3 to 2.0) | 0.687 |
| Change | −1.8 (4.1) | −3.2 (3.5) | −1.4 (−3.2 to 0.49) | 0.148 |
| ANCOVA | — | 0.8 (−0.9 to 2.3) | 0.382 | |
AW-QOL = Abdominal wall quality of life; mAAS = Modified Activity Assessment Scale (range from 1 = poor to 100 = perfect, minimally clinically important difference [MCID] = 7), VAS = visual analog scale (range from 0 = no pain to 10 = worst possible pain, MCID = 1); SD = standard deviation; ANCOVA = analysis of covariance.
One patient did not complete the one-year post-operative survey.
Synthetic mesh represents the control for mean difference calculation.
Relative risk is reported instead of difference in median as variable is categorical.
Discussion
This is the first RCT comparing clinical and patient-centered outcomes of biologic mesh and synthetic mesh for complex open VHR. Contrary to the common belief of improved outcomes, biologic mesh is associated with more major complication rates (42.4% vs. 21.6%; RR = 1.96 [95% CI = 0.94–4.08]) at one-year post-operative. Moreover, with increased rates of SSI in the early post-operative period, patients undergoing repairs with biologic mesh are at increased risk of mesh infection, re-admission, and re-operation. Although most complications were minor (Dindo-Clavien 1 or 2), nearly half of patients in our study experienced a complication demonstrating that complex VHR is associated with substantial morbidity.
It is estimated that nearly half a million VHR are performed each year in the United States, accounting for over $3.2 billion in healthcare costs [21]. Increasing costs per hernia repair have been observed as biologic mesh popularity has increased. A previous national database study found dramatic differences in estimated mean operating room services and supply costs with biologic mesh being three times more expensive than synthetic mesh for uncomplicated clean VHR ($21,106 vs. $7,067; cost ratio = 3.0, mean difference = $14,172 [95% CI = $11,451-$16,893], p < 0.001) [22]. Most of this cost difference is because of the cost of the mesh: biologic mesh on average, costs substantially more than synthetic mesh. At our institution, a 30 × 30 cm piece of biologic mesh costs the hospital $23,147 as opposed to $170 for a piece of synthetic mesh of the same size. For this reason, biologic meshes are infrequently used in uncomplicated cases. Our study further builds on these findings suggesting that the role of biologic mesh even in complicated elective cases does not seem to be justified as cost dominance (increased complications and increased costs) is likely demonstrated [23].
Although the results of our study do not support the use of biologic mesh in open elective complex ventral hernia repair, several questions remain unanswered by our pilot RCT. Primary fascial closure and placement of mesh in optimal locations are believed to be important tenets of ventral hernia surgery [24,25]. For our study, we opted to perform a retromuscular repair because placement of mesh in this anatomic layer is associated with the lowest hernia recurrence and SSI rates [26,27]. This approach can be technically challenging, may not be the preferred approach for all surgeons, and may not be indicated in some clinical settings (damaged posterior sheath, emergency setting, inability to achieve primary fascial closure, i.e., bridged repair, etc.). In these settings, a surgeon may prefer to place mesh intraperitoneal. The results of this study cannot inform the use of synthetic or biologic mesh in these situations. Additionally, there is still ongoing debate on the ideal management strategy of ventral hernias among patients undergoing minimally invasive operations in the setting of contamination.
Many medical devices introduced to the market, including meshes for VHR, are cleared through the Food and Drug Administration (FDA) 510(k) process. As opposed to the drug industry in which RCTs are required to prove safety and efficacy prior to approval, the 510(k) process allows devices to be cleared with demonstration of substantial equivalence to already marketed products. Biologic meshes, including porcine acellular dermal matrix meshes that were utilized in our study, became commercially available by showing substantial equivalence to synthetic mesh [9]. Several of these FDA-cleared devices have not been evaluated rigorously. Surgeons, healthcare workers, and patients should be aware that “FDA clearance” does not ensure that “evidence-based” safety or efficacy has been demonstrated. We advocate that new devices, including surgical meshes, should undergo RCTs prior to wide adoption.
Our present study has considerable strengths. The main strength is a study design with low risk for bias (RCT). Randomization of patients just prior to mesh placement and blinding of patients and post-operative outcome assessors to the allocation group were performed. Wide inclusion criteria were established to increase the generalizability of our findings to a broad population. In addition, a high follow-up rate was achieved [28]. Although this is typical for reports unveiling short-term outcomes, we are currently conducting three-year and five-year follow-up to assess long-term outcomes such as mesh infection, recurrence, re-operations, AW-QOL, and pain scores, with plans to publish these results.
Some limitations of our study need to be acknowledged. As a small single-center pilot trial, our study may be underpowered. Assuming the treatment effect estimates of our study are precise, an appropriately powered (α of 0.05 and a β of 80%) RCT would require at least 156 patients. As more hernias are being repaired through minimally invasive techniques, an appropriately powered study on open ventral hernia repairs would likely require collaboration between multiple high-volume centers. However, Bayesian analysis estimated a high probability of biologic mesh having worse clinical outcomes when compared to synthetic mesh. We acknowledge the importance of high follow-up rates of RCTs. All patients (87 patients, 100%) were seen in clinic for their one-month post-operative follow-up appointment. However, despite aggressive measures to avoid dropouts, 17 patients (11 in the biologic mesh group and 6 in the synthetic mesh group) were unable to be reached at one-year post-operative and were therefore excluded from the analysis of the primary outcome. Although our one-year follow up of 81.4% could be higher, the ideal follow-up of RCTs has been conventionally described as equal to or greater than 80% of the original cohort of patients [28,29]. Generalizability of our results may be limited to other centers with different patient populations. This study was performed by experts in abdominal wall reconstruction of complex patients with high risk for complications, therefore, the generalizability of our findings to surgeons with less case-volume and case-complexity may be reduced. Yet, given the relation between outcomes and volume of cases, it is reasonable to imagine surgeons with higher volume might have better outcomes. We opted to investigate the most commonly utilized biologic mesh type as this would have the greatest applicability and impact a larger number of patients. Although these outcomes may not be generalizable to other biologic or biosynthetic meshes (e.g., human and bovine acellular dermal matrix, porcine collagen intestinal submucosa derivatives, bovine pericardium derivatives, and poly-4-hydroxybutyrate), other products should demonstrate superiority to mid-density polypropylene mesh through RCTs.
Conclusions
This is the first RCT comparing clinical and patient-centered outcomes of biologic versus synthetic meshes. Regardless of the mesh type utilized, complex open ventral hernia repair is associated with a high rate of complications. Although long-term follow-up needs to be completed, the use of biologic mesh resulted in more major complications compared to mid-density polypropylene synthetic mesh.
Supplementary Material
Acknowledgments
This study was presented at the American College of Surgeons Virtual Clinical Congress in October 2020.
Funding Information
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author Disclosure Statement
No competing financial interests exist.
Supplementary Material
References
- 1. Liang MK, Holihan JL, Itani K, et al. Ventral hernia management: Expert consensus guided by systematic review. Ann Surg 2017;265:80–89 [DOI] [PubMed] [Google Scholar]
- 2. Kaufmann R, Halm JA, Eker HH, et al. Mesh versus suture repair of umbilical hernia in adults: A randomised, double-blind, controlled, multicentre trial. Lancet 2018;391:860–869 [DOI] [PubMed] [Google Scholar]
- 3. Luijendijk RW, Hop WC, Van den tol MP, et al. A comparison of suture repair with mesh repair for incisional hernia. N Engl J Med 2000;343:392–398 [DOI] [PubMed] [Google Scholar]
- 4. Burger JW, Luijendijk RW, Hop WC, et al. Long-term follow-up of a randomized controlled trial of suture versus mesh repair of incisional hernia. Ann Surg 2004;240:578–583 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. De vries reilingh TS, Van goor H, Charbon JA, et al. Repair of giant midline abdominal wall hernias: “Components separation technique” versus prosthetic repair: Interim analysis of a randomized controlled trial. World J Surg 2007;31:756–763 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Van't riet M, De vos van steenwijk PJ, Bonjer HJ, et al. Mesh repair for postoperative wound dehiscence in the presence of infection: Is absorbable mesh safer than non-absorbable mesh? Hernia 2007;11:409–413 [DOI] [PubMed] [Google Scholar]
- 7. Majumder A, Winder JS, Wen Y, et al. Comparative analysis of biologic versus synthetic mesh outcomes in contaminated hernia repairs. Surgery 2016;160:828–838 [DOI] [PubMed] [Google Scholar]
- 8. Köckerling F, Alam NN, Antoniou SA, et al. What is the evidence for the use of biologic or biosynthetic meshes in abdominal wall reconstruction? Hernia 2018;22:249–269 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Fitzgerald JF, Kumar AS. Biologic versus synthetic mesh reinforcement: What are the pros and cons? Clin Colon Rectal Surg 2014;27:140–148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Bondre IL, Holihan JL, Askenasy EP, et al. Suture, synthetic, or biologic in contaminated ventral hernia repair. J Surg Res 2016;200:488–494 [DOI] [PubMed] [Google Scholar]
- 11. Liang MK, Berger RL, Nguyen MT, et al. Outcomes with porcine acellular dermal matrix versus synthetic mesh and suture in complicated open ventral hernia repair. Surg Infect 2014;15:506–512 [DOI] [PubMed] [Google Scholar]
- 12. Nissen AT, Henn D, Moshrefi S, et al. Health-related quality of life after ventral hernia repair with biologic and synthetic mesh. Ann Plast Surg 2019;82(5S Suppl 4):S332–S338 [DOI] [PubMed] [Google Scholar]
- 13. Huerta S, Varshney A, Patel PM, et al. Biological mesh implants for abdominal hernia repair: US Food and Drug Administration approval process and systematic review of its efficacy. JAMA Surg 2016;151:374–381 [DOI] [PubMed] [Google Scholar]
- 14. Krpata DM, Schmotzer BJ, Flocke S, et al. Design and initial implementation of HerQLes: A hernia-related quality-of-life survey to assess abdominal wall function. J Am Coll Surg 2012;215:635–642 [DOI] [PubMed] [Google Scholar]
- 15. Mccarthy M, Jonasson O, Chang CH, et al. Assessment of patient functional status after surgery. J Am Coll Surg 2005;201:171–178 [DOI] [PubMed] [Google Scholar]
- 16. Cherla DV, Moses ML, Viso CP, et al. Impact of abdominal wall hernias and repair on patient quality of life. World J Surg 2018;42:19–25 [DOI] [PubMed] [Google Scholar]
- 17. Poulose BK, Roll S, Murphy JW, et al. Design and implementation of the Americas Hernia Society Quality Collaborative (AHSQC): Improving value in hernia care. Hernia 2016;20:177–189 [DOI] [PubMed] [Google Scholar]
- 18. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for Prevention of Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control 1999;27:97–132 [PubMed] [Google Scholar]
- 19. Vickers AJ, Altman DG. Statistics notes: Analysing controlled trials with baseline and follow up measurements. BMJ 2001;323:1123–1124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. R Core Team.. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2019. www.R-project.org/ (Last accessed November4, 2020)
- 21. Poulose BK, Shelton J, Phillips S, et al. Epidemiology and cost of ventral hernia repair: Making the case for hernia research. Hernia 2012;16:179–183 [DOI] [PubMed] [Google Scholar]
- 22. Totten CF, Davenport DL, Ward ND, Roth JS. Cost of ventral hernia repair using biologic or synthetic mesh. J Surg Res 2016;203:459–465 [DOI] [PubMed] [Google Scholar]
- 23. Cohen DJ, Reynolds MR. Interpreting the results of cost-effectiveness studies. J Am Coll Cardiol 2008;52:2119–2126 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Bernardi K, Olavarria OA, Holihan JL, et al. Primary fascial closure during laparoscopic ventral hernia repair improves patient quality of life: A multicenter, blinded randomized controlled trial. Ann Surg 2020;271:434–439 [DOI] [PubMed] [Google Scholar]
- 25. Bernardi K, Olavarria OA, Liang MK. Primary fascial closure during minimally invasive ventral hernia repair. JAMA Surg 2020;155:256–257 [DOI] [PubMed] [Google Scholar]
- 26. Holihan JL, Hannon C, Goodenough C, et al. Ventral hernia repair: A meta-analysis of randomized controlled trials. Surg Infect 2017;18:647–658 [DOI] [PubMed] [Google Scholar]
- 27. Holihan JL, Nguyen DH, Nguyen MT, et al. Mesh Location in open ventral hernia repair: A systematic review and network meta-analysis. World J Surg 2016;40:89–99 [DOI] [PubMed] [Google Scholar]
- 28. Fewtrell MS, Kennedy K, Singhal A, et al. How much loss to follow-up is acceptable in long-term randomised trials and prospective studies? Arch Dis Child. 2008;93:458–461 [DOI] [PubMed] [Google Scholar]
- 29. Kristman V, Manno M, Côté P. Loss to follow-up in cohort studies: How much is too much? Eur J Epidemiol 2004;19:751–760 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.