Synthetic Versus Biologic Mesh for Complex Open Ventral Hernia Repair: 3-Year Follow-Up of a Pilot Randomized Controlled Trial

Abstract

Background

Biologic mesh is often used in complex hernia repair, but there has been limited clinical evidence to date to support this practice. The aim of this study was to compare clinical and patient-reported outcomes of biologic versus synthetic mesh for complex open ventral hernia repair (OVHR) at 3 years.

Methods

Patients from a single center, randomized, controlled, pilot trial comparing biologic versus synthetic mesh in complex OVHR were followed for 3 years. The primary outcome focused on major complications, namely mesh infections, hernia recurrences, reoperations, and deaths. Secondary outcomes included surgical site infections, surgical site occurrences, and patient-reported outcomes. Outcomes were assessed using frequentist generalized linear models.

Results

A total of 87 patients (44 biologic mesh, 43 synthetic mesh) were randomized, and 61 patients (70%; 28 biologic and 33 synthetic) completed 3-year follow-up. Baseline demographics were similar in both groups. No significant differences were seen in major complications (50% vs 30%, P = .123), mesh infection (14% vs 3%, P = .144), recurrence (39% vs 24%, P = .214), reoperation (14% vs 9%, P = .531), or mortality (4% vs 0%, P = .459) between the 2 arms. A single death occurred as a result of bacteremia in a patient with hepatocellular carcinoma. Similarly, no significant differences were seen in secondary or patient-reported outcomes. Both groups demonstrated clinically important improvements in quality of life and pain scores at 3 years.

Conclusion

This study failed to find benefits with biologic mesh as opposed to synthetic mesh in complex OVHR at 3 years when comparing both clinical and patient-reported outcomes.

Introduction

More than 400 000 ventral hernia repairs (VHR) are performed each year in the United States (US), making hernias one of the most commonly encountered pathologies for the general surgeon.¹ The risk of recurrence can be higher than 50%, and the rate of complications increases with each subsequent repair.² Given this information, there has been an appropriate importance placed on improving outcomes for the first repair. Several strategies have been developed to mitigate the risk of recurrence, in particular, synthetic mesh repair over suture repair.^3–5 However, risks with synthetic mesh repair include infection, bowel obstruction, enterocutaneous fistula, and other adverse events.²

The biologic mesh procedure was developed in response to complications related to synthetic mesh repair. Biologic mesh contains scaffolding material derived from porcine, bovine, or human decellularized, collagen-rich tissue and is promoted to improve wound healing despite limited clinical data to support its use.^6–12 There were no randomized controlled trials (RCT) comparing biologic and synthetic mesh in VHR until recently, few of which report long-term data.^13–16

We previously published 1-year results of a single center, pilot RCT comparing biologic to synthetic mesh in complex (large hernias, contaminated fields, or patients with co-morbid conditions) open VHR, which demonstrated no benefit with biologic mesh, and potentially more major complications, compared to synthetic mesh.¹⁵ We now present outcomes at 3 years post-operative. We hypothesized that the biologic mesh group would have fewer major complications at 3 years post-operative.

Materials and Methods

Following institutional review board approval, we performed a single-center pilot RCT of patients undergoing open VHR in accordance with the Declaration of Helsinki, as revised in 2013. The Consolidated Standards of Reporting Trials (CONSORT) guidelines were followed. This trial was also registered on clinicaltrials.gov (NCT03091790).¹⁷ Primary analysis of early post-operative and 1-year outcomes has already been published.¹⁵ We report 3-year follow-up results in this manuscript.

Setting

This study took place at a safety net hospital in Houston, Texas.

Selection Criteria

We included adult patients (≥18 years old) undergoing complex, open VHR. Patients were deemed eligible for open repair if they had a hernia defect size >10 cm wide, extensive adhesions, or contamination. We excluded patients with an active infection (ie, acute mesh infection), patients in whom the operating surgeon normally would not place a prosthetic mesh (ie, planned second surgery such as an ostomy takedown), patients unlikely to survive more than 2 years based upon surgeon judgment, patients unlikely to follow up, and patients who did not speak either Spanish or English.

Randomization and Blinding

Following written informed consent, patients were randomly assigned to either the intervention, biologic mesh (non-cross-linked porcine acellular dermal matrix), or the control, synthetic mesh (medium-density, macroporous polypropylene). Porcine acellular dermal matrix was selected as it is one of the most commonly used biologic meshes on the market and was available at our institution. Only 1 brand of biologic mesh and 1 brand of synthetic mesh were used for the repairs. Patients were enrolled preoperatively and then randomized intra-operatively using computer-generated, variable block randomization. Prior to abdominal closure and mesh placement, the operating surgeon called the research coordinator, who was not involved in data collection, who assigned patients using sequentially numbered, opaque, sealed envelopes kept in their office. A 1:1 allocation ratio was used so that there was an equal number of patients in each group, and patients were considered randomized once the envelope was opened. The patient and outcomes assessor were blinded to the treatment group. The operating surgeon and research coordinator who allocated patients into treatment groups were not blinded.

Intervention and Control

Surgeons performing at least 50 VHR annually participated in this study. Perioperative care was standardized and included preoperative skin disinfection with chlorhexidine, multimodal pain control, prophylactic antibiotic therapy, and discharge criteria. VHR were performed using approaches recommended by surgical societies.

In general, a midline laparotomy was performed, sac contents were reduced into the abdominal cavity, and prior mesh, if present, was explanted. Posterior component separation was performed if the midline was not able to be closed without unnecessary tension or if lateral hernias were present. If primary fascial closure was unable to be performed, the defect was bridged using the hernia sac in situ or with polyglactin 910 mesh.

Posterior fascial closure was then done using running 0-polyglactin 910 suture. Mesh was placed in the retromuscular position using 0-polydioxanone sutures. Based on surgeon judgment, closed suction drains were selectively positioned anterior to the mesh. Anterior fascial closure was completed with a running 0-polydioxanone suture. All subcutaneous tissue and excess skin were excised, Scarpa’s and Camper’s fascia were approximated with 0-polyglactin 910 suture, staples were used to close the skin, and a negative pressure wound therapy device was placed.

Patients randomized to biologic mesh were treated with porcine acellular dermal matrix, whereas, patients randomized to synthetic mesh were treated with polypropylene mesh.

Outcomes

Clinical and patient-reported outcomes were assessed 3 years post-operatively. The primary outcome included major complications, a composite of mesh infections, hernia recurrences, reoperations, and deaths. Hernia recurrence was diagnosed through clinical assessment by a surgeon blinded to the randomization arm and confirmed by a computed tomography (CT) scan. Reoperation was defined as any procedure involving the fascia, mesh, or intra-peritoneal cavity. Secondary outcomes included surgical site infection, which was defined per the Centers for Disease Control and Prevention, surgical site occurrence (seroma, wound dehiscence, or hematoma), and patient-reported outcomes.¹⁸ Patient-reported outcomes included the patient’s functional status (measured using the modified Activities Assessment Scale) and a validated, hernia-specific, 12-question survey, with each question using a 10-point Likert scale.^19,20 Functional status was determined by totaling the score of the 12 questions, then normalizing the score to 100 (100 is perfect function, whereas, 1 is poor function). The minimal clinically important difference (MCID) was 5 points for a minor change and 15 points for a major change.²⁰ The remaining patient-reported outcome included pain (measured using the Visual Analog Scale), with 10 indicating extreme pain and 0 meaning no pain. One point was the MCID.²¹

Statistical Analysis

Intention-to-treat analyses were conducted using a modified Poisson regression model using a robust sandwich estimator.²² The benefit of this model is that it can estimate the relative risk rather than the odds ratio. Quantitative variables were analyzed using the Student’s t-test or the Mann-Whitney U test and are reported as mean plus or minus the standard deviation for normally distributed variables or as median (interquartile range) for non-parametric variables. Categorical variables were compared using Chi-square tests. For clinical outcomes with rare (eg, death) or no events reported, Fisher’s exact test was performed instead. Statistical significance was defined as P < .05. Patient-reported outcomes were analyzed using analysis of covariance (ANCOVA), adjusting for baseline patient-reported outcomes. All statistical analyses were performed using Stata (version 16.0).

Results

Between March 2017 to July 2019, 87 patients were randomized: 44 to biologic mesh and 43 to synthetic mesh (Figure 1). Groups were similar at baseline. Most patients were female and Hispanic, with a body mass index (BMI) > 30 kg/m². Most hernias were incisional hernias and medium or large in size (Tables 1 and 2). All repairs were performed with a retromuscular mesh placement. Three patients, 2 in the synthetic mesh group and 1 in the biologic mesh group, had a hernia too wide for primary fascial closure. The posterior or anterior fascia, or both, were bridged with a hernia sac. All 3 patients had mesh placed in the retromuscular position.

Figure 1.

The CONSORT flow diagram shows the inclusion and exclusion criteria used in the study.

Table 1.

Baseline Demographics

	Biologic mesh (n=44)	Synthetic mesh (n=43)	P value
Age (years), mean ± SD	51±9.9	51±12.1	.999
Sex, female	23 (52%)	26 (60%)	.519
Race/ethnicity			.917
Hispanic	28 (64%)	30 (70%)	-
Black	9 (21%)	7 (16%)	-
White	4 (9%)	4 (9%)	-
Other	3 (7%)	2 (5%)	-
Body mass index (kg/m2), mean ± SD	31.8±6.4	32.9±6.1	.414
Recent smoker	5 (11%)	3 (7%)	.713
Diabetes mellitus	15 (34%)	10 (23%)	.345
ASA score
1	0 (0%)	4 (9%)	-
2	23 (52%)	13 (30%)	-
3	20 (45%)	26 (60%)	-
4	1 (2%)	0 (0%)	-
Hernia type			.739
Primary	6 (14%)	4 (9%)	-
Incisional	38 (86%)	39 (91%)	-
Prior ventral hernia repair	18 (41%)	17 (40%)	.999
Hernia width, median (IQR)	6.0 (3.1, 8.9)	6.0 (4.0, 8.3)	.822
Small (<4 cm)	12 (27%)	10 (23%)	.597
Medium (4–10 cm)	22 (50%)	26 (61%)	-
Large (>10 cm)	10 (23%)	7 (16%)	-

Abbreviations: SD = standard deviation; ASA = American Society of Anesthesiologists; IQR = interquartile range

Table 2.

Interoperative Details

	Biologic mesh (n=44)	Synthetic mesh (n=43)	P value
Wound class 2–4	29 (66%)	30 (70%)	.819
VHWG 2–4	44 (100%)	43 (100%)	.999
Component separation	14 (32%)	18 (42%)	.379
Unilateral PCS	5 (11%)	11 (26%)	-
Bilateral PCS	9 (20%)	6 (14%)	-
Bilateral ACS	0	1 (2%)	-
Concomitant procedures	21 (48%)	23 (54%)	.670
Gastrointestinal	11 (25%)	7 (16%)	-
Gynecologic	4 (9%)	9 (21%)	-
Hepatobiliary	5 (11%)	5 (12%)	-
Urologic	1 (2%)	1 (2%)	-
Inguinal hernia repair	0	1 (2%)	-
Primary fascial closure	43 (98%)	41 (95%)	.999
Retromuscular mesh placement	44 (100%)	43 (100%)	.999
Operative duration (minutes) mean ±SD	183 ±86	177 ±97	.761

Abbreviations: VHWG = Ventral Hernia Working Group classification; SD = standard deviation; PCS = posterior component separation; ACS = anterior component separation

Sixty-one patients (70%) completed follow-up at 3 years. The median (interquartile range) follow-up time for all patients was 2.9 (2.3, 3.1) years.

Fourteen patients (50%) randomized to biologic mesh experienced a major complication compared to 10 patients (30%) in the synthetic mesh group (relative risk [RR] 2.34, 95% confidence interval [CI] 0.80 to 6.90, P = .123) (Table 3). In the biologic mesh arm, 11 patients (39%) had a hernia recurrence, 4 (14%) had a mesh infection, 5 (18%) required reoperation, and 1 (4%) died. Reasons for reoperation included 3 for mesh infection and 2 repairs for recurrence. In the synthetic arm, 8 patients (24%) had a hernia recurrence, 1 (3%) had a mesh infection, and 4 (12%) required reoperation. Reasons for reoperation included 2 repairs for recurrence, 1 for small bowel decompression secondary to small bowel obstruction, and 1 for mesh infection.

Table 3.

Clinical Outcomes

	Biologic mesh (n=28)	Synthetic mesh (n=33)	Relative risk (95% CI)	P value
Follow-up duration, median (IQR)	2.92 (2.49, 3.03)	2.91 (2.26, 3.16)	-	-
Major complication	14 (50%)	10 (30%)	2.34 (0.80 to 6.90)	.123
SSI	4 (14%)	2 (6%)	2.61 (0.43 to 15.7)	.296
SSO	11 (39%)	10 (30%)	1.41 (0.49 to 4.04)	.527
Seroma	5 (18%)	4 (12%)	1.58 (0.37 to 6.70)	.537
Hematoma	2 (7%)	2 (6%)	1.19 (0.15 to 9.15)	.871
Wound dehiscence	6 (21%)	5 (15%)	1.52 (0.40 to 5.81)	.535
Mesh infection	4 (14%)	1 (3%)	5.43 (0.56 to 52.53)	.144
Hernia recurrence	11 (39%)	8 (24%)	2.05 (0.66 to 6.33)	.214
Reoperation	5 (18%)	4 (12%)	1.58 (0.38 to 6.55)	.531
Mortality	1* (4%)	0 (0%)	-	.459

^*Patient with locally advanced hepatocellular carcinoma, died of sepsis secondary to Enterococcus faecalis bacteremia

Abbreviations: CI = confidence interval; SSI = surgical site infection; SSO = surgical site occurrence; IQR = interquartile range

No statistically significant differences between the 2 arms were seen among secondary outcomes. Four patients (14%) presented with a surgical site infection in the biologic arm compared to 2 patients (6%) in the synthetic arm (RR 2.61, 95% CI 0.43 to 15.7, P = .296). Eleven patients (39%) randomized to biologic mesh had a surgical site occurrence, whereas 10 (30%) patients had the same occurrence in the synthetic group (RR 1.41, 95% CI 0.49 to 4.04, P = .527).

Similarly, no statistical differences were seen in modified Activities Assessment Scale scores (Table 4). Both arms reported clinically significant improvements in functional status at 3 years, as noted by increases in scores well above the MCID of 5 points. For patients who received biologic mesh during their repair, the median functional status score improved by 42.6 points from baseline, whereas those in the synthetic mesh group improved by 51.0 points (median difference −11.4, 95% CI −25.7 to 3.0, P = .117). In addition, patients in both arms reported clinically important improvements in pain scores at 3 years with no differences between groups.

Table 4.

Clinical Outcomes

	Biologic mesh (n=28)	Synthetic mesh (n=33)	Median difference (95% CI)	P value
Functional status at baseline*	32.4 (13.0, 50.7)	37.0 (11.1, 65.7)	-	-
Functional status at three years*	75.0 (19.4, 90.7)	88.0 (62.5, 100.0)	−11.4 (−25.7 to 3.0)	.117
Pain at baseline*	7 (5, 9)	7 (4, 8)	-	-
Pain at three years*	3 (1, 8)	1 (1, 5)	1.4 (−0.44 to 3.3)	.131

^*reported as median (interquartile range)

Abbreviation: CI = confidence interval

Discussion

In this follow-up at 3 years, we failed to find benefits with biologic mesh as compared to synthetic mesh in complex open VHR for these 2 commonly used mesh types. In fact, biologic mesh was associated with more major complications (50% vs 30%, RR=2.34 [95% CI=0.80 to 6.90], P = .123) at 3 years. Although none of our clinical complication rates reached statistical significance, the percentage of each complication was higher in the biologic mesh group than the synthetic group, suggesting biologic mesh may have worse outcomes than synthetic.

As our study was designed as a pilot, it was not powered to assess differences in clinical complication rates at 3 years. It is possible that with a higher sample size, some of the clinical complications would reach statistical significance in favor of synthetic mesh. However, the findings in this study remain hypothesis-generating. A large-scale study powered to assess these clinical outcomes should be performed.

Despite this limitation, our findings contribute to the literature, as only 3 other RCTs and 2 meta-analyses have been performed to date on this topic.^{13,14,16,23,24} Additionally, our findings at 3 years are analogous to other studies, including our primary analysis, which demonstrated no short-term benefit with biologic mesh.¹⁵ Our rates of recurrence in the 2 groups are similar to other recently published RCTs. The PRICE trial (Preventing Recurrence in Clean and Contaminated Hernias Using Biologic Versus Synthetic Mesh in Ventral Hernia Repair) was an RCT assessing biologic versus synthetic mesh at 2 years and found a higher risk of recurrence with biologic than with synthetic mesh (39.7% vs 21.9%, P = .035).¹⁶ Similarly, Rosen et al¹³ also conducted an RCT with a 2-year follow-up, showing a higher rate of recurrence in the biologic mesh group. In addition, several observational studies have demonstrated similar findings of high recurrence rates and poor wound complication profiles.^6,7,11,12

Long-term recurrences are common after VHR, and each subsequent repair leads to a higher rate of recurrence, creating a vicious cycle. Additionally, patients will multiple repairs are more likely to undergo reoperation and develop a surgical site infection.² As there is a growing body of evidence showing biologic mesh has a higher rate of recurrence, using synthetic mesh over biologic mesh for most circumstances would decrease the rate of complications after repair. Additionally, repair with biologic mesh has previously been demonstrated to cost more than repair with synthetic mesh.²⁵ At our hospital, a 30 x 30 cm biologic mesh costs in excess of $23 000, whereas, the same size synthetic mesh costs less than $200.¹⁵ As 400 000 VHR are performed annually, using biologic mesh only when indicated would decrease US healthcare expenditures.¹⁶ Future research should further elucidate under which conditions biologic mesh should be used. This study confirmed the importance of investigating potential alternative methods of repair for complex ventral hernias, given the high complication rates associated with current techniques. However, great care must be taken to ensure the adoption of safe techniques.

There are several limitations to this study. First, this study is a follow-up from a small pilot trial and is under-powered. However, it contributes to the sparse literature on long-term follow-up. Second, our loss to follow-up at 3 years was not insignificant. Despite multiple attempts to reach patients, 26 were unable to be reached for follow-up and were excluded from the outcomes analysis. This lack of follow-up was likely partly due to the impact of the COVID-19 pandemic. Third, as this study focused on a single center from a safety-net hospital, the generalizability of our results may be limited. However, at least one-third of Americans are uninsured or underinsured, and the participating surgeons were high-volume surgeons with a published history of good outcomes.^26–28 It is presumable that outcomes among surgeons with a lower volume would likely be worse. Fourth, we studied 2 brands of mesh. Although we used common mesh brands, outcomes may not be generalizable to other types of mesh.

Conclusion

Patients undergoing complex open VHR continue to suffer from high complication rates. Despite the premise of biologic mesh to overcome such complications, our follow-up demonstrates no benefit in outcomes with the use of biologic as compared to synthetic mesh. However, this outcome may not apply to all mesh brands. This trial adds to a growing body of trials demonstrating a lack of long-term superiority of biologic mesh over synthetic mesh in the repair of open ventral hernias.

Funding Statement

This research was supported (in whole or in part) by HCA Healthcare and/or an HCA Healthcare-affiliated entity.