Description

BACKGROUND AND RATIONALE Repair of a complex ventral hernia with synthetic mesh in the retromuscular (sublay) position, frequently combined with posterior component separation by transversus abdominis release, is widely regarded as the durable standard of care, but it remains associated with appreciable wound morbidity and a non-trivial recurrence rate. A persistent technical concern is the interface between the prosthesis and the abdominal viscera. When the posterior rectus sheath and peritoneum cannot be re-approximated to fully exclude the mesh from the peritoneal cavity, the prosthesis may be left in contact with bowel, predisposing to dense adhesions, mesh erosion, chronic pain and, rarely, enterocutaneous fistula. Strategies to interpose a protective layer - omentum, composite or biologic meshes, or absorbable barriers - add cost or are not always feasible.

THE BONY-EDGED COMPLEX VENTRAL HERNIA A distinct and under-studied subgroup comprises hernias in which at least one margin of the fascial defect is formed by bone rather than musculo-aponeurotic tissue. These include subxiphoid and costal-margin hernias (bounded by the xiphoid and the costal cartilages), suprapubic hernias (bounded by the pubic symphysis and the superior pubic rami), and selected iliac and lumbar hernias (bounded by the iliac crest). The bony edge poses two specific problems. First, adequate prosthetic overlap and secure fixation on the bony side are difficult, because mesh cannot be sutured to fascia that does not exist. Second, the dissection plane that would normally carry the posterior layer is interrupted at the bone, so the peritoneal layer frequently cannot be closed beneath the mesh. The result is a region where the mesh is simultaneously most difficult to fix and least likely to be covered, concentrating the risk of recurrence and of mesh-related visceral complications precisely at the bony edge.

THE TECHNIQUE UNDER EVALUATION In conventional practice the hernial sac is regarded as redundant tissue and is excised. The technique evaluated in this trial reconceives the sac as a vascularized autologous resource. Rather than excising it, the surgeon preserves a broad, pedicled flap of the hernial sac and rotates it deep to the prosthesis, suturing it as a continuous neo-peritoneal layer that covers the visceral surface of the mesh and bridges the bony edge of the defect. This achieves three aims with a single autologous structure: it interposes living, well-vascularized tissue between the mesh and the bowel, plausibly reducing adhesions, mesh erosion and fistula; it provides soft-tissue coverage and a fixation buttress at the bony margin where fascial purchase is impossible; and it does so without additional prosthetic material, donor-site morbidity or cost. The technique has been applied in selected cases with encouraging operative impressions, but no controlled comparison exists. Equipoise exists: conventional sac excision remains standard practice and is not known to be inferior. A randomized trial is therefore both ethical and necessary before wider adoption.

TRIAL DESIGN This is a single-center, parallel-group, two-arm, individually randomized superiority trial with 1:1 allocation. Both arms share an identical operative approach up to the point of mesh placement, so that the contrast between them is confined to the peritonealization step. Standardization is reinforced by a written operative manual, a pre-trial surgeon training and credentialing phase, a case report form completed in theater, and intra-operative photographic documentation reviewed centrally by the steering group. All index operations are performed by, or under the direct supervision of, credentialed consultant surgeons experienced in retromuscular mesh repair and posterior component separation.

Common operative steps in both arms are: general anesthesia with institutional antibiotic and venous thromboembolism prophylaxis and standardized skin preparation; a midline or location-appropriate incision, dissection to the hernial sac, and full reduction of contents with adhesiolysis as required; definition of the fascial defect including the bony margin, with measurement of defect width and length and recording of the European Hernia Society location and size class; development of the retromuscular (retrorectus) plane, with posterior component separation by transversus abdominis release performed when required to achieve tension-free closure and applied by identical indications in both arms; and placement of a flat, medium-weight monofilament polypropylene mesh of standardized type in the retromuscular position with wide overlap and standardized suture fixation, with anterior fascial closure of the linea alba where achievable.

RANDOMIZATION, ALLOCATION CONCEALMENT AND BLINDING The allocation sequence is computer-generated by the trial statistician using random permuted blocks of variable size with 1:1 allocation, stratified by hernia location (subxiphoid/costal versus suprapubic versus iliac/lumbar) and by defect width (less than 10 cm versus 10 cm or more), because both are expected to influence the primary outcome. Block sizes are not disclosed to recruiting or operating staff. Allocation is concealed using a central, secure randomization service that releases the assignment only after the participant identifiers and stratum are entered and eligibility is confirmed; sequentially numbered, opaque, sealed envelopes prepared by the statistician, who has no clinical contact with participants, serve as a validated fallback.

Participants consent pre-operatively, but allocation is requested and revealed only intra-operatively, after the surgeon has reduced the hernia, confirmed the bony-edged defect and verified that the sac is adequate to raise a covering flap. This timing guarantees that randomization is applied only to participants who genuinely meet the anatomical criteria, eliminates pre-operative selection bias, and ensures that the two arms differ only from the point of mesh placement. The trial coordinator, not the operating surgeon, contacts the randomization service and documents the assignment.

Because the intervention is a surgical step, the operating surgeon cannot be blinded. All other parties are blinded: the participant, ward and clinic staff, the independent outcome assessors who perform the wound and recurrence evaluations, the radiologist who reads the follow-up imaging, and the statistician, who analyzes arms coded A and B. Operative notes that would reveal allocation are stored separately from the assessment record, and a standardized, identical dressing and drain regimen is used in both arms so that early wound appearance does not unblind assessors. Individual unblinding is permitted only when knowledge of allocation is essential for safe clinical management, is recorded with its reason, and is reported to the Data Safety Monitoring Board; the statistician remains blinded until database lock.

SAMPLE SIZE The sample size is driven by the primary outcome. In complex ventral-hernia repair, reported surgical site occurrence rates commonly fall in the 30 to 45 percent range, and a control-arm rate of 38 percent is assumed. A clinically meaningful absolute reduction to 18 percent is hypothesized for the intervention. For the comparison of two independent proportions at a two-sided significance level of 0.05 with 80 percent power, approximately 76 evaluable participants per arm are required. To preserve power against up to 15 percent attrition over 12 months, the target is inflated to 90 per arm, giving a total enrollment of 180 participants.

STATISTICAL ANALYSIS The primary analysis is by intention to treat, with all randomized participants analyzed in the arm to which they were allocated. The 30-day composite surgical site occurrence is compared between arms using a generalized linear mixed (logistic) model with treatment as the fixed effect of interest and the stratification factors as covariates. The treatment effect is reported as an adjusted odds ratio with a 95 percent confidence interval and a two-sided p-value, together with the absolute risk difference, the relative risk and the number needed to treat. A per-protocol analysis is performed as a sensitivity analysis, and a safety set analyzed as treated is used for harms. Binary secondary outcomes are analyzed analogously; recurrence is analyzed by Kaplan-Meier estimation and Cox regression adjusted for the stratification factors; continuous outcomes are analyzed by linear models, or by repeated-measures mixed models for outcomes measured longitudinally. Pre-specified subgroup analyses by hernia location, defect width and use of transversus abdominis release are exploratory and are tested for interaction. Missing data are handled under a missing-at-random assumption, with multiple imputation as a sensitivity analysis. One pre-planned interim analysis is performed at 50 percent recruitment using an O'Brien-Fleming boundary, and the significance level at the final analysis accounts for the alpha already spent. A full statistical analysis plan is finalized and signed before database lock and unblinding.

FOLLOW-UP AND SAFETY Participants are assessed at discharge and at 30 days, 90 days, 6 months and 12 months, with visit windows of plus or minus 7 days for the 30- and 90-day visits and plus or minus 14 days for the 6- and 12-month visits. Adverse events are solicited and recorded from the time of surgery through 12 months and are characterized by severity, seriousness, expectedness and relationship to the intervention, and graded by the Clavien-Dindo classification. Anticipated events of special interest include surgical site infection, seroma (including intra-sac seroma in the intervention arm), hematoma, wound dehiscence, mesh-related complications (exposure, infection, erosion, fistula), sac-flap necrosis, recurrence, reoperation and death. Serious adverse events are reported to the principal investigator, the research ethics committee and the Data Safety Monitoring Board within Good Clinical Practice timelines.

The trial is conducted in accordance with the Declaration of Helsinki and ICH Good Clinical Practice, was designed following the SPIRIT 2013 statement, and will be reported in accordance with the CONSORT 2010 statement. Results will be reported in full regardless of direction, and the registry record will be updated with summary results.