Biologic Mesh Underlay in Thoracoscopic Primary Repair of Congenital Diaphragmatic Hernia Confers Reduced Recurrence in Neonates: A Preliminary Report

Introduction

Surgical correction of congenital diaphragmatic hernia (CDH) in neonates has undergone a significant evolution for the past two decades. Although classically approached through laparotomy, improvement in surgical equipment and technique now allows a minimally invasive thoracoscopic approach to this defect. Initially described primarily in older children with late presentation of Bochdalek-type CDH and in Morgagni-type hernias, thoracoscopic repair (TR) of Bochdalek-type CDH has increasingly gained acceptance in select neonates.^1–3 After initial reports, Arca et al. reported increased complication/recurrence with TR. ⁴ Subsequent reports in neonates have had mixed results with regarding feasibility and recurrence rates when compared with open techniques.^5–10 In most of these reports, TR was attempted only in neonates where primary repair (PR) was a possibility.

Although both physiological and radiographic predictors of inability to primarily repair CDH using TR have been reported, factors influencing recurrence of neonatal Bochdalek-type CDH that is amenable to PR are not as well described.^8,11,12 In addition, Keijzer et al. have endorsed a low threshold to utilize nonabsorbable mesh to reduce hernia recurrence, but the use of bioprosthetic (biologic) mesh as an adjunct to primary closure has not been well described. ¹³

Within our institution, TR has been completed in select neonates for >15 years. During this period, use of a biologic mesh as reinforcement of a primarily repaired TR has been used in a subset of these patients. Most frequently the mesh is used as an underlay (i.e., abdominal side of CDH repair). The purpose of this study is to report a single institution's experience with biologic underlays as a means to reduce 5 year CDH recurrence in neonates undergoing primary TR when compared with nonreinforced primary TR.

Methods

The operative registry from a single institution (Egleston Children's Hospital, Children's Healthcare of Atlanta, Atlanta, GA) was queried for all patients undergoing repair of a Bochdalek-type CDH utilizing TR from January 1, 2003 to June 31, 2013. Children were excluded from the study cohort if PR was not completed, if TR was converted to an open technique, and/or if a biologic overlay was used to reinforce the PR. Technique and indications for TR, including the use of small intestine submucosa (SIS; Cook Biotech, Inc., West Lafayette, IN) biologic mesh as an underlay, at our institution have been previously reported. ¹⁴

In brief, after the abdominal visceral from the hemithorax are reduced, the biologic mesh is introduced through the chest and then placed on the abdominal side of the defect and unfurled. Typical closure at our institution uses a medial to lateral approach incorporating the biologic mesh in an interrupted silk suture line. A triangular stitch is used at the farthest lateral recess of the costophrenic angle, the needle is passed through a stab wound in the skin, through the intercostal space and both leaflets of diaphragm, and then back up through the skin incision (passing around the opposite site of the nearest rib).

Children were also excluded if they did not have documented follow-up with the surgeon of record or an identifiable physician within the scope of Children's Healthcare of Atlanta extending until at least 5 years after their index procedure.

Identified patient charts were then reviewed for multiple clinical and surgical variables, including age at time of surgery (days), gender, side of CDH defect (i.e., left versus right), presence of hernia sac associated with CDH, need for extracorporeal membrane oxygenation (ECMO), use of biologic mesh underlay, presence of hernia recurrence/time to recurrence, method of detecting the recurrence (i.e., clinically or radiographically), and reported difficulty with PR. Repairs were deemed “difficult” if the operating surgeon reported more than minimal tension associated with the closure or explicitly stated the repair was difficult in the operative note. Patients were then categorized according to whether a biologic mesh underlay was used in the initial repair. The primary outcome was 5-year cumulative initial hernia recurrence.

Statistical analysis was completed utilizing Statistical Package for the Social Sciences (SPSS, Inc., Chicago, IL) version 25 for Windows. Results are presented as medians with interquartile ranges or as percentage of study populations, where applicable. Mann–Whitney U tests were used for continuous variables and chi-squared tests were used for qualitative/categorical variables. A P value ≤.05 was considered statistically significant.

Results

Within the study period, 46 patients were identified who underwent a TR with either an exclusively PR or a PR with biologic mesh underlay (MR). All patients in the MR group utilized SIS biologic mesh. No other biologic meshes have been used at our institution. Three patients were lost to follow-up, leaving 43 patients to comprise the study cohort. All 3 of these patients were PR. Among the study cohort, 28 (65.1%) patients underwent PR and 15 (34.9%) underwent MR. There were no significant differences in gender, need for ECMO, day of life of repair, side of CDH, presence of hernia sac, or reported difficulty closing the defect when comparing PR and MR (Table 1). All patients except for three were diagnosed prenatally or within the first 24 hours of life. These 3 patients were diagnosed at 2 weeks, 1 month, and 3 months of age. The study cohort had an overall low requirement for ECMO, with 1 patient in each category requiring support before their TR.

Among patients undergoing PR, 3 (10.7%) had an initial recurrence within the 1st year of TR (Table 2). Three additional PR patients had an initial recurrence within 3 years of TR (21.4%) and no additional recurrences were observed in the 5-year follow-up. One patient had his recurrences repaired through laparotomy (both with SIS), 2 were repaired through thoracotomy (both without mesh), 2 were repaired laparoscopically (both with SIS), and 1 thoracoscopically (with SIS). One patient who underwent thoracotomy for treatment of recurrent CDH had two subsequent recurrences. No additional multiple recurrences were noted within the study cohort. Five of the six recurrences in the PR group were detected after the patient developed increased work of breathing and one was detected on a screening chest X-ray.

In the 15 patients who underwent MR, one recurrence was noted within the 1st year of TR (6.7%) and no additional recurrences were found in the 5-year follow-up (Table 2). The one noted recurrence, reported to be 4 mm in size, was found and repaired primarily laparoscopically while completing a fundoplication 4 months after the index procedure. This defect was both radiographically and clinically silent. Statistical analysis revealed no significant difference in 5-year initial hernia recurrence rates when comparing MR with PR.

Discussion

This study reports our institution's initial experience with utilizing biologic mesh underlay to reduce 5-year initial hernia recurrence in the setting of thoracoscopic PR of CDH. Biologic mesh underlay was then compared against patients who underwent PR only. Within the study cohort, there was a reduced rate of hernia recurrence in the patients who underwent MR. However, this difference did not reach statistical significance. This is likely due to several factors, two of which are mentioned. First, the relatively small sample size within each subgroup makes achieving significance on any clinical variable difficult. Second, the reported recurrence within the MR cohort was based upon a small defect that was clinically/radiographically silent and found only due to repeat laparoscopy performed for another indication. If this patient was removed from analysis, the difference in recurrence rates approached statistical significance (i.e., P = .06). This result suggests that a larger sample size is needed for analysis and supports the need for ongoing evaluation of this surgical technique. For transparency and the inability to project likelihood of this defect becoming clinically/radiographically apparent, this was reported as a recurrence. The true incidence and natural history of a small defect such as this is unknown.

Recent work by Criss et al. has suggested that TR of small- to moderate-sized defects was associated with decreased length of stay, time on mechanical ventilation, and time to full enteral feeds when compared with open techniques. ¹⁵ Although there was a trend toward increased hernia recurrence in the thoracoscopically repaired group in the study, this did not reach statistical significance. Although our study population is not identical in terms of defect size and type/techniques of mesh reinforcement, both cohorts represent a relatively low-risk population of children with CDH. Data points such as estimated gestational age/weight, ventilatory settings, and other physiological factors were not available for review for all patients; therefore, ability to successfully perform primary TR was used as a surrogate for a low-risk cohort but is imperfect in its comparison.

Kamran et al. has recently reported the use of a buttress/sandwich-type mesh reinforcement to reduce recurrence rates in TR. ¹⁶ Although this technique is inherently different, as the authors used a combination of permanent and bovine pericardium as their mesh reinforcement, this study is part of a growing body of evidence suggesting TR should include mesh reinforcement. Our institution's choice of mesh is based upon previous work in a porcine model as well as experience utilizing mesh underlay in larger/more complex thoracoscopic CDH repair. ¹⁷

Additional analysis correlating difficulty/tension in PR with MR versus PR approached but did not reach statistical significance (Table 1; P = .08). Identification of difficulty in primary closure for this study required the surgeon of record to report it within the operative note and is thus prone to reporting bias/error. All difficult repairs were based upon reporting more than minimal tension in the repair, and no other technical issue. Among all cases where difficulty was reported in closure only one recurrence was noted, which was in a patient who had undergone PR. Recognizing these factors, the increased rate of difficult closures within MR group would suggest that unless the biologic mesh provides protection, recurrence rates should be similar/higher. Until a more robust analysis can be completed with a larger study cohort, this is merely a postulation.

Conclusion

This study reports a preliminary experience with utilizing a biologic mesh underlay to reinforce the PR of CDH treated thoracoscopically. Although this study was not able to reach statistical significance, likely due to several factors, the reduction in recurrence warrants attention. Further research will be required to validate the findings of this study.