The “Flat Diaphragm”: Does the Degree of Curvature of the Diaphragm on Postoperative X-Ray Predict Congenital Diaphragmatic Hernia Recurrence?

Introduction

Infants with congenital diaphragmatic hernia (CDH) often experience significant morbidity related to the natural history of the disease, including neurodevelopmental delays, pulmonary insufficiency, and gastrointestinal pathologies.^1,2 However, recurrence remains one of the most dreaded complications after CDH repair. The reported incidence of recurrence ranges from 3% to 50% in the literature depending on the population being examined.^3–5 Several studies have examined factors associated with recurrence, including the need for extracorporeal membrane oxygenation (ECMO), minimally invasive operative approach, large defect size, and use of a patch repair.^6–8 In addition to these operative factors, surgeons often cite a tension-free repair as a key tenet to preventing postoperative recurrence. While this principle resonates throughout many, if not all, surgical procedures, objectively studying the “tension on” or “tightness of” a repair is challenging.

While in the operating room, the tension of a CDH repair is assessed visually by the surgeon to achieve a domed shape to the diaphragm or patch repair with sufficient overlap at the circumferential margin of the repair. While this is largely a subjective assessment, there is no real objective measure for determining the tightness of a repair intraoperatively or postoperatively. Despite this, surgeons often refer to the appearance of the diaphragmatic curvature on postoperative chest X-ray (CXR) as a surrogate for this measure with a “flat-appearing” diaphragm indicating a tighter repair with a higher associated risk of recurrence. The purpose of this study was to examine the relationship between the flatness of the diaphragm on postoperative CXR as measured by an objective instrument, the rib level of the lateral insertion of the diaphragm, and recurrence.

Materials and Methods

After approval by the Children's Healthcare of Atlanta Institutional Review Board (IRB # 16-116), a retrospective review was performed for all patients who underwent surgical repair of a CDH between January 1, 2004 and July 31, 2015 at a single, quaternary pediatric hospital. Patients with right-sided hernias (n = 37) and left-sided Morgagni hernias (n = 7), and cases with incomplete records (n = 1) were excluded. The final cohort consisted of consecutive patients with left-sided, Bochdalek hernias who underwent surgical repair. Demographics and operative details were abstracted from the medical record. The primary outcome of interest was CDH recurrence requiring surgical repair.

To objectively assess the curvature of the diaphragm on postoperative CXR, we developed a mathematical equation designed to calculate the tension of the diaphragmatic repair. Let τ equal the diaphragmatic curvature index where \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \tau = { \frac { tangential \;length } { peak \;length } } \end{align*} \end{document}

The tangential length of the diaphragm was designated as the straight-line distance between the medial and lateral insertions of the diaphragm as measured on anteroposterior CXR. The peak length was designated as the perpendicular, straight-line distance from the highest point along the diaphragm's curve to the tangential length line. A larger diaphragmatic curvature index value was indicative of a flatter appearing diaphragm on CXR (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Diagram of measurements obtained from postoperative day 0 chest X-rays, where the median diaphragmatic curvature index = tangential length/peak length.

A single pediatric radiologist (J.L.) reviewed postoperative CXRs for all patients in the cohort. When available, the immediate postoperative CXR was used for analysis. In the case that there was no postoperative day (POD) 0 CXR in the medical record, POD 1 films were used. Using the measurements function in the picture archiving and communication system imaging viewer, J.L. performed measurements for the tangential length and peak length, and τ was calculated for all patients. In addition, the level of the lateral rib insertion of the diaphragm was determined to determine if this factor had any relevance to recurrence or tightness of the diaphragmatic repair.

Demographic and clinical characteristics were summarized for the study population using counts and percentages for categorical variables and median and interquartile ranges (IQRs) for continuous variables. Characteristics were compared between patients with recurrence and patients without recurrence using chi-square tests for categorical variables and Wilcoxon rank-sum tests for continuous variables. Statistical analysis was performed using SAS 9.4 (Cary, NC).

Results

A total of 127 patients underwent surgical repair for a left-sided, Bochdalek hernia during the 11.5-year study period. Patient characteristics and operative details are summarized in Table 1. The overall recurrence rate was 21.3% (n = 27), and the mortality rate was 7.1% (n = 9). There was no significant difference in recurrence rates based on the surgical approach (laparotomy versus thoracotomy versus thoracoscopic) or type of repair (primary versus patch). The overall median time of follow-up was 1.9 years (IQR 0.7–4.1). The median peak length measurement in the recurrence and no recurrence groups was 7.40 mm [IQR (5.78, 9.43)] and 8.30 mm [IQR (5.10, 10.05)], respectively (P = .610). The median tangential length measurement in the recurrence and no recurrence groups was 47.80 mm [IQR (42.65, 52.15)] and 48.20 mm [IQR (43.35, 55.65)], respectively (P = .373). There was no difference in the calculated diaphragmatic curvature index (τ) or median lateral rib insertion level in patients with and without recurrence. There were no significant differences in gender, age at repair, ECMO requirement, or days to extubation in the recurrence and no-recurrence groups. The recurrence group had significantly longer follow-up than the no-recurrence group (Table 1).

The calculated diaphragmatic curvature index (τ) for each surgical approach category (laparotomy, thoracotomy, and thoracoscopic) and type of repair category (primary and patch) were compared, and there was no significant difference (Table 2). Within the laparotomy group and the thoracotomy group, there was no significant difference in the median diaphragmatic curvature index (τ) between patients with a recurrence and those without a recurrence. Among patients in the thoracoscopic repair group, patients with a recurrence had a median diaphragmatic curvature index (τ) equal to 5.66, whereas the no-recurrence group had a median diaphragmatic curvature index (τ) of 6.98. This difference was approaching significance with a P value of .075. Among all patients, the median lateral rib insertion level ranged from rib 7 to rib 10. There was no significant difference in the median lateral rib insertion level based on operative approach or repair type.

The median time to recurrence was 248 days [IQR (156, 364)]. Twenty patients had a single recurrence, while five patients had two recurrences, one patient had three recurrences, and one patient had four recurrences. The median diaphragmatic curvature index (τ) among patients with multiple recurrences was 7.98 [IQR (7.29, 9.67)], which was not significantly different from patients with a single recurrence.

Discussion

Recurrence is one of most feared complications following CDH repair. While several patient-level and operative factors have been identified as risk factors for recurrence, it is intuitive that a repair that is under tension is more likely to result in a recurrence. Outside of visual inspection by the surgeon while in the operating room, there is no way to objectively assess the tightness of a CDH repair. However, surgeons often use postoperative CXR images to evaluate the shape of the diaphragm after repair with a flatter diaphragm felt to represent a tighter repair. We developed a tool to objectively measure the degree of curvature of the diaphragmatic repair on postoperative CXR. Our results failed to show a difference in the calculated diaphragmatic curvature index (τ) or the lateral rib insertion level of the diaphragm in patients who experienced recurrence and those with no recurrence. In addition, there was no difference in either measurement based on the type of repair or the operative approach. These findings indicate that the appearance of the diaphragm on postoperative CXR may not be predictive of recurrence as previously thought.

Although the conclusions were variable, several studies have examined the relationship between preoperative CXR findings and clinical outcomes in infants with CDH.^9–11 To our knowledge, there are no studies investigating the relationship between postoperative CXR findings and the incidence of recurrent CDH. Studies have repeatedly demonstrated an association between the use of a diaphragmatic patch and recurrence.^6,12–14 Although patch repairs have intrinsic properties, which may be related to recurrence (i.e., inability of the synthetic material to grow with the patient, poor tissue incorporation, and lack of contractile function), this type of repair may be a surrogate for a larger defect—the innate characteristic associated with recurrence. ⁸ In support of this, Lally et al. developed a standardized classification schema for reporting defect size and using the Congenital Diaphragmatic Study Group demonstrated that all defects categorized as “D” (the largest defect type) required a patch repair. ¹⁵

When faced with a large defect, the operating surgeon must decide whether a tension-free primary repair is feasible or a patch repair is necessary. When a patch repair is undertaken, a loose patch with careful fixation to the diaphragm and/or thorax is the goal. ⁸ To prevent recurrence after patch repair, two different groups have proposed the use of a dome- or cone-shaped patch.^5,16 Although the technique used to achieve the desired patch shape in these studies was slightly different, both studies reported lower recurrence rates with the use of a dome- or cone-shaped patch when compared to traditional patch implantation techniques. Both groups attributed their results to the additional space in the abdominal cavity that the dome/cone shape provides, which allows for a tension-free suture line. Neither study commented on the appearance of the postoperative CXR when these alternative techniques are used, but based on our results, this finding likely does not have any bearing on the risk of recurrent herniation. Rather, the judgment of the surgeon in regard to the appearance of the repair and apparent lack of tension in the operating room is the factor that holds significance when discussing recurrence.

Our results suggest that there is no relationship between the appearance of the diaphragm on postoperative CXR and recurrence, namely that, there was no difference between the median diaphragmatic curvature index (τ) in patients who experienced a recurrence and those who did not experience a recurrence. This is not to suggest that the presence or absence of tension on a repair has no influence on recurrence, but that the appearance of the diaphragm on postoperative CXR does not correlate with tension. There was no difference in median diaphragmatic curvature index (τ) between the two types of repair (primary versus patch), and among the patch group, there was no difference in the median diaphragmatic curvature index (τ) among patients with and without recurrence. Whether these findings can be interpreted as there being no difference in the actual physical tension of the repair is debatable. However, it seems logical to conclude that the appearance of the curvature of diaphragm on postoperative CXR is not different among these different groups and cannot be used as a reliable predictor of recurrence. Postoperative CXRs should be obtained and used as a reference if recurrence is suspected at any point after repair. However, the utility of a CXR for assessing the tension of a repair postoperatively is less convincing.

This study is limited by its retrospective nature and the relatively small sample size. A single radiologist (J.L.) performed all measurements, so variability in measurement technique should be considered low for this study. To avoid bias, the radiologist was blinded to the patients' final outcome at the time of measurement. A major weakness of this study is the inability to test the validity of the median diaphragmatic curvature index (τ) as a measure of actual physical tension. This would require intraoperative measurements of force acting on the diaphragm to allow for comparison with calculated diaphragmatic curvature indices. The authors are unaware of a method for obtaining such measurements and this is beyond the scope of this study. Rather, the median diaphragmatic curvature index (τ) was designed to measure the flatness of the diaphragm, where a flatter appearance on CXR is presumed to represent a tighter repair.

Conclusions

A flat appearance of the diaphragm on postoperative CXR as measured by the median diaphragmatic curvature index (τ) is not associated with recurrence. The shape (flat or curved) of the diaphragm or level of rib insertion on CXR after CDH repair may not be predictive of recurrence as previously thought.