Pediatr Radiol DOI 10.1007/s00247-016-3688-6

ORIGINAL ARTICLE

Nuss bar migrations: occurrence and classification Lauren E. Binkovitz 1 & Benjamin Zendejas 2 & Christopher R. Moir 2 & Larry A. Binkovitz 1

Received: 5 April 2016 / Revised: 30 June 2016 / Accepted: 12 August 2016 # Springer-Verlag Berlin Heidelberg 2016

Abstract Background Pectus excavatum results from dorsal deviation of the sternum causing narrowing of the anterior-posterior diameter of the chest. It can result in significant cosmetic deformities and cardiopulmonary compromise if severe. The Nuss procedure is a minimally invasive technique that involves placing a thin horizontally oriented metal bar below the dorsal sternal apex for correction of the pectus deformity. Objective To identify the frequency and types of Nuss bar migrations, to present a new categorization of bar migrations, and to present examples of true migrations and pseudomigrations. Materials and methods We retrospectively reviewed the electronic medical records and all pertinent radiologic studies of 311 pediatric patients who underwent a Nuss procedure. We evaluated the frequency and type of bar migrations. Results Bar migration was demonstrated in 23 of 311 patients (7%) and occurred within a mean period of 26 days after surgery. Bar migrations were subjectively defined as deviation of the bar from the position demonstrated on the immediate postoperative radiographs and categorized as superior, inferior, rotation, lateral or flipped using a new classification system. Sixteen of the 23 migrations required re-operation. Conclusion Nuss bar migration can be diagnosed with careful evaluation of serial radiographs. Nuss bar migration has a wide variety of appearances and requires exclusion of pseudomigration resulting from changes in patient positioning between radiologic examinations. * Larry A. Binkovitz [email protected] 1

Department of Radiology, Mayo Clinic, 200 First St. S.W., Rochester, MN 55902, USA

2

Department of Surgery, Mayo Clinic, Rochester, MN, USA

Keywords Chest wall deformity . Children . Nuss bar . Pectus excavatum . Radiography

Introduction Pectus excavatum is the most common congenital chest wall deformity [1]. It has been found to be associated with numerous genetic conditions [2] but is most commonly an isolated idiopathic abnormality. It is more common in males [3] and is typically asymptomatic. However, when severe it can result in progressive cardiac dysfunction and decreased exercise tolerance [4–6]. Although often present in early childhood, it might not become clinically evident until adolescence, when the accelerated growth of puberty sometimes exacerbates a previously mild case [7–9]. It is characterized by a caved-in appearance of the chest that results from dorsal deviation of the sternum from overgrowth of costal cartilage [10] (Fig. 1). It can also result in a cosmetically important deformity, especially during adolescence, when it can lead to psychological distress related to body image [1, 10, 11]. The severity of the deformity can be quantified using the Haller index [12]. This is the ratio of the maximum transverse thoracic diameter over the minimum anteroposterior sternal-to-anterior vertebral body distance (Fig. 2). An index greater than 3.25 is considered a cutoff for surgical intervention at our institution. Surgical repair of pectus excavatum was first described by Ravitch [13] and included resection of the overgrown parasternal costochondral cartilage and preservation of the perichondrium. Because potential complications of the Ravitch procedure include persistent cartilage overgrowth or thoracic constriction if there is insufficient growth of the remaining cartilage, alternative surgical procedures were

Pediatr Radiol Fig. 1 Pectus excavatum in a 17year-old boy. a Clinical photograph shows deep dorsal displacement of the sternum with a severe caved-in appearance of the chest wall centrally. b Three-dimensional printed model viewed from below demonstrates the compression and displacement of the heart by the severe pectus deformity

developed. A minimally invasive surgery that did not require cartilage resection was pioneered by Nuss [11] in 2008. The Nuss procedure involves the subcostal insertion of a curved bar posterior to the apex of the dorsal sternal angulation. The outward pressure from the convex bar elevates the sternal depression and results in a more normal sternal configuration (Fig. 3). The bar can be anchored to the sternum, parasternal tissues or the lateral ribs in a variety of ways in order to reduce bar migration [1], and the bar is generally left in place 1–4 years. This procedure avoids some of the risks associated with the more invasive Ravitch procedure, in particular the risk of thoracic constriction, and the outcomes compare favorably to more invasive procedures. It is not without complications, however. The curved bar can become dislodged from its desired position off the apex of the sternal depression and require re-operation. The bar can become displaced along the vertical or horizontal axes or rotate. Additionally, lateral stabilizing bars sometimes disconnect from the Nuss bar. Most bar migrations occur in the early postoperative period [14] and are demonstrated with routine chest radiography.

Fig. 2 Haller index calculation. a Axial CT image in a 16-year-old boy with pectus excavatum through the point of maximum posterior sternal angulation demonstrates a Haller index of 2.9. b Axial MR of the 17-yearold boy in Fig. 1 demonstrates severe pectus deformity with a Haller Index of 40

In this paper we present the radiologic appearance of Nuss bar migrations to help radiologists accurately diagnose this complication. Additionally, variations in patient positioning between radiographs can result in apparent Nuss bar migration, and we have termed this phenomenon pseudomigration. Examples of true and pseudo migrations are included.

Materials and methods Our institutional review board approved this retrospective study, which complies with the Health Insurance Portability and Accountability Act. We included all patients who had undergone a minimally invasive pectus excavatum repair at our tertiary care center between April 1998 and June 2014. We reviewed the electronic medical record and all pertinent radiologic studies and evaluated the frequency and type of bar migrations. We defined bar migrations as deviation of the bar from the position demonstrated on the immediate postoperative

Pediatr Radiol

Fig. 3 Nuss bar procedure in a 17-year-old boy with sternum excavatum (the same boy as in Fig. 1). Because of the severity of the pectus excavatum deformity, three Nuss bars were required for repair, as seen in the intraoperative photograph. Note the marked improvement after placement of the Nuss bars

radiographs and categorized these deviations as superior, inferior, rotary, lateral or flipped using a new classification system. All preoperative and postoperative chest imaging studies were retrospectively reviewed by a pediatric radiologist with 22 years’ experience (L.A.B.). The diagnosis of migration was based on clinical and imaging studies performed in the course of each patient’s care. These included anteroposterior (AP) or posteroanterior (PA) and lateral radiographs and crosssectional studies of the chest. Postoperative images were routinely obtained to identify immediate operative complications such as pneumothorax. Additional postoperative images included daily AP chest radiographs and a single pair of PA and lateral radiographs prior to discharge. Additional chest radiographs would not be routinely obtained unless there was clinical concern for delayed bar migration or in anticipation of scheduled bar removal. Studies deemed non-pertinent included imaging that did not include the chest such as head CT or extremity films obtained for clinical reasons unrelated to the Nuss bar. These were not reviewed for purposes of this study.

Results Through medical record review we identified 311 pa tients who underwent minimally invasive surgery for pectus excavatum deformity using the Nuss procedure. The median age was 15 years (interquartile range 13–16)

and the mean Haller index was 4 (interquartile range 3.5–4.9). Mean follow-up was 3.8 years (0–16 years). The male-to-female ratio was 4:1. Bar migration was demonstrated in 23 of 311 patients (7%) and occurred within a mean period of 26 days after surgery (interquartile range = 15–70 days). The types of bar migration are described in Table 1. The most common migrations involved slippage of the bar off the dorsal sternal apex; this accounted for 43% (10/23) of our cases including six superior and four inferior migrations. More complex migrations, including flipping and rotation, were less common, found in only four patients. Results are summarized in Table 1. Sixteen of the 23 cases (70%) of bar migration required re-operation. Patients whose bars were thought to have migrated radiographically might not have had revision surgery if the cosmetic result was stable and satisfactory or if they were due for scheduled Nuss bar removal. The normal Nuss bar position after surgery is perpendicular to the long axis of the thoracic spine and parallels the thoracic vertebral body endplates (Fig. 4). In superior or inferior migration, the bar ends remained fixed along the lateral chest wall but the bar’s contact point with the sternum was displaced upward or downward, respectively (Fig. 5). In rotary migration, one or both bar ends are no longer stable, with clockwise or counterclockwise rotation of the bar with respect to the long axis of the sternum (Fig. 6). In lateral migration, neither end of the bar remains fixed and there is horizontal migration of the bar along the long axis only (Fig. 7). Bar flipping can result with loss of fixation of the lateral ends of the bar and bar rotation about

Table 1

Classification of Nuss bar migration

Migration type

Slippage off dorsal sternal apex Inferior Superior Lateral migration Rotation Flipped Stabilizer bar detachment Miscellaneous/combination Total

Lateral bar end position stable

Number of cases

Yes Yes No One or none No Not applicable Not applicable

4 6 4 3 1 1 4 23

Pediatr Radiol Fig. 4 Normal postoperative Nuss bar with horizontal orientation on both (a) posteroanterior and (b) lateral radiographs in a 15–year–old girl. There is no curving of the bar and the bar is perpendicular to the lateral chest wall and spine

Fig. 5 Superior migration (a, b), inferior migration (c, d) and pseudomigration (e, f) post Nuss bar placement. a Posteroanterior radiograph in a 7-year-old girl demonstrates superior migration 3 weeks post surgery. Note the upward curve of the bar, with the lateral ends lower than the middle of the bar. b Lateral radiograph in the same girl confirms migration, with superior deviation of the bar off the sternal apex and superior angulation of the bar with respect to the spine. c Immediately postoperative anteroposterior radiograph in a 13-year-old boy demonstrates normal bar position and orientation. No displacement is

seen. d Two months postoperative posteroanterior radiograph in the same boy demonstrates inferior migration. Note that the bar appears thinner and there is now a downward curve of the bar. The bar ends are stable, while the middle of the bar is inferiorly migrated. e Immediate postoperative lateral radiograph and (f) 1-month follow-up lateral radiograph in a 17-year-old boy demonstrate increasing superior orientation of the middle of the three Nuss bars. Note the apparent changes in lower bar position are related to differences in patient rotation and increased kyphotic positioning and not true migration. This is an example of pseudomigration

Pediatr Radiol

Fig. 8 Complex migration with Nuss bar flipping in a 17-year-old boy. Three-month follow-up posteroanterior (a) and lateral (b) radiographs demonstrate interval flipping of the Nuss bar. Note the apparent narrowing and upward curved appearance of the bar. Additionally, there is leftward migration Fig. 6 Rotary migration in a 16-year-old boy. Posteroanterior radiograph demonstrates counterclockwise rotation of the Nuss bar, with marked elevation of the left end and depression of the right end

Minimally invasive pectus excavatum repair has become the procedure of choice since its introduction by Nuss [11]. A recognized complication of this technique is bar

migration, and early reports found migration rates of approximately 5% [14, 15]. The forces created by placing a convex-oriented bar under the dorsal sternal apex can be quite large. Immediately upon placement there is a settling of the bar’s ends into a more stable position in the lateral chest wall structures. This settling can be observed intraoperatively. If the early settling position is not stable, the process might continue with the bar sliding off the apex resulting in early superior or inferior migration. This continued movement of the bar likely explains the observation that superior and inferior migrations typically occur early after bar placement, whereas lateral migration and rotation occur later. Park et al. [15] described Nuss bar migrations as upward or downward flipping, which corresponds to our

Fig. 7 Progressive lateral migration in a 9-year-old boy. a Posteroanterior radiograph 3 weeks after placement demonstrates the bar to be horizontally positioned and perpendicular to the chest wall. The bar extends farther to

the right than left, suggesting a small amount of rightward lateral positioning. b At 3 months’ follow-up, rightward migration is evident. c Three years later there is further rightward migration

its short axis. This results in flipping of the bar’s convexity (Fig. 8). Additional complications, including separation of the Nuss bar from the lateral stabilizer bar (Fig. 9), were also noted.

Discussion

Pediatr Radiol Fig. 9 Lateral stabilizing bar disconnection in a 16-year-old boy. a Immediate postoperative anteroposterior radiograph demonstrates satisfactory Nuss bar position and orientation. b Follow-up posteroanterior radiograph demonstrates separation of the bar from the right lateral stabilizer

superior and inferior migrations. In both schemas, the lateral ends of the bar remain stable and the contact point on the sternum is shifted up or down. When both ends of the Nuss bar are mobile, lateral migration (termed “sliding” by Park) or rotation about the contact point on the sternum can occur. Park et al. [15] did not describe rotation along the long axis of the Nuss bar. They did note posterior settling of the bar wherein the bar ends (termed “hinge points”) dissect into the intercostal tissues and the bar subsides, posteriorly pulling away from the sternum [15]. We did not encounter this type of migration in our study group. In our experience, the incidence of bar migration is most closely related to surgical experience, decreasing from 6–26% in the first 50 cases to less than 2% thereafter [14]. With increased surgical experience, the importance of fixation of the Nuss bar has become evident. Fixation of the bar with sternal cartilage sutures has been Fig. 10 Superior pseudomigration in a 15-year-old girl. a Posteroanterior radiograph demonstrates apparent superior migration of the Nuss bar. There is an upward curve of the bar. The expected location of the bar is indicated (blue line). b Lateral radiograph confirms normal positioning of the bar. Note that the angle between the bar (red line) and the spine (blue line) is 90° and the bar remains under the posterior sternal apex (red arrow). Arching of the lower thoracic spine and lordotic or kyphotic positioning can result in superior and inferior pseudomigration, respectively

shown to reduce early bar migration, presumably by maintaining the relationship of the bar to the dorsal sternal apex during the early settling process, reducing migration rates to 1–2% [14–16]. We found lateral stabilizing bars with fixation to the ribs to be associated with increased bar migration, which might be caused by inhibition of normal early settling of the bar in the lateral chest wall tissues. We also encountered several cases where follow-up radiographs suggested bar migration in the absence of true displacement. These resulted from variations in patient positioning between baseline and follow-up radiographs. Specifically, differences in lordotic or kyphotic positioning or arching of the lower thoracic spine can simulate inferior or superior migration (Fig. 10). Likewise, lateral leaning can create the appearance of a rotary migration (Fig. 11). Careful evaluation of the

Pediatr Radiol Fig. 11 Pseudorotation in a 16year-old boy. a Lateral radiograph demonstrates apparent rotation of the bar with the lateral bar ends at different heights (red arrows).b Posteroanterior radiograph demonstrates normal bar position with horizontal orientation. The apparent difference in bar end height on the lateral radiograph resulted from lateral leaning when the film was obtained

Nuss bar position with respect to the ribs, spine and sternum is required to prevent the misdiagnosis of migration when a pseudomigration is present.

4.

Conclusion

6.

The retrosternal bar that is used during the minimally invasive Nuss procedure for pectus excavatum can migrate in a variety of ways. Radiologists should be aware of the radiographic appearances of the various Nuss bar migrations as well as the potential for pseudomigration related to patient positioning rather than actual bar displacement. Careful evaluation of both AP or PA and lateral radiographs is often necessary to fully evaluate for Nuss bar migration and exclude pseudomigration.

7.

5.

8.

9.

10.

Compliance with ethical standards

11.

Conflicts of interest None

12.

13.

References 1. 2.

3.

Hebra A (2009) Minimally invasive repair of pectus excavatum. Semin Thorac Cardiovasc Surg 21:76–84 Kotzot D, Schwabegger AH (2009) Etiology of chest wall deformities — a genetic review for the treating physician. J Pediatr Surg 44:2004–2011 Ravitch MM (1977) Repair of pectus excavatum in children under 3 years of age: a twelve-year experience. Ann Thorac Surg 23:301

14.

15.

16.

Quigley PM, Haller JA Jr, Jelus KL et al (1996) Cardiorespiratory function before and after corrective surgery in pectus excavatum. J Pediatr 128:638–643 Peterson RJ, Young WG Jr, Godwin JD et al (1985) Noninvasive assessment of exercise cardiac function before and after pectus excavatum repair. J Thorac Cardiovasc Surg 90:251–260 Fabricius J, Davidsen HG, Hansen AT (1957) Cardiac function in funnel chest; twenty-six patients investigated by cardiac catheterization. Dan Med Bull 4:251–257 Colombani PM (2009) Preoperative assessment of chest wall deformities. Semin Thorac Cardiovasc Surg 21:58–63 Ohno K, Morotomi Y, Nakahira M et al (2003) Indications for surgical repair of funnel chest based on indices of chest wall deformity and psychological state. Surg Today 33:662–665 Fokin AA, Steuerwald NM, Ahrens WA et al (2009) Anatomical, histologic, and genetic characteristics of congenital chest wall deformities. Semin Thorac Cardiovasc Surg 21:44–57 Robicsek F, Watts LT, Fokin AA (2009) Surgical repair of pectus excavatum and carinatum. Semin Thorac Cardiovasc Surg 21:64– 75 Nuss D (2008) Minimally invasive surgical repair of pectus excavatum. Semin Pediatr Surg 17:209–217 Haller JA Jr, Kramer SS, Lietman SA (1987) Use of CT scans in selection of patients for pectus excavatum surgery: a preliminary report. J Pediatr Surg 22:904–906 Ravitch MM (1949) The operative treatment of pectus excavatum. Ann Surg 129:429–444 Gibreel W, Zendejas B, Joyce D et al (2016) Minimally invasive repairs of pectus excavatum: surgical outcomes, quality of life, and predictors of reoperation. J Am Coll Surg 222:245–252 Park HJ, Chung WJ, Lee IS et al (2008) Mechanism of bar displacement and corresponding bar fixation techniques in minimally invasive repair of pectus excavatum. J Pediatr Surg 43:74–78 Park HJ (2007) Technical innovations in the minimally invasive approach for treating pectus excavatum: a paradigm shift through six years’ experience with 630 patients. Innovations 2:25–28