Minimally invasive repair of pectus excavatum (MIRPE) using a pectus bar was first introduced in 1998 [1], and it has since become the standard surgical therapy for pectus excavatum (PE) around the world [2].

However, this surgical procedure has undergone considerable evolutionary changes to ensure that it is as safe and effective as possible. Pectus bar displacement has represented one of the most common and significant impediments to effective chest wall reconstruction [3-5]. We have been keen to combat the undesirable issues that have arisen as a result of our early lack of expertise and technical skill with this operation. Specifically, procedures for pectus bar stabilization have evolved since the early 2000s, when multiple point pericostal sutures were used to link the ribs and pectus bars. Despite the fact that substantial progress had been achieved even at that point, the bar dislocation rate remained about 3% [6]. Next, in 2006, the claw fixator, a right-angled metal blade meant to snag the rib and anchor the bar, was created and demonstrated to be effective in preventing bar flipping. The hinge plate, a metal plate with a concave cradle and a spur to embrace the pectus bar, was designed in 2011 to support the pectus bar at the hinge points (where the pectus bar enters the intercostal space) and prevent intercostal muscle stripping. With all of these innovations, we were able to lower the rate of bar displacement to 1%, but we were unable to completely eliminate it.

Finally, the bridge plate was developed as the most advanced pectus bar stabilization technique in 2013. We have found that bridge fixation using multiple bars (parallel bars, cross bars, XI bars) has rock-solid stability and has become our ultimate weapon for stabilizing the pectus bars.

In this study, we examined and assessed the most recent 10-year trends in our innovations, with a particular focus on the development of pectus bar stabilization techniques.

Patients and methods

This study was reviewed and approved by the Institutional Review Board of Catholic University of Korea (approval no., KC22RISI0779). The requirement for informed consent from individual patients was omitted because of the retrospective design of this study. From January 2013 to July 2022, 1,546 patients underwent PE correction at a single center, using a claw fixator, bride plate fixation, or hinge plate fixation for pectus bar stabilization. After excluding those who had undergone prior surgery at other institutions, cases of redo PE repair, and patients with a history of sternotomy, 1,526 patients were ultimately included in this retrospective analysis.

The diagnosis of PE or pectus carinatum was established according to the physical examination and computed tomography (CT) scan. Over the last decade, there have been significant modifications to this procedure, such as the multiple-bar approach with parallel bars and cross bars, crane-powered chest wall lifting, and pectus bar stabilization methods with claw fixators, hinge plates, and bridge plates.

The goal of this paper was to describe the 10-year trends in bar stabilization. In particular, since we stopped using hinge plates in 2019, we compared outcomes between the hinge plate and bridge plate groups. Patients undergoing primary PE repair with at least 2 pectus bar insertions utilizing the hinge plate or the bridge plate were included in the study, since bridge plate fixation was used for cases involving multiple bars (Fig. 1).

Figure 1. Chest X-rays of the patients repaired with or without a hinge plate. (A) Parallel bars with a hinge plate. (B) Cross bars with a hinge plate. (C) XI bars with a hinge plate. (D) Parallel bars without a hinge plate. (E) Cross bars without a hinge plate. (F) XI bars without a hinge plate.

As a subgroup analysis, to investigate the necessity of the hinge plate, patients were divided according to whether they received hinge plates or a combination of hinge and bridge plates (group H) or only a bridge plate (group B). In this process, 334 patients were assigned to group H and 373 to group B. The 2 groups were compared in terms of effectiveness, security, and complications. Group B had a mean age of 16.6 years (range, 10–41 years), whereas group H had a mean age of 18.2 years (range, 10–40 years), although this difference was not statistically significant. Body weight and height differed between the 2 groups, presumably because larger, heavier patients were prescribed the hinge plate more frequently. There were no statistically significant differences in any other variables between the 2 groups (Table 1).

Table 1 . Patients’ characteristics for group B (bridge plate) and group H (hinge plus bridge plate).

Characteristic	Group B (n=373)	group H ( hinge or hinge plus bridge plate)	p-value
Age (yr)	16.6 (10–41)	18.2 (10–40)	0.504
Height (cm)	168.23 (125–188)	171.76 (149–196)	0.005
Weight (kg)	52.42 (15–89)	56.75 (33–94)	0.058
Sex (male)	312 (83.6)	279 (83.5)	0.968
Symmetric type pectus excavatum	185 (49.6)	152 (45.5)	0.277

Values are presented as mean (range) or number (%)..

Postoperative complications, with particular emphasis on the bar displacement rates, were compared between group H and group B. Bar displacement was defined as the bar being tilted over 30° (flipping) on the lateral view of a chest X-ray (type 1), or a hinge point disruption due to intercostal muscle stripping with a remarkable morphological change (type 2) [7].

Surgical technique

All patients underwent MIRPE without open conversion or a hybrid technique. The principal techniques were crane pre-lifting of the sternum, a morphology-tailored approach (symmetric or asymmetric bar shaping), and bar stabilization using claw fixators, hinge plates, and bridge connection as appropriate for the period of the study. Moreover, the flare-buster and magic string techniques were used as necessary.

Operative procedure

The patients were placed in the supine position with both arms hung overhead using arm slings.

Crane sternal pre-lifting

Sternal pre-elevation using the crane technique was implemented as the initial step of repair, with sternal wiring or sternal screws (the screw crane) used later. Prior to 2017, the sternum was partially elevated in select cases, but after that, the sternum was maximally elevated above the desired level of the restored sternum in every patient, regardless of age (from 3 to 55 years old).

This was the author’s novel concept, “crane-powered PE repair,” which involves exclusively using the crane power without pectus bar turnover power, where the sternum was lifted “way up high” over the intended level of repair, and the pectus bars were simply placed at the accurate locations without undue effort. The Easy Crane System (PrimeMed, Seoul, Korea) is a sternal-specific elevating device of our design [8,9]. The crane enabled any area of the deformed chest wall to be supported firmly and steadily throughout the entire surgical procedure.

Pectoscope-guided bar passage

Bilateral incisions measuring 1.5 cm were made at both mid-axillary lines. A pectoscope (PrimeMed), which is an endoscopic instrument specifically created for pectus surgery that offers a continuous contact view to the heart and chest wall interface, was introduced through the “hinge point” located in the anterior axillary line following the formation of the subcutaneous layer pocket.

After transferring the pectoscope to the opposite side, a plastic guide (24F chest tube) was introduced into the mediastinal path. Following the guide, a morphology-tailored pectus bar was placed and flipped to sit on the deepest point of the depressed chest wall. At this point, the chest wall was already being raised with a crane, so the bar could be positioned accurately without undue effort [2].

Pectus bar stabilization

The claw fixator (Fig. 2): The claw fixator (PrimeMed) was developed in 2007 to prevent pectus bar flipping and dislocation without the need for invasive pericostal sutures [10]. The claw fixator stabilizes the bar by hooking itself to the adjacent rib. Prior to 2014, the plan was to use it for the stabilization of each single bar, but after introducing bridge plates for multiple bars, the use of the claw fixator was restricted to repairs using a single bar.

Figure 2. Claw fixator used to stabilize the pectus bar by hooking up the rib. (A) Claw fixator. (B) Skeletal model installation. (C) Chest X-ray.

The technique used to install the claw fixator was as follows: an electrocautery incision was performed at the top edge of the rib next to the pectus bar end-holes, and the claw blade was inserted into the pleural cavity through the opening. The opposite end of the claw blade was then screwed into the bar hole.

The hinge plate (Fig. 3): The hinge plate (PrimeMed) is a metal strip with a cradle in the center and a spur on the cradle’s caudal side [3]. It is intended to embrace and support the pectus bars at the hinge points. As all forces generated during PE repair primarily impinge on the hinge point, the hinge plate is designed to preserve the hinge point by avoiding intercostal muscle stripping, and the spur precisely positions the pectus bar within the cradle of the plate.

Figure 3. A hinge plate stabilizes the pectus bar by supporting it at the hinge point and preventing intercostal muscle stripping. (A) Hinge plate. (B) Skeletal model installation. (C) Chest X-ray.

To place the hinge plate, two pericostal sutures are created along the ribs above and below the hinge point. Through subcutaneous dissection, the hinge plate was pushed up to the hinge and secured with sutures at the place.

The bridge plate (Fig. 4): The bridge plate (PrimeMed) is designed to connect and stabilize multiple bars at the end-hole of the bars using a bolt-nut screwing system [5,11]. The bilateral bridge linkage to the bars holds the bars tight with each other and builds a rock-solid fortress, which makes the bars virtually unable to rotate or migrate. From a technical standpoint, this bridge plate linkage is noninvasive and straightforward. It is exclusively situated by itself in a subcutaneous and extra-thoracic location without any impingement of the ribs or other thoracic structures [11].

Figure 4. A bridge plate installed on a pectus excavatum model connecting multiple pectus bars. (A) Bridge plate. (B) Skeletal model installation. (C) Chest X-ray.

Through the incisions formed for bar insertion, an appropriately sized bridge plate was positioned in the lateral subcutaneous plane. All the bars were aligned with their end holes towards the bridge. The bolts and nuts were tightened to securely connect the bridge plate and the bars.

Cryo-intercostal nerve block and postoperative pain management strategy

Since 2022, we have added cryo-analgesia using bilateral intrathoracic intercostal nerve freezing of the third through seventh intercostal nerves. The nerves were frozen at -70°C for 2 minutes using a cryoablation device (cryoFORM; AtriCure Inc., West Chester, PA, USA) along with thoracoscopy [12,13]. In addition, the intercostal nerve is blocked using local percutaneous injections of 0.5% bupivacaine. Finally, following the procedure, a PainBuster catheter (Pain Relief System; Halyard Health Inc., Irvine, CA, USA) was implanted near the rib cage to provide a continuous infusion of local anesthetics (0.5% bupivacaine, 2 mL/hr for 30 hours). Further, patient-controlled analgesia was used in patients who tolerated opioids.

Closure

After inserting small-bore catheters to drain the pleural spaces and wound pockets, the wounds were closed in layers.

Statistical analysis

IBM SPSS ver. 24.0 (IBM Corp., Armonk, NY, USA) was utilized for the statistical analysis. All values were reported as mean (range). Comparisons were made using the unpaired t-test for continuous variables and the chi-square test for categorical variables. A propensity score matching analysis was conducted to determine the influence of potential confounding variables, such as the number of bars, the pattern of bar placement, and the crane technique, on the bar displacement rate.

The patients ranged in age from 3 to 55 years old (mean, 11.3 years). The ratio of men to women was 4 to 1. The average number of bars used per patient was 1.74, and the number of bars used tended to rise over time, as we favored utilizing multiple bars to maximize the chances of chest wall coverage. The proportion of multiple bar usage continued to rise over time, but the increase was not statistically significant (Fig. 5). To achieve “entire chest wall remodeling,” we ceased utilizing a single bar for all patients, including all young children, and in 2022, we adopted a multiple-bar strategy regardless of age.

Figure 5. Trends in the performance of the single-bar and multiple-bar approaches during the study period, showing an increasing frequency of multiple pectus bar use (red-arrowed line) and decreasing use of the single-bar approach (blue-arrowed line). The proportion of the multiple-bar approach reached 100% in 2022, when we finally stopped using the single-bar approach.

The methods used to secure the bars included the claw fixator, the hinge plate, and the bridge plate. The claw fixator was utilized in 674 cases with a single pectus bar. As we transitioned to bridge fixation, its usage progressively declined from 2013 (Fig. 6). In 2022, we abandoned single-bar repairs and shifted to a multiple-bar approach, thus making the claw fixator unnecessary, and we switched to bridge fixation for the bar stabilization technique. Before it was discontinued, the hinge plate was used in 378 cases (24.7%) from 2013 to 2019. Since 2014, we have begun using bridge plates for bar stabilization in 497 cases, with 2 bars in 355 patients (71.4%), 3 bars in 141 patients (28.2%), and 4 bars in 1 patient (0.2%).

Figure 6. Trends in the application of claw fixators, hinge plates, bridge plates, and cranes. The use of claw fixators and hinge plates decreased with time, whereas bridge plate utilization took over their functions (2020). Since we started exclusively employing the crane-powered approach (2018), hinge plates have no longer been required (2019). We stopped using claw fixators when we shifted to a multiple-bar approach (2022).

Bar displacement was identified in two patients in whom a claw fixator was used (0.1%, n=2) but there was no bar displacement with hinge plate (0%, n=0) or bridge plate (0%, n=0). Two patients treated with a claw fixator in a single-bar approach exhibited a slipped bar.

The crane technique was used throughout the study period. In an earlier period, we used it in selected patients such as heavy adults or those with a severe depression. The crane has been increasingly utilized with time, and since 2016, 51.5% of patients have been operated on utilizing a crane. Starting in 2018, we implemented crane-powered PE repair as a novel technique, and all patients, regardless of age, have since been operated on using the crane technique.

To determine the effect of the hinge plate on the stabilization of the bar in the era of bridge fixation, we compared the outcomes of using hinge plates or a combination of hinge and bridge plates (group H, n=334) to those of bridge fixation alone (group B, n=373). The pattern of bar placement consisted of parallel bars or cross bars/XI bars, and 79% (n=264) of group H and 66.8% (n=246) of group B utilized parallel bars (p=0.001). The Haller index and the asymmetry index showed no significant differences between the 2 groups. In group H, the mean size of the bars was 34.44 cm (range, 20.32–40.64 cm) compared to 33.45 cm (range, 22.86–43.18 cm) in group B (p=0.005). The number of bars used was 2.36 in group B and 2.16 in group H (p=0.001). There was no bar displacement in group H or group B. This result indicates that once the bridge connection was established, the hinge plate was not the determining factor in bar stability. Using propensity score matching, we found that potential confounding variables such as the number of bars, the pattern of bar placement, and the crane technique did not affect bar displacement or other outcomes. This result confirms the validity of our findings in this study.

The other postoperative complications occurred significantly more frequently in group H (pleural effusion: 0.3% versus 4.2%, p=0.001; wound hematoma: 0% versus 1.5%, p=0.018; percutaneous catheter drainage: 2.7% versus 7.3%, p=0.005) (Table 2). Group B had a higher incidence of pneumothorax (n=33, 8.9%) than group H (n=12, 3.6%) (p=0.004). However, the pneumothorax resolved on its own without the need for any interventions or drainage throughout the entire course of normal recovery. Group H had significantly higher rates of pleural effusion (with or without percutaneous catheter drainage) and wound hematoma (Table 2). Group H also had a significantly longer average hospital stay (5.53 days versus 6.21 days, p=0.034).

Table 2 . Outcome comparisons between group B (bridge plate) and group H (hinge plus bridge plate).

Variable	Group B (n=373)	group H ( hinge or hinge plus bridge plate)	p-value
Haller index
Preoperative	4.44 (1.42–55.70)	4.78 (1.82–22.77)	0.908
Postoperative	2.68 (1.91–4.05)	2.72 (1.75–3.79)	0.853
Asymmetric index
Preoperative	1.05 (1.00–1.31)	1.06 (1.00–1.72)	0.598
Postoperative	1.04 (1.00–10.2)	1.04 (1.00–10.20)	0.992
Pectus bar characteristics
Bar size (cm)	33.45 (22.86–43.18)	34.44 (20.32–40.64)	0.005
No. of support bars	2.36 (2–4)	2.16 (2–4)	0.001
Parallel shape	249 (66.8)	264 (79.0)	0.001
Crane application	350 (93.8)	294 (88.3)	0.009
Postoperative complications
Pleural effusion	1 (0.3)	14 (4.2)	0.001
Pneumothorax	33 (8.9)	12 (3.6)	0.004
Wound hematoma	0	5 (1.5)	0.018
Wound infection	1 (0.3)	2 (0.6)	0.499
Percutaneous catheter drainage	10 (2.7)	24 (7.3)	0.005
Bar displacement	0	0	-
Reoperation	1 (0.3)	3 (0.9)	0.267
Length of stay (day)	5.5 (4–21)	6.2 (4–24)	0.034

Values are presented as mean (range) or number (%)..

We had a single case of minor slippage of the bar, which was operated on in 2015. Four patients required reoperation, but there was no statistical significance in the difference between groups (3 in group H versus 1 in group B, p= 0.267). Among the reoperation cases, one patient in group B had a delayed hemothorax on the first postoperative day. Re-exploration revealed that a portion of the left lung was entrapped between the bar and chest wall, thus resulting in lung lacerations and hemorrhage. Thoracoscopy-assisted wedge resection of the damaged lung was performed, and the patient recovered uneventfully. The other 3 reoperations in group H were for right arm paralysis caused by compression of the brachial plexus immediately after surgery (1 in 2015, 2 in 2017). The upper bar was removed to relieve the nerve compression and the patients regained full neurologic function within a few months.

PE repair using the Nuss technique has been insufficient since it was first introduced due to a lack of experience, techniques, and appropriate devices. We and all other surgeons across the world have experienced a steep learning curve and a high rate of failure when conducting this recently developed, minimally invasive procedure [2,6,8, 10,14]. To overcome this challenge and make this heretofore unproven operation safe and effective, we have actively worked over the past 2 decades to create the necessary techniques and equipment. Moreover, a number of ground- breaking revolutionary approaches have been developed and extensively verified in the world’s largest series of pectus deformity repair. When confronted with obstacles of any kind, we adapted and developed strategies to combat them, and we devised suitable devices to implement the concepts and theories. These can be essentially categorized into 3 groups.

First, the bar stabilization strategy has been developed while creating appropriate devices to implement new ideas to eliminate any bar displacement, including even minor shakes. Bar displacement has continuously represented one of the most serious obstacles to the pectus bar repair of pectus deformities. Not only has it been the major cause of repair failure, but most importantly, it also causes catastrophic consequences [15]. During or after surgery, pectus bars are susceptible to dislocation via 3 distinct mechanisms, as explained by the author in 2010 [7]. Comparative results from other studies showed that “bar flipping” occurs in a significant proportion of cases, ranging from 1.9% to 9.2% [4,7]. In our study, we were able to drastically lower this rate by using claw fixators, hinge plates, and, finally, by making the bars immobile using bridge plates.

The claw fixators were the first less-invasive device we used to stabilize the pectus bar. We found it to be effective, and it provided us with a profound advantage over the conventional lateral stabilizers (Zimmer-Biomet, Jacksonville, FL, USA), which have been used by other surgeons worldwide without any modification or improvement for 20 years since they were first introduced. We were able to reduce the bar displacement rate with our claw fixators to 1%, which had previously been 3.4% with the Biomet stabilizers (Zimmer-Biomet) that we used in our earlier experience in 2000 [16]. However, the claw fixators did not meet our ultimate goal of a “zero” bar displacement rate.

In further endeavors to reach our goal, we subsequently developed the hinge plates. We recognized that the claw fixator would be of no assistance in protecting the hinge from heavy pressure being applied to it, because it was only designed to block bar flipping. Acute intercostal muscle stripping happened during the repair procedure or chronically over a period of several months after the repair. To resolve this problem, we invented the hinge plate to bolster the bar entry point into the thoracic cavity. By effectively strengthening the intercostal muscles at the hinge, this has significantly reduced the rate of type 3 bar dislocation [7].

Recent development of the bridge plate has made it possible to make the bar non-rotatable, thus preventing its displacement. Since 2018, there has been no bar dislocation with the bridge connection of multiple bars. In contrast to its benefits, the hinge plate was more invasive and not without complications. The use of the hinge plate could practically eliminate the hinge stripping problem, but it caused more complications such as wound hematoma or pleural space events that required catheter drainage. Initially, the hinge plate and bridge plate were employed together, but as of the beginning of 2019, with the construction of the bridge connection under the complete crane over-lifting of the sternum (the crane-powered PE repair strategy), the hinge plates no longer needed to be applied. The hinge plate was safely replaced by the bridge plate alone, as the intercostal muscle stripping issue was not observed in over 300 patients in the bridge plate group. Omitting the hinge plate and using only the bridge plate yielded superior results, with considerable advantages in terms of increased surgical efficiency and fewer perioperative problems.

Second, we introduced the crane application concept early in our practice in 2002, and we mostly employed it on adult patients with heavy chests or severe Grand Canyon-type deformities. We gradually adopted it in an increasing number of cases, and since 2017, it has been utilized in all cases without selection. The affected chest wall was lifted “way up high” above the level of the pectus bar location, and the pectus bars were safely introduced and simply placed in the accurate position without needing to apply undue pressure to the hinge points (the crane-powered approach).

Total crane pre-lifting of the depressed sternum made the entire repair surgery effortless, and in particular, it rendered the chest wall push-up with the pectus bars unnecessary. Consequently, the hinge was no longer compressed or harmed by the weight of the chest wall, and it was properly protected during surgery. Once we achieved sufficient elevation of the chest wall with solid stability of the pectus bars with bridge connections bilaterally, the hinge appeared to be well-maintained for 2–3 years postoperatively. We ultimately chose not to execute the additional procedure of installing the hinge plate, which required the use of multiple invasive pericostal sutures. The hinge plate was devised to support the hinge points and avoid intercostal muscle stripping during the procedure or postoperatively. Comparing group H and group B demonstrated that the bridge connection alone, without hinge plates, did not increase the risk of bar dislocations; we could have safely removed the hinge plates from our arsenal of bar stabilizing devices. The current analysis proved that the bridge connection of multiple bars without hinge plates did not increase the risk of bar dislocations; we have achieved our ultimate goal of a “zero” rate of bar displacement.

Lastly, there is a trend of increasingly using the multiple-bar approach for entire chest wall remodeling [17]. At the beginning of the Nuss approach, only a single bar was utilized, as the extent of chest wall coverage was not a consideration—instead, the aim was simply to lift the chest wall to relieve cardiac compression. Clearly, the single-bar approach fell well short of achieving the anatomical remodeling of the chest wall. To address this issue, we began multiple-bar repair (the parallel-bar technique) and expanded the scope of treatment [18]. Moreover, in pursuit of the highest anatomical and physiological integrity of the chest wall, we developed the cross bar (the cross-bar technique) and the cross bar plus upper bar approach (the XI technique), which allowed us to apply the entire chest wall remodeling principle [17]. Moreover, as the final touches for anatomically correct repair, we freely utilized the sandwich technique and its variants: the flare-buster technique for alleviating the lower costal flares and the magic string technique to compress focal chest wall protrusions [11,19, 20].

In conclusion, the trend of pectus bar stabilization in our practice evolved over the past decade and the results substantially changed. As we did not have any cases of bar displacement in the bridge-alone group, we could have simplified the procedure and made it safer by eliminating any invasive installation of bar-stabilizing devices, such as claw fixators and hinge plates. Our strategy gradually changed to the multiple-bar technique with the bridge plate stabilization method. Eventually, we were able to achieve the most favorable outcomes: our ultimate objective of “zero” bar displacement using the most recent method of multiple bars and the bridge fixation strategy.

Conceptualization: HJP, GMR, HKK. Data curation: HJP, HKK. Formal analysis: GMR, HKK. Methodology: HJP, GMR, HKK. Writing–original draft: GMR, HKK. Writing–review & editing: HJP, GMR, HKK.

No potential conflict of interest relevant to this article was reported.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.