Revision after prior failed pectus excavatum repair: higher risks and greater complications than primary surgery

Introduction

Severe pectus excavatum (PE) deformities can cause cardiac compression and significant symptoms, therefore surgical correction is indicated in appropriate cases (1). In experienced hands, the operation is a safe procedure and provides resolution of symptoms and improvement in the patient’s quality of life (2,3). The two most common surgical methods performed include modifications of the open Ravitch approach and the minimally invasive repair of pectus excavatum (MIRPE). In both techniques, recurrence rates after repair have been reported by individual centers from 2–37% (3-7).

The cause of recurrence is dependent on the technique used for the primary repair. With the MIRPE approach, recurrence is often due to a failed surgical technique including suboptimal bar length (Figure 1A-1C), insufficient number of bars (Figure 2A,2B), inadequate bar fixation (Figure 3A-3C), deficient placement of bars (too lateral of an entrance into the chest) or premature bar removal (5-11). Similarly, with the open techniques, recurrence is often secondary to issues related to surgical technique including over resection of cartilage and inappropriate dissection. Too early of removal or failure to use stabilizing/supporting structures can also contribute to failure. Incomplete chest wall healing with subsequent pseudoarthrosis and malunion secondary to failure of the cartilage to regenerate and heal are unique issues to open repair cases (12,13) (Figure 4A-4C). Extensive resection of cartilage at too young of an age can result in disrupted growth of the anterior chest wall termed “acquired thoracic dystrophy” (ATD) (Figure 5). Rather than a true recurrence, this is a failure of the chest to grow normally with a restricted, small physiology which requires a complex surgical reconstruction for proper repair (13). Connective tissue disorders such as Marfan Syndrome and Ehlers Danlos are also independent risk factors for complications and recurrence in both types of repairs (14).

Figure 1 Representation of improper surgical techniques for a patient with a failed prior repair. (A) Anterior-posterior chest X-ray highlighting the utilization of short bars (non-functional). (B) Chest computed tomography of the patient showing the upper implanted bar being too short to provide proper support of the defect. (C) Chest computed tomography of the lower implanted bar having the same issue.

Figure 2 Imaging of a patient with 2 bars Nuss repair that was seen for a second opinion due to persistent symptoms and dissatisfaction with cosmetic outcomes. (A) Anterior-posterior chest X-ray view is difficult for detection of any abnormality. (B) Lateral chest computerized tomography shows that the repair has incompletely corrected the defect with residual deformity and cardiac compression present below the second bar (red arrow). A third bar or lower bar is necessary to relieve cardiac compression properly.

Figure 3 Imagining of a patient with failed MIRPE due to inadequate bar fixation. A patient who had a prior failed repair with (A) anterior-posterior chest X-ray revealing bar rotation. (B) Severe bar rotation and migration is noted on the lateral chest X-ray. (C) Cross sectional chest computerized tomography shows posterolateral bar migration secondary to intercostal muscle stripping with extreme lateral bar exiting on the right (red arrow). MIRPE, minimally invasive repair of pectus excavatum.

Figure 4 Photographs and imaging of a patient with pectus excavatum recurrence after three-prior failed Ravitch procedures. (A) Anterior chest photograph verifying recurrence and the presence of previous failed repairs’ scars. (B) Intraoperative photograph shows extensive “floating” of the sternum and malunion with no distal rib attachments. (C) Cross sectional chest computerized tomography of the patient demonstrating distal sternal “floating” and extensive malunion.

Figure 5 Anterior-posterior chest X-ray of a patient who presented for correction at 42 years of age after developing an acquired thoracic dystrophy post Ravitch procedure as a 6-year-old. This chest X-ray highlights how this deformity forms a narrow upper thorax with restricted lung volume.

There are limited publications which are exclusively devoted to the repair of recurrent PE deformities, and most studies include predominantly pediatric patients (7,15,16). Revision procedures can be complicated with a greater risk due to scaring, adhesions, calcifications, and damage that has occurred to the chest wall (17,18). There may be many other factors that contribute to the complexity of revision surgery. This paper reviews our institution’s revision repairs performed after prior failed MIRPE and open repairs and discusses their intraoperative details, complications, and outcomes. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-417/rc).

Methods

A retrospective cohort study was conducted for adult patients (≥18 years) having undergone revision of a prior PE repair between January 1^st, 2010 and March 31^st, 2023 at Mayo Clinic Arizona. Patients were classified by both the type of previous procedure performed (MIRPE versus open/Ravitch versus both) and the revision procedure performed (MIRPE, hybrid MIRPE, complex hybrid reconstruction, or complex reconstruction of ATD).

Intraoperative details, outcomes, complications, and postoperative follow up of these groups were analyzed and compared. Electronic medical records were reviewed, and the collected data included patient demographics, procedure, hospitalization, complications, and postoperative follow up. Complications were classified according to Clavien-Dindo classification (19). Follow up post revision surgery involved patients with more than 6 weeks of follow up time; and post bar removal long-term follow up included patients with more than 6 months follow up time. Patients were followed in person, via phone/virtual conference, or by standardized questionnaires. Postoperative chest X-rays were obtained at week 1, week 6, 6 months and annually. In addition, when it was available, information was obtained regarding prior procedures that were performed at outside institutions.

Revision procedures were performed by a single surgeon (D.E.J.) at Mayo Clinic Arizona. All patients underwent evaluation including physical examination, and computerized tomography or magnetic resonance imaging. Haller index was calculated by dividing the maximal internal chest width by the distance between the posterior sternum and the anterior vertebral body (20). Additional testing as indicated and deemed necessary by the evaluating physician was also performed. Patients were considered for surgical correction for the following: Haller index of 3.25 or greater, cardiac compression, cardiopulmonary deficits, significant or progressing cardiopulmonary symptoms, chest wall instability and/or malunion, and psychosocial effects.

Surgical procedures

A MIRPE with technique modifications (21) was attempted in most patients, however, when recurrences following open repair were associated with osteonecrosis, malunion and/or chest wall hernia, open repair and stabilization was additionally required (Figure 6A,6B). Our technique for MIRPE has been published in its entirety (22) and is briefly described in this manuscript. If the chest deformity could not be fully elevated, a substantial deformity/asymmetry remained, or fracture occurred with elevation, open conversion was performed for a hybrid type procedure. At that time additional reconstruction including necessary osteotomy cuts of the sternum and/or costosternal attachment and stabilization was able to be performed. Anterior titanium plating was utilized when it was deemed necessary for stabilization and approximation. All ATD patients required complex reconstructions that included expansion of the anterior wall due to the thoracic dystrophy.

Figure 6 Algorithms followed to decide the best redo approach for all the cases depending mainly on the type of their prior repair. (A) Algorithm for prior MIRPE repairs. (B) Algorithm for prior open repairs. MIRPE, minimally invasive repair of pectus excavatum; ATD, acquired thoracic dystrophy.

Modified MIRPE technique

The patient was positioned in the supine position with two longitudinal gel rolls placed parallel along the spine under the back. The arms were padded and tucked at the sides. Artline monitoring was performed in all cases. All patients were administered antibiotic prophylaxis of intravenous cefazolin (unless allergy contraindicated) prior to procedure. General anesthesia was initiated, and double-lumen tube intubation performed. A transesophageal echocardiography probe was placed, and cardiac monitoring performed throughout the case which included verification of relief of all cardiac compression. Patients were prepped and draped to include exposure of groins should emergent femoral access be necessary. Cell-saver was available for cases when significant blood loss and adhesiolysis occurred.

Whenever possible, the patient’s previous incision sites were used. Otherwise, three centimeters incisions were made bilaterally following the contour of the rib at the inferolateral pectoral borders. Submuscular pockets were made by using electrocautery to elevate the pectoralis muscles along the anterior and lateral chest wall. If a patient had indwelling bars, thoracoscopic access was attempted to be obtained before removal of bars when possible. A 5-mm thoracoscopic port was placed through the right incision, after which carbon dioxide insufflation initiated to ten to eleven mmHg pressure. A 5-mm flexible endoscope (Endoeye Flex 5, Olympus America, Central Valley, PA, USA) was introduced and, if possible, a second 5-mm port was placed for the camera on the right side, superior to the diaphragm. Most patients (especially post prior MIRPE) had substantial intrathoracic adhesions, which required lysis with electrocautery and blunt dissection. Patients with lateral dense pleural adhesions that prevented any thoracoscopic visualization required a 3 to 4 cm intercostal incision (anterior thoracotomy) accessed from the pocket incision thru which direct takedown of adhesions could be performed until a port and camera could be safely inserted. Removal of indwelling bars was required before this could occur. Once space was cleared, a second port was placed above the diaphragm to continue takedown of adhesion. Transesophageal echo visualization during removal was additionally monitored for any acute changes or evidence of tamponade.

Once adhesions from the right thorax had been taken down to the point that the mediastinum was clearly visualized, forced sternal elevation was attempted. At the center of the defect, a bone clamp [Lewin Perforating Forceps (V. Mueller NL6960), CareFusion, Inc., San Diego, CA, USA] was placed into the anterior table of the sternum and connected to a Rultract retractor (Rultract Inc., Cleveland, OH, USA) which was attached to the bed on the left at the shoulder of the patient. The defect was elevated, and the mediastinum dissected across as able.

The pericardium was often densely adhered to the sternum and a combination of electrocautery and blunt dissection was used to carefully free to prevent large openings in the pericardium or inadvertent injury to the heart. Once the mediastinum was safely crossed, continued takedown of the adhesions in the left thorax can proceed from the right side or a second port can be placed on the left above the diaphragm and the camera moved to the left side to facilitate visualization. Often, a combination of different camera sites and port sites were necessary to take down all adhesions. Lung adhesions at the site of prior bar tracks were often very dense and in the case of bars that were posterolateral stripped, the subsequent anterior space allowed the lung to herniate through (Figure 7A-7C). If the lung had become trapped and was fibrotic and densely adherent to the chest wall, attempts to remove often led to significant bleeding and air leak, therefore, a wedge resection to free the lung off the chest wall was performed for these types of cases.

Figure 7 Images highlighting how bar migration can lead to lung herniation. (A) Thoracoscopic view of a Nuss bar that has stripped thru the intercostal muscle and is now exiting the chest posterior-lateral and no longer approximated against the anterior chest wall so lung can herniate between the chest wall and the bar. (B) Drawing showing posterior-lateral bar migration that occurs when the bar strips out the intercostal muscle and moves so it exits the chest lateral. The bar will no longer push anteriorly and there is often no contact in areas which allows lung herniation. (C) Proper orientation of the Nuss bar with exit out of the chest approximately 1–2 cm lateral to the sternum and with good contact to the sternum and anterior chest wall.

If the sternum elevated successfully, a modified MIRPE was performed. The extent of the defect and the involved interspaces were evaluated. The Lorenz dissector (Zimmer Biomet, Jacksonville, FL, USA) was introduced at the superior aspect of the defect through the right interspace, passed across the mediastinum, and brought out through the contralateral interspace. A #5 FiberWire (Arthrex, Inc., Naples, FL, USA) was attached to the end of the dissector and used subsequently to guide bar placement. All intercostal spaces were reinforced with a Figure-of-eight FiberWire “Hammock” stitch (Figure 8). A FiberWire (Arthrex, Inc.) was placed with either a right angle or thoracoscopic grasper to encircle the rib above and below the interspace that would hold the bar. The hammock was placed approximately one-two centimeters lateral to the bar exit site to prevent the ribs and interspace from widening out and intercostal muscle stripping (Video 1). It allows the weight of the bar to rest on the suture versus the intercostal muscles as it exits the chest. The stitch was tied under the pectoralis muscle and once the hammock was created, placement of the bars can occur.

Figure 8 A Figure-of-eight FiberWire “Hammock” stitch is placed bilaterally to encircle the rib above and below the interspace that holds the bar. It allows the weight of the bar to rest on the suture versus the intercostal muscles as it exits the chest and helps to prevent intercostal muscle stripping and posterior-lateral bar migration.

Bars were sized to keep the lateral extension of the bar around the chest to approximately 3 to 4 cm. A second bar was next inserted one to two interspaces below the first one. A third bar was placed if the patient had any residual lower defect, or the lower costal cartilages bowed inward and might possibly cause positional cardiac compression. The bar was brought to the field and the FiberWire guide was tied to the islet at the end. The bar was inserted through the right interspace in the “U” position and brought across the mediastinum and out the left interspace. The FiberWire was kept on tension to assist with guiding and pulling the bar through the left interspace. The bar was rotated into position with the sternum still elevated. The bars were circumferentially fixed bilaterally at three sites bilaterally around the rib by using FiberWire, as previously described (22). At least one medial suture fixation site was drilled through the sternum and encircled the bar. A one-eighth inch drill bit was used to create a hole in the sternum above and below the bar. Under thoracoscopic visualization, a fascial closure instrument (Karl Storz, Goleta, CA, USA) was used to pass the FiberWire up from the mediastinum to encircle the bar and sternum (Figure 8).

Hybrid MIRPE technique

In some cases, when passing the Lorenz dissector or flipping the bars, the sternum, or the ribs fracture. If this occurs, the dissector and often existing bars must be removed, and the fracture site will need to be stabilized to support the bar. Depending on the location of the fracture and the displacement, either FiberWire, or titanium plating can be used. Extension of the lateral incisions, or a midline incision may be necessary for access to placement of plating. Complete sternal fractures can be difficult to reduce and aligning the sternum in the correct position elevated before plating is critical. Once plating is completed. The case is otherwise resumed as a hybrid MIRPE with placement and security of bars in the above-described fashion.

Complex hybrid reconstruction technique

There were two classes of complex hybrid reconstructions (23): those that came with failed prior Ravitch and had chest wall instability that required restabilization and those that were solid, calcified defects that had to be cut to be lifted. For the patients that had instability, the chest wall was immediately opened and the pectoralis and upper rectus abdominus muscles lifted to expose the boney structure. All sites of malunion were resected and a variety of grafts including autograph and cadaveric bone, methylmethacrylate, mesh and plating were used to attempt stabilization and reconstruct with these patients (Figure 9A-9C) before placement of intrathoracic bars in a MIRPE fashion.

Figure 9 Imaging and photographs of a patient with a prior Ravitch procedure. (A) Cross sectional chest computed tomography demonstrating a severe form of chest malunion. (B) Intraoperative photograph of the patient’s chest prior to completing the revision. (C) Intraoperative photograph after completing a complex hybrid reconstruction procedure that involved bone reconstruction, titanium plates and mesh placement.

If patient’s chest failed to elevate with the RulTrac retractor, passage of the Lorenz dissector in the superior most aspect of the defect was attempted. If there is still no movement of the chest, conversion to open was performed. The site of fixation was identified. If the sternum was fixed, an anterior wedge was cut in a “V” at the site of angulation incorporating only the anterior table. Again, the RulTrac was used to attempt forced elevation. If elevation was still not achieved, the Lorenz dissector was then placed in the upper intercostal space and attempted to be passed across to further force upward. If there was still limited movement or fractures, the associated costal cartilages of the defect should be evaluated and resection of cartilage with a limited modified cartilage sparing Ravitch performed for only those ribs that are restricting movement. Plating is then placed on the sternum and any modified ribs to stabilize them to the sternum (Figure 10A-10C). Once elevation was achieved, pectus support bars were placed intrathoracic to serve as a support scaffold as in the modified MIRPE approach.

Figure 10 This panel highlights how plating placed on the sternum during the revision can be also attached to adjacent modified ribs to provide stabilization. (A) Pre-revision cross sectional chest computerized tomography of the same patient mentioned in Figure 4 with history of three prior Ravich procedures. (B) Intraoperative photograph of the patient after undergoing a complex hybrid reconstruction (arrow points to the upper Nuss bar). (C) Anterior-posterior chest X-ray of the patient after the reconstruction with 2 Nuss support bars and multiple titanium plates.

Reconstruction after ATD technique

The technique for ATD repair is an extensive reconstruction and is quite variable for different patients. We have published on this technique previously and the extent is beyond the scope of this paper (13). A combination of materials including mesh, methylmethacrylate, bone graft, and plating were used to expand and elevate out the chest wall in addition to placement of underlying intrathoracic MIRPE bars.

Pain management

Pain control after a revision procedure can be difficult and a multimodality approach was used. Preoperatively, patients undergo a 3-week initiation regimen of gabapentin, starting at 100 mg three times daily and increasing weekly by 100 mg three times daily until reaching 300 mg three times daily. On the morning of surgery, a premedication dose of 600 mg is administered alongside other protocol medications listed in Table 1. Intraoperative, Methadone is given to block µ-opioid receptors. In cases where chronic pain or severe post-operative pain was experienced with prior repair, ketamine is recommended both during and after the procedure. Post traumatic stress disorder is not uncommon and ketamine can be useful in managing these cases. Its dosage can be titrated to ensure adequate pain control in combination with other protocol medications (Table 1), and it has proven successful in most patients. Cryoanalgesia has been regularly employed since 2019 and has significantly improved pain outcomes (24).

Statistical analysis

Statistical analysis was computed using SPSS version 28.0 (IBM Corporation, Armonk, NY). Variables were summarized as mean ± standard deviation (SD) or median [interquartile range (IQR)] for continuous variables and count (percentage) for categorical variables. Comparisons between groups were performed using Wilcoxon rank sum test, Student’s t test, ANOVA, or χ² test as appropriate. P value <0.05 was considered significant for all analysis.

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the institutional review board of Mayo Clinic Arizona (No. 23-010389) and individual consent for this retrospective analysis was waived.

Results

In total, 190 revision cases were included, with demographics reviewed in Table 2. Regarding the initial repair procedure, 90 (47.4%) patients had previous MIRPE, 87 (45.8%) had an open repair, and 13 (6.8%) patients had both MIRPE and open repair. Furthermore, 30 (15.8%) patients had two or more prior interventions. Overall, the median time between primary and redo procedures was 4.7 years, with prior open repair patients having the longest median time between interventions (10.3 years) (Table 2).

Patients having had a prior MIRPE were much more likely to be repaired with a revision MIRPE with >82% successfully repaired. The most common reason for conversion to hybrid MIRPE was the need for stabilization and plating of a fractured rib followed by rib osteotomy, sternal osteotomy, sternal fracture, and open thoracotomy for adhesiolysis. In 87.5% of patients who required a hybrid conversion, asymmetry was present with a mean sternal tilt angle of 23 degrees. The remainder were noted to have either an age shift toward the upper limit of the cases or a severe HI. Conversely, repair with complex reconstructions were much more likely required (85.0%) in patients with a prior open repair (including those with both MIRPE and open procedures), due to the presence of malunion, other complicated needs, and as none of the ATD patients in this group had an attempted MIRPE. Figure 11 delineates the breakdown of initial case types and repairs performed.

Figure 11 Type of redo performed according to initial surgery. MIRPE, minimally invasive repair of pectus excavatum; ATD, acquired thoracic dystrophy.

Indwelling bars were removed during the revision repair in 49 cases and 97.9% of these were malpositioned (isolated posterolateral bar migration (36.7%), isolated bar rotation (16.3%), or both combined (44.9%) (Figure 12A-12C). The size of the implanted MIRPE bars were recorded in 25 cases and in 20 of these cases, shorter bars were used in the redo intervention. In the remaining five cases, longer bars were utilized with the median difference in length between bars implanted in the primary repair and in the reintervention being 2 inches (IQR, 2.0 inches).

Figure 12 Imaging of a patient with two bars failed Nuss repair due to bar migration. (A) Anterior-posterior chest X-ray, on this view, bar migration is more difficult to observe. (B) Lateral chest X-ray shows clear posterior-lateral migration of the bars. (C) Cross sectional chest computerized tomography better presents the posterolateral bar migration with intercostal muscle stripping of both upper and lower bars and lateral exit of the bars. Both bars fail to push the sternum anteriorly.

Adhesiolysis >50 min was reported in 55.3% of the cases, being more prevalent in the revision MIRPE cases (MIRPE 74.2%, hybrid MIRPE 80.0%) (Table 3). Despite this, operative times were shortest in the MIRPE approach and longest in the complex reconstruction of the ATD patients (MIRPE 3.5±1.3 hours, ATD 6.9±1.8 hours; P<0.001). Median blood loss for overall cases was 400 mL (IQR, 450 mL), with the highest estimated blood loss found in the ATD group and the lowest in the MIRPE (Table 3).

The median length of hospital stay was 5 days (IQR, 3.0 days) with the shortest being the MIRPE approach and the longest being the reconstruction of the ATD patients [MIRPE 4 days (IQR, 3.0 days); ATD 7 days (IQR, 4.0 days); P<0.001]. Among all the redo cases, 80.0% required a stay of more than three days. The primary diagnoses causing the delay in discharge were, most commonly, high chest tube drainage (27.6%) and uncontrolled pain (27.6%), followed by prolonged air leak (21.1%) (Table 4).

Major and minor post-operative complications occurring within the first 30 days are reported in Table 4. Major bleeding that required a return to the OR occurred in 5 (2.6%) patients (MIRPE 1.1%; hybrid MIRPE 13.3%; ATD 9.1%; P=0.005). Other major complications included respiratory failure requiring re-intubation in 9 (4.7%) patients with heavy occurrence in patients with ATD (31.8%; P<0.001) and pneumothorax that required chest tube insertion occurring in 15 (7.9%, P=0.36) patients without significant difference between groups. Additionally, pleural effusion requiring thoracocentesis was a common problem among all redo procedures, occurring in 13 (6.8%) patients. Two (1.1%) mortalities occurred postoperatively; one in a complex hybrid reconstruction procedure and the second in an ATD patient with both attributed to pulmonary embolism. Two patients experienced vocal cord paralysis that resolved over time. One within several days and the other over six months that was thought to be secondary to the double lumen endotracheal tube and prolonged intubation period.

For reported long-term complications, one patient with preoperative chronic pain had elective early removal of both bars by month seventeen due to continued chronic pain despite no evidence of displacement. A second patient, at 14 months, had removal of the lower of his three bars due to persistent post-operative pain that was thought to be caused by slight bar rotation. There were three cases of persistent malunion in complex reconstruction revision Ravitch cases that required reoperation and replating.

Follow up was available in 170 (89.5%) patients with a median follow up time of 7.3 years (IQR, 6.3 years). Overall, patients were satisfied, reporting good to very good cosmetic outcomes (94.1%) and improved symptoms (94.1%). Most patients had improvement of their chronic pain; however, significant pain was still reported by 8.8% at follow up (Table 5). Significant pain in the ATD group was not uncommon and higher than the other cohorts (17.6%) (Table 5). Pain was more prevalent in patients having had a prior open repair compared to prior MIRPE cases (13.6% vs. 3.6%, respectively, P=0.02).

Among follow ups, 151 patients were eligible for bar removal and 119 have had bars removed. Long-term post bar removal follow up was available in 108 patients (56.8%), with a median follow up time of 4.6 years (IQR, 3.4 years) (MIRPE 50.6%, hybrid MIRPE 46.7%, complex hybrid reconstruction 65.6%, ATD 63.6%). Several patients with prior Ravitch procedures and complex reconstructions have experienced continued issues. Two have had enough recurrence and symptom return after removal of the Nuss bars at 3 years that the Nuss bars have been replaced in both for an extended period. One complex reconstruction has fractured methylmethacrylate and is contemplating reoperation. One has had persistent malunion and replacement of fractured plating with subsequent infection that occurred over 10 years after the initial procedure. Another patient has pursued a repeat procedure with another surgeon with reported resection of additional cartilage. Two patients after revision of prior MIRPE with MIRPE have reported some cosmetic changes consistent with regression but have not pursued imaging or further evaluation.

Discussion

Recurrence of PE deformities occur after both open and MIRPE procedures and risks are based on multiple factors which differ depending on the initial repair. In our series, nearly 16% of the patients had undergone at least two prior failed attempts at repair. Understanding why a procedure failed and identifying the contributing factors to a previous procedure’s failure is critical to performing a revision and minimizing the risks of another recurrence. Some of the main causes of primary MIRPE failure include suboptimal bar length, and insufficient number of bars. These can be avoided by customizing the hardware based on every patient’s needs and by having precise pre- and intraoperative chest length measurements. Additionally, inadequate bar fixation and deficient placement of bars can be minimized by accurately implementing fixation methods mentioned earlier. As for premature bar removal, this can be avoided by giving patients at least 3 years with bars in to have the defect permanently corrected. On the contrary, causes for open primary repair failures included over resection of cartilage, which can be prevented by executing a preservative resection approach that takes small portions out as needed till sufficiency is observed and careful preservation of perichondrium. Despite best efforts, continued failure can occur when the integrity of the chest wall has been violated. Considering the increased operative complexity associated with redo cases, patients need to be properly educated about the increased surgical complications and length of recovery period. Realistic expectations of results should be frankly communicated.

There are several reasons a patient will present for revision surgery. Nearly half of the prior MIRPE patients in this cohort (Table 2) still had bars indwelling and revision was secondary to either rotation of the bars, posterolateral migration due to intercostal muscle stripping, or dissatisfaction with an incomplete repair due to an inadequate number of bars. Of the patients undergoing bar removal, 69% had a single bar for their failed repair (Table 3). At the time of revision repair, most redo patients received at least two bars with nearly 50% receiving three bars to complete the repair. In both groups (prior open repair and prior MIRPE), chronic pain was also a significant factor for pursuing reoperation (54.4% in prior MIRPE, 51.7% in prior open). Pain is always difficult to predict for improvement with an operation, however, the presence of malunion with chest wall instability or malpositioned bars were a definite source of pain that had a chance for improvement postoperatively. A standardized protocol is also of paramount importance in achieving good pain control which included gabapentin, ibuprofen, acetaminophen, and narcotics with local anesthesia through the use of cryoablation in the most recent years (Table 1). Significant pain issues were reported to persist in less than 9% at follow up and in most patients, pain was reported as overall significantly improved. It is important to note that chronic pain issues were greater in the cohort of prior open procedures when compared to prior MIRPE cases at long term follow up after bar removal. Notably, significant pain was reported in 13.6% of patients with prior failed open repairs, in contrast to 3.6% of patients with a prior failed MIRPE. There is minimal data published on long-term chronic pain after revision procedures. Emory reported on 13 patients (ages 16–54 years) who were repaired after both failed prior Ravitch and MIRPE. Chest pain was present in 52% of their patients preoperatively and all were repaired with a modified Ravitch or complex reconstruction. No patient indicated the reoperation made the condition worse and relief of symptoms was good or excellent in 92% although there was no time mentioned for the follow up interval (25).

Most publications related to redo PE procedures are small cohorts without long term follow up (4,9,10,26). There are only a few large series publications devoted exclusively to repair of recurrent pectus deformities, with most studies including pediatric patients and a limited number of adults (7,27-30). Many of these reports describe experiences with a single operative technique. The Children’s Hospital of the King’s Daughter reported on 100 patients undergoing repair for recurrent PE with MIRPE for both prior failed open and MIRPE cases (15). They too found that multiple bars were required in the revision cases. A significantly higher bar displacement rate was experienced in their series (12% Ravitch, 7.8% Nuss), especially after the prior Ravitch cases, which may be secondary to the extreme rigidity of the post repair Ravitch calcified chest, and the pressure exerted on the MIRPE bar. Hopkins also reported a 5.4% bar displacement rate in their MIRPE cases following predominantly Ravitch operation and theorized that this “reflected differential intensity and distribution of mechanical forces across the rigid and widely deformed thorax” (31). In cases of extreme rigidity, the decision was made to proceed with a hybrid approach utilizing multiple bars to balance the pressure. Although we did have one lower of three bars rotate slightly in one patient, no other cases of bar migration occurred that required surgical reintervention.

Other authors have advocated that when faced with re-operative cases, an alternative operative approach from the initial procedure should be considered, i.e., if a MIRPE approach has failed, then a Ravitch procedure should be considered for the recurrent PE. Antonoff et al. reported on 27 patients from the University of Minnesota (26). They noted that 21.1% of patients undergoing repeat open repairs and 33.3% of patients undergoing repeat MIRPE required further operative interventions. There was no need for additional repairs among patients who had open repairs after MIRPE, nor for any patients who had MIRPE after open repairs. Despite these recommendations, it should be clear that if a malunion is present from a prior open repair, performing a MIRPE will still lead to another failure when the bars are removed.

Overall, ATD is a complicated disorder that occurs when the Ravitch repair is performed at too young of an age before the chest wall has completed growth (generally <7 years). These patients have impairment of the anterior chest wall growth due to resection of cartilage, which was first described by Haller in 1996 (32). Patients with ATD should never be confused with a recurrent Ravitch as it is a complicated disorder with high risks for reconstruction. The chest wall is calcified and solid and the underlying cardiopulmonary deficits are severe and have limited reversibility. In ATD patients, elevation of the chest with a MIRPE bar alone or a simple Ravitch may not improve the patient’s underlying deficits or circumstances as expansion of the anterior chest wall may be necessary to obtain adequate space to relieve cardiac compression (13,33). Of our prior open cases, 22 patients developed ATD and required complex reconstructions. These procedures included chest wall expansion with extensive reconstruction and their postoperative course was often fraught with respiratory and cardiac instability (13,34). Of this cohort, seven patients required reintubation secondary to pulmonary edema or respiratory failure with four requiring tracheostomies due to prolonged ventilatory support. Pressure support in the initial postoperative period was not uncommon. Persistent significant pain was an issue in this group and was present preoperatively in 22.7% and continued long-term in 17.6% (Table 5).

Postoperative complications following redo interventions, as shown in Table 4, were higher than what has been published in the literature for primary procedures (3,35,36). The presence of extensive pleural and mediastinal adhesions which required lysis to place intrathoracic bars was a significant issue in more than 55% of the cases (74.2% MIRPE/80.0% hybrid MIRPE) (Table 3). Adhesiolysis increased operative times before beginning the actual pectus repair and increased bleeding risks and subsequent air leaks. Redo MIRPE cases had the longest duration of adhesiolysis due to the need to provide unrestricted thoracoscopic access. Despite this, operative times were shortest in the MIRPE approach for the revision cases, however, this was still significantly longer than a primary procedure (36-38). Redlinger similarly reported (15) higher and significant complications in their complex reconstructions after prior open procedures including two intraoperative cardiopulmonary arrests with one ending in subsequent global neurologic deficit. In a series at John Hopkins evaluating 85 repairs (median age 21.3 years), limited complications were reported however, a high incidence of pneumothorax (67%) and pleural effusion (42%) were discussed as well as two MIRPE cases complicated by hemothorax (31). They noted recurrence in seven patients after open repair (three repaired with MIRPE and four open repairs) in follow up.

In our series, we followed an algorithm approach to all these cases as described in Figure 6. It was rare that a prior open procedure, even without ATD, was repaired with a MIRPE alone (15%). Failure to identify areas of malunion or pseudoarthrosis will lead to recurrence in the future. Although the chest will elevate easily in these cases, there remains instability, and once the Nuss bars are removed, the elevation will not be sustained. Revision after open procedures were by far the most complicated and a variety of complex hybrid reconstructions were often required. This was most commonly due to the presence of malunion or pseudoarthrosis that required reopening and exposing the chest wall to excise the area, reconstruct, reapproximate, and stabilize with plating. Cadaveric bone graft, methylmethacrylate, mesh of various types, and a multitude of titanium plates were used to reconstruct patients. In cases of sternal floating or missing chest wall bone, stabilization of every rib may not be possible due to the extensive damage done to the chest wall structure. We recommend a posterior fixation to keep the anterior chest wall elevated and stable and prefer the use of intrathoracic Nuss bars over an anterior strut (Figure 13A,13B). The goal in these cases was to provide stabilization as possible with realignment of the chest in the proper position providing relief of cardiac compression. Despite attempts to stabilize malunion, it persisted in some patients and reoperation was required in some to revise plating as it will fatigue and fracture over time if the patient does not stabilize. Despite reconstruction efforts, the prior Ravitch/open cases long-term were plagued by further issues including persistent malunion, recurrence, and chronic pain. The integrity of the chest wall and the ability to heal is limited in some patients that have had over resections. Overall outcomes in this group of revision cases can be difficult to predict and the greater the prior chest wall resection and remaining instability, the higher the risk the revision repair will fail. The goal of success becomes stabilization and “better” as complete reconstruction to “normal” may be not possible.

Figure 13 A patient presenting after two failed Ravitch procedures with over resection of cartilage that failed to regenerate. (A) Cross sectional chest computerized tomography showing extensive missing cartilage and sternal floating. (B) A complex hybrid reconstruction was performed for revision repair. Cadaveric bone graft was used as interposition graft between the lateral ribs and sternum to reattach and multiple titanium plates for fixation and stabilization. Intrathoracic Nuss bars were placed in the standard fashion and secured with FiberWire to support the chest wall anteriorly.

The limitations of this study include its retrospective nature, potential selection bias, and reliance on a single surgeon’s expertise at a high-volume institution. Additionally, there was a change in postoperative pain control at the end of 2018, with the initiation of intercostal cryoablation.

Conclusions

To successfully repair a recurrent or failed PE surgery, an assessment must be made of why the patient’s repair was previously unsuccessful to prevent repeating similar errors. Both open and minimally invasive techniques have been described for repair of recurrent PE and both approaches can offer advantages in the repair of recurrent defects. Some recurrent defects may be repaired with a hybrid approach including an application of both open and minimally invasive repair techniques to achieve optimal outcomes. Regardless of the approach, repair of recurrent PE has increased technical difficulties, longer hospital stays, and higher complication rates than primary repairs.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-417/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-417/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-417/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-417/coif). D.E.J. serves as an unpaid editorial board member of Journal of Thoracic Disease from February 2023 to January 2025. D.E.J. is a consultant with IP/royalty rights under Mayo Clinic Ventures with Zimmer Biomet, Inc. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the institutional review board of Mayo Clinic Arizona (No. 23-010389) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Nuss D, Obermeyer RJ, Kelly RE Jr. Pectus excavatum from a pediatric surgeon's perspective. Ann Cardiothorac Surg 2016;5:493-500. [Crossref] [PubMed]
2. Krasopoulos G, Dusmet M, Ladas G, et al. Nuss procedure improves the quality of life in young male adults with pectus excavatum deformity. Eur J Cardiothorac Surg 2006;29:1-5. [Crossref] [PubMed]
3. Kelly RE Jr, Mellins RB, Shamberger RC, et al. Multicenter study of pectus excavatum, final report: complications, static/exercise pulmonary function, and anatomic outcomes. J Am Coll Surg 2013;217:1080-9. [Crossref] [PubMed]
4. Ellis DG, Snyder CL, Mann CM. The 're-do' chest wall deformity correction. J Pediatr Surg 1997;32:1267-71. [Crossref] [PubMed]
5. Jaroszewski DE, Ewais MM, Lackey JJ, et al. Revision of failed, recurrent or complicated pectus excavatum after Nuss, Ravitch or cardiac surgery. J Vis Surg 2016;2:74.
6. Kelly RE, Goretsky MJ, Obermeyer R, et al. Twenty-one years of experience with minimally invasive repair of pectus excavatum by the Nuss procedure in 1215 patients. Ann Surg 2010;252:1072-81. [Crossref] [PubMed]
7. Croitoru DP, Kelly RE Jr, Goretsky MJ, et al. The minimally invasive Nuss technique for recurrent or failed pectus excavatum repair in 50 patients. J Pediatr Surg 2005;40:181-6; discussion 186-7. [Crossref] [PubMed]
8. Ashfaq A, Beamer S, Ewais MM, et al. Revision of Failed Prior Nuss in Adult Patients With Pectus Excavatum. Ann Thorac Surg 2018;105:371-8. [Crossref] [PubMed]
9. Ishimaru T, Kitano Y, Uchida H, et al. Growth spurt-related recurrence after Nuss procedure. J Pediatr Surg 2009;44:E13-6. [Crossref] [PubMed]
10. Willital GH, Meier H. Cause of funnel chest recurrences---operative treatment and long-term results. Prog Pediatr Surg 1977;10:253-6.
11. Park HJ, Chung WJ, Lee IS, et al. Mechanism of bar displacement and corresponding bar fixation techniques in minimally invasive repair of pectus excavatum. J Pediatr Surg 2008;43:74-8. [Crossref] [PubMed]
12. de Beer SA, van Heurn ELW. Sternocostal instability after Ravitch repair in adolescents; 3 case-reports and a review of surgical techniques in the literature. Plast Reconstr Surg Glob Open 2020;8:e2720. [Crossref] [PubMed]
13. Jaroszewski DE, Notrica DM, McMahon LE, et al. Operative management of acquired thoracic dystrophy in adults after open pectus excavatum repair. Ann Thorac Surg 2014;97:1764-70. [Crossref] [PubMed]
14. Arn PH, Scherer LR, Haller JA Jr, et al. Outcome of pectus excavatum in patients with Marfan syndrome and in the general population. J Pediatr 1989;115:954-8. [Crossref] [PubMed]
15. Redlinger RE Jr, Kelly RE Jr, Nuss D, et al. One hundred patients with recurrent pectus excavatum repaired via the minimally invasive Nuss technique--effective in most regardless of initial operative approach. J Pediatr Surg 2011;46:1177-81. [Crossref] [PubMed]
16. Miller KA, Ostlie DJ, Wade K, et al. Minimally invasive bar repair for 'redo' correction of pectus excavatum. J Pediatr Surg 2002;37:1090-2. [Crossref] [PubMed]
17. Kocher GJ, Gstrein N, Jaroszewski DE, et al. Nuss procedure for repair of pectus excavatum after failed Ravitch procedure in adults: indications and caveats. J Thorac Dis 2016;8:1981-5. [Crossref] [PubMed]
18. de Loos ER, Daemen JHT, Coorens NA, et al. Sternal elevation by the crane technique during pectus excavatum repair: A quantitative analysis. JTCVS Tech 2021;9:167-75. [Crossref] [PubMed]
19. Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg 2009;250:187-96. [Crossref] [PubMed]
20. Haller JA Jr, Kramer SS, Lietman SA. Use of CT scans in selection of patients for pectus excavatum surgery: a preliminary report. J Pediatr Surg 1987;22:904-6. [Crossref] [PubMed]
21. Croitoru DP, Kelly RE Jr, Goretsky MJ, et al. Experience and modification update for the minimally invasive Nuss technique for pectus excavatum repair in 303 patients. J Pediatr Surg 2002;37:437-45. [Crossref] [PubMed]
22. Jaroszewski DE, Velazco CS. Minimally Invasive Pectus Excavatum Repair (MIRPE). Oper Tech Thorac Cardiovasc Surg 2018;23:198-215.
23. Johnson KN, Jaroszewski DE, Ewais M, et al. Hybrid Technique for Repair of Recurrent Pectus Excavatum After Failed Open Repair. Ann Thorac Surg 2015;99:1936-43. [Crossref] [PubMed]
24. Aly MR, Farina JM, Botros MM, et al. Minimally invasive repair of pectus excavatum in adults: a review article of presentation, workup, and surgical treatment. J Thorac Dis 2023;15:5150-73. [Crossref] [PubMed]
25. Luu TD, Kogon BE, Force SD, et al. Surgery for recurrent pectus deformities. Ann Thorac Surg 2009;88:1627-31. [Crossref] [PubMed]
26. Antonoff MB, Saltzman DA, Hess DJ, et al. Retrospective review of reoperative pectus excavatum repairs. J Pediatr Surg 2010;45:200-5. [Crossref] [PubMed]
27. Guo L, Mei J, Ding F, et al. Modified Nuss procedure in the treatment of recurrent pectus excavatum after open repair. Interact Cardiovasc Thorac Surg 2013;17:258-62. [Crossref] [PubMed]
28. Miller DL. Reoperative Pectus Repair Using Biomaterials. Ann Thorac Surg 2020;110:383-9. [Crossref] [PubMed]
29. Wang L, Zhong H, Zhang FX, et al. Minimally invasive Nuss technique allows for repair of recurrent pectus excavatum following the Ravitch procedure: report of 12 cases. Surg Today 2011;41:1156-60. [Crossref] [PubMed]
30. Pisón CJ, González AG, Pérez MA, et al. Correction of recurrent pectus excavatum post-Ravitch with the Nuss technique. Cir Pediatr 2009;22:77-80.
31. Sacco Casamassima MG, Papandria D, Goldstein SD, et al. Contemporary management of recurrent pectus excavatum. J Pediatr Surg 2015;50:1726-33. [Crossref] [PubMed]
32. Haller JA Jr, Colombani PM, Humphries CT, et al. Chest wall constriction after too extensive and too early operations for pectus excavatum. Ann Thorac Surg 1996;61:1618-24; discussion 1625. [Crossref] [PubMed]
33. Davis JT, Long FR, Adler BH, et al. Lateral thoracic expansion for Jeune syndrome: evidence of rib healing and new bone formation. Ann Thorac Surg 2004;77:445-8. [Crossref] [PubMed]
34. Sacco Casamassima MG, Goldstein SD, Salazar JH, et al. Operative management of acquired Jeune's syndrome. J Pediatr Surg 2014;49:55-60; discussion 60. [Crossref] [PubMed]
35. Jawitz OK, Raman V, Thibault D, et al. Complications after Ravitch versus Nuss repair of pectus excavatum: A Society of Thoracic Surgeons (STS) General Thoracic Surgery Database analysis. Surgery 2021;169:1493-9. [Crossref] [PubMed]
36. Brungardt JG, Chizek PW, Schropp KP. Adult pectus excavatum repair: national outcomes of the Nuss and Ravitch procedures. J Thorac Dis 2021;13:1396-402. [Crossref] [PubMed]
37. Li S, Tang ST, Tong Q, et al. Nuss repair of pectus excavatum after surgery for congenital heart disease: experience from a single institution. J Thorac Cardiovasc Surg 2014;148:657-61. [Crossref] [PubMed]
38. Tikka T, Kalkat MS, Bishay E, et al. A 20-year review of pectus surgery: an analysis of factors predictive of recurrence and outcomes. Interact Cardiovasc Thorac Surg 2016;23:908-13. [Crossref] [PubMed]