Minimization of the complications associated with bar removal after the Nuss procedure in adults

Pectus excavatum (PE) is the most common congenital chest wall deformity; it is characterized by a caved-in appearance of the anterior chest [1]. The incidence rate is approximately 0.1% with a male to female ratio of 4:1. It is associated with connective tissue disorders, neuromuscular diseases, and some genetic conditions [1, 2]. In 1998, Nuss and colleagues documented the minimally invasive repair of pectus excavatum (MIRPE), also known as the Nuss procedure, a minimally invasive method for the correction of PE [3]. The reported advantages of MIRPE include the limited number of incisions, short operative time, less blood loss, and shorter duration of hospital stay and recovery. The procedure initially involves the introduction of one or more curved stainless-steel bars behind the sternum to correct the chest wall without resection of the costal cartilages. Considering that the procedure was widely accepted, it has been used extensively in children, adolescents, and adults in the past decade [4‐9].

After more than 2 years of correction, the supporting bars need to be removed. In the past, removal of the correction bars was considered to be a simple surgery and patients were even discharged on the day of the removal. However, there have been successive reports of severe complications, including massive hemothorax and death [10‐19]. Therefore, some physicians suggested improvements to the surgical equipment and methods used for the removal to reduce surgical complications in a recent decade [11, 17‐24]. Although some recent studies have discussed the occurrence of these surgical complications, disparate results were reported in large patient cohorts [11, 18‐20, 22]. The statistical data from a large-scale questionnaire survey of physicians from the Chest Wall International Group (CWIG) who have performed such surgeries showed that the incidence of these surgical complications was underestimated and that some severe complications have not been reported [16]. Many previous studies have examined bar removal in teenagers and younger patients after about 2 years of Nuss repair; however, there have been only a few statistical analyses of correction bars inserted in adults after a long time of repair. The CWIG data revealed that more than half of the physicians felt that it was easier to perform surgeries on younger patients. Some modifications of surgical skill for pectus bar removal have been described for reducing the complications [17‐24]. Therefore, in this study, we aimed to assess the differences in our modifications on skill and the recent surgical data on pectus bar removal. The methods to decrease intraoperative and postoperative complications were also discussed.

A total of 283 patients who underwent pectus bar removal were included in this study. The male to female ratio was 260:23. Their ages at the initial Nuss repair ranged from 12 to 53, with an average of 22.8 years. In total, 221 patients underwent repair as adults (≥18 years; Group A) while 62 patients underwent repair as adolescents (12–17 years; Group B). The mean pectus bar maintenance interval was 4.3 years (range: 1.9 to 9.8 years) while the mean age upon removal of the bar(s) was 26.9 years (range: 15 to 57 years). A total of 203 patients (71.7%) had two or more bars while 80 patients (28.3%) had only one bar. The mean operative time was 66 min (range: 20 to 187 min) and the mean estimated blood loss was 11.7 mL (range: 10 mL to 100 mL). Ten patients had perioperative blood loss of 50–100 mL. The demographic and perioperative clinical features of Groups A and B are shown in Table 1. Group A had a longer period of correction for Nuss repair than Group B (P = 0.010). Overall, the complication rate was 3.2% (9/293), including pneumothorax in six patients (2.1%) and wound seroma/hematoma in three patients (1.1%). The complication rate of Groups A and B showed no significant difference (2.7% vs 4.8%, P = 0.400). Others showed no significant association between these two groups. No massive hemothorax, wound hematoma, blood transfusion, or other life-threatening events occurred during or after the operation.

Analysis of the association of other surgical factors was demonstrated in Table 2. The results showed that callus formation around the ends of the bars required a significantly longer operative time and caused more perioperative blood loss (P = 0.032 and P = 0.046, respectively). Patients with a higher BMI (> 22 kg/m²) were more likely to have a longer operative time (P = 0.048). The other factors showed no significant difference.

Pectus bar removal after the Nuss procedure is considered simple and can be performed as an outpatient procedure. Massive bleeding related to bar removal is rare; however, it may be life-threatening. Myocardial injury, lung laceration, and aortic laceration with massive bleeding after bar removal has been reported [12‐15]. Some patients have also died after bar removal; however, this statistic is underreported [16]. Considering patient safety and any unexpected major complications, pectus bar removal is suggested to be performed as an inpatient procedure.

Complications related to bleeding are a major concern during pectus bar removal, although it has only been reported in less than 1% of the cases. Bleeding from the heart, great vessels, chest wall, internal mammary artery, intercostal artery, or callus around the pectus bar has been reported [10, 18]. Patients with a history of intrathoracic infection or pericarditis have higher risks for life-threatening complications [12‐14]. Most bleeding could be controlled conservatively; however, internal control of bleeding should be considered if massive intrathoracic bleeding occurred. The other common complication after bar removal is wound seroma/hematoma with an incidence rate of approximately 2.36–11.98% [16, 18, 22]. It can also be from the callus or scar tissues of the wound. Adult patients and patients treated using more than one bar were reported to have a higher complication rate [9, 14]. The major and lethal complications that had been observed with the Nuss Procedure were related to the migration of the bar [3, 17]. The obvious displacement of the pectus bar would increase the difficulty and risk of removal of the correction plate; therefore, how to properly fix the bar is also an important issue [15, 17].

Various modified methods and their related risks have been described for reducing the complications involving the patient’s history, specially designed and customized instruments, or modifications to the procedure (see Table 3) [11, 17‐25]. The modifications in the surgical procedure could be classified into four major categories: (1) patient’s position during surgery, (2) wound incisions, (3) bending of the end of the bar before removal, and (4) bar rotation during removal [24]. The positions are classified as supine, prone [22], lateral [21], and special two-table supine positions [20, 23]. All our patients were in the supine position during the operation. We suggest adopting the supine position in a single bed (Fig. 1) which is relatively easy to prepare and can be used to smoothly cooperate with subsequent surgical procedures. Incisions are divided into unilateral (right or left side) or bilateral incision. We suggest that a bilateral incision be made on both sides of the previous Nuss procedure to better observe whether a callus has formed or if the steel wire is broken on both sides of the steel plate. If procedures are present with more than one bar insertion or bar fixed on both ends, a bilateral approach is necessary. The steel plate is suggested to be removed on the same side from which the steel plate was placed previously. The approach for bending the end of the pectus bar is divided into unilateral, bilateral, or no bending. We suggest using unilateral partial bending, i.e. bending one side of the steel plate first and then removing the steel plate from the opposite side, which can reduce the risk of injury to viscera in the thoracic cavity and pleural cavity caused by the bent steel plate. Regarding the device, the pectus removal bender from the original manufacturer would suffice. Rotation is divided into partial or no rotation. We suggest no rotation during removal to prevent unnecessary injuries or bleeding. Besides, we suggest wrapping an elastic bandage around the chest for 3 days after the surgery to reduce the dead space and prevent wound hematoma formation (Fig. 2). In addition, most of our patients were adults, and the majority of patients had more than one pectus bar placed. These were considered the risk factors associated with higher complications. In our results, the adults and number of bars removal were not risk factors.

In our experience, no major complications occurred after the procedure. There was no statistically significant correlation between the age groups and the number of pectus bar removed. Postoperative pneumothorax was found in six patients. One patient needed intraoperative pleural drainage due to one end of the bar being trapped to the lungs. Others had no clinical symptoms or an increase in the hospital stay. Pneumothorax mostly occurred due to air entering the pleural space via the wound during dissection for the bar removal. It can be resolved progressively without drainage if there are no other clinical symptoms. Otherwise, if the pneumothorax is due to the lung injury, pleural drainage should be done.

The mean intraoperative blood loss was less than 12 mL. Ten patients had perioperative blood loss of 50–100 mL. The bleeding was from the callus or newly grown vessels and was controlled carefully. No patient needed blood transfusion. We found that callus formation causes significantly more perioperative bleeding and longer operative time (Table 2). We suggest that with severe callus formation around the ends of bar, careful dissection and bleeding control be undertaken. Furthermore, to prevent traction of the mediastinal adhesions when pulling out the bar blindly, the callus covering the ends of bar should be completely removed and the bar should be removed along with the shape of the thoracic cage without rotation. Life-threatening complications such as organ injuries and major bleeding could be avoided.

The insufficient number of patients is the main limitation of this study. Other limitations include the risk of bias inherent to the retrospective design of the study and the inability to generalize the results and conclusions to other populations. Bleeding volume during surgery was also not recorded carefully. Except for a large amount of bleeding, the estimated bleeding volumes were recorded as < 10 mL. To facilitate statistics, these were calculated as 10 mL. Therefore, the average bleeding volume in this study is higher than that in other studies.

Adult patients and removal of multiple bars were not the risk factors for the occurrence of complications in our cohort. Skills in preventing intraoperative mediastinal traction, carefully controlling bleeding, and reducing the effect of dead space around wounds could minimize the risk of bleeding complications. A multicentric study or the accumulation of more cases is needed to further evaluate the risk factors of removal pectus bar(s).