Management of Sternal Wounds,Infections,and Sternal Non-Union with Plate Fixation: Result from a Single Site Experience

Patients and Methods

Technique

Sternal rigid plate fixation was performed using the following methods. For patients who were believed to have infection, the surgical approach was staged starting with one to five washouts (median of approximately 1.5) then followed by SRPF. The staged approach began with incision, drainage, and washout of the sternum, exposing the bone and existing hardware. Once exposed, any hardware that was overtly infected or loose, was removed. The sternal medullary surface was then sharply debrided with a curette and devitalized cortical bone was removed with a rongeur until healthy bleeding bone was achieved. The sternum was conservatively debrided. The sternum was largely preserved and total or subtotal sternectomy was not required on any patient. Calcium sulfate beads impregnated with antibiotic agents (vancomycin, tobramycin, and occasionally micafungin) were implanted in the wound at each washout. A wound vac and Jacob's ladder were then applied until the following procedure. The process was repeated as needed, with removal of previous antibiotic beads and placement of new ones during each washout. The end point of washouts was based on the wound appearing clean with bleeding edges to the sternum.

Removal of existing sternal cables or wires was also performed at the washouts although starting in 2020, the primary approach was to leave at least some of the hardware in place to maintain sternal reduction and stability to some degree until the final plating procedure at which time all the existing wires or cables were removed and replaced at the same setting with SRPF.

Once the wound was determined to no longer be infected, the final surgery (or only surgery in the case of patients without infection) was the internal fixation of the sternum. Internal fixation began with debridement and removal of the previously placed antibiotic beads, or removal of hardware for patients who only had primary non-union. The wound was thoroughly debrided and irrigated, and antibiotic beads were placed. The sternum was carefully reduced with a combination of techniques such as bicortical or monocortical wires, sternal reduction clamps, and point to point reduction clamps.

Once reduced, the sternum was plated with the Synthes Matrix Titanium Sternal Fixation System (DePuy Synthes, Monument, CO), with the goal of creating longer more stable constructs (Fig. 1). Plates were placed starting with a star plate, then progressively longer straight plates that extended onto the ribs. They were bent to the contour of the chest including in-plane bends. Screws were 2 mm thicker than the thickness of the bone (based on either intra-operative measurements or computed tomography prior to surgery) and were fixated using a torque-limiting automatic screwdriver. After the plates were set, any bone gaps were filled with calcium phosphate bone void filler. The pectoralis major muscle flaps were then raised and inset, with a JP drain deep to the muscle. The flaps were inset with progressive tension sutures laterally and to the midline and down to the hardware. The subcutaneous tissue and skin were then closed. Patients with SRPF and latent infections were treated with removal of plates at a time when the sternum was stable, and no additional secondary flaps were required.

OPEN IN VIEWER

FIG. 1.

Example of finished sternal rigid plate fixation.

Results

From the 97 patients included, 29 patients were CLEAN and 68 were INFECTED. From the CLEAN patients there were four with post-SRPF infections and 25 without. And of the INFECTED patients there were 11 with late or recurrent post-SRPF infection and 57 without.

Other than the categorization of wound type, patients were evaluated for risk factors and were found to be of relatively equal risk between the CLEAN and INFECTED groups (Table 1). When comparing CLEAN patients with INFECTED patients, the only difference found in comorbidities was in creatinine, where 86.21% of CLEAN patients had creatinine levels in a normal range, versus only 66.18% of INFECTED patients (p = 0.044). INFECTED patients were also found to be older than CLEAN patients, with an age of 63.15 ± 12.08 years old versus 57.76 ± 12.03 years old (p = 0.047).

The majority of patients were male (82.76% of CLEAN patients and 72.06% of INFECTED patients) and white (82.76% of CLEAN patients and 79.41% of INFECTED patients). Patients had an average BMI of 32 ± 7.4 kg/m². Eighty percent overall showed presence of hypertension, 27.59% of CLEAN and 42.65% of INFECTED had diabetes mellitus, 18% overall had COPD, and 6.9% of CLEAN and 17.65% of INFECTED had a history of thrombosis. Most patients also had a coronary artery bypass graft (44.83% of CLEAN and 41.18% of INFECTED) as the surgery that precipitated the need for sternal reconstruction.

When comparing which patients developed infection after SRPF and which patients did not, the only difference that can be seen is in smoking status (Table 2). In patients who did develop infection after SRPF compared with those who did not, 13.32% versus 6.10% were current smokers, 20% versus 46.34% were former smokers, 60% versus 47.56% were non-smokers, and 6.67% versus 0% were rarely smokers (p = 0.031).

Additionally, characteristics related to wound classification prior to surgery, and which were different between CLEAN and INFECTED were not different when looking at development of infection after SRPF (redo p = 0.001 vs. 0.515, number of washouts p < 0.001 vs. 0.481, total length of stay p < 0.001 vs. 0.362, and post-SRPF length of stay p = 0.046 vs. 0.367).

Sixteen percent (11/68) of INFECTED patients went on to develop late or recurrent infection, versus 14% (4/29) of CLEAN patients who developed infection after SRPF (p > 0.05). There were an additional three patients who had plates removed for pain or hardware failure, but those patients had no signs of infection at the time of removal. All post-SRPF infections developed within six months of surgery.

Over time, the rate of infections decreased (p < 0.001; Fig. 2). This may have correlated to increased experience as the number of cases increased year over year (p < 0.0001). Looking specifically at INFECTED patients, the same held true, with an increase in patients plated, and a decrease in the rate of recurrent infections (p = 0.010 vs. p < 0.0001; Fig. 3). Of the 11 open/infected patients who had hardware removed, the only common original infection was Pseudomonas aeruginosa in two patients, and three had cultures with no growth. Of the 15 patients who had infection after SRPF, almost all infections were one of three, Candida albicans, Pseudomonas aeruginosa, or Staphylococcus epidermidis. Only two patients had the same infection prior to and after SRPF, one with Candida albicans and one with Pseudomonas aeruginosa.

OPEN IN VIEWER

FIG. 2.

Graph of the total number of patients each year, with and without infection after sternal rigid plate fixation (SRPF). As the number of patients plated each year increases (p < 0.001), the rate of infection decreases (p < 0.0001).

OPEN IN VIEWER

FIG. 3.

Graph of the number of patients with previously open or infected wounds each year, with and without infection after sternal rigid plate fixation (SRPF). Following the pattern of overall patients, as the number of infected patients plated each year increases (p = 0.010), the rate of re-infection decreases (p < 0.0001)

Discussion

Although SRPF has been shown previously to be successful in high-risk patients, presence of active infection is usually considered a strong contraindication.^2,20 Our study, however, found no statistical difference between those who developed infection post-SRPF and those who did not when comparing pre-SRPF wound classification. In our patients, presence of infection made no difference in development of infection after sternal reconstruction with SRPF.

When looking at sternal wounds, one of the most examined outcomes is development of mediastinitis. Many studies have tried to determine the important risk factors that lead to infection. The risk factors most often seen are age, gender, obesity, smoking status, COPD, amount of time spent in the hospital, and the type of surgery that led to the need for the SRPF.^2,22,23 Our study, however, only found smoking status to be different in patients who did develop infection after SRPF and those who did not. We also found that regardless of wound classification prior to SRPF, patients were equally at high risk in those same categories, except age, where those with previously open/infected wounds were likely to be older than those with clean wounds.

Whether or not the plating was the primary sternal closure or a redo, the number of washouts prior to SRPF, and total and post-SRPF length of stay in the hospital were all significantly different when looking at wound class prior to SRPF. These can all be attributed to wound classification, because CLEAN patients would not need to have washouts prior to SRPF, and fewer procedures would generally shorten a patient's length of stay in the hospital. Although variables such as these (length of stay, operation time, and re-operation) have been shown to be risk factors for development of infection, our study did not find that to be the case.^4,22,23

Most studies have found that sternal wound infections typically occur anywhere from 1% to approximately 5%, but rates of up to 8% or 9% have been found.^{1,2,15,23–26} The overall rate of infection and subsequent hardware removal in our study was higher at 15% (14% in CLEAN patients and 16% in INFECTED patients). Even though our rate was higher than what other studies have shown, our patient population that received SRPF was selected specifically for being at high risk for wound complications, due to their comorbidities. Additionally, over time the rate of infection each year declined with increasing experience and is currently at 11% (3/27) among all cases from 2020 to 2021.

It has been estimated that up to 2% of closures with steel wiring result in sternal instability, meaning more than 150,000 patients each year deal with non-union. ⁵ If there is enough movement, the wires can even cut into the sternum, which both increases the likelihood of infection rate and decreases the likelihood of union rate. ⁶ Traditionally, complications from sternal wounds were shown to lead to mortality rates as high as 47%. ⁶ Even those that believe wire closure generally has a low rate of complication, acknowledge that there are downfalls, especially in patients with multiple comorbidities. ⁷ Comfort, stability, and even earlier union, however, are achieved by the use of SRPF.^2,14

Although some flaps see early wound closure and reduced mortality, soft tissue reconstruction also sees problems. In rectus abdominis and especially omental flaps, there is a concern for hernia formation.^2,8–11 Recurrent infection and wound dehiscence still occur with muscle flaps. ⁸ One study even found that continued pain and discomfort of the chest and shoulder were found in 51% of patients, numbness was found in 44% of patients, and instability symptoms were found in 42.5% of patients, and it was confirmed in 45% of those patients who were examined. ¹²

By using SRPF in sternal reconstructions, the surgeon is able to use pectoralis major advancement flaps and avoid the need for turnover flaps. This maintains function, minimizes donor site morbidity, and minimizes operative time. Pectoralis turnover flaps require dividing insertion of the muscle at the shoulder causing the patient to lose function, whereas the advancement flaps maintain function by repositioning the origin of the muscle to the midline. Plating allows for preservation of the pectoralis muscle while also providing sternal stability and upper girdle function and strength.

Notably, we only had one patient who did not have a healed and stable sternum when their plates were removed because of infection. However, this patient was non-compliant, not using sternal precautions, and lost to follow-up for two months and presented only when the patient noted frank purulence from his JP drain that was still in place. Even when infection and removal occurred for our patients, we were still able to avoid the functional complications associated with sternectomy and flaps that result in functional loss such as pectoralis turnover or rectus muscle flaps.

This study is significant in that it is a large sample population in terms of SRPF reconstruction. Most studies looking at SRPF include only a handful of patients,^17–19,27 with some studies larger at 40 or 50.^16,28 These studies differ from ours in that the first only included patients who were plated after wire failure, and the second only included patients who were plated primarily.^16,28 The study that look at plating after wire failure looked generally at post plating outcomes. ¹⁶ The study that looked at primary platings compared the plating outcomes to those of wire closures during the same time period. ²⁸ Our study includes both primary platings and platings after wire failures (with and without infection) and compares their post-SRPF outcomes. Although there are some studies that have a few hundred or even thousands of patients, these were focused on other methods of sternal reconstruction.^8,26 Many of these studies are also large cardiac centers or multiple hospitals, and involve multiple surgeons and procedure types.^8,16,26,28 Although it limits the size of the study, because we are only looking at the patients of one plastic surgeon, we are able to create a study with more continuity for the procedures being performed.

Conclusions

Based on the results of our study, we believe most patients can be treated with SRPF, despite the presence of infection. Even when infection does occur after SRPF, the sternum is healed and stable in almost every case. More research is needed to repeat and elucidate the cause of the unexpected elevated primary plating/clean case infection rate, and if sternal plating is a viable standard to replace soft tissue only reconstructions in sternal infection cases with sternal instability.