Intramedullary reaming and irrigation and antibiotic-loaded calcium sulfate implantation for the treatment of infection after intramedullary nailing: a retrospective study of 19 cases

Infection after intramedullary nailing is uncommon, with a reported rate of 0.9 to 3.8% [1, 2]. In a retrospective analysis of more than 1000 cases of tibial shaft fractures treated with intramedullary nailing, the infection rate after treatment of closed fractures was 1.9%, and the infection rate after treatment of open fractures was 7.7% [3, 4]. Although the infection rate is not high, if an infection is not treated in a timely manner, complications including osteomyelitis, fracture non-union, physical disability, or even systemic sepsis are inevitable.

The management of this type of infection remains controversial [5, 6]. Makridis et al. [7] described 3 stages of infection after intramedullary nailing: Stage I; 2–6 weeks after operation manifesting as cellulitis, Stage II; 2–9 months after operation, manifesting as delayed wound healing, exudation, osteonecrosis, and pathological fracture, Stage III; 9 months or longer after operation manifesting as definite osteomyelitis. Each stage has its own management protocols. For the treatment of Stage I infections, a more conservative approach is widely accepted [5, 7]_. However, opinions on the best treatments for Stage II and III infections vary greatly [1, 7, 8], especially when the fracture hasn’t healed yet. Unfortunately, to date there are no uniform and standard treatment protocol for Stage II and III intramedullary infections [5, 9]. Thus, the management of infections after intramedullary nailing tend to be experience-based, rather than evidence-based protocols.

In our experience, an intramedullary nail as a foreign fixator lacks an adequate blood supply, which means systemic antibiotics cannot reach the interface of the nail to eradicate the biofilm bacteria effectively [10]. Furthermore, delayed or inappropriate treatments allow the pathogens to spread through the whole medullary canal resulting in a diffuse infection [10]. Therefore, we believe that nail removal and surgical debridement are vital for the treatment of Stage II and III infections after intramedullary nailing, as it destroys the biofilm produced by bacteria, and thus enhances the efficiency of antibiotics which in turn improves the infection remission rate. Surgical debridement mainly includes local debridement and intramedullary nail management. Intramedullary reaming and irrigation are important components of surgical debridement, since the process eliminates the endosteal sequestra of the canal, lowers the intraosseous pressure, improves vascularization of the bone, and removes the bacterial biofilm [11]. The effectiveness of reaming and irrigation for the treatment of intramedullary nail infections and osteomyelitis has been reported in prior studies [12‐14].

After debridement, there may be remaining bacterial residue in the marrow canal or the surrounding soft tissue, and it only takes about 72 h for a bacterial biofilm to develop and become mature [15]. Due to the poor blood supply and osteonecrosis in osteomyelitis, administration of systemic antibiotics is usually not enough to eradicate the infection, and sometimes have little effect [16]. Antibiotic-impregnated calcium sulfate is an absorbable local antibiotic delivery system, and exhibits excellent osteogenesis and drug-loading properties. The advantages of antibiotic-loaded materials include more accurate positioning, higher local antibiotic concentration, less side effects, and longer treatment duration [17]. Studies have shown good outcomes when antibiotic-loaded calcium sulfate is used for the treatment of bone infections [18, 19].

Based on literature data and our experience, we have developed a protocol for the treatment of infections after intramedullary nailing: intramedullary nail removal, local debridement, medullary canal reaming and irrigation, and antibiotic-loaded calcium sulfate implantation, with or without secondary osteotomy and distraction osteogenesis. The purpose of this study is to describe this technique and report our results using the protocol for treating infections after intramedullary nailing.

The records of patients diagnosed with an infection after intramedullary nail fixation and treated at our center from 2014 to 2017 were retrospectively analyzed. The inclusion criteria for study were: 1) Infection after initial intramedullary nail fixation and the intramedullary nail was in place when admitted to the hospital; 2) Treated as per our protocol described above; 3) Minimum follow-up of 24 months. The exclusion criteria were: 1) Patients with severe liver or kidney dysfunction, cardiovascular disease, or diabetes with uncontrolled blood glucose; 2) Patient was not treated with our described method; 3) Follow-up < 24 months. Diagnosis of infection after intramedullary nailing was based on the clinical criteria described by Metsemakers et al. [20], which include the presentation of 1) a fistula, sinus, or wound breakdown, or 2) purulent drainage from the wound. Of course, we also combined the presence of positive biochemical infection markers, imaging results, and culture and histology results of tissue samples collected during surgery.

Before surgical interventions, patients were informed with the details of treatment protocols, and the informed consents were signed by patients themselves. Our study was approved by the Ethical Committee of Guangdong Second Provincial General Hospital and Nanfang Hospital.

A total of 19 patients who met the inclusion criteria were included in the analysis. There were 15 men and 4 women with a mean age of 39.0 ± 10.1 years (range, 23 to 56 years). The femur was involved in 7 patients and tibia in 12 patients, and all fractures were due to some form of trauma (traffic trauma in10 cases, falling injury in 6 cases and heavy pound injury in 3 cases). Nine patients suffered from infection on right extremity while 10 patients on left extremity. All patients presented with a sinus at their first physical examination, and preoperative radiographs indicated that 10 patients had healed fractures and 9 patients had unhealed fractures. The mean preoperative ESR was 49.4 ± 34.2 mm/h, mean CRP level was 33.1 ± 23.8 mg/L, and the mean WBC count was 12.75 ± 6.63 × 10⁹/L. Patient preoperative data are summarized in Table 1. In all cases, no antibiotics were administered until after specimens were obtained for culture during surgery.

Thirteen patients required external fixation, and 9 of these patients received segmental bone resection and bone transport after the infection markers became normal, and started bone transport 1 week later (Fig. 3) aiming to repair the bone defect. The average bone defect was 4.7 ± 1.3 cm (range, 3.3 to 7.6 cm). Four patients who did not receive segmental bone resection were treated with external fixation to assist weight bearing because bone debridement was very extensive. Postoperatively, clindamycin, cephalosporins, or quinolones were the most commonly administered intravenous antibiotics, with a mean duration of 8.3 ± 3.2 days (range, 3–14 days). Conventionally, a total course of antibiotics includes 2 weeks of for intravenous administration, and an additional route and another 4 weeks of oral administration. In our protocol, we omitted the 4 weeks of oral antibiotic administration because we deemed the extremely high concentrations and fairly long curative duration produced by the local antibiotic delivery system was enough to eradicate residual bacteria. The mean time for normalization of infection markers after surgery was 3.4 ± 1.7 weeks (range, 2 to 8 weeks).

A total of 20 strains of bacteria were isolated from 17 cases, with a positive culture rate of 89.5% (17/19). Coagulase-negative Staphylococcus (30.0%, 6/20) was the most commonly isolated pathogen, followed by methicillin-resistant Staphylococcus aureus (MRSA) (25.0%, 5/20), methicillin-sensitive Staphylococcus aureus (MSSA) (25.0%, 5/20), and Escherichia coli (15%, 3/20). Polymicrobial infections were identified in 3 patients (15.8%, 3/19). The distribution of bacterial culture results are shown in Fig. 1.

With a mean follow-up of 38.1 ± 9.4 months (range, 24 to 55 months), 18 (94.7%) patients achieved infection remission after the first surgical treatment, while 1 patient (5.3%) developed infection recurrence 3 months after surgery and underwent segmental resection and bone transport. The representative cases were presented in Figs. 2 and 3. Twelve (63.2%) patients became completely pain-free, and 16 (84.2%) patients achieved full weight bearing during the follow-up period. One patient (5.3%) experienced a re-fracture 4 months after surgery, and was successfully treated with external fixation. Prolonged aseptic drainage was the most frequent postoperative complication, and occurred in 7 patients (36.8%). In all cases, the drainage was successfully treated with regular local wound care and dressing changes. Bone transport was successful in all patients, and the mean fixation duration was 10.7 ± 4.0 months (range, 6.7 to 19.5 months). attributing to early function rehabilitation, only 1 (5.3%) case was recorded with joint stiffness after bone transport. Surgery and follow-up data are summarized in Table 2.

Infection after intramedullary nailing is defined as a confined or diffuse infection of medullary cavity caused by the invasion of pathogens during intramedullary nailing. Although uncommon, it often leads to serious consequences if not treated timely and appropriately. Today’s treatment protocols of such disease vary widely according to the severity of infection and doctor’s experience, and a wide-accepted method remains unclear. In our study, we hold a more aggressive treatment method on the management of intramedullary infection since we put the infection controlling on a higher priority.

After surgical debridement and antibiotic-loaded calcium sulfate implantation, 94.7% of patients achieved infection remission after surgical treatment, while only 1 patient developed a recurrence of infection and required a second surgery. We speculate that the high remission rate is attributed to radical debridement and the use of a local antibiotic delivery system. This method thoroughly removes infected tissues and eliminates residual bacteria due to a high local antibiotic concentration and long treatment duration. Such remarkable result was similar to the previous reports. Kanakaris et al. [23] performed intramedullary nail removal, intramedullary debridement with a RIA device, and placement of antibiotic-loaded cement rods for the treatment of 24 patients with infections. The cement rods were removed once the infections were controlled, and with mean follow-up of 21 months 23 (96%) patients had no evidence of recurrent infection. The difference between their study and ours lies in the use of a RIA device and antibiotic cement. The RIA system is a device that was initially developed to prevent fat embolism and lessen the systemic inflammatory process after reaming the femur in nailing procedures [24, 25]. Due to its versatility, it has been expanded to the treatment of long bone osteomyelitis [26, 27]. Unfortunately, the RIA system is not available in our country; however, the method we used produces similar results as the use of a RIA system. The other difference between the studies is the topical antibiotic carrier. Although antibiotic-loaded cement used in medullary infection have been well-demonstrated in former studies [7, 23, 27], one of the shortcomings of such material was non-absorbable after implantation, thus required a second procedure to remove it because leaving the rods in place can increase the risk of recurrent infection. Additionally, the antibiotic level curve produced by such antibiotic cement rod has been shown to be unstable, which might lead to a sharp decrease of antibiotic level several days after implantation, resulting in incomplete eradication of the pathogens, or even becoming a nidus for bacterial colonization. In another case series, Qiang et al. [28] treated infections after intramedullary nailing in 19 patients with nail removal, reaming and irrigation, and antibiotic-loaded cement implantation. All 19 patients achieved remission from infection, although there was 1 case of non-union and 1 patient ultimately required amputation due to severe trauma. However, The shortcoming in his study was similar. As the PMMA cement was absorbable, a second surgery was necessary for cement removal.

Nine of our 19 patients were treated with additional segmental bone resection and bone transport, and all of patients achieved infection remission, and the lengths of the involved limbs were well-restored. Although our outcomes were satisfactory, the best protocol for the management of non-healing fractures is not clear and treatments are primarily based on the experiences of individual surgeons. Those who prefer to retain the nails believe that fracture healing is more important, and that management of intramedullary infection can be postponed until after bone union. On the other hand, some surgeons believe infection control should take priority. Thus, based on different concepts treatment protocols mainly include 1) local debridement and antibiotic administration and retaining the nail until after bone union, 2) nail removal, re-reaming, and replacement of a larger diameter intramedullary nail or a resorbable antibiotic coated nail [10, 23], and 3) nail removal, segmental bone resection, reaming and irrigation, and bone defect reconstruction with bone transport [29] or the Masquelet technique [30, 31]. We are inclined to a more aggressive treatment protocol, and in our opinion retaining the nail or replacing it with another internal fixation device after debridement is not suitable for infection control, because the residual pathogens and its biofilms might lead a higher potential for infection recurrence and treatment failure [6]. External fixation avoids this shortcoming and remains the recommendation for treatment of bone infections [6, 11]. Additionally, a persistent infection or infection recurrence might prevent the bone healing process [32], and even lead to diffuse osteomyelitis and resulting disability and amputation. External fixation combined with segmental bone resection and transport as a mature and efficient technique can be used to manage infection, nonunion and bone defects, and deformity at the same time.

Local antibiotic-carriers were proposed in 1970s [33, 34], and are currently recommended as a bone substitute in the management of bone defects, or as a local antibiotic carrier in the case of bone infection. In our study, we used an injectable antibiotic-loaded calcium sulfate, which overcomes the shortcomings of PMMA cement, such as difficulty in intramedullary placement, the need for a second surgery for removal, and an unstable antibiotic release curve. As an absorbable antibiotic carrier, calcium sulfate has a stable antibiotic release curve, and can maintain the local antibiotic level higher than the MIC for 6–8 weeks. The local concentration is 100 to 1000 times higher than the antibiotic levels resulting from intravenous administration [35], which is sufficiently high to penetrate the bacterial biofilm. Furthermore, calcium sulfate exhibits a similar microstructure to cancellous bone and after being absorbed a network structure remains and trabecular bone can be observed under a light microscope, which contributes to the growth and migration of blood vessels and bone cells [36, 37]. To the best of our knowledge, there are no other studies reporting the use of antibiotic-loaded calcium sulfate to treat the medullary canal after removal of an intramedullary nail because of an infection.

Prolonged aseptic drainage was the most frequent complication in our study, with a relatively high rate of 36.8%, with prior studies reporting rates ranging from 4.2 to 33% [38‐40]. In our experience, poor soft tissue coverage, scar formation, and excessive calcium sulfate implantation may be the reasons for the high incidence of postoperative exudation. Although this type of aseptic exudation is not a sign of infection, management is important as a persistently wet gauze can increase the risk of a wound infection. Generally, routine treatment of prolonged drainage includes regular dressing and wound care. Other effective methods to prevent prolonged aseptic drainage may include good soft tissue coverage and reduction of the amount of calcium sulfate implanted. Other complications of the treatment of intramedullary nail infection include hypercalcemia, debridement-related fracture, and post-operative pain and joint stiffness; however, these complications were rare in our series.

There are limitations of this study that need to be considered. Firstly, our outcomes were not compared with those of other surgical methods. In addition, some detailed patient data were not available because of the retrospective nature of the study, and this might influence the understanding of outcomes. Finally, our patients were heterogeneous, with different sites of infection (tibia or femur), infection with bone union or non-union, and receiving bone transport or not, all of which inevitably can lead to more complex outcomes. However, we have to point out that the emphasis of our study was to introduce an effective method to eliminate infection after intramedullary nailing, and from this perspective all patients received the same management and overall the outcomes were good.

Intramedullary nail removal, medullary reaming and irrigation, and antibiotic-loaded calcium sulfate implantation seems effective in the treatment of infection after intramedullary nailing. Additional prospective studies with larger case numbers are necessary to confirm our findings.