Surgical debridement in long bone chronic osteomyelitis: is wide tumour-like resection necessary?

Adequate debridement

Extent of adequate debridement is carried out based on imaging review. For CM type I cases, intramedullary nails are removed, canal reamed, and screw holes over drilled.

Access to the medullary canal is created for debridement of metaphyseal areas. A cloaca, if present, is extended to allow access as mentioned. Using multiple drill holes (2.7 mm), the channel is outlined and completed with an osteotome. Once with access, curettes and/or a burr are used, and reach is confirmed under fluoroscopy. At times, a dental mirror is useful to visualize difficult areas. This channel is created along the length of infected bone, and kept to 5 to 7 mm wide in long windows (> 10 cm), 7 to 10 mm in shorter windows (up to 10 cm), and 10 to 15 mm when muscle is used for biological dead space management.¹⁰ This process, compared to wide segmental resection, is tedious and time-consuming and risks an iatrogenic fracture. However, if done properly, extensive resection is avoided, maintaining skeletal stability.

The same is done for CM type II and III cases. The channel is created by connecting screw holes. Surfaces with biofilm are carefully decorticated using curettes, chisels, or burr. Previous screw tracts are over drilled. Less vital areas (i.e. minimal or no punctate bleeding in sclerotic bone) may be left behind to avoid destabilization, and are perforated with small drill holes to hopefully aid revascularization. Except for the paprika sign, bone viability is assessed using “point-of-care testing”, similar to Venter et al¹¹ (bone colour, texture, etc).

For CM type IV cases, the above mentioned techniques were also employed, with the addition of skeletal stabilization. Wide, tumour-like resection was not done in cases included in this study.

High-volume saline (without antiseptic) was used for irrigation, often using pulsatile lavage for metaphyseal intramedullary surfaces and screw holes to dislodge debris. Calcium sulfate (Stimulan Biocomposites, UK) or calcium sulfate plus hydroxyapatite (CERAMENT-G; Bonesupport, Sweden) were used as local antibiotic carriers and dead space management. In some cases, antibiotic coated intramedullary nails were used, and one case used bioactive glass.

Primary closure of soft-tissue was done if possible. Otherwise, plastic surgeons cover the defect with local or free flaps. Patients initially receive individualized postoperative systemic intravenous and then oral antibiotics.

Outcomes per cm type

A total of ten patients (18.5%) were CM type I: six previously with an intramedullary nail, three from hematogenous spread, and one after knee arthroscopy. Mean follow-up was 34.8 months. Removal of nail and reaming was done where applicable. Otherwise, access to the canal was created for debridement using the technique described. One case required a temporary circular external fixator after debridement (due to the extent of bone window created), and was stabilized with a hydrogel-coated intramedullary nail as a planned second-stage procedure.

One case had a sinus four months after debridement. However, during follow-up, the sinus has closed, and no further intervention was done. No patient from this group reported complications (Table IV).

While there were no CM type II OM reported, there were 39 cases of CM type III (72%). A total of 36 cases were fracture-related, three from hematogenous spread. Mean follow-up was 26 months. Implants were removed, cortical surfaces and screw holes debrided, with selected cases requiring access to medullary canal. In all, 28 cases used Stimulan, while seven used CERAMENT-G. Local antibiotics were not used in three cases; in one case, BonAlive (BonAlive Biomaterials, Finland) was used, in another case wherein the surgeon felt that local antibiotics were not required (frame extension to femur), and in one case, Stimulan was just introduced in our unit and the surgeon had reservations using it. Overall, 11 cases required a flap for soft-tissue coverage.

In all, seven CM type III cases had recurrence, wherein four had repeat surgery. Two cases had repeat debridement, local antibiotics, and a flap. The third case had a recurrent culture negative OM of the proximal radius, but was positive on histology. All three cases had resolution of symptoms on last follow-up. The fourth case had recurrence 791 days from debridement, and underwent repeat debridement, local antibiotics, and a flap. However, on follow-up, the patient had occasional discharge over the proximal tibia; repeat surgery was offered, but the patient chose conservative management (Table IV).

Three other cases of CM type III had recurrences but did not undergo repeat debridement. One case was advised amputation but refused. The other case had recurrence eight months after surgery, medically managed with antibiotics. The last case had intermittent discharge, but had poor follow-up and compliance.

In all, five cases had complications; three had sterile leak from local antibiotics, which eventually resolved, and two had sensory deficits (one around the knee, and one over the distribution of the superficial peroneal nerve).

There were five cases of CM type IV, four of which were fracture-related and one hematogenous spread. Mean follow-up was 42 months. One case was an infected segmental tibia fracture, and underwent removal of nail and reaming of canal, frame fixation, plus intramedullary antibiotic pellets. No wide resection was done. The second case was an infected subtrochanteric fracture nonunion, whereby an intramedullary nail was removed, reaming done, plus intramedullary antibiotic pellets. However, the limb was stable after nail removal and no wide resection was done, no fixation applied and nonunion eventually consolidated. The third case underwent debridement and exchange tibia nailing with a gentamicin-coated nail (Synthes Expert Tibia Nail PROtect; Synthes, Switzerland), and no wide resection was also done. These three cases had no recurrences nor complications on last follow-up.

The fourth case was an infected nonunion of a humerus fracture, which underwent removal of plate, debridement of cortex, over drilling of screw holes, and local antibiotics. The fracture was stabilized with a frame, and no wide resection was done. The fracture united with no recurrence of infection. However, the patient developed pin site OM due to the frame, and underwent debridement and local antibiotics. There was no recurrence of infection on last follow-up (Figure 2). The last case was a pathological femur fracture (hematogenous OM; intravenous drug user). The patient underwent debridement, reaming, local antibiotics, and frame fixation. However, the patient then required re-fixation with a frame ten months after first surgery after falling off a motorcycle. No recurrence of infection at two years and eight months after re-fixation.

Fig. 2

a) 39-year-old female type A host with infected nonunion of humerus fracture after open reduction and internal fixation. b) Intraoperative radiograph after plate removal, local antibiotics, and stabilization with frame. No wide resection was done. c) Interim radiograph showing good alignment and partial resorption of local antibiotics. d) Post removal of frame showing good union, no clinical signs and symptoms of recurrence of infection at 18 months after surgery.

Discussion

Surgical debridement of COM has seen significant changes in practice through time. Avicenna in the ninth century highlighted the need for eradication of dead bone through “deep excavation.” If this is not enough, then amputation should be done.¹² Significant developments through time included the use and eventual denouncement of cautery as treatment, and drainage and resection of dead bone.

Like Avicenna, the Orr treatment of pyogenic OM in 1936 emphasized total eradication of infected bone as much as possible.¹³ Total eradication was achieved through a generous incision, saucerization, and removal of all diseased bone. However, the Orr treatment did not come without complications. Prolonged immobilization in the cast (after significant bone resection) resulted in joint stiffness and contractures. In addition, immobilization did not promote circulation which further delayed healing. Hence, the treatment had various modifications including a longer and narrower corticotomy (to access the medullary canal) which provided sufficient bone stability for earlier mobilization.

Rowling et al¹⁴ in 1959 reported 58 cases of COM managed through a “sufficiently radical surgical attack” combined with soft-tissue reconstruction and antimicrobial therapy. In this series, the authors emphasized that “complete excision is essential,” and doing otherwise results in treatment failure. Similarly, many other authors presented their results anchored on “radical” debridement. McNally et al³ in 1993 reported 37 patients via a two-stage technique, emphasizing “radical debridement of all infected bone” and achieved union in 34. In “all cases, the excision of bone was radical with no attempt at a limited resection.” Bone with “abnormal appearance” were removed until bleeding was noted. Areas that did not bleed were re-examined and debrided to bleeding bone.

As such, wide tumour-like resection has been the prevalent paradigm in the management of OM for some decades now and, despite the difficulty in determining the limits of resection, is promoted by many authors, including Tiemann et al¹⁵ (2009), Forsberg et al² (2011), Panteli et al⁵ (2016), and Barakat et al¹ (2019).

Morelli et al¹⁶ reports that via the masquelet technique 91.1% of OM patients were in remission on last follow-up. Yalikun et al,¹⁷ on the other hand, reported an 11.4% recurrence of post-traumatic tibial OM treated with Ilizarov bone transport. These recurrence rates are comparable to the results of our series. However, excessive bone resection and the need for reconstruction comes with complications. Bone reconstruction usually is through distraction osteogenesis with external fixators or via masquelet technique.⁶ Liu et al¹⁸ reported that majority who underwent Ilizarov or monorail fixation suffered many complications during treatment. Common complications included pin site infection, axial deviation, soft-tissue incarceration, and delated union of docking site, and the most common major complication was joint stiffness. Morelli et al¹⁶ reviewed complications of the Masquelet technique. The technique requires two stages, which may not be suitable for certain populations. In addition, two-staged techniques increase morbidity and have cost implications to healthcare systems. The Masquelet technique for bone defects secondary to infection had a higher risk of complications, seen in almost half of cases (49.6%). Complications include superficial and deep infections, failure of either or both of the staged procedure, hardware failure, soft-tissue healing concerns, malalignment, joint stiffness, and refractures. In our study, two patients (4%) required surgery for complications. The most common complication was sterile leak from local antibiotics in three patients, which eventually resolved. Two reported mild sensory deficit after surgery. One had pin site OM secondary to external fixation of the humerus, and one required re-fixation of the femur after falling off a motorcycle.

Because of these, along with recent significant advancements in high concentration local antibiotic delivery and an MDT approach, adequate debridement has steadily challenged the need for wide, tumour-like resection in all cases of long bone COM. Adequate debridement is removal of enough bone to be able to debride all gross infection. Resection of devitalized bone is limited to gain access, which maintains primary stability of the bone. A channel is created as access for medullary debridement, and if a cloacae is present, it is extended to gain access as mentioned. In this technique, surgeons no longer look for punctate bleeding, or “paprika sign”, as a marker of healthy bone, meaning devitalized bone may be left behind.

Simpson et al⁴ in 2001 determined outcomes in patients with OM with varying debridement margins: wide resection with a margin greater than 5 mm, marginal resection less than 5 mm, and intralesional debridement. The wide resection group had no recurrence of OM versus 28% in the marginal excision group. In Simpson et al’s study, MRI was used to plan the extent of debridement but the extent of resection was decided during surgery. The paprika sign was used to determine viability of bone. In the marginal excision group, 50% of CM type B hosts had recurrence while there was no recurrence among type A hosts. Hence, the authors stated that they have shifted their practice towards marginal excision especially for type A hosts. Our study, with almost six of ten patients being type B hosts, reports a 15% recurrence rate compared to 28% in Simpson et al’s marginal excision group. With the use of local antibiotics, we suggest that adequate excision even in type B hosts may be sufficient.

Wide, tumour-like resection in search of the paprika sign was used to ensure local blood supply, hence delivery of systemic antibiotics. With the advent of dissolvable customized local antibiotic delivery systems with high minimum inhibitory concentration levels up to six weeks, the argument for wide resection is much weaker.

Single-staged management of OM is arguably the current gold standard when achievable. Successful outcomes of single-staged management of OM are made possible with advancements in local antibiotic delivery, dead space management, and soft-tissue coverage, a concept reported for a number of years now by McNally et al.⁷ This has been the experience of our unit, with an overall resolution rate of 90.7%. Segmental resections were more frequently performed in the early years, and with time has transitioned to most cases being treated with adequate debridement. Segmental resection still, however, has a role when indicated, such as in cases of grossly infected and unstable nonunion.

Overall, 17 studies showed a pooled non-recurrence rate of 90.9% via debridement and local antibiotics, similar to our unit’s results using this strategy.¹⁹ Ferguson et al²⁰ used biodegradable calcium sulphate for dead space management and delivery of local antibiotics with a recurrence rate of 9.2% and eventual resolution rate of 97.9%. Our unit’s use of calcium sulfate had a recurrence rate of 15% and overall success rate of 90.7% yielded comparable results.

Prophylactic stabilization avoids fractures after resection of infected bone, particularly when less than 70% of the cortical volume remains.¹⁰ Chan et al²¹ reported approximately 15% of their patients with type II and III OM benefited from prophylactic fixation. In our technique, access to the medullary canal is done through a trough small enough to maintain structural stability. Fortunately, in our series, no patient sustained a pathological fracture.

In our study, eight patients presented with recurrence of symptoms after surgical debridement. Garcia Del Pozo et al²² determined risk factors for recurrence after treatment of long bone COM. Significant risk factors were duration of OM greater than three months, exposure of bone, and “treatment other than surgical debridement with muscular flap.” Interestingly, only one of our eight recurrences presented with bone exposure. On the other hand, seven of the eight recurrences in our series had a duration of greater than three months.

Limitations

Our study has the inherent limitations of retrospective observational studies. There was no control group that underwent wide resection, which might have provided more insight on the benefits and disadvantages of adequate debridement. There is also the risk of selection bias, although in the same period of this study, our unit performed only four segmental resections for appropriate cases.

In conclusion, with the advent of local antibiotic delivery systems, the argument for wide, tumour-like resection during debridement of long bone COM is much weaker. Adequate debridement of COM achieves success through detailed preoperative evaluation and planning via MDT approach. We are proposing a discussion within the surgical community to re-evaluate the extent of bony debridement.