Journal of Orthopaedic Surgery 2011;19(3):341-5
Non-vascularised fibular graft as an intramedullary strut for infected non-union of the humerus G Gopisankar, Arockiaraj SV Justin, Manasseh Nithyananth, Vinoo M Cherian, VN Lee Department of Orthopaedics, Christian Medical College, Vellore, Tamilnadu, India
ABSTRACT Purpose. To review outcomes of 7 patients who underwent revision surgery for infected non-union of the humerus using a fibular graft as an intramedullary strut. Methods. Records of 7 men aged 29 to 59 (mean, 40) years with humeral diaphyseal infected nonunion who underwent fixation using a compression plate and a non-vascularised fibular graft as an intramedullary strut were reviewed. The mean number of previous surgeries was 2.7 (range, 2–4). Three of the patients had active draining sinuses previously. Their C-reactive protein levels were normal and tissue cultures negative. The remaining 4 patients had active draining sinuses. They first underwent implant removal and debridement. Tissue cultures confirmed infection in 3 of them. The mean duration between debridement and the index surgery was 5 (range, 3–10) months. Results. The mean length of the fibulae harvested was 13 (range, 12–15) cm. All 7 non-unions healed. The
mean time to healing was 5.4 (range, 4–8) months. The mean follow-up period was 15 (range, 13–24) months. All patients had weakness of the extensor hallucis longus, which improved to near normal at month 3. There was no donor-site morbidity. Three patients with active infection at presentation underwent repeat surgery. Two of them had wound washouts, and their non-unions went on to heal successfully; one underwent implant removal after union due to an active sinus. Six of the patients returned to their pre-injury activity level, and one endured a brachial plexus injury. Conclusion. Fixation using a compression plate and a non-vascularised fibular graft as an intramedullary strut achieved good outcome for infected non-union of the humerus despite prior multiple failed surgeries. Key words: fractures, ununited; humeral fractures; surgical wound infection; treatment failure
INTRODUCTION Most of non-unions of humeral diaphyseal fractures
Address correspondence and reprint requests to: Dr Manasseh Nithyananth, Department of Orthopaedics, Unit I, Christian Medical College and Hospital, Vellore, Tamilnadu, 632004, India. E-mail:
[email protected]
Journal of Orthopaedic Surgery
342 G Gopisankar et al.
are atrophic and usually heal following conventional treatment (open reduction and internal fixation with bone grafting).1,2 Management of non-union is challenging, especially in patients with wound infection following multiple failed surgeries and in the presence of scarred, adherent soft tissues, and inadequate bone stock. With the advent of locked compression plates, stabilisation can be achieved in osteoporotic bones. Nonetheless, in severely osteoporotic bones, internal fixation may fail even with locked compression plates.3 For patients with humeral non-union who also have multiple risk factors for failure of internal fixation, the use of a fibular graft as an intramedullary strut is an alternative.4–6 We reviewed outcomes of 7 patients who underwent revision surgery for infected nonunion of the humerus using a fibular graft as an intramedullary strut. MATERIALS AND METHODS Records of 7 men aged 29 to 59 (mean, 40) years with humeral diaphyseal infected non-union (Fig. 1) who underwent fixation using a compression plate and a non-vascularised fibular graft as an intramedullary strut between 2005 and 2009 were reviewed (Table). The mean number of previous surgeries was 2.7 (range, 2–4). Three of the patients had active draining sinuses 6 to 12 months earlier but not at presentation. Their C-reactive protein levels were normal and tissue cultures negative. The remaining 4 patients
had active draining sinuses. They first underwent implant removal and debridement. Tissue cultures confirmed infection in 3 of them. In one patient with methicillin-resistant Staphylococcus aureus, intravenous teicoplanin for 3 weeks was given. In another with methicillin-resistant Staphylococcus epidermidis, antibiotic-impregnated bone cement (4 g of vancomycin mixed with 40 g of cement) was used; 15 beads were made in 2 strings and placed in the wound. Patients were reviewed regularly after debridement. When their C-reactive protein level was normal and the wounds had healed well, the index surgery was performed. The mean duration between debridement and the index surgery was 5 (range, 3–10) months. The skin was incised through the anterolateral approach, which was modified depending on previous incisions and soft-tissue scarring. Intervening fibrous tissue was excised. The medullary canal was reestablished in both fragments. Serial reaming of the proximal and distal fragments was then performed with rigid Kuntscher nail reamers. Bone-holding forceps were used to hold the fragments especially near the fracture site so as to prevent splintering of osteoporotic bone. The length of the fibular graft needed was measured by adding the depth of the last reamer inserted into both fragments. An ipsilateral non-vascularised fibular graft was harvested by a standard technique. The diameter of the fibular graft was larger than that of the last reamer. An oscillating saw was used to reshape and reduce the diameter of the fibular graft to
Figure 1 Infected non-unions of the humeral shaft with failed internal fixation.
Vol. 19 No. 3, December 2011 Non-vascularised fibular graft as an intramedullary strut for infected non-union of the humerus 343 Table Patient characteristics and outcome Sex/age (years)
Patients with active draining sinuses M/59 M/51 M/30 M/36 Patients with active draining sinuses previously M/30 M/45 M/29
No. of Drainage- Tissue culture during previous free period debridement surgeries (months)
2 3
3 3
3 4
10 4
3 2 2
10 12 6
Nil Methicillin-resistant Staphylococcus epidermidis Enterobacter Methicillin-resistant Staphylococcus aureus
-
1 mm less than the last reamer size. The centre of the graft was marked. The fibular graft was then inserted into the proximal fragment, with the distal 4 to 5 cm exposed. The distal fragment was reduced onto the fibular graft protruding from the proximal fragment. Care was taken to avoid fracturing the fibular graft or distal fragment. Once the distal fragment was reduced, the fracture site was distracted. The fibular graft was then pushed into the distal fragment, until the central mark was at the fracture site. A contoured 4.5 mm narrow dynamic compression plate was fixed with 4 or 5 bicortical screws proximally and distally. Axial compression was applied manually prior to fixation. Autogenous iliac crest bone grafts were placed around the non-union area. The wound was then closed. Postoperative drains were retained for 48 hours. For the 3 patients with positive cultures, specific antibiotics were given for 6 weeks. For the 4 patients with no positive culture, antibiotics were given for 10 days until suture removal. An arm sling was used; plaster immobilisation was avoided. Mobilisation exercises were started after 2 days. Patients were followed up every 3 months until bone union. RESULTS The mean length of the fibulae harvested was 13 (range, 12–15) cm. All 7 non-unions healed (Figs. 2 and 3). The mean time to healing was 5.4 (range,
Tissue culture during fibular grafting
Duration of Complications antibiotic treatment after fibular grafting
Nil Methicillin-resistant Staphylococcus epidermidis Enterobacter Methicillin-sensitive Staphylococcus aureus
10 days 6 weeks
Washout
6 weeks 6 weeks
Washout Sinus
Nil Nil Nil
10 days 10 days 10 days
-
4–8) months. The mean follow-up period was 15 (range, 13–24) months. All patients had weakness of the extensor hallucis longus, which improved to near normal at month 3. There was no donor-site morbidity. Three patients with active infection at presentation underwent repeat surgery. The duration of antibiotic use after debridement was suboptimal (usually at least 6 weeks is advocated for osteomyelitis). At the time of index surgery, tissue cultures were positive although there was no macroscopic evidence of infection. A more rigorous assessment for infection should have been performed to confirm absence of infection before the index surgery. Two of them had wound washouts, and their non-unions went on to heal successfully. The third patient underwent implant removal after union owing to an active sinus. Six of the patients returned to their pre-injury activity level, and one endured a brachial plexus injury. DISCUSSION Treatment modalities for humeral non-unions include casting, bracing, compression plating, intramedullary nailing, and external fixation.7,8 In compression plate fixation, the screw purchase depends on the quality of the bone.9 Screw purchase may be inadequate owing to osteoporosis or previous surgery.10 Although locked compression plates provide improved pull-out
Journal of Orthopaedic Surgery
344 G Gopisankar et al.
(a)
(b)
Figure 2 (a) An infected non-union of the humeral shaft for 18 months after 4 previous surgeries; (b) one year after fixation showing good bone union.
(a)
(b)
Figure 3 (a) An infected non-union of the humeral shaft after debridement; (b) 2 years after fixation showing good bone union.
strength and fixation, implant failure can still occur.3 The intramedullary fibular graft provides good purchase for screws and adequate fixation.4 Rigid stabilisation is necessary for non-unions. In one series, 17 of 18 patients achieved bone union after fixation using the fibular allografts and locked compression plates for proximal humeral non-unions.6 Good outcome was also achieved in 6 patients after fixation with fibular grafts, when those with poor soft tissues and infection were excluded.5 In selected patients, adequate stability can be obtained with non-locked
plates and fibular grafts. In a biomechanical study of fixation using a locked compression plate with and without fibular strut grafting for proximal humerus osteotomy with a gap on the medial side,11 none of the constructs augmented with the fibular graft failed, as opposed to 6 of 8 constructs not so augmented failed. The fibular graft can increase the pull-out strength even for locked compression plates. In arthroplasty surgery, >10 neutrophils per high power field in a frozen section is indicative of infection.12 Non-vascularised grafts have less
Vol. 19 No. 3, December 2011 Non-vascularised fibular graft as an intramedullary strut for infected non-union of the humerus 345
resistance to infection. If infection remains, redebridement and local antibiotic delivery should be performed. In our study, repeat surgeries could have been avoided if the absence of infection had been confirmed before the index surgery. In our study, postoperative immobilisation with plaster could be avoided owing to adequate
stabilisation after fibular strut grafting. In patients with multiple surgeries, the joints are already stiff. Any further immobilisation leads to more stiffness and suboptimal outcome. The limitations of this study were its retrospective nature and the absence of functional assessment before and after surgery.
REFERENCES 1. Epps CH Jr, Grant RE. Fractures of the shaft of the humerus. In: Rockwood CA Jr. Green DP, Bucholz RW, editors. Rockwood and Green’s fractures in adults. 3rd ed, Vol 1. Philadelphia: JB Lippincott: 1991:843–69. 2. Jupiter JB. Complex non-union of the humeral diaphysis. Treatment with a medial approach, an anterior plate, and a vascularized fibular graft. J Bone Joint Surg Am 1990;72:701–7. 3. Micic ID, Kim KC, Shin DJ, Shin SJ, Kim PT, Park IH, et al. Analysis of early failure of the locking compression plate in osteoporotic proximal humerus fractures. J Orthop Sci 2009;14:596–601. 4. Daniel AJ, Sundararaj GD. Treatment of complex nonunion of the humeral shaft. J Orthop Surg (Hong Kong) 1996;4:79–82. 5. Vidyadhara S, Vamsi K, Rao SK, Gnanadoss JJ, Pandian S. Use of intramedullary fibular strut graft: a novel adjunct to plating in the treatment of osteoporotic humeral shaft nonunion. Int Orthop 2009;33:1009–14. 6. Badman BL, Mighell M, Kalandiak SP, Prasarn M. Proximal humeral nonunions treated with fixed-angle locked plating and an intramedullary strut allograft. J Orthop Trauma 2009;23:173–9. 7. Dhar SA, Kawoosa AA, Butt MF, Ali MF, Mir MR, Halwai MA. Acute invaginating docking for infected non-unions of the humerus. J Orthop Surg (Hong Kong) 2008;16:290–4. 8. Brinker MR, O’Connor DP, Crouch CC, Mehlhoff TL, Bennett JB. Ilizarov treatment of infected nonunions of the distal humerus after failure of internal fixation: an outcomes study. J Orthop Trauma 2007;21:178–84. 9. Schatzker J, Sanderson R, Murnaghan JP. The holding power of orthopedic screws in vivo. Clin Orthop Relat Res 1975;108:115–26. 10. Martinez AA, Cuenca J, Herrera A. Two-plate fixation for humeral shaft non-unions. J Orthop Surg (Hong Kong) 2009;17:135– 8. 11. Bouliane M, Begum F, Beaupre L, Carey JP, Adeeb S, Chow R. Proximal humerus fracture fixation: locking plate construct +/- intramedullary fibular allograft. Paper presented at American Academy of Orthopaedic Surgeons Annual Meeting; 2011 February 15-19; San Diego, California, USA. 12. Lonner JH, Desai P, Dicesare PE, Steiner G, Zuckerman JD. The reliability of analysis of intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J Bone Joint Surg Am 1996;78:1553–8.
